Spinach carries the most cadmium of the common leafy greens and head cabbage among the least; lead is usually low in monitored retail leaves but rises sharply near mining, industry, roads and contaminated irrigation water.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →2FY2018-FY2020 TDS Elements Analytical ResultsFDA Total Diet StudyRead source →3Heavy metal accumulation in leafy vegetables grown in industrial areas under varying levels of pollutionBangladesh Journal of Agricultural ResearchRead source →
-
Metals of concern:
Cadmium decides most leafy-green questions; lead matters near pollution sources; arsenic is high in some aquatic leaves.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →4Scientific Opinion of the Panel on Contaminants in the Food Chain on a request from the European Commission on cadmium in foodThe EFSA JournalRead source →5The First Hong Kong Total Diet Study: Inorganic ArsenicCentre for Food Safety, Food and Environmental Hygiene Department, Government of the Hong Kong Special Administrative RegionRead source →
-
Main drivers:
Leaf type, growing site and irrigation water, then surface dust that washing can remove.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →6Mercury pollution in vegetables, grains and soils from areas surrounding coal-fired power plantsScientific ReportsRead source →7Concentrations and health risks of selected elements in leafy vegetables: a comparison between roadside open-air markets and large stores in Johannesburg, South AfricaNot reportedRead source →
-
Practical framing:
Ask for lot cadmium and lead results by leaf on a fresh-weight basis, with the growing site.2FY2018-FY2020 TDS Elements Analytical ResultsFDA Total Diet StudyRead source →8Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →
Leafy greens overview
Cadmium is the metal that most often decides a leafy-green question, and the leaf type matters more than any other single factor in the monitoring data.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →2FY2018-FY2020 TDS Elements Analytical ResultsFDA Total Diet StudyRead source → Across more than 2,200 WHO GEMS/Food spinach results the median was about 41 µg/kg fresh weight, against 19–20 µg/kg for leaf lettuce, 11–12 µg/kg for head lettuce and 2–4 µg/kg for head cabbage.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source → The FDA Total Diet Study found the same order in US market composites from 2018 to 2020: spinach 200, leaf lettuce 57, iceberg lettuce 27 and cabbage 5.1 µg/kg as medians.2FY2018-FY2020 TDS Elements Analytical ResultsFDA Total Diet StudyRead source → European monitoring summarised by EFSA gave spinach a mean of 61.5 µg/kg and other leafy vegetables 23.1 µg/kg, with 3.9% of spinach samples above the EU limit.4Scientific Opinion of the Panel on Contaminants in the Food Chain on a request from the European Commission on cadmium in foodThe EFSA JournalRead source →
Lead is usually low in routinely monitored retail leaves.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →2FY2018-FY2020 TDS Elements Analytical ResultsFDA Total Diet StudyRead source → Most head lettuce and cabbage results in GEMS/Food were below detection, most FDA lettuce and cabbage composites were below a 4 µg/kg reporting limit, and EFSA’s European leafy-vegetable median was 10 µg/kg.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →2FY2018-FY2020 TDS Elements Analytical ResultsFDA Total Diet StudyRead source →9Scientific Opinion on Lead in FoodEFSA Journal 2010;8(4):1570Read source → The EU limits are 0.30 mg/kg (300 µg/kg) lead for leaf vegetables and 0.10 mg/kg cadmium, raised to 0.20 mg/kg for spinach and similar leaves and fresh herbs.8Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →
Field studies near mining, smelting, coal power plants, wastewater irrigation or heavy traffic report leaves tens to hundreds of times more contaminated than market monitoring.3Heavy metal accumulation in leafy vegetables grown in industrial areas under varying levels of pollutionBangladesh Journal of Agricultural ResearchRead source →6Mercury pollution in vegetables, grains and soils from areas surrounding coal-fired power plantsScientific ReportsRead source →7Concentrations and health risks of selected elements in leafy vegetables: a comparison between roadside open-air markets and large stores in Johannesburg, South AfricaNot reportedRead source →10Is it healthy urban agriculture? Human exposure to potentially toxic elements in urban gardens from Andalusia, SpainEnvironmental Science and Pollution ResearchRead source → Many of those studies report on a dry-weight basis, which by itself raises concentrations about ten-fold, and several fail checks against their own methods, so the page shows them without comparing them.11Food survey: Levels and potential health risks of chromium, lead, zinc and copper content in fruits and vegetables consumed in AlgeriaFood and Chemical ToxicologyRead source →12Heavy Metals in Commonly Consumed Root and Leafy Vegetables in Dhaka City, Bangladesh, and Assessment of Associated Public Health RisksEnvironmental Systems ResearchRead source →13Enrichment Characteristics and Health Risk Assessment of Heavy Metals in Soil-Crop SystemsKSCE Journal of Civil EngineeringRead source → For a buyer the supported priorities are the growing site and irrigation water, the leaf type, and lot results for cadmium and lead reported on a fresh-weight basis.6Mercury pollution in vegetables, grains and soils from areas surrounding coal-fired power plantsScientific ReportsRead source →8Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →14Uptake of Arsenic by Irrigated Vegetables and Cooked Food Products in Burkina FasoFrontiers in WaterRead source →
Leafy greens and their preparations
This page covers edible leaves eaten as vegetables, each kept as its own material: spinach, leaf, head and romaine lettuce, head cabbage, Chinese cabbage, pak choi, kale, chard, amaranth, water spinach, mustard greens, watercress, lamb’s lettuce, rocket and leafy herbs such as coriander, parsley, mint and dill.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →15Evaluation of essential, toxic and potentially toxic elements in leafy vegetables grown in the Canary IslandsToxicsRead source →16The First Hong Kong Total Diet Study: Metallic ContaminantsCentre for Food Safety, Food and Environmental Hygiene Department, Government of the Hong Kong Special Administrative RegionRead source → Cooked, frozen, canned, fermented and dried leaves are kept apart from raw leaves, because preparation changes water content and can add or remove metals.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →14Uptake of Arsenic by Irrigated Vegetables and Cooked Food Products in Burkina FasoFrontiers in WaterRead source →
The basis of a result matters as much as the leaf. Fresh leaves are mostly water, about 92% in the Algerian spinach and salad samples, so a dry-weight result is roughly ten times the fresh-weight value for the same leaf.11Food survey: Levels and potential health risks of chromium, lead, zinc and copper content in fruits and vegetables consumed in AlgeriaFood and Chemical ToxicologyRead source → The page never converts between fresh and dry weight, and results whose basis is not stated stay separate.17Nutrients and non-essential metals in darkibor kale grown at urban and rural farms: a pilot studyPLOS ONERead source →11Food survey: Levels and potential health risks of chromium, lead, zinc and copper content in fruits and vegetables consumed in AlgeriaFood and Chemical ToxicologyRead source → Seaweed sold as sea lettuce, tea and medicinal herb leaves, mixed baby foods that contain spinach, Brussels sprouts, roots and stems are outside this scope.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →2FY2018-FY2020 TDS Elements Analytical ResultsFDA Total Diet StudyRead source →18Total Diet Study of metals and other elements in foodFood and Environment Research Agency report for the UK Food Standards Agency, Fera report 15/06, project FS102081Read source →
Heavy metal contamination profile of leafy greens
The table holds every leafy-green result extracted from 98 sampling families, including WHO GEMS/Food submissions summarised by leaf, region and sampling type, the FDA and UK total diet studies, Hong Kong’s total diet study and every field and market survey in the library.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →2FY2018-FY2020 TDS Elements Analytical ResultsFDA Total Diet StudyRead source →15Evaluation of essential, toxic and potentially toxic elements in leafy vegetables grown in the Canary IslandsToxicsRead source →16The First Hong Kong Total Diet Study: Metallic ContaminantsCentre for Food Safety, Food and Environmental Hygiene Department, Government of the Hong Kong Special Administrative RegionRead source →18Total Diet Study of metals and other elements in foodFood and Environment Research Agency report for the UK Food Standards Agency, Fera report 15/06, project FS102081Read source → Each row is one statistic for one leaf, group and metal on its stated basis; ratings place a mean, median or sample value against a pooled HMI reference for the same leaf and basis.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →2FY2018-FY2020 TDS Elements Analytical ResultsFDA Total Diet StudyRead source →17Nutrients and non-essential metals in darkibor kale grown at urban and rural farms: a pilot studyPLOS ONERead source →
Non-detects stay as intervals from zero to the reporting limit and are never read as zero.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →2FY2018-FY2020 TDS Elements Analytical ResultsFDA Total Diet StudyRead source → Rows that fail a check against their own paper, and values read from a figure rather than printed, stay visible with the reason and are not compared.7Concentrations and health risks of selected elements in leafy vegetables: a comparison between roadside open-air markets and large stores in Johannesburg, South AfricaNot reportedRead source →12Heavy Metals in Commonly Consumed Root and Leafy Vegetables in Dhaka City, Bangladesh, and Assessment of Associated Public Health RisksEnvironmental Systems ResearchRead source →13Enrichment Characteristics and Health Risk Assessment of Heavy Metals in Soil-Crop SystemsKSCE Journal of Civil EngineeringRead source →
Comparison methods and scope
Each row is one statistic a source reports for one leaf, subgroup and metal; a mean, median and range from the same samples are separate rows. Ratings place a reported mean, median or sample result against a pooled HMI reference for the same leaf and basis (fresh or dry weight) built from every compatible sampling family. Same-study comparisons set a site or irrigation group against its own control or study total. Held rows failed a check against their own paper and are shown but not compared.
First displayed analyte: Lead: 524 findings from 76 sources. Each row retains its statistic and measurement basis.
Filter findings
2123 of 2123 findings · 12 analytes
About this table
| Food form | Collection scope | Finding | Comparison | Evidence |
|---|---|---|---|---|
48 findings · 6 studies | ||||
| Amaranth leaves (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010Fresh weight | median. | ||
| Chinese cabbage (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Head cabbage (cooked)WHO GEMS/Food | WHO region AFRO; random sampling2018Fresh weight | median. | ||
| Head cabbage (cooked)WHO GEMS/Food | WHO region EURO; random sampling2009–2019Fresh weight | median. | ||
| Head cabbage (cooked)WHO GEMS/Food | WHO region PAHO; random sampling2000–2017Fresh weight | median. | ||
| Head cabbage (cooked)WHO GEMS/Food | WHO region SEARO2011Fresh weight | median. | ||
| Head cabbage (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Head cabbage (cooked)WHO GEMS/Food | WHO region WPRO; targeted sampling2008Fresh weight | median. | ||
| Head lettuce (cooked)WHO GEMS/Food | WHO region SEARO2009–2010Fresh weight | median. | ||
| Head lettuce (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010Fresh weight | median. | ||
| Kale (cooked)WHO GEMS/Food | WHO region EURO; random sampling2009–2017Fresh weight | median. | ||
| Kale (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Leaf lettuce (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Leafy vegetables, other or unspecified (cooked)WHO GEMS/Food | WHO region AFRO; random sampling2018Fresh weight | median. | ||
| Leafy vegetables, other or unspecified (cooked)WHO GEMS/Food | WHO region PAHO; random sampling2000–2017Fresh weight | median. | ||
| Leafy vegetables, other or unspecified (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Leafy vegetables, other or unspecified (cooked)WHO GEMS/Food | WHO region WPRO; targeted sampling1998Fresh weight | median. | ||
| Mustard greens (cooked)WHO GEMS/Food | WHO region SEARO2011Fresh weight | median. | ||
| Mustard greens (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Pak choi (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2014Fresh weight | median. | ||
| Spinach (cooked)WHO GEMS/Food | WHO region EURO; random sampling2007–2019Fresh weight | median. | ||
| Spinach (cooked)WHO GEMS/Food | WHO region PAHO; random sampling2000–2017Fresh weight | median. | ||
| Spinach (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Water spinach (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010Fresh weight | median. | ||
| Watercress (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Collard greens (cooked)FDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Kale (cooked)FDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Amaranth leaves (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Chinese cabbage (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Chinese kale (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Choy sum (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Head cabbage (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | not detected. | ||
| Lettuce (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Lettuce (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Mustard greens (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Pak choi (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Spinach (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Water spinach (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Watercress (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Leafy vegetables, other or unspecified (cooked)Mahaffey et al. 1975 | US FDA Total Diet Study FY 19731968–1974Basis not stated | mean. | ||
| Head cabbage (cooked)UK TDS 2014 (Fera 2015) | UK TDS 2014 category 11_622014Fresh weight | sample value. | ||
| Other green vegetables (mixed category) (cooked)UK TDS 2014 (Fera 2015) | UK TDS 2014 category 11_682014Fresh weight | below LOD. | ||
| Chinese cabbage (cooked)Wang et al. 2020 | Jilin ProvinceCollection date unreportedBasis not stated | mean. | ||
| Chinese cabbage (cooked)Wang et al. 2020 | Other AreaCollection date unreportedBasis not stated | mean. | ||
| Chinese cabbage (cooked)Wang et al. 2020 | Songhua River AreaCollection date unreportedBasis not stated | mean. | ||
| Leek (cooked)Wang et al. 2020 | Jilin ProvinceCollection date unreportedBasis not stated | mean. | ||
| Leek (cooked)Wang et al. 2020 | Other AreaCollection date unreportedBasis not stated | mean. | ||
| Leek (cooked)Wang et al. 2020 | Songhua River AreaCollection date unreportedBasis not stated | mean. | ||
436 findings · 73 studies · 28 spread figures folded in | ||||
| Chinese cabbageWHO GEMS/Food | WHO region PAHO; random sampling2014–2016Fresh weight | median. | ||
| Chinese cabbageWHO GEMS/Food | WHO region SEARO; random sampling2012Fresh weight | median. | ||
| Chinese cabbageWHO GEMS/Food | WHO region WPRO; random sampling2003–2013Fresh weight | median. | ||
| Head cabbageWHO GEMS/Food | WHO region EURO; random sampling2003–2019Fresh weight | median. | ||
| Head cabbageWHO GEMS/Food | WHO region EURO; targeted sampling1999–2001Fresh weight | median. | ||
| Head cabbageWHO GEMS/Food | WHO region EURO; unknown2003–2017Fresh weight | median. | ||
| Head cabbageWHO GEMS/Food | WHO region PAHO; random sampling2003–2017Fresh weight | median. | ||
| Head cabbageWHO GEMS/Food | WHO region SEARO; random sampling2012–2014Fresh weight | median. | ||
| Head cabbageWHO GEMS/Food | WHO region SEARO; unknown2012Fresh weight | median. | ||
| Head cabbageWHO GEMS/Food | WHO region WPRO; random sampling2003–2013Fresh weight | median. | ||
| Head lettuceWHO GEMS/Food | WHO region EURO; random sampling1997–1999Dry weight | median. | ||
| Head lettuceWHO GEMS/Food | WHO region EURO; random sampling2003–2019Fresh weight | median. | ||
| Head lettuceWHO GEMS/Food | WHO region EURO; unknown2007–2017Fresh weight | median. | ||
| Head lettuceWHO GEMS/Food | WHO region PAHO; random sampling2000–2017Fresh weight | median. | ||
| Head lettuceWHO GEMS/Food | WHO region WPRO; random sampling2003–2014Fresh weight | median. | ||
| Head lettuceWHO GEMS/Food | WHO region WPRO; targeted sampling1998–2008Fresh weight | median. | ||
| KaleWHO GEMS/Food | WHO region EURO; random sampling2003–2019Fresh weight | median. | ||
| KaleWHO GEMS/Food | WHO region EURO; unknown2005–2018Fresh weight | median. | ||
| KaleWHO GEMS/Food | WHO region PAHO; random sampling2014–2016Fresh weight | median. | ||
| KaleWHO GEMS/Food | WHO region SEARO; random sampling2012–2013Fresh weight | median. | ||
| KaleWHO GEMS/Food | WHO region SEARO; unknown2012Fresh weight | median. | ||
| KaleWHO GEMS/Food | WHO region WPRO; targeted sampling2001Fresh weight | median. | ||
| Lamb's lettuceWHO GEMS/Food | WHO region EURO; random sampling2003–2019Fresh weight | median. | ||
| Lamb's lettuceWHO GEMS/Food | WHO region EURO; unknown2006–2017Fresh weight | median. | ||
| Lamb's lettuceWHO GEMS/Food | WHO region SEARO; random sampling2013–2014Fresh weight | median. | ||
| Lamb's lettuceWHO GEMS/Food | WHO region WPRO; random sampling2013Fresh weight | median. | ||
| Leaf lettuceWHO GEMS/Food | WHO region EURO; random sampling2003–2019Fresh weight | median. | ||
| Leaf lettuceWHO GEMS/Food | WHO region EURO; unknown2003–2018Fresh weight | median. | ||
| Leaf lettuceWHO GEMS/Food | WHO region PAHO; random sampling2003–2017Fresh weight | median. | ||
| Leaf lettuceWHO GEMS/Food | WHO region SEARO; random sampling2012Fresh weight | median. | ||
| Leaf lettuceWHO GEMS/Food | WHO region WPRO; random sampling2007–2009Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region EURO; random sampling2003–2019Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region EURO; unknown2003–2017Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region WPRO; random sampling2005–2009Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region WPRO; targeted sampling2001Fresh weight | median. | ||
| Mustard greensWHO GEMS/Food | WHO region EURO; unknown2015–2016Fresh weight | median. | ||
| Mustard greensWHO GEMS/Food | WHO region PAHO; random sampling2016Fresh weight | median. | ||
| Mustard greensWHO GEMS/Food | WHO region SEARO; random sampling2014Fresh weight | median. | ||
| Mustard greensWHO GEMS/Food | WHO region WPRO; random sampling2005–2009Fresh weight | median. | ||
| Mustard greensWHO GEMS/Food | WHO region WPRO; targeted sampling2002Fresh weight | median. | ||
| Pak choiWHO GEMS/Food | WHO region EURO; random sampling2003–2019Fresh weight | median. | ||
| Pak choiWHO GEMS/Food | WHO region EURO; unknown2003–2018Fresh weight | median. | ||
| Pak choiWHO GEMS/Food | WHO region PAHO; random sampling2014–2016Fresh weight | median. | ||
| Pak choiWHO GEMS/Food | WHO region SEARO; random sampling2012Fresh weight | median. | ||
| Pak choiWHO GEMS/Food | WHO region WPRO; random sampling2007–2009Fresh weight | median. | ||
| Pak choiWHO GEMS/Food | WHO region WPRO; targeted sampling2001Fresh weight | median. | ||
| SpinachWHO GEMS/Food | WHO region EURO; random sampling2003–2019Fresh weight | median. | ||
| SpinachWHO GEMS/Food | WHO region EURO; targeted sampling2000Fresh weight | median. | ||
| SpinachWHO GEMS/Food | WHO region EURO; unknown2003–2018Fresh weight | median. | ||
| SpinachWHO GEMS/Food | WHO region PAHO; random sampling2005–2016Fresh weight | median. | ||
| SpinachWHO GEMS/Food | WHO region SEARO; random sampling2012Fresh weight | median. | ||
| SpinachWHO GEMS/Food | WHO region WPRO; random sampling2003–2016Fresh weight | median. | ||
| WatercressWHO GEMS/Food | WHO region EURO; random sampling2019–2020Fresh weight | median. | ||
| WatercressWHO GEMS/Food | WHO region EURO; unknown2003–2008Fresh weight | median. | ||
| WatercressWHO GEMS/Food | WHO region PAHO; random sampling2016Fresh weight | median. | ||
| WatercressWHO GEMS/Food | WHO region SEARO; random sampling2012Fresh weight | median. | ||
| Head cabbageFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Head lettuceFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Leaf lettuceFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| SpinachFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| BasilAbdolahpour et al. 2023 | Babol2021Fresh weight | mean. | ||
| ParsleyAbdolahpour et al. 2023 | Babol2021Fresh weight | mean. | ||
| SpinachAbdolahpour et al. 2023 | Babol2021Fresh weight | mean. | ||
| SpinachAbubakar et al. 2025 | Site ICollection date unreportedBasis not stated | mean. | ||
| Amaranth leavesAdebayo et al. 2020 | Railway Quarters irrigation farmCollection date unreportedDry weight | mean. | ||
| Coriander leavesAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| Coriander leavesAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| Coriander leavesAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| Coriander leavesAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | mean. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | mean. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | mean. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | mean. | ||
| LettuceAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| LettuceAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| LettuceAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| LettuceAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| LeekAli 2024 | FarmerYear not reportedDry weight | mean (figure-read). | ||
| LeekAli 2024 | FarmerYear not reportedDry weight | mean. | ||
| LeekAli 2024 | MarketYear not reportedDry weight | mean. | ||
| LeekAli 2024 | MarketYear not reportedDry weight | mean (figure-read). | ||
| LeekAli 2024 | RestaurantYear not reportedDry weight | mean (figure-read). | ||
| LeekAli 2024 | RestaurantYear not reportedDry weight | mean (figure-read). | ||
| Head cabbageAlimohammadi et al. 2018 | Autumn 20162016Fresh weight | mean. | ||
| Head cabbageAlimohammadi et al. 2018 | Summer 20162016Fresh weight | mean. | ||
| LettuceAlimohammadi et al. 2018 | Autumn 20162016Fresh weight | mean. | ||
| LettuceAlimohammadi et al. 2018 | Summer 20162016Fresh weight | mean. | ||
| Head cabbageAmetepey et al. 2018 | Central and Aboabo MarketsCollection date unreportedDry weight | below LOD. | ||
| Head cabbageArmand et al. 2026 | 10 retail vegetable storesCollection date unreportedFresh weight | mean. | ||
| Head cabbageArmand et al. 2026 | 10 retail vegetable storesCollection date unreportedFresh weight | range. | ||
| LettuceArmand et al. 2026 | 10 retail vegetable storesCollection date unreportedFresh weight | mean. | ||
| LettuceArmand et al. 2026 | 10 retail vegetable storesCollection date unreportedFresh weight | range. | ||
| LettuceBambara et al. 2023 | Loumbila market gardenCollection date unreportedDry weight | mean. | ||
| Spring onion leavesBambara et al. 2023 | Loumbila market gardenCollection date unreportedDry weight | mean. | ||
| DillBora et al. 2022 | Vegetable and fruit market2020–2021Fresh weight | mean. | ||
| Head cabbageBora et al. 2022 | Amateur farmers2020–2021Fresh weight | below LOD. | ||
| Head cabbageBora et al. 2022 | Amateur farmers2020–2021Fresh weight | below LOD. | ||
| Head cabbageBora et al. 2022 | Amateur farmers2020–2021Fresh weight | below LOD. | ||
| Head cabbageBora et al. 2022 | Vegetable and fruit market2020–2021Fresh weight | below LOD. | ||
| Head cabbageBora et al. 2022 | Vegetable and fruit market2020–2021Fresh weight | below LOD. | ||
| Head cabbageBora et al. 2022 | Vegetable and fruit market2020–2021Fresh weight | mean. | ||
| LettuceBora et al. 2022 | Amateur farmers2020–2021Fresh weight | mean. | ||
| LettuceBora et al. 2022 | Amateur farmers2020–2021Fresh weight | mean. | ||
| LettuceBora et al. 2022 | Amateur farmers2020–2021Fresh weight | mean. | ||
| LettuceBora et al. 2022 | Vegetable and fruit market2020–2021Fresh weight | mean. | ||
| LettuceBora et al. 2022 | Vegetable and fruit market2020–2021Fresh weight | mean. | ||
| LettuceBora et al. 2022 | Vegetable and fruit market2020–2021Fresh weight | mean. | ||
| LovageBora et al. 2022 | Amateur farmers2020–2021Fresh weight | below LOD. | ||
| ParsleyBora et al. 2022 | Amateur farmers2020–2021Fresh weight | below LOD. | ||
| ParsleyBora et al. 2022 | Vegetable and fruit market2020–2021Fresh weight | below LOD. | ||
| TarragonBora et al. 2022 | Vegetable and fruit market2020–2021Fresh weight | below LOD. | ||
| Leafy vegetables, other or unspecifiedChen et al. 2025 | Chongqing retail/market samples2017–2021Basis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedChen et al. 2025 | Chongqing retail/market samples2017–2021Basis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedChen et al. 2025 | Chongqing retail/market samples2017–2021Basis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedCherfi et al. 2014 | Two wholesale markets supplying2012Dry weight | mean. | ||
| SpinachCherfi et al. 2014 | Two wholesale markets supplying2012Dry weight | mean. | ||
| Amaranth leavesChiutula et al. 2025 | Soche WWTP effluent-irrigatedYear not reportedDry weight | not detected. | ||
| Chinese cabbageChiutula et al. 2025 | Soche WWTP effluent-irrigatedYear not reportedDry weight | not detected. | ||
| Mustard greensChiutula et al. 2025 | Soche WWTP effluent-irrigatedYear not reportedDry weight | mean. | ||
| Pumpkin leavesChiutula et al. 2025 | Soche WWTP effluent-irrigatedYear not reportedDry weight | not detected. | ||
| Rape greensChiutula et al. 2025 | Soche WWTP effluent-irrigatedYear not reportedDry weight | not detected. | ||
| Sweet potato leavesChiutula et al. 2025 | Soche WWTP effluent-irrigatedYear not reportedDry weight | mean. | ||
| Beet leavesChmielewski et al. 2025 | 16 allotment gardens2022–2023Dry weight | mean. | ||
| Celery leavesChmielewski et al. 2025 | 16 allotment gardens2022–2023Dry weight | mean. | ||
| ParsleyChmielewski et al. 2025 | 16 allotment gardens2022–2023Dry weight | mean. | ||
| LettuceDearing et al. 2025 | Hawke's Bay market gardens2024Fresh weight | max. | ||
| Leafy vegetables, other or unspecifiedEFSA 2010 | EU monitoring data 2003-20082003–2008Basis not stated | 5th percentile. | ||
| Leafy vegetables, other or unspecifiedEFSA 2010 | EU monitoring data 2003-20082003–2008Basis not stated | 95th percentile. | ||
| Leafy vegetables, other or unspecifiedEFSA 2010 | EU monitoring data 2003-20082003–2008Basis not stated | mean.Same samples: 0 µg/kg (5th percentile); 115 µg/kg (95th percentile); 11,300 µg/kg (maximum) | ||
| Leafy vegetables, other or unspecifiedEFSA 2010 | EU monitoring data 2003-20082003–2008Basis not stated | median. | ||
| Leafy vegetables, other or unspecifiedEFSA 2010 | EU monitoring data 2003-20082003–2008Basis not stated | max. | ||
| SpinachElbagermi et al. 2012 | Misurata City local suppliers and2010Dry weight | mean. | ||
| LettuceEmmanuel 2025 | Sharada2025Basis not stated | mean. | ||
| LettuceEmmanuel 2025 | Wudil2025Basis not stated | mean. | ||
| SpinachEmmanuel 2025 | Nomans-Land2025Basis not stated | mean. | ||
| Amaranth leavesFatai et al. 2024 | Vegetable farms2022Basis not stated | mean. | ||
| Head cabbageFonge et al. 2021 | SantaCollection date unreportedDry weight | mean. | ||
| Coriander leavesGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | mean.Same samples: 746–1,180 µg/kg (range); 781–1,280 µg/kg (range); 940–1,180 µg/kg (range); 866–1,140 µg/kg (range); 836–1,160 µg/kg (range); 776–1,180 µg/kg (range); 900–1,160 µg/kg (range); 870–8,170 µg/kg (range); 849–1,140 µg/kg (range); 936–1,170 µg/kg (range) | ||
| Coriander leavesGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | range. | ||
| Fenugreek leavesGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | mean. | ||
| Fenugreek leavesGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | range. | ||
| Garden cressGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | mean. | ||
| Garden cressGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | range. | ||
| Leafy vegetables, other or unspecifiedGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | mean. | ||
| LeekGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | mean. | ||
| LeekGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | range. | ||
| MintGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | mean. | ||
| MintGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | range. | ||
| ParsleyGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | mean. | ||
| ParsleyGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | range. | ||
| Radish greens (tissue not stated)Ghasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | mean. | ||
| Radish greens (tissue not stated)Ghasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | range. | ||
| SpinachGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | mean. | ||
| SpinachGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | range. | ||
| Spring onion leavesGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | mean. | ||
| Spring onion leavesGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | range. | ||
| TarragonGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | mean. | ||
| TarragonGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | range. | ||
| Head cabbageGurmu et al. 2025 | Akaki Rivers irrigated farmlandYear not reportedDry weight | mean. | ||
| Head cabbageHabu et al. 2025 | Tamburawa wastewater-irrigatedCollection date unreportedBasis not stated | mean. | ||
| Head cabbageHabu et al. 2025 | Tamburawa wastewater-irrigatedCollection date unreportedBasis not stated | not detected. | ||
| LettuceHabu et al. 2025 | Tamburawa wastewater-irrigatedCollection date unreportedBasis not stated | not detected. | ||
| LettuceHabu et al. 2025 | Tamburawa wastewater-irrigatedCollection date unreportedBasis not stated | not detected. | ||
| SpinachHabu et al. 2025 | Tamburawa wastewater-irrigatedCollection date unreportedBasis not stated | not detected. | ||
| SpinachHabu et al. 2025 | Tamburawa wastewater-irrigatedCollection date unreportedBasis not stated | not detected. | ||
| Spring onion leavesHabu et al. 2025 | Tamburawa wastewater-irrigatedCollection date unreportedBasis not stated | not detected. | ||
| LettuceHeshmati et al. 2020 | Hamadan Province2017–2018Dry weight | mean.Same samples: 2–70 µg/kg (range) | ||
| LettuceHeshmati et al. 2020 | Hamadan Province2017–2018Fresh weight | mean. | ||
| LettuceHeshmati et al. 2020 | Hamadan Province2017–2018Fresh weight | range. | ||
| Amaranth leavesHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | 75th percentile. | ||
| Amaranth leavesHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | median.Same samples: 13,800 µg/kg (75th percentile); 38,400 µg/kg (maximum); 12,700 µg/kg (75th percentile); 29,800 µg/kg (maximum) | ||
| Amaranth leavesHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | not detected. | ||
| Amaranth leavesHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | min. | ||
| Amaranth leavesHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | max. | ||
| Head cabbageHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | 75th percentile. | ||
| Head cabbageHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | median. | ||
| Head cabbageHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | not detected. | ||
| Head cabbageHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | min. | ||
| Head cabbageHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | max. | ||
| Malabar spinachHossen et al. 2026 | farm-82024Fresh weight | mean. | ||
| Chinese cabbageHuang et al. 2014 | Zhejiang province2012Fresh weight | mean. | ||
| Head cabbageHuang et al. 2014 | Zhejiang province2012Fresh weight | mean. | ||
| Pak choiHuang et al. 2014 | Zhejiang province2012Fresh weight | mean. | ||
| Rape greensHuang et al. 2014 | Zhejiang province2012Fresh weight | mean. | ||
| SpinachHuang et al. 2014 | Zhejiang province2012Fresh weight | mean. | ||
| Chinese cabbageJiang et al. 2020 | SZD village near a battery factory2018Basis not stated | mean. | ||
| Pak choiJiang et al. 2020 | SZD village near a battery factory2018Basis not stated | mean. | ||
| SpinachJiang et al. 2020 | SZD village near a battery factory2018Basis not stated | mean. | ||
| Kenaf leavesJoseph et al. 2025 | Gobbiya Dam irrigated farmlandCollection date unreportedDry weight | mean. | ||
| SpinachJoseph et al. 2025 | Gobbiya Dam irrigated farmlandCollection date unreportedDry weight | mean. | ||
| Swiss chardJurkovic et al. 2025 | 1 leaves (flooded site 1)2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | 4 leaves (flooded site 4)2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | 6 leaves (flooded site 6)2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | 8 leaves (flooded site 8)2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | all chard samples2024Basis not stated | mean. | ||
| Swiss chardJurkovic et al. 2025 | Control 2 leaves2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | Control 4 leaves (control site 4)2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | Control 7 leaves (control site 7)2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | Control 8 leaves (control site 8)2024Basis not stated | sample value. | ||
| Alternanthera sessilis leavesKananke et al. 2026 | BandaragamaCollection date unreportedBasis not stated | mean. | ||
| Alternanthera sessilis leavesKananke et al. 2026 | KolonnawaCollection date unreportedBasis not stated | mean. | ||
| Alternanthera sessilis leavesKananke et al. 2026 | KottawaCollection date unreportedBasis not stated | mean. | ||
| Alternanthera sessilis leavesKananke et al. 2026 | PiliyandalaCollection date unreportedBasis not stated | mean. | ||
| Alternanthera sessilis leavesKananke et al. 2026 | WellampitiyaCollection date unreportedBasis not stated | mean. | ||
| Centella asiatica leavesKananke et al. 2026 | BandaragamaCollection date unreportedBasis not stated | mean. | ||
| Centella asiatica leavesKananke et al. 2026 | KolonnawaCollection date unreportedBasis not stated | mean. | ||
| Centella asiatica leavesKananke et al. 2026 | KottawaCollection date unreportedBasis not stated | mean. | ||
| Centella asiatica leavesKananke et al. 2026 | PiliyandalaCollection date unreportedBasis not stated | mean. | ||
| Centella asiatica leavesKananke et al. 2026 | WellampitiyaCollection date unreportedBasis not stated | mean. | ||
| KaleKim et al. 2024 | All six farms, field samples2019Fresh weight | mean.Same samples: 5.95–217 µg/kg (range) | ||
| KaleKim et al. 2024 | All six farms, field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | All six farms, field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm R1 (Rural), field samples2019Fresh weight | mean.Same samples: 6.82–12.7 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm R1 (Rural), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm R1 (Rural), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm R2 (Rural), field samples2019Fresh weight | mean.Same samples: 8.64–13.4 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm R2 (Rural), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm R2 (Rural), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm R3 (Rural), field samples2019Fresh weight | mean.Same samples: 5.95–36.8 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm R3 (Rural), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm R3 (Rural), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm R4 (Rural), field samples2019Fresh weight | mean.Same samples: 12.1–15.4 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm R4 (Rural), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm R4 (Rural), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm U1 (Urban), field samples2019Fresh weight | mean.Same samples: 91.6–217 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm U1 (Urban), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm U1 (Urban), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm U2 (Urban), field samples2019Fresh weight | mean.Same samples: 82.7–121 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm U2 (Urban), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm U2 (Urban), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Rural farms R1-R42019Fresh weight | mean. | ||
| KaleKim et al. 2024 | Urban farms U1-U22019Fresh weight | mean. | ||
| Amaranth leavesKulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| Amaranth leavesKulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| Bathua (lamb's quarters)Kulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| Bathua (lamb's quarters)Kulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| Fenugreek leavesKulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| Fenugreek leavesKulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| Mustard greensKulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| Mustard greensKulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| Mustard greensKulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| Mustard greensKulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| Radish leavesKulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| Radish leavesKulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| SpinachKulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| SpinachKulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| Head cabbageLi et al. 2021 | Four major producing cities of2018Fresh weight | mean. | ||
| Head cabbageLi et al. 2021 | Wuding (as attributed in text)2018Fresh weight | sample value. | ||
| Head cabbageLiang et al. 2019 | BeijingCollection date unreportedDry weight | mean. | ||
| Mustard greensLiu et al. 2023 | MustardCollection date unreportedBasis not stated | mean. | ||
| Head lettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Lamb's lettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| LettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| RocketMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Romaine lettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| SpinachMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Swiss chardMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| WatercressMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Fenugreek leavesMawari et al. 2022 | Farms in ChandrahalCollection date unreportedBasis not stated | sample value. | ||
| Head cabbageMawari et al. 2022 | Farms in ChandrahalCollection date unreportedBasis not stated | sample value. | ||
| SpinachMohod 2015 | Grown near Amba NallaCollection date unreportedDry weight | mean. | ||
| Amaranth leavesNaser et al. 2018 | High-level pollution: KalakoirCollection date unreportedDry weight | mean. | ||
| Amaranth leavesNaser et al. 2018 | High-level pollution: KalakoirCollection date unreportedDry weight | mean. | ||
| Amaranth leavesNaser et al. 2018 | Low-level pollution: BARICollection date unreportedDry weight | mean. | ||
| Amaranth leavesNaser et al. 2018 | Low-level pollution: BARICollection date unreportedDry weight | mean. | ||
| Amaranth leavesNaser et al. 2018 | Medium-level pollution: ZorunCollection date unreportedDry weight | mean. | ||
| Amaranth leavesNaser et al. 2018 | Medium-level pollution: ZorunCollection date unreportedDry weight | mean. | ||
| SpinachNaser et al. 2018 | High-level pollution: KalakoirCollection date unreportedDry weight | mean. | ||
| SpinachNaser et al. 2018 | Low-level pollution: BARICollection date unreportedDry weight | mean. | ||
| SpinachNaser et al. 2018 | Medium-level pollution: ZorunCollection date unreportedDry weight | mean. | ||
| KaleNgodhe et al. 2018 | All five plots around KIWASCO2018Basis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | FuzhouCollection date unreportedBasis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | GanzhouCollection date unreportedBasis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | JiujiangCollection date unreportedBasis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | Leafy vegetablesCollection date unreportedBasis not stated | range. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | Leafy vegetablesCollection date unreportedBasis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | ShangraoCollection date unreportedBasis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | YichunCollection date unreportedBasis not stated | mean. | ||
| Head cabbageOkonofua et al. 2024 | Dry Season2019Basis not stated | mean. | ||
| Head cabbageOkonofua et al. 2024 | Dry season Control2019Basis not stated | mean. | ||
| Head cabbageOkonofua et al. 2024 | Rainy Season2019Basis not stated | mean. | ||
| Head cabbageOkonofua et al. 2024 | Rainy season Control2019Basis not stated | below LOD. | ||
| Amaranth leavesOrosun et al. 2023 | Irrigated farmlands along the2021Dry weight | mean. | ||
| Head cabbageOrosun et al. 2023 | Irrigated farmlands along the2021Dry weight | mean. | ||
| Leafy vegetables, other or unspecifiedOsuolale et al. 2025 | Ondo Metropolis market/farm2018Dry weight | below LOD. | ||
| Head cabbagePal et al. 2017 | Domestic-wastewater-irrigatedCollection date unreportedBasis not stated | mean. | ||
| SpinachPal et al. 2017 | Domestic-wastewater-irrigatedCollection date unreportedBasis not stated | mean. | ||
| Amaranth leavesPan et al. 2016 | Zhejiang province marketed2014Fresh weight | mean. | ||
| Coriander leavesPan et al. 2016 | Zhejiang province marketed2014Fresh weight | mean. | ||
| Romaine lettucePan et al. 2016 | Zhejiang province marketed2014Fresh weight | mean. | ||
| SpinachPan et al. 2016 | Zhejiang province marketed2014Fresh weight | mean. | ||
| Water spinachPan et al. 2016 | Zhejiang province marketed2014Fresh weight | mean. | ||
| Head cabbageRempelos et al. 2023 | Crop protection CON (CP)2004–2009Fresh weight | mean. | ||
| Head cabbageRempelos et al. 2023 | Crop protection ORG (OP)2004–2009Fresh weight | mean. | ||
| Head cabbageRempelos et al. 2023 | Fertilizer type CON (NPK)2004–2009Fresh weight | mean. | ||
| Head cabbageRempelos et al. 2023 | Fertilizer type ORG (FYM)2004–2009Fresh weight | mean. | ||
| Head cabbageRempelos et al. 2023 | Pre-crop Spring Beans2004–2009Fresh weight | mean. | ||
| Head cabbageRempelos et al. 2023 | Pre-crop Winter Barley2004–2009Fresh weight | mean. | ||
| Head lettuceRempelos et al. 2023 | Crop protection CON (CP)2004–2009Fresh weight | mean. | ||
| Head lettuceRempelos et al. 2023 | Crop protection ORG (OP)2004–2009Fresh weight | mean. | ||
| Head lettuceRempelos et al. 2023 | Fertilizer type CON (NPK)2004–2009Fresh weight | mean. | ||
| Head lettuceRempelos et al. 2023 | Fertilizer type ORG (FYM)2004–2009Fresh weight | mean. | ||
| Head lettuceRempelos et al. 2023 | Pre-crop Spring Beans2004–2009Fresh weight | mean. | ||
| Head lettuceRempelos et al. 2023 | Pre-crop Winter Barley2004–2009Fresh weight | mean. | ||
| ChivesRomero-Crespo et al. 2023 | Ponce Enriquez gold-mining area2021Fresh weight | below LOD. | ||
| Coriander leavesRomero-Crespo et al. 2023 | Ponce Enriquez gold-mining area2021Fresh weight | below LOD. | ||
| LettuceRomero-Crespo et al. 2023 | Ponce Enriquez gold-mining area2021Fresh weight | below LOD. | ||
| LettuceRossini-Oliva et al. 2024 | City urban gardens2021–2023Dry weight | mean. | ||
| LettuceRossini-Oliva et al. 2024 | Peri-urban garden2021–2023Dry weight | mean. | ||
| LettuceRossini-Oliva et al. 2024 | Seville local market (Mk)2021–2023Dry weight | mean. | ||
| LettuceRossini-Oliva et al. 2024 | City urban garden MI2 (Seville)2021–2023Fresh weight | max. | ||
| Swiss chardRossini-Oliva et al. 2024 | City urban gardens2021–2023Dry weight | mean. | ||
| Swiss chardRossini-Oliva et al. 2024 | Mining-area urban gardens2021–2023Dry weight | mean.Same samples: 160 µg/kg (maximum) | ||
| Swiss chardRossini-Oliva et al. 2024 | Peri-urban garden2021–2023Dry weight | mean. | ||
| Swiss chardRossini-Oliva et al. 2024 | Seville local market (Mk)2021–2023Dry weight | mean. | ||
| Swiss chardRossini-Oliva et al. 2024 | Mining-area urban gardens2021–2023Fresh weight | max. | ||
| ArumSadee 2022 | Local markets2021Dry weight | mean. | ||
| Garden cressSadee 2022 | Local markets2021Dry weight | mean. | ||
| MallowSadee 2022 | Local markets2021Dry weight | mean. | ||
| MintSadee 2022 | Local markets2021Dry weight | mean. | ||
| SpinachSadee 2022 | Local markets2021Dry weight | mean. | ||
| Swiss chardSadee 2022 | Local markets2021Dry weight | mean. | ||
| SpinachSalawu et al. 2015 | Samaru irrigation area2014Basis not stated | mean. | ||
| DillSaleem et al. 2025 | Town Square Farmer's MarketYear not reportedDry weight | mean. | ||
| KaleSaleem et al. 2025 | Town Square Farmer's MarketYear not reportedDry weight | mean. | ||
| SpinachSaleem et al. 2025 | Town Square Farmer's MarketYear not reportedDry weight | mean. | ||
| BasilSalehipour et al. 2015 | Wholesaler outlets andCollection date unreportedFresh weight | 95% UCL of mean. | ||
| LettuceSalehipour et al. 2015 | Wholesaler outlets andCollection date unreportedFresh weight | 95% UCL of mean. | ||
| MintSalehipour et al. 2015 | Wholesaler outlets andCollection date unreportedFresh weight | 95% UCL of mean. | ||
| SpinachSalhotra et al. 2017 | Jagdalpur market supply2017Dry weight | mean. | ||
| Head cabbageSamma et al. 2024 | Bogura districtCollection date unreportedDry weight | mean.Same samples: 23,300 µg/kg (maximum) | ||
| Head cabbageSamma et al. 2024 | Bogura districtCollection date unreportedDry weight | min. | ||
| Head cabbageSamma et al. 2024 | Bogura districtCollection date unreportedDry weight | max. | ||
| Head cabbageSee et al. 2025 | Pasar Borong Selangor wholesale2022Fresh weight | not detected. | ||
| Mustard greensSee et al. 2025 | Pasar Borong Selangor wholesale2022Fresh weight | not detected. | ||
| Pak choiSee et al. 2025 | Pasar Borong Selangor wholesale2022Fresh weight | not detected. | ||
| SpinachSee et al. 2025 | Pasar Borong Selangor wholesale2022Fresh weight | not detected. | ||
| Coriander leavesSharma et al. 2016 | Site 1: groundwater-irrigatedCollection date unreportedDry weight | mean. | ||
| Coriander leavesSharma et al. 2016 | Site 1: groundwater-irrigatedCollection date unreportedDry weight | max. | ||
| Coriander leavesSharma et al. 2016 | Site 2: groundwater-irrigatedCollection date unreportedDry weight | mean. | ||
| Coriander leavesSharma et al. 2016 | Site 2: groundwater-irrigatedCollection date unreportedDry weight | max. | ||
| Coriander leavesSharma et al. 2016 | Site 3:Collection date unreportedDry weight | mean. | ||
| Coriander leavesSharma et al. 2016 | Site 3:Collection date unreportedDry weight | max. | ||
| Fenugreek leavesSharma et al. 2016 | Site 1: groundwater-irrigatedCollection date unreportedDry weight | mean. | ||
| Fenugreek leavesSharma et al. 2016 | Site 1: groundwater-irrigatedCollection date unreportedDry weight | max. | ||
| Fenugreek leavesSharma et al. 2016 | Site 2: groundwater-irrigatedCollection date unreportedDry weight | mean. | ||
| Fenugreek leavesSharma et al. 2016 | Site 2: groundwater-irrigatedCollection date unreportedDry weight | max. | ||
| Fenugreek leavesSharma et al. 2016 | Site 3:Collection date unreportedDry weight | mean. | ||
| Fenugreek leavesSharma et al. 2016 | Site 3:Collection date unreportedDry weight | max. | ||
| MintSharma et al. 2016 | Site 1: groundwater-irrigatedCollection date unreportedDry weight | mean. | ||
| MintSharma et al. 2016 | Site 1: groundwater-irrigatedCollection date unreportedDry weight | max. | ||
| MintSharma et al. 2016 | Site 2: groundwater-irrigatedCollection date unreportedDry weight | mean. | ||
| MintSharma et al. 2016 | Site 2: groundwater-irrigatedCollection date unreportedDry weight | max. | ||
| MintSharma et al. 2016 | Site 3:Collection date unreportedDry weight | mean. | ||
| MintSharma et al. 2016 | Site 3:Collection date unreportedDry weight | max. | ||
| SpinachSharma et al. 2016 | Site 1: groundwater-irrigatedCollection date unreportedDry weight | mean. | ||
| SpinachSharma et al. 2016 | Site 1: groundwater-irrigatedCollection date unreportedDry weight | max. | ||
| SpinachSharma et al. 2016 | Site 2: groundwater-irrigatedCollection date unreportedDry weight | mean. | ||
| SpinachSharma et al. 2016 | Site 2: groundwater-irrigatedCollection date unreportedDry weight | max. | ||
| SpinachSharma et al. 2016 | Site 3:Collection date unreportedDry weight | mean. | ||
| SpinachSharma et al. 2016 | Site 3:Collection date unreportedDry weight | max. | ||
| BasilSouri et al. 2019 | Shahre Rey wastewater-irrigated2011Dry weight | mean. | ||
| BasilSouri et al. 2019 | Shahre Rey wastewater-irrigated2011Dry weight | mean. | ||
| Garden cressSouri et al. 2019 | Shahre Rey wastewater-irrigated2011Dry weight | mean. | ||
| Garden cressSouri et al. 2019 | Shahre Rey wastewater-irrigated2011Dry weight | mean. | ||
| Bristly ox tongueStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Bristly ox tongueStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Cat's earStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Dandelion greensStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| English plantainStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| English plantainStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| LettuceStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| MallowStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Nasturtium (Tropaeolum)Stark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Sweet fennelStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Sweet fennelStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| SpinachSuleiman and Ibrahim 2025 | Three irrigation farmsCollection date unreportedBasis not stated | mean. | ||
| Coriander leavesSultana et al. 2022 | 1st phase (early harvest)Year not reportedDry weight | not detected. | ||
| Coriander leavesSultana et al. 2022 | 2nd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| Coriander leavesSultana et al. 2022 | 3rd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| Coriander leavesSultana et al. 2022 | 4th phase (late harvest)Year not reportedDry weight | not detected. | ||
| Coriander leavesSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | mean. | ||
| Coriander leavesSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | range. | ||
| Head cabbageSultana et al. 2022 | 1st phase (early harvest)Year not reportedDry weight | not detected. | ||
| Head cabbageSultana et al. 2022 | 2nd phase (mid-time harvest)Year not reportedDry weight | not detected. | ||
| Head cabbageSultana et al. 2022 | 3rd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| Head cabbageSultana et al. 2022 | 4th phase (late harvest)Year not reportedDry weight | not detected. | ||
| Head cabbageSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | mean. | ||
| Head cabbageSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | range. | ||
| MintSultana et al. 2022 | 1st phase (early harvest)Year not reportedDry weight | not detected. | ||
| MintSultana et al. 2022 | 2nd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| MintSultana et al. 2022 | 3rd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| MintSultana et al. 2022 | 4th phase (late harvest)Year not reportedDry weight | sample value. | ||
| MintSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | mean. | ||
| MintSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | range. | ||
| Mustard greensSultana et al. 2022 | 1st phase (early harvest)Year not reportedDry weight | not detected. | ||
| Mustard greensSultana et al. 2022 | 2nd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| Mustard greensSultana et al. 2022 | 3rd phase (mid-time harvest)Year not reportedDry weight | not detected. | ||
| Mustard greensSultana et al. 2022 | 4th phase (late harvest)Year not reportedDry weight | sample value. | ||
| Mustard greensSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | mean. | ||
| Mustard greensSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | range. | ||
| SpinachSultana et al. 2022 | 1st phase (early harvest)Year not reportedDry weight | not detected. | ||
| SpinachSultana et al. 2022 | 2nd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| SpinachSultana et al. 2022 | 3rd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| SpinachSultana et al. 2022 | 4th phase (late harvest)Year not reportedDry weight | not detected. | ||
| SpinachSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | mean. | ||
| SpinachSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | range. | ||
| Head cabbageTweij 2018 | Iranian green cabbage2017Basis not stated | sample value. | ||
| Head cabbageTweij 2018 | Iranian green lettuce-type cabbage2017Basis not stated | sample value. | ||
| Head cabbageTweij 2018 | Iranian purple cabbage2017Basis not stated | sample value. | ||
| Head cabbageTweij 2018 | Iraqi green cabbage2017Basis not stated | sample value. | ||
| Senna obtusifolia leavesUdiba et al. 2020 | Dareta village, all five plotsCollection date unreportedBasis not stated | range. | ||
| Senna obtusifolia leavesUdiba et al. 2020 | Plot 1Collection date unreportedBasis not stated | mean. | ||
| Senna obtusifolia leavesUdiba et al. 2020 | Plot 2Collection date unreportedBasis not stated | mean. | ||
| Senna obtusifolia leavesUdiba et al. 2020 | Plot 3Collection date unreportedBasis not stated | mean. | ||
| Senna obtusifolia leavesUdiba et al. 2020 | Plot 4Collection date unreportedBasis not stated | mean. | ||
| Senna obtusifolia leavesUdiba et al. 2020 | Plot 5Collection date unreportedBasis not stated | mean. | ||
| Senna obtusifolia leavesUdiba et al. 2020 | Zaria (control)Collection date unreportedBasis not stated | mean. | ||
| LettuceUK TDS 2014 (Fera 2015) | UK TDS 2014 category 11_652014Fresh weight | sample value. | ||
| LettuceVela et al. 2023 | Quito wholesale open marketsCollection date unreportedDry weight | mean. | ||
| LettuceVela et al. 2023 | Table II Pb lettuce row 1Collection date unreportedDry weight | sample value. | ||
| LettuceVela et al. 2023 | Table II Pb lettuce row 2Collection date unreportedDry weight | sample value. | ||
| LettuceVela et al. 2023 | Table II Pb lettuce row 3Collection date unreportedDry weight | sample value. | ||
| LettuceVela et al. 2023 | Table II Pb lettuce row 4Collection date unreportedDry weight | sample value. | ||
| Leafy vegetables, other or unspecifiedWang et al. 2019 | Guangzhou food safety risk2014–2017Basis not stated | 95th percentile. | ||
| Leafy vegetables, other or unspecifiedWang et al. 2019 | Guangzhou food safety risk2014–2017Basis not stated | mean.Same samples: 143 µg/kg (95th percentile) | ||
| Leafy vegetables, other or unspecifiedWang et al. 2019 | Guangzhou food safety risk2014–2017Basis not stated | median. | ||
| Alternanthera sessilis leavesWijeyaratne et al. 2021 | Non-organic cultivation sites2019Dry weight | mean. | ||
| Alternanthera sessilis leavesWijeyaratne et al. 2021 | Non-organic cultivation sites2019Dry weight | mean. | ||
| Alternanthera sessilis leavesWijeyaratne et al. 2021 | Organic cultivation sites2019Dry weight | mean. | ||
| Alternanthera sessilis leavesWijeyaratne et al. 2021 | Organic cultivation sites2019Dry weight | mean. | ||
| BasilWiniarska-Mieczan et al. 2023 | Lublin groceriesCollection date unreportedFresh weight | mean. | ||
| Celery leavesWiniarska-Mieczan et al. 2023 | Lublin groceriesCollection date unreportedFresh weight | mean. | ||
| Coriander leavesWiniarska-Mieczan et al. 2023 | Lublin groceriesCollection date unreportedFresh weight | mean. | ||
| DillWiniarska-Mieczan et al. 2023 | Lublin groceriesCollection date unreportedFresh weight | mean. | ||
| LovageWiniarska-Mieczan et al. 2023 | Lublin groceriesCollection date unreportedFresh weight | mean. | ||
| ParsleyWiniarska-Mieczan et al. 2023 | Lublin groceriesCollection date unreportedFresh weight | mean. | ||
| WatercressWiniarska-Mieczan et al. 2023 | Lublin groceriesCollection date unreportedFresh weight | mean. | ||
| Chinese cabbageWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| ChivesWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| Head cabbageWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| LettuceWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| Pak choiWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| Romaine lettuceWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| SpinachWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Cattle manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Chemical fertilizer (NPK)2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Chicken manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Control (no fertilizer)2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Sheep manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Turkey manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Vermicompost2017Basis not stated | mean. | ||
3 findings · 1 study | ||||
| Head cabbage (canned)WHO GEMS/Food | WHO region PAHO; random sampling2014Fresh weight | median. | ||
| Mustard greens (canned)WHO GEMS/Food | WHO region PAHO; random sampling2016Fresh weight | median. | ||
| Spinach (canned)WHO GEMS/Food | WHO region PAHO; random sampling2014–2016Fresh weight | median. | ||
4 findings · 1 study | ||||
| Head cabbage (fermented)WHO GEMS/Food | WHO region EURO; random sampling2004–2019Fresh weight | median. | ||
| Head cabbage (fermented)WHO GEMS/Food | WHO region EURO; unknown2004–2017Fresh weight | median. | ||
| Head cabbage (fermented)WHO GEMS/Food | WHO region PAHO; random sampling2014–2016Fresh weight | median. | ||
| Head cabbage (fermented)WHO GEMS/Food | WHO region WPRO; random sampling2008Fresh weight | median. | ||
2 findings · 1 study | ||||
| Spinach (frozen)WHO GEMS/Food | WHO region EURO; random sampling2003–2018Fresh weight | median. | ||
| Spinach (frozen)WHO GEMS/Food | WHO region EURO; unknown2006–2017Fresh weight | median. | ||
3 findings · 1 study | ||||
| Basil (dried)Winiarska-Mieczan et al. 2023 | Lublin groceries (dried herbs)Collection date unreportedFresh weight | mean. | ||
| Coriander leaves (dried)Winiarska-Mieczan et al. 2023 | Lublin groceries (dried herbs)Collection date unreportedFresh weight | mean. | ||
| Lovage (dried)Winiarska-Mieczan et al. 2023 | Lublin groceries (dried herbs)Collection date unreportedFresh weight | mean. | ||
47 findings · 6 studies | ||||
| Amaranth leaves (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010Fresh weight | median. | ||
| Chinese cabbage (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Head cabbage (cooked)WHO GEMS/Food | WHO region AFRO; random sampling2018Fresh weight | median. | ||
| Head cabbage (cooked)WHO GEMS/Food | WHO region EURO; random sampling2004–2019Fresh weight | median. | ||
| Head cabbage (cooked)WHO GEMS/Food | WHO region PAHO; random sampling2003–2017Fresh weight | median. | ||
| Head cabbage (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Head cabbage (cooked)WHO GEMS/Food | WHO region WPRO; targeted sampling2008Fresh weight | median. | ||
| Head lettuce (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010Fresh weight | median. | ||
| Kale (cooked)WHO GEMS/Food | WHO region EURO; random sampling2009–2017Fresh weight | median. | ||
| Kale (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Leaf lettuce (cooked)WHO GEMS/Food | WHO region PAHO; random sampling2019–2020Fresh weight | median. | ||
| Leaf lettuce (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Leafy vegetables, other or unspecified (cooked)WHO GEMS/Food | WHO region AFRO; random sampling2018Fresh weight | median. | ||
| Leafy vegetables, other or unspecified (cooked)WHO GEMS/Food | WHO region EURO; random sampling2004Fresh weight | median. | ||
| Leafy vegetables, other or unspecified (cooked)WHO GEMS/Food | WHO region PAHO; random sampling2003–2017Fresh weight | median. | ||
| Leafy vegetables, other or unspecified (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Leafy vegetables, other or unspecified (cooked)WHO GEMS/Food | WHO region WPRO; targeted sampling1998Fresh weight | median. | ||
| Mustard greens (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Pak choi (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2014Fresh weight | median. | ||
| Spinach (cooked)WHO GEMS/Food | WHO region EURO; random sampling2007–2019Fresh weight | median. | ||
| Spinach (cooked)WHO GEMS/Food | WHO region PAHO; random sampling2003–2017Fresh weight | median. | ||
| Spinach (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Water spinach (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010Fresh weight | median. | ||
| Watercress (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Collard greens (cooked)FDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Kale (cooked)FDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Amaranth leaves (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Chinese cabbage (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Chinese kale (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Choy sum (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Head cabbage (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Lettuce (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Lettuce (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Mustard greens (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Pak choi (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Spinach (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Water spinach (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Watercress (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Leafy vegetables, other or unspecified (cooked)Mahaffey et al. 1975 | US FDA Total Diet Study FY 19731968–1974Basis not stated | mean. | ||
| Head cabbage (cooked)UK TDS 2014 (Fera 2015) | UK TDS 2014 category 11_622014Fresh weight | sample value. | ||
| Other green vegetables (mixed category) (cooked)UK TDS 2014 (Fera 2015) | UK TDS 2014 category 11_682014Fresh weight | below LOD. | ||
| Chinese cabbage (cooked)Wang et al. 2020 | Jilin ProvinceCollection date unreportedBasis not stated | mean. | ||
| Chinese cabbage (cooked)Wang et al. 2020 | Other AreaCollection date unreportedBasis not stated | mean. | ||
| Chinese cabbage (cooked)Wang et al. 2020 | Songhua River AreaCollection date unreportedBasis not stated | mean. | ||
| Leek (cooked)Wang et al. 2020 | Jilin ProvinceCollection date unreportedBasis not stated | mean. | ||
| Leek (cooked)Wang et al. 2020 | Other AreaCollection date unreportedBasis not stated | mean. | ||
| Leek (cooked)Wang et al. 2020 | Songhua River AreaCollection date unreportedBasis not stated | mean. | ||
400 findings · 68 studies · 31 spread figures folded in | ||||
| Chinese cabbageWHO GEMS/Food | WHO region EURO; random sampling2003–2006Fresh weight | median. | ||
| Chinese cabbageWHO GEMS/Food | WHO region WPRO; random sampling2002–2011Fresh weight | median. | ||
| Head cabbageWHO GEMS/Food | WHO region EURO; random sampling2003–2019Fresh weight | median. | ||
| Head cabbageWHO GEMS/Food | WHO region EURO; targeted sampling1999–2006Fresh weight | median. | ||
| Head cabbageWHO GEMS/Food | WHO region EURO; unknown2003–2017Fresh weight | median. | ||
| Head cabbageWHO GEMS/Food | WHO region PAHO; random sampling1999–2018Fresh weight | median. | ||
| Head cabbageWHO GEMS/Food | WHO region WPRO; random sampling2000–2010Fresh weight | median. | ||
| Head lettuceWHO GEMS/Food | WHO region EURO; random sampling1997–1999Dry weight | median. | ||
| Head lettuceWHO GEMS/Food | WHO region EURO; random sampling2003–2019Fresh weight | median. | ||
| Head lettuceWHO GEMS/Food | WHO region EURO; targeted sampling2005–2007Fresh weight | median. | ||
| Head lettuceWHO GEMS/Food | WHO region EURO; unknown2005–2017Fresh weight | median. | ||
| Head lettuceWHO GEMS/Food | WHO region PAHO; random sampling2003–2017Fresh weight | median. | ||
| Head lettuceWHO GEMS/Food | WHO region WPRO; random sampling2003–2014Fresh weight | median. | ||
| Head lettuceWHO GEMS/Food | WHO region WPRO; targeted sampling1994–2008Fresh weight | median. | ||
| KaleWHO GEMS/Food | WHO region EURO; random sampling2003–2019Fresh weight | median. | ||
| KaleWHO GEMS/Food | WHO region EURO; targeted sampling2003–2006Fresh weight | median. | ||
| KaleWHO GEMS/Food | WHO region EURO; unknown2003–2018Fresh weight | median. | ||
| KaleWHO GEMS/Food | WHO region PAHO; random sampling2015–2016Fresh weight | median. | ||
| KaleWHO GEMS/Food | WHO region WPRO; targeted sampling2001Fresh weight | median. | ||
| Lamb's lettuceWHO GEMS/Food | WHO region EURO; random sampling2009–2019Fresh weight | median. | ||
| Lamb's lettuceWHO GEMS/Food | WHO region EURO; unknown2015–2017Fresh weight | median. | ||
| Leaf lettuceWHO GEMS/Food | WHO region EMRO; unknown2014–2019Fresh weight | median. | ||
| Leaf lettuceWHO GEMS/Food | WHO region EURO; random sampling2003–2019Fresh weight | median. | ||
| Leaf lettuceWHO GEMS/Food | WHO region EURO; targeted sampling2003–2006Fresh weight | median. | ||
| Leaf lettuceWHO GEMS/Food | WHO region EURO; unknown2003–2018Fresh weight | median. | ||
| Leaf lettuceWHO GEMS/Food | WHO region PAHO; random sampling1999–2018Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region EMRO; unknown2019Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region EURO; random sampling2003–2019Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region EURO; targeted sampling2005–2006Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region EURO; unknown2003–2017Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region WPRO; random sampling2002–2011Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region WPRO; targeted sampling2001Fresh weight | median. | ||
| Mustard greensWHO GEMS/Food | WHO region EURO; random sampling2005Fresh weight | median. | ||
| Mustard greensWHO GEMS/Food | WHO region EURO; targeted sampling2003Fresh weight | median. | ||
| Mustard greensWHO GEMS/Food | WHO region EURO; unknown2015–2016Fresh weight | median. | ||
| Mustard greensWHO GEMS/Food | WHO region WPRO; random sampling2005Fresh weight | median. | ||
| Mustard greensWHO GEMS/Food | WHO region WPRO; targeted sampling2002Fresh weight | median. | ||
| Pak choiWHO GEMS/Food | WHO region EURO; random sampling2003–2019Fresh weight | median. | ||
| Pak choiWHO GEMS/Food | WHO region EURO; unknown2003–2018Fresh weight | median. | ||
| Pak choiWHO GEMS/Food | WHO region WPRO; targeted sampling2001Fresh weight | median. | ||
| SpinachWHO GEMS/Food | WHO region EMRO; unknown2014–2019Fresh weight | median. | ||
| SpinachWHO GEMS/Food | WHO region EURO; random sampling2003–2019Fresh weight | median. | ||
| SpinachWHO GEMS/Food | WHO region EURO; targeted sampling2000–2004Fresh weight | median. | ||
| SpinachWHO GEMS/Food | WHO region EURO; unknown2003–2018Fresh weight | median. | ||
| SpinachWHO GEMS/Food | WHO region PAHO; random sampling2005–2018Fresh weight | median. | ||
| SpinachWHO GEMS/Food | WHO region WPRO; random sampling2003Fresh weight | median. | ||
| WatercressWHO GEMS/Food | WHO region EURO; random sampling2004–2019Fresh weight | median. | ||
| Head cabbageFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Head lettuceFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Leaf lettuceFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| SpinachFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| BasilAbdolahpour et al. 2023 | Babol2021Fresh weight | mean. | ||
| ParsleyAbdolahpour et al. 2023 | Babol2021Fresh weight | mean. | ||
| SpinachAbdolahpour et al. 2023 | Babol2021Fresh weight | mean. | ||
| SpinachAbubakar et al. 2025 | Site ICollection date unreportedBasis not stated | mean. | ||
| Amaranth leavesAdebayo et al. 2020 | Railway Quarters irrigation farmCollection date unreportedDry weight | mean. | ||
| Coriander leavesAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| Coriander leavesAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| Coriander leavesAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| Coriander leavesAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | mean. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | mean. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | mean. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | mean. | ||
| LettuceAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| LettuceAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| LettuceAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| LettuceAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| Head cabbageAlimohammadi et al. 2018 | Autumn 20162016Fresh weight | mean. | ||
| Head cabbageAlimohammadi et al. 2018 | Summer 20162016Fresh weight | mean. | ||
| LettuceAlimohammadi et al. 2018 | Autumn 20162016Fresh weight | mean. | ||
| LettuceAlimohammadi et al. 2018 | Summer 20162016Fresh weight | mean. | ||
| Leafy vegetables, other or unspecifiedAltunay et al. 2023 | Greengrocer, Sivas, TurkeyCollection date unreportedBasis not stated | mean. | ||
| SpinachAltunay et al. 2023 | Greengrocer, Sivas, TurkeyCollection date unreportedBasis not stated | mean. | ||
| Head cabbageAmetepey et al. 2018 | Central and Aboabo MarketsCollection date unreportedDry weight | mean.Same samples: 40–70 µg/kg (range) | ||
| Head cabbageAmetepey et al. 2018 | Central and Aboabo MarketsCollection date unreportedDry weight | range. | ||
| Head cabbageArmand et al. 2026 | 10 retail vegetable storesCollection date unreportedFresh weight | mean. | ||
| Head cabbageArmand et al. 2026 | 10 retail vegetable storesCollection date unreportedFresh weight | range. | ||
| LettuceArmand et al. 2026 | 10 retail vegetable storesCollection date unreportedFresh weight | mean. | ||
| LettuceArmand et al. 2026 | 10 retail vegetable storesCollection date unreportedFresh weight | range. | ||
| DillBora et al. 2022 | Vegetable and fruit market2020–2021Fresh weight | below LOD. | ||
| Head cabbageBora et al. 2022 | Amateur farmers2020–2021Fresh weight | below LOD. | ||
| Head cabbageBora et al. 2022 | Amateur farmers2020–2021Fresh weight | below LOD. | ||
| Head cabbageBora et al. 2022 | Amateur farmers2020–2021Fresh weight | below LOD. | ||
| Head cabbageBora et al. 2022 | Vegetable and fruit market2020–2021Fresh weight | below LOD. | ||
| Head cabbageBora et al. 2022 | Vegetable and fruit market2020–2021Fresh weight | below LOD. | ||
| Head cabbageBora et al. 2022 | Vegetable and fruit market2020–2021Fresh weight | mean. | ||
| LettuceBora et al. 2022 | Amateur farmers2020–2021Fresh weight | mean. | ||
| LettuceBora et al. 2022 | Amateur farmers2020–2021Fresh weight | mean. | ||
| LettuceBora et al. 2022 | Amateur farmers2020–2021Fresh weight | mean. | ||
| LettuceBora et al. 2022 | Vegetable and fruit market2020–2021Fresh weight | mean. | ||
| LettuceBora et al. 2022 | Vegetable and fruit market2020–2021Fresh weight | mean. | ||
| LettuceBora et al. 2022 | Vegetable and fruit market2020–2021Fresh weight | mean. | ||
| LovageBora et al. 2022 | Amateur farmers2020–2021Fresh weight | below LOD. | ||
| ParsleyBora et al. 2022 | Amateur farmers2020–2021Fresh weight | below LOD. | ||
| ParsleyBora et al. 2022 | Vegetable and fruit market2020–2021Fresh weight | below LOD. | ||
| TarragonBora et al. 2022 | Vegetable and fruit market2020–2021Fresh weight | below LOD. | ||
| Coriander leavesCantoral et al. 2024 | Retail outlets2022–2023Basis not stated | mean. | ||
| Head cabbageCantoral et al. 2024 | Retail outlets2022–2023Basis not stated | mean. | ||
| Romaine lettuceCantoral et al. 2024 | Retail outlets2022–2023Basis not stated | mean. | ||
| SpinachCantoral et al. 2024 | Retail outlets2022–2023Basis not stated | mean. | ||
| Swiss chardCantoral et al. 2024 | Retail outlets2022–2023Basis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedChen et al. 2025 | Chongqing retail/market samples2017–2021Basis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedChen et al. 2025 | Chongqing retail/market samples2017–2021Basis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedChen et al. 2025 | Chongqing retail/market samples2017–2021Basis not stated | mean. | ||
| Amaranth leavesChiutula et al. 2025 | Soche WWTP effluent-irrigatedYear not reportedDry weight | mean. | ||
| Amaranth leavesChiutula et al. 2025 | Soche WWTP effluent-irrigatedYear not reportedDry weight | max. | ||
| Chinese cabbageChiutula et al. 2025 | Soche WWTP effluent-irrigatedYear not reportedDry weight | mean. | ||
| Mustard greensChiutula et al. 2025 | Soche WWTP effluent-irrigatedYear not reportedDry weight | mean. | ||
| Pumpkin leavesChiutula et al. 2025 | Soche WWTP effluent-irrigatedYear not reportedDry weight | mean. | ||
| Rape greensChiutula et al. 2025 | Soche WWTP effluent-irrigatedYear not reportedDry weight | mean. | ||
| Sweet potato leavesChiutula et al. 2025 | Soche WWTP effluent-irrigatedYear not reportedDry weight | not detected. | ||
| Beet leavesChmielewski et al. 2025 | 16 allotment gardens2022–2023Dry weight | mean. | ||
| Celery leavesChmielewski et al. 2025 | 16 allotment gardens2022–2023Dry weight | mean. | ||
| ParsleyChmielewski et al. 2025 | 16 allotment gardens2022–2023Dry weight | mean. | ||
| LettuceDearing et al. 2025 | Hawke's Bay market gardens2024Fresh weight | median.Same samples: 74 µg/kg (maximum); 93 µg/kg (maximum); 28 µg/kg (maximum) | ||
| LettuceDearing et al. 2025 | Hawke's Bay market gardens2024Fresh weight | max. | ||
| Pak choiDearing et al. 2025 | Hawke's Bay market gardens2024Fresh weight | max. | ||
| ParsleyDearing et al. 2025 | Hawke's Bay market gardens2024Fresh weight | median. | ||
| ParsleyDearing et al. 2025 | Hawke's Bay market gardens2024Fresh weight | max. | ||
| Water spinachDearing et al. 2025 | Hawke's Bay market gardens2024Fresh weight | sample value. | ||
| Leafy vegetables, other or unspecifiedEFSA 2009 | EU Member State monitoring2003–2007Basis not stated | 5th percentile. | ||
| Leafy vegetables, other or unspecifiedEFSA 2009 | EU Member State monitoring2003–2007Basis not stated | 95th percentile. | ||
| Leafy vegetables, other or unspecifiedEFSA 2009 | EU Member State monitoring2003–2007Basis not stated | mean.Same samples: 0.5 µg/kg (5th percentile); 73 µg/kg (95th percentile); 200 µg/kg (maximum); 186 µg/kg (95th percentile); 325 µg/kg (maximum) | ||
| Leafy vegetables, other or unspecifiedEFSA 2009 | EU Member State monitoring2003–2007Basis not stated | median. | ||
| Leafy vegetables, other or unspecifiedEFSA 2009 | EU Member State monitoring2003–2007Basis not stated | max. | ||
| SpinachEFSA 2009 | EU Member State monitoring2003–2007Basis not stated | 5th percentile. | ||
| SpinachEFSA 2009 | EU Member State monitoring2003–2007Basis not stated | 95th percentile. | ||
| SpinachEFSA 2009 | EU Member State monitoring2003–2007Basis not stated | mean. | ||
| SpinachEFSA 2009 | EU Member State monitoring2003–2007Basis not stated | median. | ||
| SpinachEFSA 2009 | EU Member State monitoring2003–2007Basis not stated | max. | ||
| Leafy vegetables, other or unspecifiedEFSA 2012 | EU/EEA official monitoring and2003–2011Basis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedEFSA 2012 | EU/EEA official monitoring and2003–2011Basis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedEFSA 2012 | EU/EEA official monitoring and2003–2011Basis not stated | mean. | ||
| SpinachElbagermi et al. 2012 | Misurata City local suppliers and2010Dry weight | mean. | ||
| LettuceEmmanuel 2025 | Sharada2025Basis not stated | mean. | ||
| LettuceEmmanuel 2025 | Wudil2025Basis not stated | mean. | ||
| SpinachEmmanuel 2025 | Nomans-Land2025Basis not stated | mean. | ||
| Amaranth leavesFatai et al. 2024 | Vegetable farms2022Basis not stated | mean. | ||
| Head cabbageFonge et al. 2021 | SantaCollection date unreportedDry weight | mean. | ||
| Head cabbageGurmu et al. 2025 | Akaki Rivers irrigated farmlandYear not reportedDry weight | mean. | ||
| LettuceHeshmati et al. 2020 | Hamadan Province2017–2018Dry weight | mean.Same samples: up to 26 µg/kg (range maximum) | ||
| LettuceHeshmati et al. 2020 | Hamadan Province2017–2018Fresh weight | mean. | ||
| LettuceHeshmati et al. 2020 | Hamadan Province2017–2018Fresh weight | range. | ||
| Amaranth leavesHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | 75th percentile. | ||
| Amaranth leavesHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | median.Same samples: 1,770 µg/kg (75th percentile); 50 µg/kg (minimum); 5,320 µg/kg (maximum); 640 µg/kg (75th percentile); 98 µg/kg (minimum); 1,110 µg/kg (maximum) | ||
| Amaranth leavesHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | not detected. | ||
| Amaranth leavesHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | min. | ||
| Amaranth leavesHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | max. | ||
| Head cabbageHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | 75th percentile. | ||
| Head cabbageHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | median. | ||
| Head cabbageHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | not detected. | ||
| Head cabbageHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | min. | ||
| Head cabbageHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | max. | ||
| Malabar spinachHossen et al. 2026 | farm-82024Fresh weight | mean. | ||
| Chinese cabbageHuang et al. 2014 | Zhejiang province2012Fresh weight | mean. | ||
| Head cabbageHuang et al. 2014 | Zhejiang province2012Fresh weight | mean. | ||
| Pak choiHuang et al. 2014 | Zhejiang province2012Fresh weight | mean. | ||
| Rape greensHuang et al. 2014 | Zhejiang province2012Fresh weight | mean. | ||
| SpinachHuang et al. 2014 | Zhejiang province2012Fresh weight | mean. | ||
| Chinese cabbageJiang et al. 2020 | SZD village near a battery factory2018Basis not stated | mean. | ||
| Pak choiJiang et al. 2020 | SZD village near a battery factory2018Basis not stated | mean. | ||
| SpinachJiang et al. 2020 | SZD village near a battery factory2018Basis not stated | mean. | ||
| Kenaf leavesJoseph et al. 2025 | Gobbiya Dam irrigated farmlandCollection date unreportedDry weight | mean. | ||
| SpinachJoseph et al. 2025 | Gobbiya Dam irrigated farmlandCollection date unreportedDry weight | mean. | ||
| Swiss chardJurkovic et al. 2025 | 1 leaves (flooded site 1)2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | 4 leaves (flooded site 4)2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | 6 leaves (flooded site 6)2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | 8 leaves (flooded site 8)2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | all chard samples2024Basis not stated | mean. | ||
| Swiss chardJurkovic et al. 2025 | Control 2 leaves2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | Control 4 leaves (control site 4)2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | Control 7 leaves (control site 7)2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | Control 8 leaves (control site 8)2024Basis not stated | sample value. | ||
| Alternanthera sessilis leavesKananke et al. 2026 | BandaragamaCollection date unreportedBasis not stated | mean. | ||
| Alternanthera sessilis leavesKananke et al. 2026 | KolonnawaCollection date unreportedBasis not stated | mean. | ||
| Alternanthera sessilis leavesKananke et al. 2026 | KottawaCollection date unreportedBasis not stated | mean. | ||
| Alternanthera sessilis leavesKananke et al. 2026 | PiliyandalaCollection date unreportedBasis not stated | mean. | ||
| Alternanthera sessilis leavesKananke et al. 2026 | WellampitiyaCollection date unreportedBasis not stated | mean. | ||
| Centella asiatica leavesKananke et al. 2026 | BandaragamaCollection date unreportedBasis not stated | mean. | ||
| Centella asiatica leavesKananke et al. 2026 | KolonnawaCollection date unreportedBasis not stated | mean. | ||
| Centella asiatica leavesKananke et al. 2026 | KottawaCollection date unreportedBasis not stated | mean. | ||
| Centella asiatica leavesKananke et al. 2026 | PiliyandalaCollection date unreportedBasis not stated | mean. | ||
| Centella asiatica leavesKananke et al. 2026 | WellampitiyaCollection date unreportedBasis not stated | mean. | ||
| KaleKim et al. 2024 | All six farms, field samples2019Fresh weight | mean.Same samples: 8.37–47.2 µg/kg (range) | ||
| KaleKim et al. 2024 | All six farms, field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | All six farms, field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm R1 (Rural), field samples2019Fresh weight | mean.Same samples: 8.37–9.19 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm R1 (Rural), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm R1 (Rural), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm R2 (Rural), field samples2019Fresh weight | mean.Same samples: 35.4–42.2 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm R2 (Rural), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm R2 (Rural), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm R3 (Rural), field samples2019Fresh weight | mean.Same samples: 12.2–16.2 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm R3 (Rural), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm R3 (Rural), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm R4 (Rural), field samples2019Fresh weight | mean.Same samples: 17.4–20.9 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm R4 (Rural), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm R4 (Rural), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm U1 (Urban), field samples2019Fresh weight | mean.Same samples: 37.9–47.2 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm U1 (Urban), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm U1 (Urban), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm U2 (Urban), field samples2019Fresh weight | mean.Same samples: 14.7–20.8 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm U2 (Urban), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm U2 (Urban), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Rural farms R1-R42019Fresh weight | mean. | ||
| KaleKim et al. 2024 | Urban farms U1-U22019Fresh weight | mean. | ||
| Amaranth leavesKulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| Amaranth leavesKulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| Bathua (lamb's quarters)Kulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| Bathua (lamb's quarters)Kulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| Fenugreek leavesKulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| Fenugreek leavesKulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| Mustard greensKulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| Mustard greensKulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| Mustard greensKulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| Mustard greensKulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| Radish leavesKulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| Radish leavesKulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| SpinachKulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| SpinachKulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| Head cabbageLi et al. 2021 | Four major producing cities of2018Fresh weight | mean. | ||
| Head cabbageLi et al. 2021 | Wuding (as attributed in text)2018Fresh weight | sample value. | ||
| Head cabbageLiang et al. 2019 | BeijingCollection date unreportedDry weight | mean. | ||
| Mustard greensLiu et al. 2023 | MustardCollection date unreportedBasis not stated | mean. | ||
| Head lettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Lamb's lettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| LettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| RocketMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Romaine lettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| SpinachMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Swiss chardMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| WatercressMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Fenugreek leavesMawari et al. 2022 | Farms in ChandrahalCollection date unreportedBasis not stated | sample value. | ||
| Head cabbageMawari et al. 2022 | Farms in ChandrahalCollection date unreportedBasis not stated | sample value. | ||
| SpinachMohod 2015 | Grown near Amba NallaCollection date unreportedDry weight | mean. | ||
| Amaranth leavesNaser et al. 2018 | High-level pollution: KalakoirCollection date unreportedDry weight | mean. | ||
| Amaranth leavesNaser et al. 2018 | High-level pollution: KalakoirCollection date unreportedDry weight | mean. | ||
| Amaranth leavesNaser et al. 2018 | Low-level pollution: BARICollection date unreportedDry weight | mean. | ||
| Amaranth leavesNaser et al. 2018 | Low-level pollution: BARICollection date unreportedDry weight | mean. | ||
| Amaranth leavesNaser et al. 2018 | Medium-level pollution: ZorunCollection date unreportedDry weight | mean. | ||
| Amaranth leavesNaser et al. 2018 | Medium-level pollution: ZorunCollection date unreportedDry weight | mean. | ||
| SpinachNaser et al. 2018 | High-level pollution: KalakoirCollection date unreportedDry weight | mean. | ||
| SpinachNaser et al. 2018 | Low-level pollution: BARICollection date unreportedDry weight | mean. | ||
| SpinachNaser et al. 2018 | Medium-level pollution: ZorunCollection date unreportedDry weight | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | FuzhouCollection date unreportedBasis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | GanzhouCollection date unreportedBasis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | JiujiangCollection date unreportedBasis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | Leafy vegetablesCollection date unreportedBasis not stated | range. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | Leafy vegetablesCollection date unreportedBasis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | ShangraoCollection date unreportedBasis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | YichunCollection date unreportedBasis not stated | mean. | ||
| Head cabbageOkonofua et al. 2024 | Dry Season2019Basis not stated | mean. | ||
| Head cabbageOkonofua et al. 2024 | Dry season Control2019Basis not stated | mean. | ||
| Head cabbageOkonofua et al. 2024 | Rainy Season2019Basis not stated | mean. | ||
| Head cabbageOkonofua et al. 2024 | Rainy season Control2019Basis not stated | mean. | ||
| Amaranth leavesOrosun et al. 2023 | Irrigated farmlands along the2021Dry weight | mean. | ||
| Head cabbageOrosun et al. 2023 | Irrigated farmlands along the2021Dry weight | mean. | ||
| Leafy vegetables, other or unspecifiedOsuolale et al. 2025 | Ondo Metropolis market/farm2018Dry weight | mean.Same samples: 29 µg/kg (minimum); 49 µg/kg (maximum) | ||
| Leafy vegetables, other or unspecifiedOsuolale et al. 2025 | Ondo Metropolis market/farm2018Dry weight | median. | ||
| Leafy vegetables, other or unspecifiedOsuolale et al. 2025 | Ondo Metropolis market/farm2018Dry weight | min. | ||
| Leafy vegetables, other or unspecifiedOsuolale et al. 2025 | Ondo Metropolis market/farm2018Dry weight | max. | ||
| Head cabbagePal et al. 2017 | Domestic-wastewater-irrigatedCollection date unreportedBasis not stated | mean. | ||
| SpinachPal et al. 2017 | Domestic-wastewater-irrigatedCollection date unreportedBasis not stated | mean. | ||
| Amaranth leavesPan et al. 2016 | Zhejiang province marketed2014Fresh weight | mean. | ||
| Coriander leavesPan et al. 2016 | Zhejiang province marketed2014Fresh weight | mean. | ||
| Pak choiPan et al. 2016 | Zhejiang province marketed2014Fresh weight | mean. | ||
| Romaine lettucePan et al. 2016 | Zhejiang province marketed2014Fresh weight | mean. | ||
| Water spinachPan et al. 2016 | Zhejiang province marketed2014Fresh weight | mean. | ||
| Head cabbageRempelos et al. 2023 | Crop protection CON (CP)2004–2009Fresh weight | mean. | ||
| Head cabbageRempelos et al. 2023 | Crop protection ORG (OP)2004–2009Fresh weight | mean. | ||
| Head cabbageRempelos et al. 2023 | Fertilizer type CON (NPK)2004–2009Fresh weight | mean. | ||
| Head cabbageRempelos et al. 2023 | Fertilizer type ORG (FYM)2004–2009Fresh weight | mean. | ||
| Head cabbageRempelos et al. 2023 | Pre-crop Spring Beans2004–2009Fresh weight | mean. | ||
| Head cabbageRempelos et al. 2023 | Pre-crop Winter Barley2004–2009Fresh weight | mean. | ||
| Head lettuceRempelos et al. 2023 | Crop protection CON (CP)2004–2009Fresh weight | mean. | ||
| Head lettuceRempelos et al. 2023 | Crop protection ORG (OP)2004–2009Fresh weight | mean. | ||
| Head lettuceRempelos et al. 2023 | Fertilizer type CON (NPK)2004–2009Fresh weight | mean. | ||
| Head lettuceRempelos et al. 2023 | Fertilizer type ORG (FYM)2004–2009Fresh weight | mean. | ||
| Head lettuceRempelos et al. 2023 | Pre-crop Spring Beans2004–2009Fresh weight | mean. | ||
| Head lettuceRempelos et al. 2023 | Pre-crop Winter Barley2004–2009Fresh weight | mean. | ||
| ChivesRomero-Crespo et al. 2023 | Ponce Enriquez gold-mining area2021Fresh weight | sample value. | ||
| Coriander leavesRomero-Crespo et al. 2023 | Ponce Enriquez gold-mining area2021Fresh weight | sample value. | ||
| LettuceRomero-Crespo et al. 2023 | Ponce Enriquez gold-mining area2021Fresh weight | sample value. | ||
| LettuceRossini-Oliva et al. 2024 | City urban gardens2021–2023Dry weight | mean. | ||
| LettuceRossini-Oliva et al. 2024 | Peri-urban garden2021–2023Dry weight | mean. | ||
| LettuceRossini-Oliva et al. 2024 | Seville local market (Mk)2021–2023Dry weight | mean. | ||
| LettuceRossini-Oliva et al. 2024 | Peri-urban garden2021–2023Fresh weight | sample value. | ||
| Swiss chardRossini-Oliva et al. 2024 | City urban gardens2021–2023Dry weight | mean. | ||
| Swiss chardRossini-Oliva et al. 2024 | Mining-area urban gardens2021–2023Dry weight | mean. | ||
| Swiss chardRossini-Oliva et al. 2024 | Peri-urban garden2021–2023Dry weight | mean. | ||
| Swiss chardRossini-Oliva et al. 2024 | Seville local market (Mk)2021–2023Dry weight | mean. | ||
| ArumSadee 2022 | Local markets2021Dry weight | below LOD. | ||
| Garden cressSadee 2022 | Local markets2021Dry weight | mean. | ||
| MallowSadee 2022 | Local markets2021Dry weight | mean. | ||
| MintSadee 2022 | Local markets2021Dry weight | below LOD. | ||
| SpinachSadee 2022 | Local markets2021Dry weight | mean. | ||
| Swiss chardSadee 2022 | Local markets2021Dry weight | mean. | ||
| SpinachSalawu et al. 2015 | Samaru irrigation area2014Basis not stated | mean. | ||
| DillSaleem et al. 2025 | Town Square Farmer's MarketYear not reportedDry weight | mean. | ||
| KaleSaleem et al. 2025 | Town Square Farmer's MarketYear not reportedDry weight | mean. | ||
| SpinachSaleem et al. 2025 | Town Square Farmer's MarketYear not reportedDry weight | mean. | ||
| SpinachSalhotra et al. 2017 | Jagdalpur market supply2017Dry weight | mean. | ||
| Head cabbageSee et al. 2025 | Pasar Borong Selangor wholesale2022Fresh weight | not detected. | ||
| Mustard greensSee et al. 2025 | Pasar Borong Selangor wholesale2022Fresh weight | not detected. | ||
| Pak choiSee et al. 2025 | Pasar Borong Selangor wholesale2022Fresh weight | not detected. | ||
| SpinachSee et al. 2025 | Pasar Borong Selangor wholesale2022Fresh weight | not detected. | ||
| Coriander leavesSharma et al. 2016 | Site 1: groundwater-irrigatedCollection date unreportedDry weight | mean. | ||
| Coriander leavesSharma et al. 2016 | Site 1: groundwater-irrigatedCollection date unreportedDry weight | max. | ||
| Coriander leavesSharma et al. 2016 | Site 2: groundwater-irrigatedCollection date unreportedDry weight | mean. | ||
| Coriander leavesSharma et al. 2016 | Site 2: groundwater-irrigatedCollection date unreportedDry weight | max. | ||
| Coriander leavesSharma et al. 2016 | Site 3:Collection date unreportedDry weight | mean. | ||
| Coriander leavesSharma et al. 2016 | Site 3:Collection date unreportedDry weight | max. | ||
| Fenugreek leavesSharma et al. 2016 | Site 1: groundwater-irrigatedCollection date unreportedDry weight | mean. | ||
| Fenugreek leavesSharma et al. 2016 | Site 1: groundwater-irrigatedCollection date unreportedDry weight | max. | ||
| Fenugreek leavesSharma et al. 2016 | Site 2: groundwater-irrigatedCollection date unreportedDry weight | mean. | ||
| Fenugreek leavesSharma et al. 2016 | Site 2: groundwater-irrigatedCollection date unreportedDry weight | max. | ||
| Fenugreek leavesSharma et al. 2016 | Site 3:Collection date unreportedDry weight | mean. | ||
| Fenugreek leavesSharma et al. 2016 | Site 3:Collection date unreportedDry weight | max. | ||
| MintSharma et al. 2016 | Site 1: groundwater-irrigatedCollection date unreportedDry weight | mean. | ||
| MintSharma et al. 2016 | Site 1: groundwater-irrigatedCollection date unreportedDry weight | max. | ||
| MintSharma et al. 2016 | Site 2: groundwater-irrigatedCollection date unreportedDry weight | mean. | ||
| MintSharma et al. 2016 | Site 2: groundwater-irrigatedCollection date unreportedDry weight | max. | ||
| MintSharma et al. 2016 | Site 3:Collection date unreportedDry weight | mean. | ||
| MintSharma et al. 2016 | Site 3:Collection date unreportedDry weight | max. | ||
| SpinachSharma et al. 2016 | Site 1: groundwater-irrigatedCollection date unreportedDry weight | mean. | ||
| SpinachSharma et al. 2016 | Site 1: groundwater-irrigatedCollection date unreportedDry weight | max. | ||
| SpinachSharma et al. 2016 | Site 2: groundwater-irrigatedCollection date unreportedDry weight | mean. | ||
| SpinachSharma et al. 2016 | Site 2: groundwater-irrigatedCollection date unreportedDry weight | max. | ||
| SpinachSharma et al. 2016 | Site 3:Collection date unreportedDry weight | mean. | ||
| SpinachSharma et al. 2016 | Site 3:Collection date unreportedDry weight | max. | ||
| Leafy vegetables, other or unspecifiedSong et al. 2017 | China, 31 provinces, 2011-20152011–2015Basis not stated | mean. | ||
| BasilSouri et al. 2019 | Shahre Rey wastewater-irrigated2011Dry weight | mean. | ||
| BasilSouri et al. 2019 | Shahre Rey wastewater-irrigated2011Dry weight | mean. | ||
| Garden cressSouri et al. 2019 | Shahre Rey wastewater-irrigated2011Dry weight | mean. | ||
| Garden cressSouri et al. 2019 | Shahre Rey wastewater-irrigated2011Dry weight | mean. | ||
| Bristly ox tongueStark et al. 2019 | Oakland2014–2016Dry weight | sample value. | ||
| Bristly ox tongueStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Cat's earStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Dandelion greensStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| English plantainStark et al. 2019 | Oakland2014–2016Dry weight | sample value. | ||
| English plantainStark et al. 2019 | Oakland2014–2016Dry weight | sample value. | ||
| LettuceStark et al. 2019 | Oakland2014–2016Dry weight | sample value. | ||
| LettuceStark et al. 2019 | Oakland2014–2016Fresh weight | sample value. | ||
| MallowStark et al. 2019 | Oakland2014–2016Dry weight | sample value. | ||
| Nasturtium (Tropaeolum)Stark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Sweet fennelStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Sweet fennelStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| SpinachSuleiman and Ibrahim 2025 | Three irrigation farmsCollection date unreportedBasis not stated | mean. | ||
| Coriander leavesSultana et al. 2022 | 1st phase (early harvest)Year not reportedDry weight | sample value. | ||
| Coriander leavesSultana et al. 2022 | 2nd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| Coriander leavesSultana et al. 2022 | 3rd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| Coriander leavesSultana et al. 2022 | 4th phase (late harvest)Year not reportedDry weight | sample value. | ||
| Coriander leavesSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | mean.Same samples: 600–1,800 µg/kg (range); up to 400 µg/kg (range maximum); up to 500 µg/kg (range maximum); up to 900 µg/kg (range maximum) | ||
| Coriander leavesSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | range. | ||
| Head cabbageSultana et al. 2022 | 1st phase (early harvest)Year not reportedDry weight | not detected. | ||
| Head cabbageSultana et al. 2022 | 2nd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| Head cabbageSultana et al. 2022 | 3rd phase (mid-time harvest)Year not reportedDry weight | not detected. | ||
| Head cabbageSultana et al. 2022 | 4th phase (late harvest)Year not reportedDry weight | not detected. | ||
| Head cabbageSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | mean. | ||
| Head cabbageSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | range. | ||
| MintSultana et al. 2022 | 1st phase (early harvest)Year not reportedDry weight | sample value. | ||
| MintSultana et al. 2022 | 2nd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| MintSultana et al. 2022 | 3rd phase (mid-time harvest)Year not reportedDry weight | not detected. | ||
| MintSultana et al. 2022 | 4th phase (late harvest)Year not reportedDry weight | not detected. | ||
| MintSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | mean. | ||
| MintSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | range. | ||
| Mustard greensSultana et al. 2022 | 1st phase (early harvest)Year not reportedDry weight | sample value. | ||
| Mustard greensSultana et al. 2022 | 2nd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| Mustard greensSultana et al. 2022 | 3rd phase (mid-time harvest)Year not reportedDry weight | not detected. | ||
| Mustard greensSultana et al. 2022 | 4th phase (late harvest)Year not reportedDry weight | not detected. | ||
| Mustard greensSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | mean. | ||
| Mustard greensSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | range. | ||
| SpinachSultana et al. 2022 | 1st phase (early harvest)Year not reportedDry weight | sample value. | ||
| SpinachSultana et al. 2022 | 2nd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| SpinachSultana et al. 2022 | 3rd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| SpinachSultana et al. 2022 | 4th phase (late harvest)Year not reportedDry weight | not detected. | ||
| SpinachSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | mean. | ||
| SpinachSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | range. | ||
| Head cabbageTweij 2018 | Iranian green cabbage2017Basis not stated | sample value. | ||
| Head cabbageTweij 2018 | Iranian green lettuce-type cabbage2017Basis not stated | sample value. | ||
| Head cabbageTweij 2018 | Iranian purple cabbage2017Basis not stated | sample value. | ||
| Head cabbageTweij 2018 | Iraqi green cabbage2017Basis not stated | sample value. | ||
| LettuceUK TDS 2014 (Fera 2015) | UK TDS 2014 category 11_652014Fresh weight | sample value. | ||
| SpinachUS House Subcommittee 2021 | US baby-food manufacturer2018–2019Basis not stated | sample value. | ||
| SpinachUS House Subcommittee 2021 | US baby-food manufacturer2018–2019Basis not stated | sample value. | ||
| SpinachUS House Subcommittee 2021 | US baby-food manufacturer2018–2019Basis not stated | sample value. | ||
| SpinachUS House Subcommittee 2021 | US baby-food manufacturer2018–2019Basis not stated | sample value. | ||
| SpinachUS House Subcommittee 2021 | US baby-food manufacturer2018–2019Basis not stated | sample value. | ||
| LettuceVela et al. 2023 | Quito wholesale open marketsCollection date unreportedDry weight | mean. | ||
| LettuceVela et al. 2023 | Table II Cd lettuce row 1Collection date unreportedDry weight | sample value. | ||
| LettuceVela et al. 2023 | Table II Cd lettuce row 2Collection date unreportedDry weight | sample value. | ||
| LettuceVela et al. 2023 | Table II Cd lettuce row 3Collection date unreportedDry weight | sample value. | ||
| LettuceVela et al. 2023 | Table II Cd lettuce row 4Collection date unreportedDry weight | sample value. | ||
| Alternanthera sessilis leavesWijeyaratne et al. 2021 | Non-organic cultivation sites2019Dry weight | mean. | ||
| Alternanthera sessilis leavesWijeyaratne et al. 2021 | Non-organic cultivation sites2019Dry weight | mean. | ||
| Alternanthera sessilis leavesWijeyaratne et al. 2021 | Organic cultivation sites2019Dry weight | mean. | ||
| Alternanthera sessilis leavesWijeyaratne et al. 2021 | Organic cultivation sites2019Dry weight | mean. | ||
| BasilWiniarska-Mieczan et al. 2023 | Lublin groceriesCollection date unreportedFresh weight | mean. | ||
| Celery leavesWiniarska-Mieczan et al. 2023 | Lublin groceriesCollection date unreportedFresh weight | mean. | ||
| Coriander leavesWiniarska-Mieczan et al. 2023 | Lublin groceriesCollection date unreportedFresh weight | mean. | ||
| DillWiniarska-Mieczan et al. 2023 | Lublin groceriesCollection date unreportedFresh weight | mean. | ||
| LovageWiniarska-Mieczan et al. 2023 | Lublin groceriesCollection date unreportedFresh weight | mean. | ||
| ParsleyWiniarska-Mieczan et al. 2023 | Lublin groceriesCollection date unreportedFresh weight | mean. | ||
| WatercressWiniarska-Mieczan et al. 2023 | Lublin groceriesCollection date unreportedFresh weight | mean. | ||
| Chinese cabbageWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| ChivesWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| Head cabbageWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| LettuceWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| Pak choiWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| Romaine lettuceWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| SpinachWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Cattle manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Chemical fertilizer (NPK)2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Chicken manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Control (no fertilizer)2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Sheep manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Turkey manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Vermicompost2017Basis not stated | mean. | ||
3 findings · 1 study | ||||
| Head cabbage (fermented)WHO GEMS/Food | WHO region EURO; random sampling2003–2019Fresh weight | median. | ||
| Head cabbage (fermented)WHO GEMS/Food | WHO region EURO; targeted sampling2003Fresh weight | median. | ||
| Head cabbage (fermented)WHO GEMS/Food | WHO region EURO; unknown2017Fresh weight | median. | ||
5 findings · 1 study | ||||
| Head cabbage (frozen)WHO GEMS/Food | WHO region EURO; random sampling2003–2007Fresh weight | median. | ||
| Kale (frozen)WHO GEMS/Food | WHO region EURO; random sampling2003–2006Fresh weight | median. | ||
| Leafy vegetables, other or unspecified (frozen)WHO GEMS/Food | WHO region EURO; random sampling2003–2005Fresh weight | median. | ||
| Spinach (frozen)WHO GEMS/Food | WHO region EURO; random sampling2003–2018Fresh weight | median. | ||
| Spinach (frozen)WHO GEMS/Food | WHO region EURO; unknown2017Fresh weight | median. | ||
3 findings · 1 study | ||||
| Kale (canned)WHO GEMS/Food | WHO region EURO; random sampling2005–2007Fresh weight | median. | ||
| Leafy vegetables, other or unspecified (canned)WHO GEMS/Food | WHO region EURO; random sampling2004Fresh weight | median. | ||
| Spinach (canned)WHO GEMS/Food | WHO region EURO; random sampling2007Fresh weight | median. | ||
4 findings · 2 studies | ||||
| Beet leaves (dried)Cardini et al. 2025 | Retail powder from Danish internetCollection date unreportedBasis not stated | sample value. | ||
| Basil (dried)Winiarska-Mieczan et al. 2023 | Lublin groceries (dried herbs)Collection date unreportedFresh weight | mean. | ||
| Coriander leaves (dried)Winiarska-Mieczan et al. 2023 | Lublin groceries (dried herbs)Collection date unreportedFresh weight | mean. | ||
| Lovage (dried)Winiarska-Mieczan et al. 2023 | Lublin groceries (dried herbs)Collection date unreportedFresh weight | mean. | ||
217 findings · 33 studies · 10 spread figures folded in | ||||
| Chinese cabbageWHO GEMS/Food | WHO region EURO; random sampling2014–2023Fresh weight | median. | ||
| Chinese cabbageWHO GEMS/Food | WHO region EURO; unknown2022Fresh weight | median. | ||
| Head cabbageWHO GEMS/Food | WHO region EURO; random sampling2007–2023Fresh weight | median. | ||
| Head cabbageWHO GEMS/Food | WHO region EURO; targeted sampling2020–2023Fresh weight | median. | ||
| Head cabbageWHO GEMS/Food | WHO region EURO; unknown2011–2022Fresh weight | median. | ||
| Head cabbageWHO GEMS/Food | WHO region European Union2004–2014Fresh weight | median. | ||
| Head cabbageWHO GEMS/Food | WHO region European Union2000–2012Fresh weight | median. | ||
| Head cabbageWHO GEMS/Food | WHO region European Union; unknown2002–2012Fresh weight | median. | ||
| Head cabbageWHO GEMS/Food | WHO region WPRO; random sampling2004–2009Fresh weight | median. | ||
| Head lettuceWHO GEMS/Food | WHO region EURO; random sampling2010–2023Fresh weight | median. | ||
| Head lettuceWHO GEMS/Food | WHO region EURO; unknown2022Fresh weight | median. | ||
| Head lettuceWHO GEMS/Food | WHO region European Union2003–2014Fresh weight | median. | ||
| Head lettuceWHO GEMS/Food | WHO region WPRO; random sampling2004–2014Fresh weight | median. | ||
| Head lettuceWHO GEMS/Food | WHO region WPRO; targeted sampling1998–2008Fresh weight | median. | ||
| KaleWHO GEMS/Food | WHO region EURO; random sampling2010–2023Fresh weight | median. | ||
| KaleWHO GEMS/Food | WHO region EURO; unknown2004–2022Fresh weight | median. | ||
| KaleWHO GEMS/Food | WHO region European Union2004–2013Fresh weight | median. | ||
| KaleWHO GEMS/Food | WHO region European Union; unknown2005–2006Fresh weight | median. | ||
| KaleWHO GEMS/Food | WHO region WPRO; targeted sampling2001Fresh weight | median. | ||
| Lamb's lettuceWHO GEMS/Food | WHO region EURO; random sampling2008–2014Fresh weight | median. | ||
| Lamb's lettuceWHO GEMS/Food | WHO region European Union2003–2014Fresh weight | median. | ||
| Lamb's lettuceWHO GEMS/Food | WHO region European Union; unknown2004–2006Fresh weight | median. | ||
| Leaf lettuceWHO GEMS/Food | WHO region EURO; random sampling2009–2023Fresh weight | median. | ||
| Leaf lettuceWHO GEMS/Food | WHO region EURO; targeted sampling2022–2023Fresh weight | median. | ||
| Leaf lettuceWHO GEMS/Food | WHO region EURO; unknown2009–2022Fresh weight | median. | ||
| Leaf lettuceWHO GEMS/Food | WHO region European Union2003–2014Fresh weight | median. | ||
| Leaf lettuceWHO GEMS/Food | WHO region European Union; unknown2003–2012Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region EURO; random sampling2009–2023Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region EURO; unknown2004–2022Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region European Union2003–2013Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region European Union; unknown2002–2012Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region WPRO; random sampling2005Fresh weight | median. | ||
| Mustard greensWHO GEMS/Food | WHO region WPRO; random sampling2005Fresh weight | median. | ||
| Mustard greensWHO GEMS/Food | WHO region WPRO; targeted sampling2002Fresh weight | median. | ||
| Pak choiWHO GEMS/Food | WHO region EURO; random sampling2007–2015Fresh weight | median. | ||
| Pak choiWHO GEMS/Food | WHO region European Union2003–2014Fresh weight | median. | ||
| Pak choiWHO GEMS/Food | WHO region European Union2012Fresh weight | median. | ||
| Pak choiWHO GEMS/Food | WHO region European Union; unknown2005Fresh weight | median. | ||
| SpinachWHO GEMS/Food | WHO region EURO; unknown2008Dry weight | median. | ||
| SpinachWHO GEMS/Food | WHO region EURO; random sampling2009–2023Fresh weight | median. | ||
| SpinachWHO GEMS/Food | WHO region EURO; unknown2007–2017Fresh weight | median. | ||
| SpinachWHO GEMS/Food | WHO region European Union2003–2014Fresh weight | median. | ||
| SpinachWHO GEMS/Food | WHO region European Union2000Fresh weight | median. | ||
| SpinachWHO GEMS/Food | WHO region European Union; unknown2003–2012Fresh weight | median. | ||
| SpinachWHO GEMS/Food | WHO region WPRO; random sampling2013–2014Fresh weight | median. | ||
| WatercressWHO GEMS/Food | WHO region EURO; random sampling2019–2020Fresh weight | median. | ||
| Head cabbageFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Head lettuceFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Leaf lettuceFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| SpinachFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Coriander leavesAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| Coriander leavesAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| Coriander leavesAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| Coriander leavesAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | mean. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | mean. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | mean. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | mean. | ||
| LettuceAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| LettuceAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| LettuceAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| LettuceAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| HelenchaAhmed et al. 2017 | all three areas2015Dry weight | range. | ||
| HelenchaAhmed et al. 2017 | Banglabazar2015Dry weight | mean. | ||
| HelenchaAhmed et al. 2017 | Chandra2015Dry weight | mean. | ||
| HelenchaAhmed et al. 2017 | Kashimpur2015Dry weight | mean. | ||
| Leafy vegetables, other or unspecifiedAhmed et al. 2017 | all three areas2015Dry weight | mean. | ||
| Leafy vegetables, other or unspecifiedAhmed et al. 2017 | beds with soil As 6.85-8.51 mg/kg2015Dry weight | max. | ||
| Taro leavesAhmed et al. 2017 | Banglabazar2015Dry weight | mean. | ||
| Taro leavesAhmed et al. 2017 | Chandra2015Dry weight | mean. | ||
| Taro leavesAhmed et al. 2017 | Kashimpur2015Dry weight | mean. | ||
| Water spinachAhmed et al. 2017 | Banglabazar2015Dry weight | mean. | ||
| Water spinachAhmed et al. 2017 | Chandra2015Dry weight | mean. | ||
| Water spinachAhmed et al. 2017 | Kashimpur2015Dry weight | mean. | ||
| Head cabbageAlimohammadi et al. 2018 | Autumn 20162016Fresh weight | mean. | ||
| Head cabbageAlimohammadi et al. 2018 | Summer 20162016Fresh weight | mean. | ||
| LettuceAlimohammadi et al. 2018 | Autumn 20162016Fresh weight | mean. | ||
| LettuceAlimohammadi et al. 2018 | Summer 20162016Fresh weight | mean. | ||
| DillBora et al. 2022 | Vegetable and fruit market2020–2021Fresh weight | mean. | ||
| Head cabbageBora et al. 2022 | Amateur farmers2020–2021Fresh weight | below LOD. | ||
| Head cabbageBora et al. 2022 | Amateur farmers2020–2021Fresh weight | below LOD. | ||
| Head cabbageBora et al. 2022 | Amateur farmers2020–2021Fresh weight | below LOD. | ||
| Head cabbageBora et al. 2022 | Vegetable and fruit market2020–2021Fresh weight | below LOD. | ||
| Head cabbageBora et al. 2022 | Vegetable and fruit market2020–2021Fresh weight | below LOD. | ||
| Head cabbageBora et al. 2022 | Vegetable and fruit market2020–2021Fresh weight | mean. | ||
| LettuceBora et al. 2022 | Amateur farmers2020–2021Fresh weight | below LOD. | ||
| LettuceBora et al. 2022 | Amateur farmers2020–2021Fresh weight | below LOD. | ||
| LettuceBora et al. 2022 | Amateur farmers2020–2021Fresh weight | mean. | ||
| LettuceBora et al. 2022 | Vegetable and fruit market2020–2021Fresh weight | below LOD. | ||
| LettuceBora et al. 2022 | Vegetable and fruit market2020–2021Fresh weight | below LOD. | ||
| LettuceBora et al. 2022 | Vegetable and fruit market2020–2021Fresh weight | mean. | ||
| LovageBora et al. 2022 | Amateur farmers2020–2021Fresh weight | below LOD. | ||
| ParsleyBora et al. 2022 | Amateur farmers2020–2021Fresh weight | below LOD. | ||
| ParsleyBora et al. 2022 | Vegetable and fruit market2020–2021Fresh weight | below LOD. | ||
| TarragonBora et al. 2022 | Vegetable and fruit market2020–2021Fresh weight | below LOD. | ||
| Leafy vegetables, other or unspecifiedChen et al. 2025 | Chongqing retail/market samples2017–2021Basis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedChen et al. 2025 | Chongqing retail/market samples2017–2021Basis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedChen et al. 2025 | Chongqing retail/market samples2017–2021Basis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedDai et al. 2024 | Chongqing city food safety2018–2023Basis not stated | 5th percentile. | ||
| Leafy vegetables, other or unspecifiedDai et al. 2024 | Chongqing city food safety2018–2023Basis not stated | 95th percentile. | ||
| Leafy vegetables, other or unspecifiedDai et al. 2024 | Chongqing city food safety2018–2023Basis not stated | mean.Same samples: 1.4 µg/kg (5th percentile); 65 µg/kg (95th percentile) | ||
| Leafy vegetables, other or unspecifiedDai et al. 2024 | Chongqing city food safety2018–2023Basis not stated | median. | ||
| Head cabbageFonge et al. 2021 | SantaCollection date unreportedDry weight | mean. | ||
| Head cabbageGurmu et al. 2025 | Akaki Rivers irrigated farmlandYear not reportedDry weight | mean. | ||
| Malabar spinachHossen et al. 2026 | farm-82024Fresh weight | mean. | ||
| Chinese cabbageHuang et al. 2014 | Zhejiang province2012Fresh weight | mean. | ||
| Head cabbageHuang et al. 2014 | Zhejiang province2012Fresh weight | mean. | ||
| Pak choiHuang et al. 2014 | Zhejiang province2012Fresh weight | mean. | ||
| Rape greensHuang et al. 2014 | Zhejiang province2012Fresh weight | mean. | ||
| SpinachHuang et al. 2014 | Zhejiang province2012Fresh weight | mean. | ||
| Chinese cabbageJiang et al. 2020 | SZD village near a battery factory2018Basis not stated | mean. | ||
| Pak choiJiang et al. 2020 | SZD village near a battery factory2018Basis not stated | mean. | ||
| SpinachJiang et al. 2020 | SZD village near a battery factory2018Basis not stated | mean. | ||
| Kenaf leavesJoseph et al. 2025 | Gobbiya Dam irrigated farmlandCollection date unreportedDry weight | mean. | ||
| SpinachJoseph et al. 2025 | Gobbiya Dam irrigated farmlandCollection date unreportedDry weight | mean. | ||
| KaleKim et al. 2024 | All six farms, field samples2019Fresh weight | mean.Same samples: 10–34.5 µg/kg (range) | ||
| KaleKim et al. 2024 | All six farms, field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | All six farms, field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm R1 (Rural), field samples2019Fresh weight | mean.Same samples: 10–10 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm R1 (Rural), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm R1 (Rural), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm R2 (Rural), field samples2019Fresh weight | mean.Same samples: 10–10 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm R2 (Rural), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm R2 (Rural), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm R3 (Rural), field samples2019Fresh weight | mean.Same samples: 10–10 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm R3 (Rural), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm R3 (Rural), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm R4 (Rural), field samples2019Fresh weight | mean.Same samples: 10–10 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm R4 (Rural), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm R4 (Rural), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm U1 (Urban), field samples2019Fresh weight | mean.Same samples: 10–13.5 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm U1 (Urban), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm U1 (Urban), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm U2 (Urban), field samples2019Fresh weight | mean.Same samples: 21.4–34.5 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm U2 (Urban), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm U2 (Urban), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Rural farms R1-R42019Fresh weight | mean. | ||
| KaleKim et al. 2024 | Urban farms U1-U22019Fresh weight | mean. | ||
| Head cabbageLi et al. 2021 | Four major producing cities of2018Fresh weight | mean.Same samples: 50–400 µg/kg (range) | ||
| Head cabbageLi et al. 2021 | Four major producing cities of2018Fresh weight | range. | ||
| Head cabbageLiang et al. 2019 | BeijingCollection date unreportedDry weight | mean. | ||
| Mustard greensLiu et al. 2023 | MustardCollection date unreportedBasis not stated | mean. | ||
| Head lettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Lamb's lettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| LettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| RocketMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Romaine lettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| SpinachMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Swiss chardMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| WatercressMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Fenugreek leavesMawari et al. 2022 | Farms in ChandrahalCollection date unreportedBasis not stated | sample value. | ||
| Head cabbageMawari et al. 2022 | Farms in ChandrahalCollection date unreportedBasis not stated | sample value. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | FuzhouCollection date unreportedBasis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | GanzhouCollection date unreportedBasis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | JiujiangCollection date unreportedBasis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | Leafy vegetablesCollection date unreportedBasis not stated | range. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | Leafy vegetablesCollection date unreportedBasis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | ShangraoCollection date unreportedBasis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | YichunCollection date unreportedBasis not stated | mean. | ||
| Amaranth leavesOrosun et al. 2023 | Irrigated farmlands along the2021Dry weight | mean. | ||
| Head cabbageOrosun et al. 2023 | Irrigated farmlands along the2021Dry weight | mean. | ||
| Coriander leavesPan et al. 2016 | Zhejiang province marketed2014Fresh weight | mean. | ||
| Pak choiPan et al. 2016 | Zhejiang province marketed2014Fresh weight | mean. | ||
| ChivesRomero-Crespo et al. 2023 | Ponce Enriquez gold-mining area2021Fresh weight | sample value. | ||
| Coriander leavesRomero-Crespo et al. 2023 | Ponce Enriquez gold-mining area2021Fresh weight | sample value. | ||
| LettuceRomero-Crespo et al. 2023 | Ponce Enriquez gold-mining area2021Fresh weight | sample value. | ||
| LettuceRossini-Oliva et al. 2024 | City urban gardens2021–2023Dry weight | mean. | ||
| LettuceRossini-Oliva et al. 2024 | Peri-urban garden2021–2023Dry weight | mean. | ||
| LettuceRossini-Oliva et al. 2024 | Seville local market (Mk)2021–2023Dry weight | mean. | ||
| Swiss chardRossini-Oliva et al. 2024 | City urban gardens2021–2023Dry weight | mean. | ||
| Swiss chardRossini-Oliva et al. 2024 | Mining-area urban gardens2021–2023Dry weight | mean. | ||
| Swiss chardRossini-Oliva et al. 2024 | Peri-urban garden2021–2023Dry weight | mean. | ||
| Swiss chardRossini-Oliva et al. 2024 | Seville local market (Mk)2021–2023Dry weight | mean. | ||
| ArumSadee 2022 | Local markets2021Dry weight | mean. | ||
| Garden cressSadee 2022 | Local markets2021Dry weight | mean. | ||
| MallowSadee 2022 | Local markets2021Dry weight | mean. | ||
| MintSadee 2022 | Local markets2021Dry weight | mean. | ||
| SpinachSadee 2022 | Local markets2021Dry weight | mean. | ||
| Swiss chardSadee 2022 | Local markets2021Dry weight | mean. | ||
| DillSaleem et al. 2025 | Town Square Farmer's MarketYear not reportedDry weight | mean. | ||
| KaleSaleem et al. 2025 | Town Square Farmer's MarketYear not reportedDry weight | mean. | ||
| SpinachSaleem et al. 2025 | Town Square Farmer's MarketYear not reportedDry weight | mean. | ||
| BasilSalehipour et al. 2015 | Wholesaler outlets andCollection date unreportedFresh weight | not detected. | ||
| LettuceSalehipour et al. 2015 | Wholesaler outlets andCollection date unreportedFresh weight | not detected. | ||
| MintSalehipour et al. 2015 | Wholesaler outlets andCollection date unreportedFresh weight | not detected. | ||
| BasilSouri et al. 2019 | Shahre Rey wastewater-irrigated2011Dry weight | mean. | ||
| BasilSouri et al. 2019 | Shahre Rey wastewater-irrigated2011Dry weight | mean. | ||
| Garden cressSouri et al. 2019 | Shahre Rey wastewater-irrigated2011Dry weight | mean. | ||
| Garden cressSouri et al. 2019 | Shahre Rey wastewater-irrigated2011Dry weight | mean. | ||
| Bristly ox tongueStark et al. 2019 | Oakland2014–2016Dry weight | sample value. | ||
| Bristly ox tongueStark et al. 2019 | Oakland2014–2016Dry weight | sample value. | ||
| Cat's earStark et al. 2019 | Oakland2014–2016Dry weight | sample value. | ||
| Dandelion greensStark et al. 2019 | Oakland2014–2016Dry weight | sample value. | ||
| English plantainStark et al. 2019 | Oakland2014–2016Dry weight | sample value. | ||
| English plantainStark et al. 2019 | Oakland2014–2016Dry weight | sample value. | ||
| LettuceStark et al. 2019 | Oakland2014–2016Dry weight | sample value. | ||
| MallowStark et al. 2019 | Oakland2014–2016Dry weight | sample value. | ||
| Nasturtium (Tropaeolum)Stark et al. 2019 | Oakland2014–2016Dry weight | sample value. | ||
| Sweet fennelStark et al. 2019 | Oakland2014–2016Dry weight | sample value. | ||
| Sweet fennelStark et al. 2019 | Oakland2014–2016Dry weight | sample value. | ||
| LettuceUK TDS 2014 (Fera 2015) | UK TDS 2014 category 11_652014Fresh weight | sample value. | ||
| Chinese cabbageWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| ChivesWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| Head cabbageWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| LettuceWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| Pak choiWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| Romaine lettuceWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| SpinachWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Cattle manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Chemical fertilizer (NPK)2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Chicken manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Control (no fertilizer)2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Sheep manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Turkey manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Vermicompost2017Basis not stated | mean. | ||
2 findings · 1 study | ||||
| Chinese cabbage (fermented)WHO GEMS/Food | WHO region EURO; random sampling2022Fresh weight | median. | ||
| Head cabbage (fermented)WHO GEMS/Food | WHO region EURO; random sampling2008–2015Fresh weight | median. | ||
4 findings · 1 study | ||||
| Chinese cabbage (frozen)WHO GEMS/Food | WHO region EURO; random sampling2014–2015Fresh weight | median. | ||
| Kale (frozen)WHO GEMS/Food | WHO region EURO; random sampling2018Fresh weight | median. | ||
| Leafy vegetables, other or unspecified (frozen)WHO GEMS/Food | WHO region EURO; random sampling2014–2017Fresh weight | median. | ||
| Spinach (frozen)WHO GEMS/Food | WHO region EURO; random sampling2009–2022Fresh weight | median. | ||
26 findings · 6 studies | ||||
| Head cabbage (cooked)WHO GEMS/Food | WHO region AFRO; random sampling2018Fresh weight | median. | ||
| Head cabbage (cooked)WHO GEMS/Food | WHO region EURO; random sampling2008–2015Fresh weight | median. | ||
| Head cabbage (cooked)WHO GEMS/Food | WHO region European Union2008–2014Fresh weight | median. | ||
| Head cabbage (cooked)WHO GEMS/Food | WHO region European Union; unknown2002–2003Fresh weight | median. | ||
| Head cabbage (cooked)WHO GEMS/Food | WHO region WPRO; targeted sampling2008Fresh weight | median. | ||
| Kale (cooked)WHO GEMS/Food | WHO region EURO; random sampling2008–2015Fresh weight | median. | ||
| Kale (cooked)WHO GEMS/Food | WHO region European Union2008–2014Fresh weight | median. | ||
| Leafy vegetables, other or unspecified (cooked)WHO GEMS/Food | WHO region AFRO; random sampling2018Fresh weight | median. | ||
| Leafy vegetables, other or unspecified (cooked)WHO GEMS/Food | WHO region WPRO; targeted sampling1998Fresh weight | median. | ||
| Pak choi (cooked)WHO GEMS/Food | WHO region EURO; random sampling2008Fresh weight | median. | ||
| Pak choi (cooked)WHO GEMS/Food | WHO region European Union2008Fresh weight | median. | ||
| Pak choi (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2014Fresh weight | median. | ||
| Spinach (cooked)WHO GEMS/Food | WHO region EURO; random sampling2008–2023Fresh weight | median. | ||
| Spinach (cooked)WHO GEMS/Food | WHO region European Union2007–2014Fresh weight | median. | ||
| Collard greens (cooked)FDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Kale (cooked)FDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Spinach (cooked)Clair-Caliot et al. 2021 | Market spinach leaves from market2017–2018Fresh weight | mean. | ||
| Leafy vegetables, other or unspecified (cooked)Mahaffey et al. 1975 | US FDA Total Diet Study FY 19731968–1974Basis not stated | below LOD. | ||
| Head cabbage (cooked)UK TDS 2014 (Fera 2015) | UK TDS 2014 category 11_622014Fresh weight | below LOD. | ||
| Other green vegetables (mixed category) (cooked)UK TDS 2014 (Fera 2015) | UK TDS 2014 category 11_682014Fresh weight | below LOD. | ||
| Chinese cabbage (cooked)Wang et al. 2020 | Jilin ProvinceCollection date unreportedBasis not stated | mean. | ||
| Chinese cabbage (cooked)Wang et al. 2020 | Other AreaCollection date unreportedBasis not stated | mean. | ||
| Chinese cabbage (cooked)Wang et al. 2020 | Songhua River AreaCollection date unreportedBasis not stated | mean. | ||
| Leek (cooked)Wang et al. 2020 | Jilin ProvinceCollection date unreportedBasis not stated | mean. | ||
| Leek (cooked)Wang et al. 2020 | Other AreaCollection date unreportedBasis not stated | mean. | ||
| Leek (cooked)Wang et al. 2020 | Songhua River AreaCollection date unreportedBasis not stated | mean. | ||
3 findings · 1 study | ||||
| Head cabbage (dried)WHO GEMS/Food | WHO region EURO; random sampling2014Fresh weight | median. | ||
| Leafy vegetables, other or unspecified (dried)WHO GEMS/Food | WHO region EURO; random sampling2016–2017Fresh weight | median. | ||
| Leafy vegetables, other or unspecified (dried)WHO GEMS/Food | WHO region EURO; targeted sampling2015Fresh weight | median. | ||
14 findings · 2 studies | ||||
| Amaranth leaves (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2022Fresh weight | median. | ||
| Chinese cabbage (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Head cabbage (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Head lettuce (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010Fresh weight | median. | ||
| Kale (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Leaf lettuce (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Leafy vegetables, other or unspecified (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Mustard greens (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Pak choi (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010Fresh weight | median. | ||
| Spinach (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2022Fresh weight | median. | ||
| Water spinach (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2022Fresh weight | median. | ||
| Watercress (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Head cabbage (cooked)UK TDS 2014 (Fera 2015) | UK TDS 2014 category 11_622014Fresh weight | below LOD. | ||
| Other green vegetables (mixed category) (cooked)UK TDS 2014 (Fera 2015) | UK TDS 2014 category 11_682014Fresh weight | below LOD. | ||
31 findings · 3 studies | ||||
| Head cabbageWHO GEMS/Food | WHO region EURO; random sampling2010Dry weight | median. | ||
| Head cabbageWHO GEMS/Food | WHO region EURO; random sampling2007–2016Fresh weight | median. | ||
| KaleWHO GEMS/Food | WHO region EURO; random sampling2010Dry weight | median. | ||
| KaleWHO GEMS/Food | WHO region EURO; random sampling2010–2016Fresh weight | median. | ||
| KaleWHO GEMS/Food | WHO region WPRO; targeted sampling2000Fresh weight | median. | ||
| Leaf lettuceWHO GEMS/Food | WHO region EURO; random sampling2010Dry weight | median. | ||
| Leaf lettuceWHO GEMS/Food | WHO region EURO; random sampling2009–2022Fresh weight | median. | ||
| Leaf lettuceWHO GEMS/Food | WHO region EURO; unknown2017Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region EURO; random sampling2010Dry weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region EURO; random sampling2007–2023Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region EURO; unknown2016–2017Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region WPRO; random sampling2005Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region WPRO; targeted sampling2004Fresh weight | median. | ||
| Pak choiWHO GEMS/Food | WHO region WPRO; random sampling2020–2022Fresh weight | median. | ||
| SpinachWHO GEMS/Food | WHO region EURO; random sampling2010Dry weight | median. | ||
| SpinachWHO GEMS/Food | WHO region EURO; random sampling2010–2019Fresh weight | median. | ||
| SpinachWHO GEMS/Food | WHO region WPRO; random sampling2020Fresh weight | median. | ||
| Water spinachWHO GEMS/Food | WHO region WPRO; random sampling2020Fresh weight | median. | ||
| Amaranth leavesHong Kong TDS 2012 | 1st HKTDS2010–2011Basis not stated | mean. | ||
| Chinese cabbageHong Kong TDS 2012 | 1st HKTDS2010–2011Basis not stated | mean. | ||
| Chinese kaleHong Kong TDS 2012 | 1st HKTDS2010–2011Basis not stated | mean. | ||
| Choy sumHong Kong TDS 2012 | 1st HKTDS2010–2011Basis not stated | mean. | ||
| Head cabbageHong Kong TDS 2012 | 1st HKTDS2010–2011Basis not stated | not detected. | ||
| LettuceHong Kong TDS 2012 | 1st HKTDS2010–2011Basis not stated | mean. | ||
| LettuceHong Kong TDS 2012 | 1st HKTDS2010–2011Basis not stated | not detected. | ||
| Mustard greensHong Kong TDS 2012 | 1st HKTDS2010–2011Basis not stated | mean. | ||
| Pak choiHong Kong TDS 2012 | 1st HKTDS2010–2011Basis not stated | mean. | ||
| SpinachHong Kong TDS 2012 | 1st HKTDS2010–2011Basis not stated | mean. | ||
| Water spinachHong Kong TDS 2012 | 1st HKTDS2010–2011Basis not stated | mean. | ||
| WatercressHong Kong TDS 2012 | 1st HKTDS2010–2011Basis not stated | mean. | ||
| LettuceUK TDS 2014 (Fera 2015) | UK TDS 2014 category 11_652014Fresh weight | below LOD. | ||
1 finding · 1 study | ||||
| Head cabbage (canned)WHO GEMS/Food | WHO region WPRO; random sampling2020Fresh weight | median. | ||
1 finding · 1 study | ||||
| Leafy vegetables, other or unspecified (dried)WHO GEMS/Food | WHO region EURO; random sampling2017Fresh weight | median. | ||
2 findings · 1 study | ||||
| Leafy vegetables, other or unspecified (frozen)WHO GEMS/Food | WHO region EURO; random sampling2017Fresh weight | median. | ||
| Spinach (frozen)WHO GEMS/Food | WHO region EURO; random sampling2019Fresh weight | median. | ||
3 findings · 1 study | ||||
| Leaf lettuceWHO GEMS/Food | WHO region EURO; random sampling2022Fresh weight | median. | ||
| Leaf lettuceWHO GEMS/Food | WHO region EURO; unknown2017Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region EURO; unknown2016–2017Fresh weight | median. | ||
170 findings · 25 studies · 16 spread figures folded in | ||||
| Head cabbageWHO GEMS/Food | WHO region EURO; random sampling2004–2020Fresh weight | median. | ||
| Head cabbageWHO GEMS/Food | WHO region EURO; targeted sampling2000Fresh weight | median. | ||
| Head cabbageWHO GEMS/Food | WHO region EURO; unknown2011Fresh weight | median. | ||
| Head cabbageWHO GEMS/Food | WHO region WPRO; random sampling2004Fresh weight | median. | ||
| Head lettuceWHO GEMS/Food | WHO region EURO; random sampling2010–2017Fresh weight | median. | ||
| Head lettuceWHO GEMS/Food | WHO region WPRO; random sampling1998–2014Fresh weight | median. | ||
| Head lettuceWHO GEMS/Food | WHO region WPRO; targeted sampling1998–2008Fresh weight | median. | ||
| KaleWHO GEMS/Food | WHO region EURO; random sampling2004–2016Fresh weight | median. | ||
| KaleWHO GEMS/Food | WHO region EURO; targeted sampling2001Fresh weight | median. | ||
| KaleWHO GEMS/Food | WHO region WPRO; targeted sampling2001Fresh weight | median. | ||
| Lamb's lettuceWHO GEMS/Food | WHO region EURO; random sampling2006–2017Fresh weight | median. | ||
| Leaf lettuceWHO GEMS/Food | WHO region EURO; random sampling2004–2020Fresh weight | median. | ||
| Leaf lettuceWHO GEMS/Food | WHO region EURO; unknown2011–2017Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region EURO; random sampling2005–2017Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region EURO; targeted sampling2001Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region EURO; unknown2011–2017Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region WPRO; random sampling2005Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region WPRO; targeted sampling2001Fresh weight | median. | ||
| Mustard greensWHO GEMS/Food | WHO region WPRO; random sampling1998–2005Fresh weight | median. | ||
| Mustard greensWHO GEMS/Food | WHO region WPRO; targeted sampling2002Fresh weight | median. | ||
| Pak choiWHO GEMS/Food | WHO region EURO; random sampling2005–2020Fresh weight | median. | ||
| Pak choiWHO GEMS/Food | WHO region WPRO; random sampling1998Fresh weight | median. | ||
| Pak choiWHO GEMS/Food | WHO region WPRO; targeted sampling2001Fresh weight | median. | ||
| SpinachWHO GEMS/Food | WHO region EURO; random sampling2005–2020Fresh weight | median. | ||
| SpinachWHO GEMS/Food | WHO region EURO; targeted sampling2000Fresh weight | median. | ||
| SpinachWHO GEMS/Food | WHO region EURO; unknown2011–2016Fresh weight | median. | ||
| WatercressWHO GEMS/Food | WHO region EURO; random sampling2011Fresh weight | median. | ||
| Head cabbageFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Head lettuceFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Leaf lettuceFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| SpinachFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| SpinachAbubakar et al. 2025 | Site ICollection date unreportedBasis not stated | mean. | ||
| Coriander leavesAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| Coriander leavesAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| Coriander leavesAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| Coriander leavesAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | mean. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | mean. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | mean. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | mean. | ||
| LettuceAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| LettuceAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| LettuceAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| LettuceAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| Leafy vegetables, other or unspecifiedChen et al. 2025 | Chongqing retail/market samples2017–2021Basis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedChen et al. 2025 | Chongqing retail/market samples2017–2021Basis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedChen et al. 2025 | Chongqing retail/market samples2017–2021Basis not stated | mean. | ||
| Coriander leavesGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | mean.Same samples: 25–88.3 µg/kg (range); 25.5–90.9 µg/kg (range); 10–98.6 µg/kg (range); 14.3–91.5 µg/kg (range); 10.8–89.8 µg/kg (range); 10–83.9 µg/kg (range); 29.8–63.7 µg/kg (range); 10–69.3 µg/kg (range); 24.4–83.1 µg/kg (range); 33–92.9 µg/kg (range) | ||
| Coriander leavesGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | range. | ||
| Fenugreek leavesGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | mean. | ||
| Fenugreek leavesGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | range. | ||
| Garden cressGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | mean. | ||
| Garden cressGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | range. | ||
| Leafy vegetables, other or unspecifiedGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | mean. | ||
| LeekGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | mean. | ||
| LeekGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | range. | ||
| MintGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | mean. | ||
| MintGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | range. | ||
| ParsleyGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | mean. | ||
| ParsleyGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | range. | ||
| Radish greens (tissue not stated)Ghasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | mean. | ||
| Radish greens (tissue not stated)Ghasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | range. | ||
| SpinachGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | mean. | ||
| SpinachGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | range. | ||
| Spring onion leavesGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | mean. | ||
| Spring onion leavesGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | range. | ||
| TarragonGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | mean. | ||
| TarragonGhasemidehkordi et al. 2018 | Five agricultural sites2014Dry weight | range. | ||
| Head cabbageGurmu et al. 2025 | Akaki Rivers irrigated farmlandYear not reportedDry weight | mean. | ||
| Chinese cabbageHuang et al. 2014 | Zhejiang province2012Fresh weight | mean. | ||
| Head cabbageHuang et al. 2014 | Zhejiang province2012Fresh weight | mean. | ||
| Pak choiHuang et al. 2014 | Zhejiang province2012Fresh weight | mean. | ||
| Rape greensHuang et al. 2014 | Zhejiang province2012Fresh weight | mean. | ||
| SpinachHuang et al. 2014 | Zhejiang province2012Fresh weight | mean. | ||
| Chinese cabbageJiang et al. 2020 | SZD village near a battery factory2018Basis not stated | mean. | ||
| Pak choiJiang et al. 2020 | SZD village near a battery factory2018Basis not stated | mean. | ||
| SpinachJiang et al. 2020 | SZD village near a battery factory2018Basis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedKao et al. 2026 | Taiwan food samples for2017–2022Fresh weight | mean. | ||
| SpinachKurniawati et al. 2021 | Five traditional markets inCollection date unreportedFresh weight | mean. | ||
| Water spinachKurniawati et al. 2021 | Five traditional markets inCollection date unreportedFresh weight | mean. | ||
| Amaranth leavesLi et al. 2017 | Control: grocery store >55 km from2015Fresh weight | mean. | ||
| Amaranth leavesLi et al. 2017 | Location A12015Fresh weight | mean. | ||
| Amaranth leavesLi et al. 2017 | Location A22015Fresh weight | mean. | ||
| Amaranth leavesLi et al. 2017 | Location B12015Fresh weight | mean. | ||
| Amaranth leavesLi et al. 2017 | Location B12015Fresh weight | mean. | ||
| Amaranth leavesLi et al. 2017 | Location B22015Fresh weight | mean. | ||
| Amaranth leavesLi et al. 2017 | Location B22015Fresh weight | mean. | ||
| Amaranth leavesLi et al. 2017 | Location B32015Fresh weight | mean. | ||
| Amaranth leavesLi et al. 2017 | Location B32015Fresh weight | mean. | ||
| Amaranth leavesLi et al. 2017 | Location B42015Fresh weight | mean. | ||
| Amaranth leavesLi et al. 2017 | Location B42015Fresh weight | mean. | ||
| LettuceLi et al. 2017 | Control: grocery store >55 km from2015Fresh weight | mean. | ||
| LettuceLi et al. 2017 | Location A12015Fresh weight | mean. | ||
| LettuceLi et al. 2017 | Location A22015Fresh weight | mean. | ||
| LettuceLi et al. 2017 | Location B12015Fresh weight | mean. | ||
| LettuceLi et al. 2017 | Location B12015Fresh weight | mean. | ||
| LettuceLi et al. 2017 | Location B22015Fresh weight | mean. | ||
| LettuceLi et al. 2017 | Location B22015Fresh weight | mean. | ||
| LettuceLi et al. 2017 | Location B32015Fresh weight | mean. | ||
| LettuceLi et al. 2017 | Location B32015Fresh weight | mean. | ||
| LettuceLi et al. 2017 | Location B42015Fresh weight | mean. | ||
| LettuceLi et al. 2017 | Location B42015Fresh weight | mean. | ||
| Water spinachLi et al. 2017 | Control: grocery store >55 km from2015Fresh weight | mean. | ||
| Water spinachLi et al. 2017 | Location A12015Fresh weight | mean. | ||
| Water spinachLi et al. 2017 | Location A22015Fresh weight | mean. | ||
| Water spinachLi et al. 2017 | Location B12015Fresh weight | mean. | ||
| Water spinachLi et al. 2017 | Location B22015Fresh weight | mean. | ||
| Water spinachLi et al. 2017 | Location B32015Fresh weight | mean. | ||
| Water spinachLi et al. 2017 | Location B42015Fresh weight | mean. | ||
| Head cabbageLiang et al. 2019 | BeijingCollection date unreportedDry weight | mean. | ||
| Mustard greensLiu et al. 2023 | MustardCollection date unreportedBasis not stated | mean. | ||
| Head lettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Lamb's lettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| LettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| RocketMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Romaine lettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| SpinachMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Swiss chardMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| WatercressMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Fenugreek leavesMawari et al. 2022 | Farms in ChandrahalCollection date unreportedBasis not stated | sample value. | ||
| Head cabbageMawari et al. 2022 | Farms in ChandrahalCollection date unreportedBasis not stated | sample value. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | FuzhouCollection date unreportedBasis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | GanzhouCollection date unreportedBasis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | JiujiangCollection date unreportedBasis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | Leafy vegetablesCollection date unreportedBasis not stated | range. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | Leafy vegetablesCollection date unreportedBasis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | ShangraoCollection date unreportedBasis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | YichunCollection date unreportedBasis not stated | mean. | ||
| Head cabbageOkonofua et al. 2024 | Dry Season2019Basis not stated | below LOD. | ||
| Head cabbageOkonofua et al. 2024 | Dry season Control2019Basis not stated | mean. | ||
| Head cabbageOkonofua et al. 2024 | Rainy Season2019Basis not stated | not detected. | ||
| Head cabbageOkonofua et al. 2024 | Rainy season Control2019Basis not stated | mean. | ||
| Water spinachPan et al. 2016 | Zhejiang province marketed2014Fresh weight | mean. | ||
| DillSaleem et al. 2025 | Town Square Farmer's MarketYear not reportedDry weight | mean. | ||
| KaleSaleem et al. 2025 | Town Square Farmer's MarketYear not reportedDry weight | mean. | ||
| SpinachSaleem et al. 2025 | Town Square Farmer's MarketYear not reportedDry weight | mean. | ||
| Bristly ox tongueStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Bristly ox tongueStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Cat's earStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Dandelion greensStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| English plantainStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| English plantainStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| LettuceStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| MallowStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Nasturtium (Tropaeolum)Stark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Sweet fennelStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Sweet fennelStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| LettuceUK TDS 2014 (Fera 2015) | UK TDS 2014 category 11_652014Fresh weight | below LOD. | ||
| HouttuyniaWang et al. 2021 | Danzhai2019Dry weight | mean.Same samples: 138–541 µg/kg (range); 169–659 µg/kg (range); 21–83 µg/kg (range) | ||
| HouttuyniaWang et al. 2021 | Danzhai2019Dry weight | mean. | ||
| HouttuyniaWang et al. 2021 | Danzhai2019Dry weight | range. | ||
| HouttuyniaWang et al. 2021 | Danzhai2019Dry weight | range. | ||
| HouttuyniaWang et al. 2021 | Zhijin2019Dry weight | mean.Same samples: 25–52 µg/kg (range); 25–61 µg/kg (range); 4–8 µg/kg (range) | ||
| HouttuyniaWang et al. 2021 | Zhijin2019Dry weight | mean. | ||
| HouttuyniaWang et al. 2021 | Zhijin2019Dry weight | range. | ||
| HouttuyniaWang et al. 2021 | Zhijin2019Dry weight | range. | ||
| HouttuyniaWang et al. 2021 | Danzhai2019Fresh weight | mean. | ||
| HouttuyniaWang et al. 2021 | Danzhai2019Fresh weight | range. | ||
| HouttuyniaWang et al. 2021 | Zhijin2019Fresh weight | mean. | ||
| HouttuyniaWang et al. 2021 | Zhijin2019Fresh weight | range. | ||
| Chinese cabbageWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| ChivesWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| Head cabbageWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| LettuceWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| Pak choiWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| Romaine lettuceWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| SpinachWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Cattle manure2017Basis not stated | not detected. | ||
| SpinachYaman Türkkan and Kibar 2022 | Chemical fertilizer (NPK)2017Basis not stated | not detected. | ||
| SpinachYaman Türkkan and Kibar 2022 | Chicken manure2017Basis not stated | not detected. | ||
| SpinachYaman Türkkan and Kibar 2022 | Control (no fertilizer)2017Basis not stated | not detected. | ||
| SpinachYaman Türkkan and Kibar 2022 | Sheep manure2017Basis not stated | not detected. | ||
| SpinachYaman Türkkan and Kibar 2022 | Turkey manure2017Basis not stated | not detected. | ||
| SpinachYaman Türkkan and Kibar 2022 | Vermicompost2017Basis not stated | not detected. | ||
11 findings · 4 studies | ||||
| Head cabbage (cooked)WHO GEMS/Food | WHO region AFRO; random sampling2018Fresh weight | median. | ||
| Head cabbage (cooked)WHO GEMS/Food | WHO region WPRO; targeted sampling2008Fresh weight | median. | ||
| Leafy vegetables, other or unspecified (cooked)WHO GEMS/Food | WHO region AFRO; random sampling2018Fresh weight | median. | ||
| Leafy vegetables, other or unspecified (cooked)WHO GEMS/Food | WHO region WPRO; targeted sampling1998Fresh weight | median. | ||
| Pak choi (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2014Fresh weight | median. | ||
| Spinach (cooked)WHO GEMS/Food | WHO region EURO; random sampling2007–2009Fresh weight | median. | ||
| Collard greens (cooked)FDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Kale (cooked)FDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Leafy vegetables, other or unspecified (cooked)Mahaffey et al. 1975 | US FDA Total Diet Study FY 19731968–1974Basis not stated | not detected. | ||
| Head cabbage (cooked)UK TDS 2014 (Fera 2015) | UK TDS 2014 category 11_622014Fresh weight | below LOD. | ||
| Other green vegetables (mixed category) (cooked)UK TDS 2014 (Fera 2015) | UK TDS 2014 category 11_682014Fresh weight | below LOD. | ||
1 finding · 1 study | ||||
| Head cabbage (fermented)WHO GEMS/Food | WHO region EURO; random sampling2004–2020Fresh weight | median. | ||
6 findings · 1 study · 6 spread figures folded in | ||||
| HouttuyniaWang et al. 2021 | Danzhai2019Dry weight | mean.Same samples: 0.3–1.39 µg/kg (range); 0.2–1.5 µg/kg (range); 0.05–0.21 µg/kg (range) | ||
| HouttuyniaWang et al. 2021 | Danzhai2019Dry weight | mean. | ||
| HouttuyniaWang et al. 2021 | Danzhai2019Dry weight | range. | ||
| HouttuyniaWang et al. 2021 | Danzhai2019Dry weight | range. | ||
| HouttuyniaWang et al. 2021 | Zhijin2019Dry weight | mean.Same samples: 0.16–0.53 µg/kg (range); 0.1–0.6 µg/kg (range); 0.02–0.08 µg/kg (range) | ||
| HouttuyniaWang et al. 2021 | Zhijin2019Dry weight | mean. | ||
| HouttuyniaWang et al. 2021 | Zhijin2019Dry weight | range. | ||
| HouttuyniaWang et al. 2021 | Zhijin2019Dry weight | range. | ||
| HouttuyniaWang et al. 2021 | Danzhai2019Fresh weight | mean. | ||
| HouttuyniaWang et al. 2021 | Danzhai2019Fresh weight | range. | ||
| HouttuyniaWang et al. 2021 | Zhijin2019Fresh weight | mean. | ||
| HouttuyniaWang et al. 2021 | Zhijin2019Fresh weight | range. | ||
234 findings · 40 studies · 17 spread figures folded in | ||||
| Head cabbageWHO GEMS/Food | WHO region EURO; targeted sampling2000Fresh weight | median. | ||
| SpinachWHO GEMS/Food | WHO region EURO; targeted sampling2000Fresh weight | median. | ||
| Head cabbageFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Head lettuceFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Leaf lettuceFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| SpinachFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| SpinachAbubakar et al. 2025 | Site ICollection date unreportedBasis not stated | mean. | ||
| Amaranth leavesAdebayo et al. 2020 | Railway Quarters irrigation farmCollection date unreportedDry weight | mean. | ||
| Coriander leavesAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| Coriander leavesAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| Coriander leavesAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| Coriander leavesAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | mean. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | mean. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | mean. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | mean. | ||
| LettuceAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| LettuceAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| LettuceAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| LettuceAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| LeekAli 2024 | FarmerYear not reportedDry weight | mean (figure-read). | ||
| LeekAli 2024 | FarmerYear not reportedDry weight | mean. | ||
| LeekAli 2024 | MarketYear not reportedDry weight | mean. | ||
| LeekAli 2024 | MarketYear not reportedDry weight | mean (figure-read). | ||
| LeekAli 2024 | RestaurantYear not reportedDry weight | mean (figure-read). | ||
| LeekAli 2024 | RestaurantYear not reportedDry weight | mean (figure-read). | ||
| Head cabbageAlimohammadi et al. 2018 | Autumn 20162016Fresh weight | mean. | ||
| Head cabbageAlimohammadi et al. 2018 | Summer 20162016Fresh weight | mean. | ||
| LettuceAlimohammadi et al. 2018 | Autumn 20162016Fresh weight | mean. | ||
| LettuceAlimohammadi et al. 2018 | Summer 20162016Fresh weight | mean. | ||
| Head cabbageArmand et al. 2026 | 10 retail vegetable storesCollection date unreportedFresh weight | mean. | ||
| Head cabbageArmand et al. 2026 | 10 retail vegetable storesCollection date unreportedFresh weight | range. | ||
| LettuceArmand et al. 2026 | 10 retail vegetable storesCollection date unreportedFresh weight | mean. | ||
| LettuceArmand et al. 2026 | 10 retail vegetable storesCollection date unreportedFresh weight | range. | ||
| LettuceBambara et al. 2023 | Loumbila market gardenCollection date unreportedDry weight | mean. | ||
| Spring onion leavesBambara et al. 2023 | Loumbila market gardenCollection date unreportedDry weight | mean. | ||
| LettuceDearing et al. 2025 | Hawke's Bay market gardens2024Fresh weight | median.Same samples: 610 µg/kg (maximum); 370 µg/kg (maximum) | ||
| LettuceDearing et al. 2025 | Hawke's Bay market gardens2024Fresh weight | max. | ||
| ParsleyDearing et al. 2025 | Hawke's Bay market gardens2024Fresh weight | median. | ||
| ParsleyDearing et al. 2025 | Hawke's Bay market gardens2024Fresh weight | max. | ||
| Water spinachDearing et al. 2025 | Hawke's Bay market gardens2024Fresh weight | sample value. | ||
| SpinachElbagermi et al. 2012 | Misurata City local suppliers and2010Dry weight | mean. | ||
| LettuceEmmanuel 2025 | Sharada2025Basis not stated | mean. | ||
| LettuceEmmanuel 2025 | Wudil2025Basis not stated | mean. | ||
| SpinachEmmanuel 2025 | Nomans-Land2025Basis not stated | mean. | ||
| Amaranth leavesFatai et al. 2024 | Vegetable farms2022Basis not stated | mean. | ||
| Head cabbageFonge et al. 2021 | SantaCollection date unreportedDry weight | mean. | ||
| Leafy vegetables, other or unspecifiedHan et al. 2024 | Zhejiang Province2018–2019Basis not stated | 95th percentile. | ||
| Leafy vegetables, other or unspecifiedHan et al. 2024 | Zhejiang Province2018–2019Basis not stated | mean.Same samples: 231 µg/kg (95th percentile); 7–990 µg/kg (range) | ||
| Leafy vegetables, other or unspecifiedHan et al. 2024 | Zhejiang Province2018–2019Basis not stated | median. | ||
| Leafy vegetables, other or unspecifiedHan et al. 2024 | Zhejiang Province2018–2019Basis not stated | range. | ||
| Head cabbageImongben et al. 2026 | Five Kaduna Metropolis marketsCollection date unreportedBasis not stated | mean. | ||
| LettuceImongben et al. 2026 | Five Kaduna Metropolis marketsCollection date unreportedBasis not stated | mean. | ||
| SpinachImongben et al. 2026 | Five Kaduna Metropolis marketsCollection date unreportedBasis not stated | mean. | ||
| Chinese cabbageJiang et al. 2020 | SZD village near a battery factory2018Basis not stated | mean. | ||
| Pak choiJiang et al. 2020 | SZD village near a battery factory2018Basis not stated | mean. | ||
| SpinachJiang et al. 2020 | SZD village near a battery factory2018Basis not stated | mean. | ||
| Kenaf leavesJoseph et al. 2025 | Gobbiya Dam irrigated farmlandCollection date unreportedDry weight | mean. | ||
| SpinachJoseph et al. 2025 | Gobbiya Dam irrigated farmlandCollection date unreportedDry weight | mean. | ||
| Swiss chardJurkovic et al. 2025 | 1 leaves (flooded site 1)2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | 4 leaves (flooded site 4)2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | 6 leaves (flooded site 6)2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | 8 leaves (flooded site 8)2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | all chard samples2024Basis not stated | mean. | ||
| Swiss chardJurkovic et al. 2025 | Control 2 leaves2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | Control 4 leaves (control site 4)2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | Control 7 leaves (control site 7)2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | Control 8 leaves (control site 8)2024Basis not stated | sample value. | ||
| Alternanthera sessilis leavesKananke et al. 2026 | BandaragamaCollection date unreportedBasis not stated | mean. | ||
| Alternanthera sessilis leavesKananke et al. 2026 | KolonnawaCollection date unreportedBasis not stated | mean. | ||
| Alternanthera sessilis leavesKananke et al. 2026 | KottawaCollection date unreportedBasis not stated | mean. | ||
| Alternanthera sessilis leavesKananke et al. 2026 | PiliyandalaCollection date unreportedBasis not stated | mean. | ||
| Alternanthera sessilis leavesKananke et al. 2026 | WellampitiyaCollection date unreportedBasis not stated | mean. | ||
| Centella asiatica leavesKananke et al. 2026 | BandaragamaCollection date unreportedBasis not stated | mean. | ||
| Centella asiatica leavesKananke et al. 2026 | KolonnawaCollection date unreportedBasis not stated | mean. | ||
| Centella asiatica leavesKananke et al. 2026 | KottawaCollection date unreportedBasis not stated | mean. | ||
| Centella asiatica leavesKananke et al. 2026 | PiliyandalaCollection date unreportedBasis not stated | mean. | ||
| Centella asiatica leavesKananke et al. 2026 | WellampitiyaCollection date unreportedBasis not stated | mean. | ||
| KaleKim et al. 2024 | All six farms, field samples2019Fresh weight | mean.Same samples: 50–151 µg/kg (range) | ||
| KaleKim et al. 2024 | All six farms, field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | All six farms, field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm R1 (Rural), field samples2019Fresh weight | mean.Same samples: 50–50 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm R1 (Rural), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm R1 (Rural), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm R2 (Rural), field samples2019Fresh weight | mean.Same samples: 123–147 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm R2 (Rural), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm R2 (Rural), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm R3 (Rural), field samples2019Fresh weight | mean.Same samples: 50–80.6 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm R3 (Rural), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm R3 (Rural), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm R4 (Rural), field samples2019Fresh weight | mean.Same samples: 74.5–77.6 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm R4 (Rural), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm R4 (Rural), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm U1 (Urban), field samples2019Fresh weight | mean.Same samples: 50–104 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm U1 (Urban), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm U1 (Urban), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm U2 (Urban), field samples2019Fresh weight | mean.Same samples: 99.3–151 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm U2 (Urban), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm U2 (Urban), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Rural farms R1-R42019Fresh weight | mean. | ||
| KaleKim et al. 2024 | Urban farms U1-U22019Fresh weight | mean. | ||
| Amaranth leavesKulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| Amaranth leavesKulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| Bathua (lamb's quarters)Kulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| Bathua (lamb's quarters)Kulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| Fenugreek leavesKulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| Fenugreek leavesKulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| Mustard greensKulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| Mustard greensKulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| Mustard greensKulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| Mustard greensKulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| Radish leavesKulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| Radish leavesKulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| SpinachKulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| SpinachKulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| Head lettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Lamb's lettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| LettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| RocketMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Romaine lettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| SpinachMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Swiss chardMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| WatercressMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Amaranth leavesNaser et al. 2018 | High-level pollution: KalakoirCollection date unreportedDry weight | mean. | ||
| Amaranth leavesNaser et al. 2018 | High-level pollution: KalakoirCollection date unreportedDry weight | mean. | ||
| Amaranth leavesNaser et al. 2018 | Low-level pollution: BARICollection date unreportedDry weight | mean. | ||
| Amaranth leavesNaser et al. 2018 | Low-level pollution: BARICollection date unreportedDry weight | mean. | ||
| Amaranth leavesNaser et al. 2018 | Medium-level pollution: ZorunCollection date unreportedDry weight | mean. | ||
| Amaranth leavesNaser et al. 2018 | Medium-level pollution: ZorunCollection date unreportedDry weight | mean. | ||
| SpinachNaser et al. 2018 | High-level pollution: KalakoirCollection date unreportedDry weight | mean. | ||
| SpinachNaser et al. 2018 | Low-level pollution: BARICollection date unreportedDry weight | mean. | ||
| SpinachNaser et al. 2018 | Medium-level pollution: ZorunCollection date unreportedDry weight | mean. | ||
| Head cabbageOkonofua et al. 2024 | Dry Season2019Basis not stated | mean. | ||
| Head cabbageOkonofua et al. 2024 | Dry season Control2019Basis not stated | mean. | ||
| Head cabbageOkonofua et al. 2024 | Rainy Season2019Basis not stated | mean. | ||
| Head cabbageOkonofua et al. 2024 | Rainy season Control2019Basis not stated | mean. | ||
| Bitter leafOsuolale et al. 2025 | Ondo Metropolis market/farm2018Dry weight | max. | ||
| Leafy vegetables, other or unspecifiedOsuolale et al. 2025 | Ondo Metropolis market/farm2018Dry weight | mean. | ||
| Leafy vegetables, other or unspecifiedOsuolale et al. 2025 | Ondo Metropolis market/farm2018Dry weight | median. | ||
| Leafy vegetables, other or unspecifiedOsuolale et al. 2025 | Ondo Metropolis market/farm2018Dry weight | not detected. | ||
| Leafy vegetables, other or unspecifiedOsuolale et al. 2025 | Ondo Metropolis market/farm2018Dry weight | min. | ||
| Leafy vegetables, other or unspecifiedOsuolale et al. 2025 | Ondo Metropolis market/farm2018Dry weight | max. | ||
| LettuceOwusu Adoma et al. 2024 | AH (Ahodwo)Collection date unreportedBasis not stated | below LOD. | ||
| LettuceOwusu Adoma et al. 2024 | BSECollection date unreportedBasis not stated | mean. | ||
| LettuceOwusu Adoma et al. 2024 | BSGSCollection date unreportedBasis not stated | mean. | ||
| LettuceOwusu Adoma et al. 2024 | KT (Kentinkrono)Collection date unreportedBasis not stated | below LOD. | ||
| LettuceOwusu Adoma et al. 2024 | mean of four sitesCollection date unreportedBasis not stated | mean. | ||
| Head cabbagePal et al. 2017 | Domestic-wastewater-irrigatedCollection date unreportedBasis not stated | mean.Same samples: 10–3,800 µg/kg (range) | ||
| SpinachPal et al. 2017 | Domestic-wastewater-irrigatedCollection date unreportedBasis not stated | mean. | ||
| SpinachPal et al. 2017 | Domestic-wastewater-irrigatedCollection date unreportedBasis not stated | range. | ||
| Head cabbageRempelos et al. 2023 | Crop protection CON (CP)2004–2009Fresh weight | mean. | ||
| Head cabbageRempelos et al. 2023 | Crop protection ORG (OP)2004–2009Fresh weight | mean. | ||
| Head cabbageRempelos et al. 2023 | Fertilizer type CON (NPK)2004–2009Fresh weight | mean. | ||
| Head cabbageRempelos et al. 2023 | Fertilizer type ORG (FYM)2004–2009Fresh weight | mean. | ||
| Head cabbageRempelos et al. 2023 | Pre-crop Spring Beans2004–2009Fresh weight | mean. | ||
| Head cabbageRempelos et al. 2023 | Pre-crop Winter Barley2004–2009Fresh weight | mean. | ||
| Head lettuceRempelos et al. 2023 | Crop protection CON (CP)2004–2009Fresh weight | mean. | ||
| Head lettuceRempelos et al. 2023 | Crop protection ORG (OP)2004–2009Fresh weight | mean. | ||
| Head lettuceRempelos et al. 2023 | Fertilizer type CON (NPK)2004–2009Fresh weight | mean. | ||
| Head lettuceRempelos et al. 2023 | Fertilizer type ORG (FYM)2004–2009Fresh weight | mean. | ||
| Head lettuceRempelos et al. 2023 | Pre-crop Spring Beans2004–2009Fresh weight | mean. | ||
| Head lettuceRempelos et al. 2023 | Pre-crop Winter Barley2004–2009Fresh weight | mean. | ||
| ChivesRomero-Crespo et al. 2023 | Ponce Enriquez gold-mining area2021Fresh weight | sample value. | ||
| Coriander leavesRomero-Crespo et al. 2023 | Ponce Enriquez gold-mining area2021Fresh weight | sample value. | ||
| LettuceRomero-Crespo et al. 2023 | Ponce Enriquez gold-mining area2021Fresh weight | sample value. | ||
| LettuceRossini-Oliva et al. 2024 | City urban gardens2021–2023Dry weight | mean. | ||
| LettuceRossini-Oliva et al. 2024 | Peri-urban garden2021–2023Dry weight | mean. | ||
| LettuceRossini-Oliva et al. 2024 | Seville local market (Mk)2021–2023Dry weight | mean. | ||
| Swiss chardRossini-Oliva et al. 2024 | City urban gardens2021–2023Dry weight | mean. | ||
| Swiss chardRossini-Oliva et al. 2024 | Mining-area urban gardens2021–2023Dry weight | mean. | ||
| Swiss chardRossini-Oliva et al. 2024 | Peri-urban garden2021–2023Dry weight | mean. | ||
| Swiss chardRossini-Oliva et al. 2024 | Seville local market (Mk)2021–2023Dry weight | mean. | ||
| DillSaleem et al. 2025 | Town Square Farmer's MarketYear not reportedDry weight | mean. | ||
| KaleSaleem et al. 2025 | Town Square Farmer's MarketYear not reportedDry weight | mean. | ||
| SpinachSaleem et al. 2025 | Town Square Farmer's MarketYear not reportedDry weight | mean. | ||
| BasilSalehipour et al. 2015 | Wholesaler outlets andCollection date unreportedFresh weight | 95% UCL of mean. | ||
| LettuceSalehipour et al. 2015 | Wholesaler outlets andCollection date unreportedFresh weight | 95% UCL of mean. | ||
| MintSalehipour et al. 2015 | Wholesaler outlets andCollection date unreportedFresh weight | 95% UCL of mean. | ||
| BasilSouri et al. 2019 | Shahre Rey wastewater-irrigated2011Dry weight | mean. | ||
| BasilSouri et al. 2019 | Shahre Rey wastewater-irrigated2011Dry weight | mean. | ||
| Garden cressSouri et al. 2019 | Shahre Rey wastewater-irrigated2011Dry weight | mean. | ||
| Garden cressSouri et al. 2019 | Shahre Rey wastewater-irrigated2011Dry weight | mean. | ||
| Bristly ox tongueStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Bristly ox tongueStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Cat's earStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Dandelion greensStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| English plantainStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| English plantainStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| LettuceStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| MallowStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Nasturtium (Tropaeolum)Stark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Sweet fennelStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Sweet fennelStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| SpinachSuleiman and Ibrahim 2025 | Three irrigation farmsCollection date unreportedBasis not stated | mean. | ||
| Coriander leavesSultana et al. 2022 | 1st phase (early harvest)Year not reportedDry weight | sample value. | ||
| Coriander leavesSultana et al. 2022 | 2nd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| Coriander leavesSultana et al. 2022 | 3rd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| Coriander leavesSultana et al. 2022 | 4th phase (late harvest)Year not reportedDry weight | sample value. | ||
| Coriander leavesSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | mean.Same samples: 1,700–4,500 µg/kg (range); up to 2,700 µg/kg (range maximum); up to 3,700 µg/kg (range maximum); 1,500–6,500 µg/kg (range); up to 1,700 µg/kg (range maximum) | ||
| Coriander leavesSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | range. | ||
| Head cabbageSultana et al. 2022 | 1st phase (early harvest)Year not reportedDry weight | not detected. | ||
| Head cabbageSultana et al. 2022 | 2nd phase (mid-time harvest)Year not reportedDry weight | not detected. | ||
| Head cabbageSultana et al. 2022 | 3rd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| Head cabbageSultana et al. 2022 | 4th phase (late harvest)Year not reportedDry weight | not detected. | ||
| Head cabbageSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | mean. | ||
| Head cabbageSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | range. | ||
| MintSultana et al. 2022 | 1st phase (early harvest)Year not reportedDry weight | sample value. | ||
| MintSultana et al. 2022 | 2nd phase (mid-time harvest)Year not reportedDry weight | not detected. | ||
| MintSultana et al. 2022 | 3rd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| MintSultana et al. 2022 | 4th phase (late harvest)Year not reportedDry weight | sample value. | ||
| MintSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | mean. | ||
| MintSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | range. | ||
| Mustard greensSultana et al. 2022 | 1st phase (early harvest)Year not reportedDry weight | sample value. | ||
| Mustard greensSultana et al. 2022 | 2nd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| Mustard greensSultana et al. 2022 | 3rd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| Mustard greensSultana et al. 2022 | 4th phase (late harvest)Year not reportedDry weight | sample value. | ||
| Mustard greensSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | mean. | ||
| Mustard greensSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | range. | ||
| SpinachSultana et al. 2022 | 1st phase (early harvest)Year not reportedDry weight | sample value. | ||
| SpinachSultana et al. 2022 | 2nd phase (mid-time harvest)Year not reportedDry weight | not detected. | ||
| SpinachSultana et al. 2022 | 3rd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| SpinachSultana et al. 2022 | 4th phase (late harvest)Year not reportedDry weight | sample value. | ||
| SpinachSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | mean. | ||
| SpinachSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | range. | ||
| LettuceUK TDS 2014 (Fera 2015) | UK TDS 2014 category 11_652014Fresh weight | sample value. | ||
| Chinese cabbageWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| ChivesWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| Head cabbageWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| LettuceWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| Pak choiWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| Romaine lettuceWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| SpinachWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Cattle manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Chemical fertilizer (NPK)2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Chicken manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Control (no fertilizer)2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Sheep manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Turkey manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Vermicompost2017Basis not stated | mean. | ||
16 findings · 3 studies | ||||
| Collard greens (cooked)FDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Kale (cooked)FDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Amaranth leaves (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Chinese cabbage (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Chinese kale (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Choy sum (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Head cabbage (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Lettuce (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Lettuce (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Mustard greens (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Pak choi (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Spinach (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Water spinach (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Watercress (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Head cabbage (cooked)UK TDS 2014 (Fera 2015) | UK TDS 2014 category 11_622014Fresh weight | sample value. | ||
| Other green vegetables (mixed category) (cooked)UK TDS 2014 (Fera 2015) | UK TDS 2014 category 11_682014Fresh weight | sample value. | ||
257 findings · 52 studies · 17 spread figures folded in | ||||
| Head cabbageWHO GEMS/Food | WHO region EURO; targeted sampling2000Fresh weight | median. | ||
| SpinachWHO GEMS/Food | WHO region EURO; targeted sampling2000Fresh weight | median. | ||
| Head cabbageFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Head lettuceFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Leaf lettuceFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| SpinachFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| SpinachAbubakar et al. 2025 | Site ICollection date unreportedBasis not stated | mean. | ||
| Amaranth leavesAdebayo et al. 2020 | Railway Quarters irrigation farmCollection date unreportedDry weight | mean. | ||
| Coriander leavesAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| Coriander leavesAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| Coriander leavesAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| Coriander leavesAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | mean. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | mean. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | mean. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | mean. | ||
| LettuceAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| LettuceAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| LettuceAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| LettuceAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| Radish leavesAhmed et al. 2019 | Dry season2015–2017Dry weight | mean. | ||
| Head cabbageAlimohammadi et al. 2018 | Autumn 20162016Fresh weight | mean. | ||
| Head cabbageAlimohammadi et al. 2018 | Summer 20162016Fresh weight | mean. | ||
| LettuceAlimohammadi et al. 2018 | Autumn 20162016Fresh weight | mean. | ||
| LettuceAlimohammadi et al. 2018 | Summer 20162016Fresh weight | mean. | ||
| Head cabbageAmetepey et al. 2018 | Central and Aboabo MarketsCollection date unreportedDry weight | mean.Same samples: 30–470 µg/kg (range) | ||
| Head cabbageAmetepey et al. 2018 | Central and Aboabo MarketsCollection date unreportedDry weight | range. | ||
| Head cabbageArmand et al. 2026 | 10 retail vegetable storesCollection date unreportedFresh weight | mean. | ||
| Head cabbageArmand et al. 2026 | 10 retail vegetable storesCollection date unreportedFresh weight | range. | ||
| LettuceArmand et al. 2026 | 10 retail vegetable storesCollection date unreportedFresh weight | mean. | ||
| LettuceArmand et al. 2026 | 10 retail vegetable storesCollection date unreportedFresh weight | range. | ||
| LettuceBambara et al. 2023 | Loumbila market gardenCollection date unreportedDry weight | mean. | ||
| Spring onion leavesBambara et al. 2023 | Loumbila market gardenCollection date unreportedDry weight | mean. | ||
| Leafy vegetables, other or unspecifiedCherfi et al. 2014 | Two wholesale markets supplying2012Dry weight | mean. | ||
| SpinachCherfi et al. 2014 | Two wholesale markets supplying2012Dry weight | mean. | ||
| Amaranth leavesChiutula et al. 2025 | Soche WWTP effluent-irrigatedYear not reportedDry weight | mean. | ||
| Chinese cabbageChiutula et al. 2025 | Soche WWTP effluent-irrigatedYear not reportedDry weight | mean. | ||
| Mustard greensChiutula et al. 2025 | Soche WWTP effluent-irrigatedYear not reportedDry weight | mean. | ||
| Pumpkin leavesChiutula et al. 2025 | Soche WWTP effluent-irrigatedYear not reportedDry weight | mean. | ||
| Rape greensChiutula et al. 2025 | Soche WWTP effluent-irrigatedYear not reportedDry weight | mean. | ||
| Sweet potato leavesChiutula et al. 2025 | Soche WWTP effluent-irrigatedYear not reportedDry weight | mean. | ||
| LettuceEmmanuel 2025 | Sharada2025Basis not stated | mean. | ||
| LettuceEmmanuel 2025 | Wudil2025Basis not stated | mean. | ||
| SpinachEmmanuel 2025 | Nomans-Land2025Basis not stated | mean. | ||
| Amaranth leavesFatai et al. 2024 | Vegetable farms2022Basis not stated | mean. | ||
| Head cabbageFonge et al. 2021 | SantaCollection date unreportedDry weight | mean. | ||
| Head cabbageGurmu et al. 2025 | Akaki Rivers irrigated farmlandYear not reportedDry weight | below LOD. | ||
| Head cabbageHabu et al. 2025 | Tamburawa wastewater-irrigatedCollection date unreportedBasis not stated | mean. | ||
| Head cabbageHabu et al. 2025 | Tamburawa wastewater-irrigatedCollection date unreportedBasis not stated | mean. | ||
| LettuceHabu et al. 2025 | Tamburawa wastewater-irrigatedCollection date unreportedBasis not stated | mean. | ||
| LettuceHabu et al. 2025 | Tamburawa wastewater-irrigatedCollection date unreportedBasis not stated | mean. | ||
| SpinachHabu et al. 2025 | Tamburawa wastewater-irrigatedCollection date unreportedBasis not stated | mean. | ||
| SpinachHabu et al. 2025 | Tamburawa wastewater-irrigatedCollection date unreportedBasis not stated | mean. | ||
| Spring onion leavesHabu et al. 2025 | Tamburawa wastewater-irrigatedCollection date unreportedBasis not stated | mean. | ||
| Amaranth leavesHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | 75th percentile. | ||
| Amaranth leavesHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | median.Same samples: 3,590 µg/kg (75th percentile); 3,590 µg/kg (maximum); 2,140 µg/kg (75th percentile); 7,430 µg/kg (maximum) | ||
| Amaranth leavesHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | not detected. | ||
| Amaranth leavesHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | max. | ||
| Head cabbageHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | 75th percentile. | ||
| Head cabbageHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | median. | ||
| Head cabbageHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | not detected. | ||
| Head cabbageHossain et al. 2016 | Nationwide district samplesCollection date unreportedDry weight | max. | ||
| Malabar spinachHossen et al. 2026 | farm-82024Fresh weight | mean. | ||
| Head cabbageImongben et al. 2026 | Five Kaduna Metropolis marketsCollection date unreportedBasis not stated | mean. | ||
| LettuceImongben et al. 2026 | Five Kaduna Metropolis marketsCollection date unreportedBasis not stated | mean. | ||
| SpinachImongben et al. 2026 | Five Kaduna Metropolis marketsCollection date unreportedBasis not stated | mean. | ||
| Chinese cabbageJiang et al. 2020 | SZD village near a battery factory2018Basis not stated | mean. | ||
| Pak choiJiang et al. 2020 | SZD village near a battery factory2018Basis not stated | mean. | ||
| SpinachJiang et al. 2020 | SZD village near a battery factory2018Basis not stated | mean. | ||
| Kenaf leavesJoseph et al. 2025 | Gobbiya Dam irrigated farmlandCollection date unreportedDry weight | mean. | ||
| SpinachJoseph et al. 2025 | Gobbiya Dam irrigated farmlandCollection date unreportedDry weight | mean. | ||
| Swiss chardJurkovic et al. 2025 | 1 leaves (flooded site 1)2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | 4 leaves (flooded site 4)2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | 6 leaves (flooded site 6)2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | 8 leaves (flooded site 8)2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | all chard samples2024Basis not stated | mean. | ||
| Swiss chardJurkovic et al. 2025 | Control 2 leaves2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | Control 4 leaves (control site 4)2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | Control 7 leaves (control site 7)2024Basis not stated | sample value. | ||
| Swiss chardJurkovic et al. 2025 | Control 8 leaves (control site 8)2024Basis not stated | sample value. | ||
| Alternanthera sessilis leavesKananke et al. 2026 | BandaragamaCollection date unreportedBasis not stated | mean. | ||
| Alternanthera sessilis leavesKananke et al. 2026 | KolonnawaCollection date unreportedBasis not stated | mean. | ||
| Alternanthera sessilis leavesKananke et al. 2026 | KottawaCollection date unreportedBasis not stated | mean. | ||
| Alternanthera sessilis leavesKananke et al. 2026 | PiliyandalaCollection date unreportedBasis not stated | mean. | ||
| Alternanthera sessilis leavesKananke et al. 2026 | WellampitiyaCollection date unreportedBasis not stated | mean. | ||
| Centella asiatica leavesKananke et al. 2026 | BandaragamaCollection date unreportedBasis not stated | mean. | ||
| Centella asiatica leavesKananke et al. 2026 | KolonnawaCollection date unreportedBasis not stated | mean. | ||
| Centella asiatica leavesKananke et al. 2026 | KottawaCollection date unreportedBasis not stated | mean. | ||
| Centella asiatica leavesKananke et al. 2026 | PiliyandalaCollection date unreportedBasis not stated | mean. | ||
| Centella asiatica leavesKananke et al. 2026 | WellampitiyaCollection date unreportedBasis not stated | mean. | ||
| KaleKim et al. 2024 | All six farms, field samples2019Fresh weight | mean.Same samples: 17.4–324 µg/kg (range) | ||
| KaleKim et al. 2024 | All six farms, field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | All six farms, field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm R1 (Rural), field samples2019Fresh weight | mean.Same samples: 17.4–43.2 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm R1 (Rural), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm R1 (Rural), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm R2 (Rural), field samples2019Fresh weight | mean.Same samples: 23.9–34.8 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm R2 (Rural), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm R2 (Rural), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm R3 (Rural), field samples2019Fresh weight | mean.Same samples: 32.8–73.2 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm R3 (Rural), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm R3 (Rural), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm R4 (Rural), field samples2019Fresh weight | mean.Same samples: 56.7–72 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm R4 (Rural), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm R4 (Rural), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm U1 (Urban), field samples2019Fresh weight | mean.Same samples: 69–146 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm U1 (Urban), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm U1 (Urban), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm U2 (Urban), field samples2019Fresh weight | mean.Same samples: 173–324 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm U2 (Urban), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm U2 (Urban), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Rural farms R1-R42019Fresh weight | mean. | ||
| KaleKim et al. 2024 | Urban farms U1-U22019Fresh weight | mean. | ||
| Amaranth leavesKulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| Amaranth leavesKulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| Bathua (lamb's quarters)Kulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| Bathua (lamb's quarters)Kulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| Fenugreek leavesKulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| Fenugreek leavesKulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| Mustard greensKulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| Mustard greensKulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| Mustard greensKulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| Mustard greensKulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| Radish leavesKulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| Radish leavesKulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| SpinachKulshrestha 2022 | Local venders2015–2016Dry weight | mean. | ||
| SpinachKulshrestha 2022 | Walled garden2015–2016Dry weight | mean. | ||
| SpinachKurniawati et al. 2021 | Five traditional markets inCollection date unreportedFresh weight | mean. | ||
| Water spinachKurniawati et al. 2021 | Five traditional markets inCollection date unreportedFresh weight | mean. | ||
| Head cabbageLi et al. 2021 | Four major producing cities of2018Fresh weight | mean. | ||
| Head cabbageLi et al. 2021 | Four major producing cities of2018Fresh weight | range. | ||
| Head cabbageLiang et al. 2019 | BeijingCollection date unreportedDry weight | mean. | ||
| Mustard greensLiu et al. 2023 | MustardCollection date unreportedBasis not stated | mean. | ||
| Head lettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Lamb's lettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| LettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| RocketMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Romaine lettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| SpinachMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Swiss chardMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| WatercressMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Amaranth leavesNaser et al. 2018 | High-level pollution: KalakoirCollection date unreportedDry weight | mean. | ||
| Amaranth leavesNaser et al. 2018 | High-level pollution: KalakoirCollection date unreportedDry weight | mean. | ||
| Amaranth leavesNaser et al. 2018 | Low-level pollution: BARICollection date unreportedDry weight | mean. | ||
| Amaranth leavesNaser et al. 2018 | Low-level pollution: BARICollection date unreportedDry weight | mean. | ||
| Amaranth leavesNaser et al. 2018 | Medium-level pollution: ZorunCollection date unreportedDry weight | mean. | ||
| Amaranth leavesNaser et al. 2018 | Medium-level pollution: ZorunCollection date unreportedDry weight | mean. | ||
| SpinachNaser et al. 2018 | High-level pollution: KalakoirCollection date unreportedDry weight | mean. | ||
| SpinachNaser et al. 2018 | Low-level pollution: BARICollection date unreportedDry weight | mean. | ||
| SpinachNaser et al. 2018 | Medium-level pollution: ZorunCollection date unreportedDry weight | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | FuzhouCollection date unreportedBasis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | GanzhouCollection date unreportedBasis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | JiujiangCollection date unreportedBasis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | Leafy vegetablesCollection date unreportedBasis not stated | range. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | Leafy vegetablesCollection date unreportedBasis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | ShangraoCollection date unreportedBasis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedNie et al. 2023 | YichunCollection date unreportedBasis not stated | mean. | ||
| Amaranth leavesOrosun et al. 2023 | Irrigated farmlands along the2021Dry weight | mean. | ||
| Head cabbageOrosun et al. 2023 | Irrigated farmlands along the2021Dry weight | mean. | ||
| LettuceOwusu Adoma et al. 2024 | AH (Ahodwo)Collection date unreportedBasis not stated | below LOD. | ||
| LettuceOwusu Adoma et al. 2024 | BSECollection date unreportedBasis not stated | below LOD. | ||
| LettuceOwusu Adoma et al. 2024 | BSGSCollection date unreportedBasis not stated | below LOD. | ||
| LettuceOwusu Adoma et al. 2024 | KT (Kentinkrono)Collection date unreportedBasis not stated | below LOD. | ||
| Head cabbagePal et al. 2017 | Domestic-wastewater-irrigatedCollection date unreportedBasis not stated | mean. | ||
| SpinachPal et al. 2017 | Domestic-wastewater-irrigatedCollection date unreportedBasis not stated | mean. | ||
| SpinachPan et al. 2016 | Zhejiang province marketed2014Fresh weight | mean. | ||
| Head cabbageReczajska et al. 2005 | Commercial plantations2001Fresh weight | 90th percentile. | ||
| Head cabbageReczajska et al. 2005 | Commercial plantations2001Fresh weight | mean. | ||
| Head cabbageReczajska et al. 2005 | Commercial plantations2001Fresh weight | median. | ||
| Head cabbageReczajska et al. 2005 | Commercial plantations2001Fresh weight | range. | ||
| ChivesRomero-Crespo et al. 2023 | Ponce Enriquez gold-mining area2021Fresh weight | sample value. | ||
| Coriander leavesRomero-Crespo et al. 2023 | Ponce Enriquez gold-mining area2021Fresh weight | sample value. | ||
| LettuceRomero-Crespo et al. 2023 | Ponce Enriquez gold-mining area2021Fresh weight | sample value. | ||
| LettuceRossini-Oliva et al. 2024 | City urban garden MI2 (Seville)2021–2023Basis not stated | max. | ||
| LettuceRossini-Oliva et al. 2024 | City urban gardens2021–2023Dry weight | mean. | ||
| LettuceRossini-Oliva et al. 2024 | Peri-urban garden2021–2023Dry weight | mean. | ||
| LettuceRossini-Oliva et al. 2024 | Seville local market (Mk)2021–2023Dry weight | mean. | ||
| Swiss chardRossini-Oliva et al. 2024 | City urban gardens2021–2023Dry weight | mean. | ||
| Swiss chardRossini-Oliva et al. 2024 | Mining-area urban gardens2021–2023Dry weight | mean. | ||
| Swiss chardRossini-Oliva et al. 2024 | Peri-urban garden2021–2023Dry weight | mean. | ||
| Swiss chardRossini-Oliva et al. 2024 | Seville local market (Mk)2021–2023Dry weight | mean. | ||
| ArumSadee 2022 | Local markets2021Dry weight | mean. | ||
| Garden cressSadee 2022 | Local markets2021Dry weight | mean. | ||
| MallowSadee 2022 | Local markets2021Dry weight | mean. | ||
| MintSadee 2022 | Local markets2021Dry weight | mean. | ||
| SpinachSadee 2022 | Local markets2021Dry weight | mean. | ||
| Swiss chardSadee 2022 | Local markets2021Dry weight | mean. | ||
| DillSaleem et al. 2025 | Town Square Farmer's MarketYear not reportedDry weight | mean. | ||
| KaleSaleem et al. 2025 | Town Square Farmer's MarketYear not reportedDry weight | mean. | ||
| SpinachSaleem et al. 2025 | Town Square Farmer's MarketYear not reportedDry weight | mean. | ||
| Head cabbageSamma et al. 2024 | Bogura districtCollection date unreportedDry weight | mean. | ||
| Head cabbageSamma et al. 2024 | Bogura districtCollection date unreportedDry weight | below LOD. | ||
| Head cabbageSamma et al. 2024 | Bogura districtCollection date unreportedDry weight | max. | ||
| Head cabbageSee et al. 2025 | Pasar Borong Selangor wholesale2022Fresh weight | not detected. | ||
| Mustard greensSee et al. 2025 | Pasar Borong Selangor wholesale2022Fresh weight | not detected. | ||
| Pak choiSee et al. 2025 | Pasar Borong Selangor wholesale2022Fresh weight | not detected. | ||
| SpinachSee et al. 2025 | Pasar Borong Selangor wholesale2022Fresh weight | not detected. | ||
| BasilSouri et al. 2019 | Shahre Rey wastewater-irrigated2011Dry weight | mean. | ||
| BasilSouri et al. 2019 | Shahre Rey wastewater-irrigated2011Dry weight | mean. | ||
| Garden cressSouri et al. 2019 | Shahre Rey wastewater-irrigated2011Dry weight | mean. | ||
| Garden cressSouri et al. 2019 | Shahre Rey wastewater-irrigated2011Dry weight | mean. | ||
| Bristly ox tongueStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Bristly ox tongueStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Cat's earStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Dandelion greensStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| English plantainStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| English plantainStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| LettuceStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| MallowStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Nasturtium (Tropaeolum)Stark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Sweet fennelStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| Sweet fennelStark et al. 2019 | Oakland2014–2016Dry weight | below LOD. | ||
| SpinachSuleiman and Ibrahim 2025 | Three irrigation farmsCollection date unreportedBasis not stated | mean. | ||
| Coriander leavesSultana et al. 2022 | 1st phase (early harvest)Year not reportedDry weight | sample value. | ||
| Coriander leavesSultana et al. 2022 | 2nd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| Coriander leavesSultana et al. 2022 | 3rd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| Coriander leavesSultana et al. 2022 | 4th phase (late harvest)Year not reportedDry weight | sample value. | ||
| Coriander leavesSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | mean.Same samples: 1,500–6,500 µg/kg (range); up to 2,400 µg/kg (range maximum); 1,400–16,700 µg/kg (range); 2,700–7,700 µg/kg (range); 1,400–3,300 µg/kg (range) | ||
| Coriander leavesSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | range. | ||
| Head cabbageSultana et al. 2022 | 1st phase (early harvest)Year not reportedDry weight | sample value. | ||
| Head cabbageSultana et al. 2022 | 2nd phase (mid-time harvest)Year not reportedDry weight | not detected. | ||
| Head cabbageSultana et al. 2022 | 3rd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| Head cabbageSultana et al. 2022 | 4th phase (late harvest)Year not reportedDry weight | sample value. | ||
| Head cabbageSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | mean. | ||
| Head cabbageSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | range. | ||
| MintSultana et al. 2022 | 1st phase (early harvest)Year not reportedDry weight | sample value. | ||
| MintSultana et al. 2022 | 2nd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| MintSultana et al. 2022 | 3rd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| MintSultana et al. 2022 | 4th phase (late harvest)Year not reportedDry weight | sample value. | ||
| MintSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | mean. | ||
| MintSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | range. | ||
| Mustard greensSultana et al. 2022 | 1st phase (early harvest)Year not reportedDry weight | sample value. | ||
| Mustard greensSultana et al. 2022 | 2nd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| Mustard greensSultana et al. 2022 | 3rd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| Mustard greensSultana et al. 2022 | 4th phase (late harvest)Year not reportedDry weight | sample value. | ||
| Mustard greensSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | mean. | ||
| Mustard greensSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | range. | ||
| SpinachSultana et al. 2022 | 1st phase (early harvest)Year not reportedDry weight | sample value. | ||
| SpinachSultana et al. 2022 | 2nd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| SpinachSultana et al. 2022 | 3rd phase (mid-time harvest)Year not reportedDry weight | sample value. | ||
| SpinachSultana et al. 2022 | 4th phase (late harvest)Year not reportedDry weight | sample value. | ||
| SpinachSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | mean. | ||
| SpinachSultana et al. 2022 | All four harvesting phasesYear not reportedDry weight | range. | ||
| LettuceUK TDS 2014 (Fera 2015) | UK TDS 2014 category 11_652014Fresh weight | sample value. | ||
| Alternanthera sessilis leavesWijeyaratne et al. 2021 | Non-organic cultivation sites2019Dry weight | mean. | ||
| Alternanthera sessilis leavesWijeyaratne et al. 2021 | Non-organic cultivation sites2019Dry weight | mean. | ||
| Alternanthera sessilis leavesWijeyaratne et al. 2021 | Organic cultivation sites2019Dry weight | mean. | ||
| Alternanthera sessilis leavesWijeyaratne et al. 2021 | Organic cultivation sites2019Dry weight | mean. | ||
| Chinese cabbageWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| ChivesWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| Head cabbageWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| LettuceWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| Pak choiWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| Romaine lettuceWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| SpinachWu 2024 | Jiaozuo marketsCollection date unreportedFresh weight | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Cattle manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Chemical fertilizer (NPK)2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Chicken manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Control (no fertilizer)2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Sheep manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Turkey manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Vermicompost2017Basis not stated | mean. | ||
4 findings · 2 studies | ||||
| Collard greens (cooked)FDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Kale (cooked)FDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Head cabbage (cooked)UK TDS 2014 (Fera 2015) | UK TDS 2014 category 11_622014Fresh weight | below LOD. | ||
| Other green vegetables (mixed category) (cooked)UK TDS 2014 (Fera 2015) | UK TDS 2014 category 11_682014Fresh weight | below LOD. | ||
50 findings · 7 studies | ||||
| Head cabbageWHO GEMS/Food | WHO region EURO; targeted sampling2000Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region EURO; random sampling2007–2009Fresh weight | median. | ||
| SpinachWHO GEMS/Food | WHO region EURO; targeted sampling2000Fresh weight | median. | ||
| Coriander leavesAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| Coriander leavesAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| Coriander leavesAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| Coriander leavesAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | mean. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | mean. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | mean. | ||
| Leafy vegetables, other or unspecifiedAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | mean. | ||
| LettuceAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| LettuceAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| LettuceAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| LettuceAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| MintAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| Spring onion leavesAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Open-air markets (unwashed, OMUW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Open-air markets (washed, OMW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Supermarkets (washed, SMW)Collection date unreportedDry weight | sample value. | ||
| Swiss chardAdhikari et al. 2024 | Vegetable markets (washed, VMW)Collection date unreportedDry weight | sample value. | ||
| Head lettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Lamb's lettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| LettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| RocketMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Romaine lettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| SpinachMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Swiss chardMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| WatercressMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Premna latifolia leavesNarayana Murthy and Padala 2024 | BorivankaCollection date unreportedBasis not stated | mean. | ||
| Premna latifolia leavesNarayana Murthy and Padala 2024 | KusumapuramCollection date unreportedBasis not stated | mean. | ||
| Premna latifolia leavesNarayana Murthy and Padala 2024 | VarakaCollection date unreportedBasis not stated | mean. | ||
| Head cabbageSee et al. 2025 | Pasar Borong Selangor wholesale2022Fresh weight | mean. | ||
| Mustard greensSee et al. 2025 | Pasar Borong Selangor wholesale2022Fresh weight | mean. | ||
| Pak choiSee et al. 2025 | Pasar Borong Selangor wholesale2022Fresh weight | mean. | ||
| SpinachSee et al. 2025 | Pasar Borong Selangor wholesale2022Fresh weight | mean. | ||
| LettuceUK TDS 2014 (Fera 2015) | UK TDS 2014 category 11_652014Fresh weight | sample value. | ||
| SpinachYaman Türkkan and Kibar 2022 | Cattle manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Chemical fertilizer (NPK)2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Chicken manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Control (no fertilizer)2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Sheep manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Turkey manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Vermicompost2017Basis not stated | mean. | ||
21 findings · 4 studies | ||||
| Spinach (cooked)WHO GEMS/Food | WHO region EURO; random sampling2007–2009Fresh weight | median. | ||
| Amaranth leaves (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Chinese cabbage (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Chinese kale (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Choy sum (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Head cabbage (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Lettuce (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Lettuce (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Mustard greens (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Pak choi (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Spinach (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Water spinach (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Watercress (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Head cabbage (cooked)UK TDS 2014 (Fera 2015) | UK TDS 2014 category 11_622014Fresh weight | sample value. | ||
| Other green vegetables (mixed category) (cooked)UK TDS 2014 (Fera 2015) | UK TDS 2014 category 11_682014Fresh weight | sample value. | ||
| Chinese cabbage (cooked)Wang et al. 2020 | Jilin ProvinceCollection date unreportedBasis not stated | mean. | ||
| Chinese cabbage (cooked)Wang et al. 2020 | Other AreaCollection date unreportedBasis not stated | mean. | ||
| Chinese cabbage (cooked)Wang et al. 2020 | Songhua River AreaCollection date unreportedBasis not stated | mean. | ||
| Leek (cooked)Wang et al. 2020 | Jilin ProvinceCollection date unreportedBasis not stated | mean. | ||
| Leek (cooked)Wang et al. 2020 | Other AreaCollection date unreportedBasis not stated | mean. | ||
| Leek (cooked)Wang et al. 2020 | Songhua River AreaCollection date unreportedBasis not stated | mean. | ||
28 findings · 3 studies | ||||
| Amaranth leaves (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010Fresh weight | median. | ||
| Chinese cabbage (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Head cabbage (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Head lettuce (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010Fresh weight | median. | ||
| Kale (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Leaf lettuce (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Leafy vegetables, other or unspecified (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Leafy vegetables, other or unspecified (cooked)WHO GEMS/Food | WHO region WPRO; targeted sampling1998Fresh weight | median. | ||
| Mustard greens (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Pak choi (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010Fresh weight | median. | ||
| Spinach (cooked)WHO GEMS/Food | WHO region EURO; random sampling2007–2009Fresh weight | median. | ||
| Spinach (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Water spinach (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010Fresh weight | median. | ||
| Watercress (cooked)WHO GEMS/Food | WHO region WPRO; random sampling2010–2011Fresh weight | median. | ||
| Amaranth leaves (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Chinese cabbage (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | not detected. | ||
| Chinese kale (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Choy sum (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | not detected. | ||
| Head cabbage (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | not detected. | ||
| Lettuce (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Lettuce (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Mustard greens (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Pak choi (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Spinach (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Water spinach (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Watercress (cooked)Hong Kong TDS 2013 | Hong Kong retail purchases2010–2011Basis not stated | mean. | ||
| Head cabbage (cooked)UK TDS 2014 (Fera 2015) | UK TDS 2014 category 11_622014Fresh weight | below LOD. | ||
| Other green vegetables (mixed category) (cooked)UK TDS 2014 (Fera 2015) | UK TDS 2014 category 11_682014Fresh weight | below LOD. | ||
29 findings · 5 studies · 2 spread figures folded in | ||||
| Head lettuceWHO GEMS/Food | WHO region WPRO; targeted sampling1998Fresh weight | median. | ||
| KaleWHO GEMS/Food | WHO region WPRO; targeted sampling2001Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region EURO; random sampling2007–2009Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region WPRO; random sampling2005Fresh weight | median. | ||
| Leafy vegetables, other or unspecifiedWHO GEMS/Food | WHO region WPRO; targeted sampling2001Fresh weight | median. | ||
| Mustard greensWHO GEMS/Food | WHO region WPRO; random sampling2005Fresh weight | median. | ||
| Mustard greensWHO GEMS/Food | WHO region WPRO; targeted sampling2002Fresh weight | median. | ||
| Pak choiWHO GEMS/Food | WHO region WPRO; targeted sampling2001Fresh weight | median. | ||
| Head lettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Lamb's lettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| LettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| RocketMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Romaine lettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| SpinachMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Swiss chardMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| WatercressMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| LettuceUK TDS 2014 (Fera 2015) | UK TDS 2014 category 11_652014Fresh weight | below LOD. | ||
| SpinachYaman Türkkan and Kibar 2022 | Cattle manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Chemical fertilizer (NPK)2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Chicken manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Control (no fertilizer)2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Sheep manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Turkey manure2017Basis not stated | mean. | ||
| SpinachYaman Türkkan and Kibar 2022 | Vermicompost2017Basis not stated | mean. | ||
| Leafy vegetables, other or unspecifiedYe et al. 2026 | Zhejiang Province2018–2019Dry weight | iqr. | ||
| Leafy vegetables, other or unspecifiedYe et al. 2026 | Zhejiang Province2018–2019Dry weight | 95th percentile. | ||
| Leafy vegetables, other or unspecifiedYe et al. 2026 | Zhejiang Province2018–2019Dry weight | mean.Same samples: 210 µg/kg (95th percentile); 1,140 µg/kg (maximum) | ||
| Leafy vegetables, other or unspecifiedYe et al. 2026 | Zhejiang Province2018–2019Dry weight | median. | ||
| Leafy vegetables, other or unspecifiedYe et al. 2026 | Zhejiang Province2018–2019Dry weight | not detected. | ||
| Leafy vegetables, other or unspecifiedYe et al. 2026 | Zhejiang Province2018–2019Dry weight | min. | ||
| Leafy vegetables, other or unspecifiedYe et al. 2026 | Zhejiang Province2018–2019Dry weight | max. | ||
2 findings · 1 study | ||||
| Collard greens (cooked)FDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Kale (cooked)FDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
32 findings · 4 studies · 7 spread figures folded in | ||||
| Head cabbageFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Head lettuceFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| Leaf lettuceFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| SpinachFDA TDS FY2018–FY2020 | United States national collections2018–2020Fresh weight | median. | ||
| KaleKim et al. 2024 | All six farms, field samples2019Fresh weight | mean.Same samples: 5–15.2 µg/kg (range) | ||
| KaleKim et al. 2024 | All six farms, field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | All six farms, field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm R1 (Rural), field samples2019Fresh weight | mean.Same samples: 5–5 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm R1 (Rural), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm R1 (Rural), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm R2 (Rural), field samples2019Fresh weight | mean.Same samples: 5–5 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm R2 (Rural), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm R2 (Rural), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm R3 (Rural), field samples2019Fresh weight | mean.Same samples: 5–5 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm R3 (Rural), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm R3 (Rural), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm R4 (Rural), field samples2019Fresh weight | mean.Same samples: 5–5 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm R4 (Rural), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm R4 (Rural), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm U1 (Urban), field samples2019Fresh weight | mean.Same samples: 5–5 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm U1 (Urban), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm U1 (Urban), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Farm U2 (Urban), field samples2019Fresh weight | mean.Same samples: 7–15.2 µg/kg (range) | ||
| KaleKim et al. 2024 | Farm U2 (Urban), field samples2019Fresh weight | median. | ||
| KaleKim et al. 2024 | Farm U2 (Urban), field samples2019Fresh weight | range. | ||
| KaleKim et al. 2024 | Rural farms R1-R42019Fresh weight | mean. | ||
| KaleKim et al. 2024 | Urban farms U1-U22019Fresh weight | mean. | ||
| Head lettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Lamb's lettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| LettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| RocketMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Romaine lettuceMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| SpinachMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Swiss chardMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| WatercressMartin-Leon et al. 2023 | Large supermarkets2016–2017Fresh weight | median. | ||
| Head cabbageOkonofua et al. 2024 | Dry Season2019Basis not stated | not detected. | ||
| Head cabbageOkonofua et al. 2024 | Dry season Control2019Basis not stated | not detected. | ||
| Head cabbageOkonofua et al. 2024 | Rainy Season2019Basis not stated | not detected. | ||
| Head cabbageOkonofua et al. 2024 | Rainy season Control2019Basis not stated | not detected. | ||
Shared dataset methods and interpretation
Source summaries retain their original statistic, material, basis and censoring. Each source drawer identifies methods and overlapping samples. No matched reference or pooled distribution is inferred.
HMI amaranth leaves (cooked) Pb reference (fresh weight) · 1 sampling family
Reference and scope
HMI amaranth leaves (cooked) Pb reference (fresh weight) · 1 sampling family. Amaranth leaves (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 62.5 | 62.5 | 62.5 |
| Alternative weighting | 62.5 | 62.5 | 62.5 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for amaranth leaves (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-amaranth-leaves-cooked-fresh-weight:82ff8ca23965
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI chinese cabbage Pb reference (fresh weight) · 1 sampling family
Reference and scope
HMI chinese cabbage Pb reference (fresh weight) · 1 sampling family. Chinese cabbage; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤9 | 4.53–14.8 (bounded) | 22 |
| Alternative weighting | ≤9 | 4.53–14.8 (bounded) | 22 |
14 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 14 | 100% |
Methods
HMI pools subgroup medians for chinese cabbage on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-chinese-cabbage-fresh-weight:4f71370fad03
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI chinese cabbage (cooked) Pb reference (fresh weight) · 1 sampling family
Reference and scope
HMI chinese cabbage (cooked) Pb reference (fresh weight) · 1 sampling family. Chinese cabbage (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤2 | 1–2 (bounded) | 2 |
| Alternative weighting | ≤2 | 1–2 (bounded) | 2 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for chinese cabbage (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-chinese-cabbage-cooked-fresh-weight:6bacebd30036
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head cabbage Pb reference (fresh weight) · 3 sampling families
Reference and scope
HMI head cabbage Pb reference (fresh weight) · 3 sampling families. Head cabbage; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤4 | ≤9 | 295 |
| Alternative weighting | ≤4 | ≤9 | 295 |
54 reference units; 3 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 43 | 33% |
| FDA TDS FY2018–FY2020 | 1 | 33% |
| Armand et al. 2026 | 10 | 33% |
Methods
HMI pools subgroup medians for head cabbage on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-head-cabbage-fresh-weight:c38390d4d7fd
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head cabbage (canned) Pb reference (fresh weight) · 1 sampling family
Reference and scope
HMI head cabbage (canned) Pb reference (fresh weight) · 1 sampling family. Head cabbage (canned); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 367 | 367 | 367 |
| Alternative weighting | 367 | 367 | 367 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for head cabbage (canned) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-head-cabbage-canned-fresh-weight:99f99aa4efee
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head cabbage (cooked) Pb reference (fresh weight) · 1 sampling family
Reference and scope
HMI head cabbage (cooked) Pb reference (fresh weight) · 1 sampling family. Head cabbage (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤3 | ≤5 | 7.8 |
| Alternative weighting | ≤3 | ≤5 | 7.8 |
31 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 31 | 100% |
Methods
HMI pools subgroup medians for head cabbage (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-head-cabbage-cooked-fresh-weight:fb24c7e95338
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head cabbage (fermented) Pb reference (fresh weight) · 1 sampling family
Reference and scope
HMI head cabbage (fermented) Pb reference (fresh weight) · 1 sampling family. Head cabbage (fermented); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 6.04–10 (bounded) | 11.1–12.4 (bounded) | 16.4–17.6 (bounded) |
| Alternative weighting | 6.04–10 (bounded) | 11.1–12.4 (bounded) | 16.4–17.6 (bounded) |
20 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 20 | 100% |
Methods
HMI pools subgroup medians for head cabbage (fermented) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-head-cabbage-fermented-fresh-weight:777cd441f04a
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head lettuce Pb reference (dry weight) · 1 sampling family
Reference and scope
HMI head lettuce Pb reference (dry weight) · 1 sampling family. Head lettuce; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤100 | ≤100 | ≤100 |
| Alternative weighting | ≤100 | ≤100 | ≤100 |
3 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 3 | 100% |
Methods
HMI pools subgroup medians for head lettuce on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-head-lettuce-dry-weight:6dc1c47a0776
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head lettuce Pb reference (fresh weight) · 3 sampling families
Reference and scope
HMI head lettuce Pb reference (fresh weight) · 3 sampling families. Head lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤2 | ≤4 | 2–5 (bounded) |
| Alternative weighting | ≤2 | ≤4 | 2–5 (bounded) |
42 reference units; 3 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 40 | 33% |
| FDA TDS FY2018–FY2020 | 1 | 33% |
| Martin-Leon et al. 2023 | 1 | 33% |
Methods
HMI pools subgroup medians for head lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-head-lettuce-fresh-weight:49b693d6ae16
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head lettuce (cooked) Pb reference (fresh weight) · 1 sampling family
Reference and scope
HMI head lettuce (cooked) Pb reference (fresh weight) · 1 sampling family. Head lettuce (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 5 | 5 | 5 |
| Alternative weighting | 5 | 5 | 5 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for head lettuce (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-head-lettuce-cooked-fresh-weight:896f3e6810c6
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI kale Pb reference (fresh weight) · 2 sampling families
Reference and scope
HMI kale Pb reference (fresh weight) · 2 sampling families. Kale; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 5.95–8.29 (bounded) | 12.6–15 (bounded) | 20.3–23 (bounded) |
| Alternative weighting | 5.58–7.91 (bounded) | 12.6–13.7 (bounded) | 20–22 (bounded) |
60 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 22 | 50% |
| Kim et al. 2024 | 38 | 50% |
Methods
HMI pools subgroup medians for kale on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-kale-fresh-weight:224f40b3027f
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI kale (cooked) Pb reference (fresh weight) · 2 sampling families
Reference and scope
HMI kale (cooked) Pb reference (fresh weight) · 2 sampling families. Kale (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤4 | 2.55–4.55 (bounded) | 8.53 |
| Alternative weighting | ≤4 | 2.55–4.55 (bounded) | 8.53 |
12 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 11 | 50% |
| FDA TDS FY2018–FY2020 | 1 | 50% |
Methods
HMI pools subgroup medians for kale (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-kale-cooked-fresh-weight:52862efb6a75
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI lamb's lettuce Pb reference (fresh weight) · 2 sampling families
Reference and scope
HMI lamb's lettuce Pb reference (fresh weight) · 2 sampling families. Lamb's lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 6.4 | 6.4 | 23 |
| Alternative weighting | 6.4 | 6.4 | 23 |
20 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 19 | 50% |
| Martin-Leon et al. 2023 | 1 | 50% |
Methods
HMI pools subgroup medians for lamb's lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-lamb-s-lettuce-fresh-weight:d81f3cbe9187
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leaf lettuce Pb reference (fresh weight) · 2 sampling families
Reference and scope
HMI leaf lettuce Pb reference (fresh weight) · 2 sampling families. Leaf lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤4 | ≤4 | ≤10 |
| Alternative weighting | ≤4 | ≤4 | ≤10 |
35 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 34 | 50% |
| FDA TDS FY2018–FY2020 | 1 | 50% |
Methods
HMI pools subgroup medians for leaf lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-leaf-lettuce-fresh-weight:c5b846bc54d7
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leaf lettuce (cooked) Pb reference (fresh weight) · 1 sampling family
Reference and scope
HMI leaf lettuce (cooked) Pb reference (fresh weight) · 1 sampling family. Leaf lettuce (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 10 | 11 | 12 |
| Alternative weighting | 10 | 11 | 12 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for leaf lettuce (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-leaf-lettuce-cooked-fresh-weight:7c1720d54ea2
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leafy vegetables, other or unspecified Pb reference (fresh weight) · 1 sampling family
Reference and scope
HMI leafy vegetables, other or unspecified Pb reference (fresh weight) · 1 sampling family. Leafy vegetables, other or unspecified; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤16 | 5–24 (bounded) | 11–47 (bounded) |
| Alternative weighting | ≤16 | 5–24 (bounded) | 11–47 (bounded) |
21 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 21 | 100% |
Methods
HMI pools subgroup medians for leafy vegetables, other or unspecified on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-leafy-vegetables-other-or-unspecified-fresh-weight:0c53d6fe4623
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leafy vegetables, other or unspecified (cooked) Pb reference (fresh weight) · 1 sampling family
Reference and scope
HMI leafy vegetables, other or unspecified (cooked) Pb reference (fresh weight) · 1 sampling family. Leafy vegetables, other or unspecified (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤7 | 3.75–7 (bounded) | 6.25–8.5 (bounded) |
| Alternative weighting | ≤7 | 3.75–7 (bounded) | 6.25–8.5 (bounded) |
20 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 20 | 100% |
Methods
HMI pools subgroup medians for leafy vegetables, other or unspecified (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-leafy-vegetables-other-or-unspecified-cooked-fresh-weight:fcdb058a04d4
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI mustard greens Pb reference (fresh weight) · 1 sampling family
Reference and scope
HMI mustard greens Pb reference (fresh weight) · 1 sampling family. Mustard greens; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 33 | 42.5 | 84 |
| Alternative weighting | 33 | 42.5 | 84 |
6 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 6 | 100% |
Methods
HMI pools subgroup medians for mustard greens on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-mustard-greens-fresh-weight:2cf9df8d5ac0
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI mustard greens (canned) Pb reference (fresh weight) · 1 sampling family
Reference and scope
HMI mustard greens (canned) Pb reference (fresh weight) · 1 sampling family. Mustard greens (canned); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 61.4 | 61.4 | 61.4 |
| Alternative weighting | 61.4 | 61.4 | 61.4 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for mustard greens (canned) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-mustard-greens-canned-fresh-weight:dd2ad78daf1f
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI mustard greens (cooked) Pb reference (fresh weight) · 1 sampling family
Reference and scope
HMI mustard greens (cooked) Pb reference (fresh weight) · 1 sampling family. Mustard greens (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 6 | 13.5 | 21 |
| Alternative weighting | 6 | 13.5 | 21 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for mustard greens (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-mustard-greens-cooked-fresh-weight:446f98e4083b
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI pak choi Pb reference (fresh weight) · 1 sampling family
Reference and scope
HMI pak choi Pb reference (fresh weight) · 1 sampling family. Pak choi; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤5 | 0.955–9.69 (bounded) | 3.37–13.8 (bounded) |
| Alternative weighting | ≤5 | 0.955–9.69 (bounded) | 3.37–13.8 (bounded) |
24 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 24 | 100% |
Methods
HMI pools subgroup medians for pak choi on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-pak-choi-fresh-weight:76c1bd31d322
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI pak choi (cooked) Pb reference (fresh weight) · 1 sampling family
Reference and scope
HMI pak choi (cooked) Pb reference (fresh weight) · 1 sampling family. Pak choi (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤5 | 6.5–9 (bounded) | 13 |
| Alternative weighting | ≤5 | 6.5–9 (bounded) | 13 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for pak choi (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-pak-choi-cooked-fresh-weight:0a7d16d949f8
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach Pb reference (fresh weight) · 3 sampling families
Reference and scope
HMI spinach Pb reference (fresh weight) · 3 sampling families. Spinach; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 6.2 | 12–12.8 (bounded) | 12.8 |
| Alternative weighting | 6.2 | 12–12.8 (bounded) | 12.8 |
41 reference units; 3 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 39 | 33% |
| FDA TDS FY2018–FY2020 | 1 | 33% |
| Martin-Leon et al. 2023 | 1 | 33% |
Methods
HMI pools subgroup medians for spinach on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-spinach-fresh-weight:c8ab166bf052
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach (canned) Pb reference (fresh weight) · 1 sampling family
Reference and scope
HMI spinach (canned) Pb reference (fresh weight) · 1 sampling family. Spinach (canned); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 7.35 | 10.2 | 30.6 |
| Alternative weighting | 7.35 | 10.2 | 30.6 |
3 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 3 | 100% |
Methods
HMI pools subgroup medians for spinach (canned) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-spinach-canned-fresh-weight:01e56c2f4666
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach (cooked) Pb reference (fresh weight) · 1 sampling family
Reference and scope
HMI spinach (cooked) Pb reference (fresh weight) · 1 sampling family. Spinach (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 5–7.5 (bounded) | 13.7 | 18.7 |
| Alternative weighting | 5–7.5 (bounded) | 13.7 | 18.7 |
32 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 32 | 100% |
Methods
HMI pools subgroup medians for spinach (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-spinach-cooked-fresh-weight:6a5cd661a517
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach (frozen) Pb reference (fresh weight) · 1 sampling family
Reference and scope
HMI spinach (frozen) Pb reference (fresh weight) · 1 sampling family. Spinach (frozen); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤14.5 | 0.0298–18.8 (bounded) | 12–26.3 (bounded) |
| Alternative weighting | ≤14.5 | 0.0298–18.8 (bounded) | 12–26.3 (bounded) |
16 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 16 | 100% |
Methods
HMI pools subgroup medians for spinach (frozen) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-spinach-frozen-fresh-weight:2cc05488357c
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI water spinach (cooked) Pb reference (fresh weight) · 1 sampling family
Reference and scope
HMI water spinach (cooked) Pb reference (fresh weight) · 1 sampling family. Water spinach (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 44.5 | 44.5 | 44.5 |
| Alternative weighting | 44.5 | 44.5 | 44.5 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for water spinach (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-water-spinach-cooked-fresh-weight:dca987b0cc6f
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI watercress Pb reference (fresh weight) · 2 sampling families
Reference and scope
HMI watercress Pb reference (fresh weight) · 2 sampling families. Watercress; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 5.8–7.5 (bounded) | 7.5 | 9.75–16 (bounded) |
| Alternative weighting | 5.8–7.5 (bounded) | 7.5 | 9.75–16 (bounded) |
5 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 4 | 50% |
| Martin-Leon et al. 2023 | 1 | 50% |
Methods
HMI pools subgroup medians for watercress on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-watercress-fresh-weight:7de57a64d9ed
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI watercress (cooked) Pb reference (fresh weight) · 1 sampling family
Reference and scope
HMI watercress (cooked) Pb reference (fresh weight) · 1 sampling family. Watercress (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 21 | 151 | 280 |
| Alternative weighting | 21 | 151 | 280 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for watercress (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-watercress-cooked-fresh-weight:76c844b253c4
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI collard greens (cooked) Pb reference (fresh weight) · 1 sampling family
Reference and scope
HMI collard greens (cooked) Pb reference (fresh weight) · 1 sampling family. Collard greens (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤4 | ≤4 | ≤4 |
| Alternative weighting | ≤4 | ≤4 | ≤4 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| FDA TDS FY2018–FY2020 | 1 | 100% |
Methods
HMI pools subgroup medians for collard greens (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-collard-greens-cooked-fresh-weight:5aa5b98147be
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI amaranth leaves Pb reference (dry weight) · 1 sampling family
Reference and scope
HMI amaranth leaves Pb reference (dry weight) · 1 sampling family. Amaranth leaves; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 4020 | 6800 | 13800 |
| Alternative weighting | 4020 | 6800 | 13800 |
26 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Hossain et al. 2016 | 26 | 100% |
Methods
HMI pools subgroup medians for amaranth leaves on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-amaranth-leaves-dry-weight:545a4cfa4760
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI coriander leaves Pb reference (dry weight) · 1 sampling family
Reference and scope
HMI coriander leaves Pb reference (dry weight) · 1 sampling family. Coriander leaves; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 28.9 | 32.7 | 41.5 |
| Alternative weighting | 28.9 | 32.7 | 41.5 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 100% |
Methods
HMI pools subgroup medians for coriander leaves on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-coriander-leaves-dry-weight:a8d214cee124
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leafy vegetables, other or unspecified Pb reference (dry weight) · 1 sampling family
Reference and scope
HMI leafy vegetables, other or unspecified Pb reference (dry weight) · 1 sampling family. Leafy vegetables, other or unspecified; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 20800 | 20800 | 20800 |
| Alternative weighting | 20800 | 20800 | 20800 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Cherfi et al. 2014 | 1 | 100% |
Methods
HMI pools subgroup medians for leafy vegetables, other or unspecified on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-leafy-vegetables-other-or-unspecified-dry-weight:aa656c439551
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI lettuce Pb reference (dry weight) · 3 sampling families
Reference and scope
HMI lettuce Pb reference (dry weight) · 3 sampling families. Lettuce; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤30.8 | 28.1–3910 (bounded) | 4150–4330 (bounded) |
| Alternative weighting | ≤30.8 | 28.1–3910 (bounded) | 4150–4330 (bounded) |
9 reference units; 3 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 33% |
| Stark et al. 2019 | 1 | 33% |
| Vela et al. 2023 | 4 | 33% |
Methods
HMI pools subgroup medians for lettuce on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-lettuce-dry-weight:b300694a4371
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI mint Pb reference (dry weight) · 1 sampling family
Reference and scope
HMI mint Pb reference (dry weight) · 1 sampling family. Mint; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 29.8 | 32.5 | 37.7 |
| Alternative weighting | 29.8 | 32.5 | 37.7 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 100% |
Methods
HMI pools subgroup medians for mint on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-mint-dry-weight:020f703cd971
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spring onion leaves Pb reference (dry weight) · 1 sampling family
Reference and scope
HMI spring onion leaves Pb reference (dry weight) · 1 sampling family. Spring onion leaves; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 22.5 | 26.4 | 31.4 |
| Alternative weighting | 22.5 | 26.4 | 31.4 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 100% |
Methods
HMI pools subgroup medians for spring onion leaves on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-spring-onion-leaves-dry-weight:8f01cb829186
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI swiss chard Pb reference (dry weight) · 1 sampling family
Reference and scope
HMI swiss chard Pb reference (dry weight) · 1 sampling family. Swiss chard; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 31.7 | 33.6 | 36.7 |
| Alternative weighting | 31.7 | 33.6 | 36.7 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 100% |
Methods
HMI pools subgroup medians for swiss chard on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-swiss-chard-dry-weight:000844d71597
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI lettuce Pb reference (fresh weight) · 5 sampling families
Reference and scope
HMI lettuce Pb reference (fresh weight) · 5 sampling families. Lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤6.9 | 4–10 (bounded) | 73.5–111 (bounded) |
| Alternative weighting | ≤6.9 | 4–10 (bounded) | 26–100 (bounded) |
21 reference units; 5 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Armand et al. 2026 | 10 | 20% |
| Dearing et al. 2025 | 8 | 20% |
| Romero-Crespo et al. 2023 | 1 | 20% |
| UK TDS 2014 (Fera 2015) | 1 | 20% |
| Martin-Leon et al. 2023 | 1 | 20% |
Methods
HMI pools subgroup medians for lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
- Armand et al. 2026
- Dearing et al. 2025
- Romero-Crespo et al. 2023
- UK TDS 2014 (Fera 2015)
- Martin-Leon et al. 2023
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-lettuce-fresh-weight:9a7e6bcfd20a
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head cabbage Pb reference (dry weight) · 1 sampling family
Reference and scope
HMI head cabbage Pb reference (dry weight) · 1 sampling family. Head cabbage; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 3380 | 5950 | 12300 |
| Alternative weighting | 3380 | 5950 | 12300 |
12 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Hossain et al. 2016 | 12 | 100% |
Methods
HMI pools subgroup medians for head cabbage on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-head-cabbage-dry-weight:92d92c9acc50
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach Pb reference (dry weight) · 1 sampling family
Reference and scope
HMI spinach Pb reference (dry weight) · 1 sampling family. Spinach; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 16700 | 16700 | 16700 |
| Alternative weighting | 16700 | 16700 | 16700 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Cherfi et al. 2014 | 1 | 100% |
Methods
HMI pools subgroup medians for spinach on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-spinach-dry-weight:17772612b27d
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI rocket Pb reference (fresh weight) · 1 sampling family
Reference and scope
HMI rocket Pb reference (fresh weight) · 1 sampling family. Rocket; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 13.1 | 13.1 | 13.1 |
| Alternative weighting | 13.1 | 13.1 | 13.1 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for rocket on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-rocket-fresh-weight:65cacbd30ea3
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI romaine lettuce Pb reference (fresh weight) · 1 sampling family
Reference and scope
HMI romaine lettuce Pb reference (fresh weight) · 1 sampling family. Romaine lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 4.9 | 4.9 | 4.9 |
| Alternative weighting | 4.9 | 4.9 | 4.9 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for romaine lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-romaine-lettuce-fresh-weight:fb7e210693ac
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI swiss chard Pb reference (fresh weight) · 1 sampling family
Reference and scope
HMI swiss chard Pb reference (fresh weight) · 1 sampling family. Swiss chard; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 9.3 | 9.3 | 9.3 |
| Alternative weighting | 9.3 | 9.3 | 9.3 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for swiss chard on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-swiss-chard-fresh-weight:49c6d5494c76
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI coriander leaves Pb reference (fresh weight) · 1 sampling family
Reference and scope
HMI coriander leaves Pb reference (fresh weight) · 1 sampling family. Coriander leaves; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤100 | ≤100 | ≤100 |
| Alternative weighting | ≤100 | ≤100 | ≤100 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Romero-Crespo et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for coriander leaves on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-coriander-leaves-fresh-weight:bab943d1330b
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI chives Pb reference (fresh weight) · 1 sampling family
Reference and scope
HMI chives Pb reference (fresh weight) · 1 sampling family. Chives; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤100 | ≤100 | ≤100 |
| Alternative weighting | ≤100 | ≤100 | ≤100 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Romero-Crespo et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for chives on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-chives-fresh-weight:3bcc9d02e59e
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI mallow Pb reference (dry weight) · 1 sampling family
Reference and scope
HMI mallow Pb reference (dry weight) · 1 sampling family. Mallow; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤3940 | ≤3940 | ≤3940 |
| Alternative weighting | ≤3940 | ≤3940 | ≤3940 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 1 | 100% |
Methods
HMI pools subgroup medians for mallow on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-mallow-dry-weight:4989c125bc89
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI bristly ox tongue Pb reference (dry weight) · 1 sampling family
Reference and scope
HMI bristly ox tongue Pb reference (dry weight) · 1 sampling family. Bristly ox tongue; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤3830 | ≤3830 | ≤3830 |
| Alternative weighting | ≤3830 | ≤3830 | ≤3830 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 2 | 100% |
Methods
HMI pools subgroup medians for bristly ox tongue on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-bristly-ox-tongue-dry-weight:512f9b92ed8e
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI cat's ear Pb reference (dry weight) · 1 sampling family
Reference and scope
HMI cat's ear Pb reference (dry weight) · 1 sampling family. Cat's ear; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤3910 | ≤3910 | ≤3910 |
| Alternative weighting | ≤3910 | ≤3910 | ≤3910 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 1 | 100% |
Methods
HMI pools subgroup medians for cat's ear on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-cat-s-ear-dry-weight:e88e7a33fac5
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI dandelion greens Pb reference (dry weight) · 1 sampling family
Reference and scope
HMI dandelion greens Pb reference (dry weight) · 1 sampling family. Dandelion greens; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤3990 | ≤3990 | ≤3990 |
| Alternative weighting | ≤3990 | ≤3990 | ≤3990 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 1 | 100% |
Methods
HMI pools subgroup medians for dandelion greens on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-dandelion-greens-dry-weight:cca6eda0b70f
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI english plantain Pb reference (dry weight) · 1 sampling family
Reference and scope
HMI english plantain Pb reference (dry weight) · 1 sampling family. English plantain; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤3830 | ≤3840 | ≤3850 |
| Alternative weighting | ≤3830 | ≤3840 | ≤3850 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 2 | 100% |
Methods
HMI pools subgroup medians for english plantain on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-english-plantain-dry-weight:07a3d13d5508
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI nasturtium (tropaeolum) Pb reference (dry weight) · 1 sampling family
Reference and scope
HMI nasturtium (tropaeolum) Pb reference (dry weight) · 1 sampling family. Nasturtium (Tropaeolum); Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤3830 | ≤3830 | ≤3830 |
| Alternative weighting | ≤3830 | ≤3830 | ≤3830 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 1 | 100% |
Methods
HMI pools subgroup medians for nasturtium (tropaeolum) on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-nasturtium-tropaeolum-dry-weight:86a6df5e8936
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI sweet fennel Pb reference (dry weight) · 1 sampling family
Reference and scope
HMI sweet fennel Pb reference (dry weight) · 1 sampling family. Sweet fennel; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤3920 | ≤3960 | ≤3990 |
| Alternative weighting | ≤3920 | ≤3960 | ≤3990 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 2 | 100% |
Methods
HMI pools subgroup medians for sweet fennel on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-pb-sweet-fennel-dry-weight:c8013a0c563a
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI amaranth leaves (cooked) Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI amaranth leaves (cooked) Cd reference (fresh weight) · 1 sampling family. Amaranth leaves (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 56.5 | 56.5 | 56.5 |
| Alternative weighting | 56.5 | 56.5 | 56.5 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for amaranth leaves (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-amaranth-leaves-cooked-fresh-weight:faafcdd57686
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI chinese cabbage Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI chinese cabbage Cd reference (fresh weight) · 1 sampling family. Chinese cabbage; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 12 | 22.5 | 30 |
| Alternative weighting | 12 | 22.5 | 30 |
6 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 6 | 100% |
Methods
HMI pools subgroup medians for chinese cabbage on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-chinese-cabbage-fresh-weight:cef41dc399ed
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI chinese cabbage (cooked) Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI chinese cabbage (cooked) Cd reference (fresh weight) · 1 sampling family. Chinese cabbage (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 48 | 67 | 86 |
| Alternative weighting | 48 | 67 | 86 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for chinese cabbage (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-chinese-cabbage-cooked-fresh-weight:928d24dd87f0
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head cabbage Cd reference (fresh weight) · 3 sampling families
Reference and scope
HMI head cabbage Cd reference (fresh weight) · 3 sampling families. Head cabbage; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 4–5 (bounded) | 5.1 | 64 |
| Alternative weighting | 4–5 (bounded) | 5.1 | 64 |
49 reference units; 3 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 38 | 33% |
| FDA TDS FY2018–FY2020 | 1 | 33% |
| Armand et al. 2026 | 10 | 33% |
Methods
HMI pools subgroup medians for head cabbage on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-head-cabbage-fresh-weight:ac4a41f47833
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head cabbage (cooked) Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI head cabbage (cooked) Cd reference (fresh weight) · 1 sampling family. Head cabbage (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 3 | 4.5 | 6 |
| Alternative weighting | 3 | 4.5 | 6 |
30 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 30 | 100% |
Methods
HMI pools subgroup medians for head cabbage (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-head-cabbage-cooked-fresh-weight:12e64aec411c
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head cabbage (fermented) Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI head cabbage (fermented) Cd reference (fresh weight) · 1 sampling family. Head cabbage (fermented); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1.89–2 (bounded) | 2.37–3 (bounded) | 3.5–4 (bounded) |
| Alternative weighting | 1.89–2 (bounded) | 2.37–3 (bounded) | 3.5–4 (bounded) |
17 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 17 | 100% |
Methods
HMI pools subgroup medians for head cabbage (fermented) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-head-cabbage-fermented-fresh-weight:599c7a59f1fa
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head cabbage (frozen) Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI head cabbage (frozen) Cd reference (fresh weight) · 1 sampling family. Head cabbage (frozen); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 2–4.5 (bounded) | 4.5–6 (bounded) | 6.5–7.5 (bounded) |
| Alternative weighting | 2–4.5 (bounded) | 4.5–6 (bounded) | 6.5–7.5 (bounded) |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 4 | 100% |
Methods
HMI pools subgroup medians for head cabbage (frozen) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-head-cabbage-frozen-fresh-weight:b5034e8891b0
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head lettuce Cd reference (dry weight) · 1 sampling family
Reference and scope
HMI head lettuce Cd reference (dry weight) · 1 sampling family. Head lettuce; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤50 | ≤50 | ≤50 |
| Alternative weighting | ≤50 | ≤50 | ≤50 |
3 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 3 | 100% |
Methods
HMI pools subgroup medians for head lettuce on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-head-lettuce-dry-weight:c5f57c7d9c7f
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head lettuce Cd reference (fresh weight) · 3 sampling families
Reference and scope
HMI head lettuce Cd reference (fresh weight) · 3 sampling families. Head lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 4 | 15–19.8 (bounded) | 27 |
| Alternative weighting | 4 | 15–19.8 (bounded) | 27 |
36 reference units; 3 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 34 | 33% |
| FDA TDS FY2018–FY2020 | 1 | 33% |
| Martin-Leon et al. 2023 | 1 | 33% |
Methods
HMI pools subgroup medians for head lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-head-lettuce-fresh-weight:6835dba952f7
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head lettuce (cooked) Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI head lettuce (cooked) Cd reference (fresh weight) · 1 sampling family. Head lettuce (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 14.5 | 14.5 | 14.5 |
| Alternative weighting | 14.5 | 14.5 | 14.5 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for head lettuce (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-head-lettuce-cooked-fresh-weight:e5f557a43c69
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI kale Cd reference (fresh weight) · 2 sampling families
Reference and scope
HMI kale Cd reference (fresh weight) · 2 sampling families. Kale; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 9.02–9.27 (bounded) | 15.8 | 22.3 |
| Alternative weighting | 9.02–9.35 (bounded) | 16 | 23.6 |
57 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 19 | 50% |
| Kim et al. 2024 | 38 | 50% |
Methods
HMI pools subgroup medians for kale on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-kale-fresh-weight:9ed9fd46a092
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI kale (canned) Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI kale (canned) Cd reference (fresh weight) · 1 sampling family. Kale (canned); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 20 | 27 | 34 |
| Alternative weighting | 20 | 27 | 34 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for kale (canned) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-kale-canned-fresh-weight:92428362dd30
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI kale (cooked) Cd reference (fresh weight) · 2 sampling families
Reference and scope
HMI kale (cooked) Cd reference (fresh weight) · 2 sampling families. Kale (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 4.65 | 15.5 | 15.5 |
| Alternative weighting | 4.65 | 15.5 | 15.5 |
12 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 11 | 50% |
| FDA TDS FY2018–FY2020 | 1 | 50% |
Methods
HMI pools subgroup medians for kale (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-kale-cooked-fresh-weight:54f4742eeb01
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI kale (frozen) Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI kale (frozen) Cd reference (fresh weight) · 1 sampling family. Kale (frozen); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 22 | 23 | 28.8 |
| Alternative weighting | 22 | 23 | 28.8 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 4 | 100% |
Methods
HMI pools subgroup medians for kale (frozen) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-kale-frozen-fresh-weight:509f5c742de8
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI lamb's lettuce Cd reference (fresh weight) · 2 sampling families
Reference and scope
HMI lamb's lettuce Cd reference (fresh weight) · 2 sampling families. Lamb's lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤1.1 | 1.1 | 1.1–2 (bounded) |
| Alternative weighting | ≤1.1 | 1.1 | 1.1–2 (bounded) |
12 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 11 | 50% |
| Martin-Leon et al. 2023 | 1 | 50% |
Methods
HMI pools subgroup medians for lamb's lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-lamb-s-lettuce-fresh-weight:99f855720100
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leaf lettuce Cd reference (fresh weight) · 2 sampling families
Reference and scope
HMI leaf lettuce Cd reference (fresh weight) · 2 sampling families. Leaf lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 30.8 | 57 | 57 |
| Alternative weighting | 30.8 | 57 | 57 |
38 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 37 | 50% |
| FDA TDS FY2018–FY2020 | 1 | 50% |
Methods
HMI pools subgroup medians for leaf lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-leaf-lettuce-fresh-weight:4c5ec51bbaeb
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leaf lettuce (cooked) Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI leaf lettuce (cooked) Cd reference (fresh weight) · 1 sampling family. Leaf lettuce (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤8.5 | 16–20.5 (bounded) | 35.5 |
| Alternative weighting | ≤8.5 | 16–20.5 (bounded) | 35.5 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 4 | 100% |
Methods
HMI pools subgroup medians for leaf lettuce (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-leaf-lettuce-cooked-fresh-weight:3ab16ae0ba84
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leafy vegetables, other or unspecified Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI leafy vegetables, other or unspecified Cd reference (fresh weight) · 1 sampling family. Leafy vegetables, other or unspecified; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 3–9 (bounded) | 10–13.4 (bounded) | 14–20 (bounded) |
| Alternative weighting | 3–9 (bounded) | 10–13.4 (bounded) | 14–20 (bounded) |
23 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 23 | 100% |
Methods
HMI pools subgroup medians for leafy vegetables, other or unspecified on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-leafy-vegetables-other-or-unspecified-fresh-weight:9c0cbcfc34a6
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leafy vegetables, other or unspecified (canned) Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI leafy vegetables, other or unspecified (canned) Cd reference (fresh weight) · 1 sampling family. Leafy vegetables, other or unspecified (canned); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 7 | 7 | 7 |
| Alternative weighting | 7 | 7 | 7 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for leafy vegetables, other or unspecified (canned) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-leafy-vegetables-other-or-unspecified-canned-fresh-weight:4e3adc44c96e
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leafy vegetables, other or unspecified (cooked) Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI leafy vegetables, other or unspecified (cooked) Cd reference (fresh weight) · 1 sampling family. Leafy vegetables, other or unspecified (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 25 | 27 | 30.5 |
| Alternative weighting | 25 | 27 | 30.5 |
19 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 19 | 100% |
Methods
HMI pools subgroup medians for leafy vegetables, other or unspecified (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-leafy-vegetables-other-or-unspecified-cooked-fresh-weight:770e8dbaf05d
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leafy vegetables, other or unspecified (frozen) Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI leafy vegetables, other or unspecified (frozen) Cd reference (fresh weight) · 1 sampling family. Leafy vegetables, other or unspecified (frozen); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 0.2 | 36.6 | 73 |
| Alternative weighting | 0.2 | 36.6 | 73 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for leafy vegetables, other or unspecified (frozen) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-leafy-vegetables-other-or-unspecified-frozen-fresh-weight:779498d02552
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI mustard greens Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI mustard greens Cd reference (fresh weight) · 1 sampling family. Mustard greens; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 50 | 57.5–67.5 (bounded) | 65–85 (bounded) |
| Alternative weighting | 50 | 57.5–67.5 (bounded) | 65–85 (bounded) |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for mustard greens on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-mustard-greens-fresh-weight:2c6aa08ac428
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI mustard greens (cooked) Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI mustard greens (cooked) Cd reference (fresh weight) · 1 sampling family. Mustard greens (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 39 | 41.5 | 44 |
| Alternative weighting | 39 | 41.5 | 44 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for mustard greens (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-mustard-greens-cooked-fresh-weight:21fad878238d
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI pak choi Cd reference (fresh weight) · 2 sampling families
Reference and scope
HMI pak choi Cd reference (fresh weight) · 2 sampling families. Pak choi; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 8.18–10 (bounded) | 15 | 29 |
| Alternative weighting | 8.18–10 (bounded) | 15 | 29 |
24 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 17 | 50% |
| Dearing et al. 2025 | 7 | 50% |
Methods
HMI pools subgroup medians for pak choi on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-pak-choi-fresh-weight:bfbf8badaa4a
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI pak choi (cooked) Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI pak choi (cooked) Cd reference (fresh weight) · 1 sampling family. Pak choi (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤5 | 23.8–26.3 (bounded) | 47.5 |
| Alternative weighting | ≤5 | 23.8–26.3 (bounded) | 47.5 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for pak choi (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-pak-choi-cooked-fresh-weight:270a50a74230
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach Cd reference (fresh weight) · 3 sampling families
Reference and scope
HMI spinach Cd reference (fresh weight) · 3 sampling families. Spinach; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 21.5 | 51 | 200 |
| Alternative weighting | 21.5 | 51 | 200 |
32 reference units; 3 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 30 | 33% |
| FDA TDS FY2018–FY2020 | 1 | 33% |
| Martin-Leon et al. 2023 | 1 | 33% |
Methods
HMI pools subgroup medians for spinach on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-spinach-fresh-weight:de6057f9c85a
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach (canned) Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI spinach (canned) Cd reference (fresh weight) · 1 sampling family. Spinach (canned); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 130 | 130 | 130 |
| Alternative weighting | 130 | 130 | 130 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for spinach (canned) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-spinach-canned-fresh-weight:41f6c744fa4a
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach (cooked) Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI spinach (cooked) Cd reference (fresh weight) · 1 sampling family. Spinach (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 65.9 | 81.7 | 131 |
| Alternative weighting | 65.9 | 81.7 | 131 |
30 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 30 | 100% |
Methods
HMI pools subgroup medians for spinach (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-spinach-cooked-fresh-weight:18f3c3d27be3
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach (frozen) Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI spinach (frozen) Cd reference (fresh weight) · 1 sampling family. Spinach (frozen); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 60.3 | 72.3 | 83 |
| Alternative weighting | 60.3 | 72.3 | 83 |
16 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 16 | 100% |
Methods
HMI pools subgroup medians for spinach (frozen) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-spinach-frozen-fresh-weight:3d61a464a0d8
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI water spinach (cooked) Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI water spinach (cooked) Cd reference (fresh weight) · 1 sampling family. Water spinach (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 34.5 | 34.5 | 34.5 |
| Alternative weighting | 34.5 | 34.5 | 34.5 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for water spinach (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-water-spinach-cooked-fresh-weight:4a2f7de14bf7
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI watercress Cd reference (fresh weight) · 2 sampling families
Reference and scope
HMI watercress Cd reference (fresh weight) · 2 sampling families. Watercress; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1.9 | 7.95 | 26 |
| Alternative weighting | 1.9 | 7.95 | 26 |
3 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 50% |
| Martin-Leon et al. 2023 | 1 | 50% |
Methods
HMI pools subgroup medians for watercress on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-watercress-fresh-weight:61308a63cf57
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI watercress (cooked) Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI watercress (cooked) Cd reference (fresh weight) · 1 sampling family. Watercress (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 35 | 39 | 43 |
| Alternative weighting | 35 | 39 | 43 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for watercress (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-watercress-cooked-fresh-weight:d499bcef8de5
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI collard greens (cooked) Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI collard greens (cooked) Cd reference (fresh weight) · 1 sampling family. Collard greens (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 22 | 22 | 22 |
| Alternative weighting | 22 | 22 | 22 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| FDA TDS FY2018–FY2020 | 1 | 100% |
Methods
HMI pools subgroup medians for collard greens (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-collard-greens-cooked-fresh-weight:312d8ecd9967
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI amaranth leaves Cd reference (dry weight) · 1 sampling family
Reference and scope
HMI amaranth leaves Cd reference (dry weight) · 1 sampling family. Amaranth leaves; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 345 | 885 | 1590 |
| Alternative weighting | 345 | 885 | 1590 |
33 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Hossain et al. 2016 | 33 | 100% |
Methods
HMI pools subgroup medians for amaranth leaves on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-amaranth-leaves-dry-weight:b5045f01cd62
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI coriander leaves Cd reference (dry weight) · 2 sampling families
Reference and scope
HMI coriander leaves Cd reference (dry weight) · 2 sampling families. Coriander leaves; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1050 | 1880 | 5150 |
| Alternative weighting | 1050 | 1880 | 5150 |
8 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 50% |
| Sultana et al. 2022 | 4 | 50% |
Methods
HMI pools subgroup medians for coriander leaves on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-coriander-leaves-dry-weight:445f056b5d07
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI lettuce Cd reference (dry weight) · 3 sampling families
Reference and scope
HMI lettuce Cd reference (dry weight) · 3 sampling families. Lettuce; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1010 | 1970 | 5380 |
| Alternative weighting | 1010 | 1970 | 5380 |
9 reference units; 3 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 33% |
| Stark et al. 2019 | 1 | 33% |
| Vela et al. 2023 | 4 | 33% |
Methods
HMI pools subgroup medians for lettuce on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-lettuce-dry-weight:f8ed81225472
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI mint Cd reference (dry weight) · 2 sampling families
Reference and scope
HMI mint Cd reference (dry weight) · 2 sampling families. Mint; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 150–160 (bounded) | 1300 | 9810 |
| Alternative weighting | 150–160 (bounded) | 1300 | 9810 |
8 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 50% |
| Sultana et al. 2022 | 4 | 50% |
Methods
HMI pools subgroup medians for mint on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-mint-dry-weight:aa9961761257
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spring onion leaves Cd reference (dry weight) · 1 sampling family
Reference and scope
HMI spring onion leaves Cd reference (dry weight) · 1 sampling family. Spring onion leaves; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 2010 | 3050 | 4850 |
| Alternative weighting | 2010 | 3050 | 4850 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 100% |
Methods
HMI pools subgroup medians for spring onion leaves on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-spring-onion-leaves-dry-weight:d2cbe1e5e81c
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI swiss chard Cd reference (dry weight) · 1 sampling family
Reference and scope
HMI swiss chard Cd reference (dry weight) · 1 sampling family. Swiss chard; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 2500 | 3430 | 4440 |
| Alternative weighting | 2500 | 3430 | 4440 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 100% |
Methods
HMI pools subgroup medians for swiss chard on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-swiss-chard-dry-weight:210425668531
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI lettuce Cd reference (fresh weight) · 5 sampling families
Reference and scope
HMI lettuce Cd reference (fresh weight) · 5 sampling families. Lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 19 | 25 | 72 |
| Alternative weighting | 19 | 25 | 67.5 |
24 reference units; 5 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Armand et al. 2026 | 10 | 20% |
| Dearing et al. 2025 | 11 | 20% |
| Romero-Crespo et al. 2023 | 1 | 20% |
| UK TDS 2014 (Fera 2015) | 1 | 20% |
| Martin-Leon et al. 2023 | 1 | 20% |
Methods
HMI pools subgroup medians for lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
- Armand et al. 2026
- Dearing et al. 2025
- Romero-Crespo et al. 2023
- UK TDS 2014 (Fera 2015)
- Martin-Leon et al. 2023
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-lettuce-fresh-weight:ce54b5f719ae
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head cabbage Cd reference (dry weight) · 2 sampling families
Reference and scope
HMI head cabbage Cd reference (dry weight) · 2 sampling families. Head cabbage; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤19.9 | 297 | 436 |
| Alternative weighting | ≤19.9 | 297 | 436 |
21 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Hossain et al. 2016 | 17 | 50% |
| Sultana et al. 2022 | 4 | 50% |
Methods
HMI pools subgroup medians for head cabbage on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-head-cabbage-dry-weight:652c26a605ea
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI parsley Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI parsley Cd reference (fresh weight) · 1 sampling family. Parsley; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 10.3 | 12.5 | 20.5 |
| Alternative weighting | 12 | 12.5 | 28 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Dearing et al. 2025 | 4 | 100% |
Methods
HMI pools subgroup medians for parsley on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-parsley-fresh-weight:71fffd09a506
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI water spinach Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI water spinach Cd reference (fresh weight) · 1 sampling family. Water spinach; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 19 | 19 | 19 |
| Alternative weighting | 19 | 19 | 19 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Dearing et al. 2025 | 1 | 100% |
Methods
HMI pools subgroup medians for water spinach on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-water-spinach-fresh-weight:446bcd3bb7c9
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach Cd reference (dry weight) · 1 sampling family
Reference and scope
HMI spinach Cd reference (dry weight) · 1 sampling family. Spinach; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 250–260 (bounded) | 550 | 750 |
| Alternative weighting | 250–260 (bounded) | 550 | 750 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Sultana et al. 2022 | 4 | 100% |
Methods
HMI pools subgroup medians for spinach on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-spinach-dry-weight:2e794ccba095
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI mustard greens Cd reference (dry weight) · 1 sampling family
Reference and scope
HMI mustard greens Cd reference (dry weight) · 1 sampling family. Mustard greens; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤19.9 | 200–210 (bounded) | 400 |
| Alternative weighting | ≤19.9 | 200–210 (bounded) | 400 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Sultana et al. 2022 | 4 | 100% |
Methods
HMI pools subgroup medians for mustard greens on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-mustard-greens-dry-weight:94691ba5a976
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI rocket Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI rocket Cd reference (fresh weight) · 1 sampling family. Rocket; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 9.3 | 9.3 | 9.3 |
| Alternative weighting | 9.3 | 9.3 | 9.3 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for rocket on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-rocket-fresh-weight:094a51d4360c
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI romaine lettuce Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI romaine lettuce Cd reference (fresh weight) · 1 sampling family. Romaine lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 11.9 | 11.9 | 11.9 |
| Alternative weighting | 11.9 | 11.9 | 11.9 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for romaine lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-romaine-lettuce-fresh-weight:30051a13e7e4
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI swiss chard Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI swiss chard Cd reference (fresh weight) · 1 sampling family. Swiss chard; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 10.5 | 10.5 | 10.5 |
| Alternative weighting | 10.5 | 10.5 | 10.5 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for swiss chard on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-swiss-chard-fresh-weight:2638b6ec4b00
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI coriander leaves Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI coriander leaves Cd reference (fresh weight) · 1 sampling family. Coriander leaves; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 550 | 550 | 550 |
| Alternative weighting | 550 | 550 | 550 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Romero-Crespo et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for coriander leaves on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-coriander-leaves-fresh-weight:19d51413bbf4
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI chives Cd reference (fresh weight) · 1 sampling family
Reference and scope
HMI chives Cd reference (fresh weight) · 1 sampling family. Chives; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1010 | 1010 | 1010 |
| Alternative weighting | 1010 | 1010 | 1010 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Romero-Crespo et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for chives on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-chives-fresh-weight:42c598e8e354
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI mallow Cd reference (dry weight) · 1 sampling family
Reference and scope
HMI mallow Cd reference (dry weight) · 1 sampling family. Mallow; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 709 | 709 | 709 |
| Alternative weighting | 709 | 709 | 709 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 1 | 100% |
Methods
HMI pools subgroup medians for mallow on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-mallow-dry-weight:72f869a272cc
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI bristly ox tongue Cd reference (dry weight) · 1 sampling family
Reference and scope
HMI bristly ox tongue Cd reference (dry weight) · 1 sampling family. Bristly ox tongue; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤383 | 430–621 (bounded) | 859 |
| Alternative weighting | ≤383 | 430–621 (bounded) | 859 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 2 | 100% |
Methods
HMI pools subgroup medians for bristly ox tongue on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-bristly-ox-tongue-dry-weight:d8f4db6b2774
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI cat's ear Cd reference (dry weight) · 1 sampling family
Reference and scope
HMI cat's ear Cd reference (dry weight) · 1 sampling family. Cat's ear; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤391 | ≤391 | ≤391 |
| Alternative weighting | ≤391 | ≤391 | ≤391 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 1 | 100% |
Methods
HMI pools subgroup medians for cat's ear on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-cat-s-ear-dry-weight:909e41c558db
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI dandelion greens Cd reference (dry weight) · 1 sampling family
Reference and scope
HMI dandelion greens Cd reference (dry weight) · 1 sampling family. Dandelion greens; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤400 | ≤400 | ≤400 |
| Alternative weighting | ≤400 | ≤400 | ≤400 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 1 | 100% |
Methods
HMI pools subgroup medians for dandelion greens on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-dandelion-greens-dry-weight:a669edf8d83e
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI english plantain Cd reference (dry weight) · 1 sampling family
Reference and scope
HMI english plantain Cd reference (dry weight) · 1 sampling family. English plantain; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 500 | 503 | 506 |
| Alternative weighting | 500 | 503 | 506 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 2 | 100% |
Methods
HMI pools subgroup medians for english plantain on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-english-plantain-dry-weight:fb2e89d154d3
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI nasturtium (tropaeolum) Cd reference (dry weight) · 1 sampling family
Reference and scope
HMI nasturtium (tropaeolum) Cd reference (dry weight) · 1 sampling family. Nasturtium (Tropaeolum); Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤383 | ≤383 | ≤383 |
| Alternative weighting | ≤383 | ≤383 | ≤383 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 1 | 100% |
Methods
HMI pools subgroup medians for nasturtium (tropaeolum) on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-nasturtium-tropaeolum-dry-weight:12db6e9b4cfe
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI sweet fennel Cd reference (dry weight) · 1 sampling family
Reference and scope
HMI sweet fennel Cd reference (dry weight) · 1 sampling family. Sweet fennel; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤392 | ≤396 | ≤399 |
| Alternative weighting | ≤392 | ≤396 | ≤399 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 2 | 100% |
Methods
HMI pools subgroup medians for sweet fennel on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cd-sweet-fennel-dry-weight:3f7009ff2b28
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI chinese cabbage tAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI chinese cabbage tAs reference (fresh weight) · 1 sampling family. Chinese cabbage; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤2.3 | 1.23–9 (bounded) | 1.9–11.6 (bounded) |
| Alternative weighting | ≤2.3 | 1.23–9 (bounded) | 1.9–11.6 (bounded) |
9 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 9 | 100% |
Methods
HMI pools subgroup medians for chinese cabbage on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-chinese-cabbage-fresh-weight:f1e3e98aa8ea
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI chinese cabbage (fermented) tAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI chinese cabbage (fermented) tAs reference (fresh weight) · 1 sampling family. Chinese cabbage (fermented); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 6.2 | 6.2 | 6.2 |
| Alternative weighting | 6.2 | 6.2 | 6.2 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for chinese cabbage (fermented) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-chinese-cabbage-fermented-fresh-weight:7a1aa8490303
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI chinese cabbage (frozen) tAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI chinese cabbage (frozen) tAs reference (fresh weight) · 1 sampling family. Chinese cabbage (frozen); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1.19 | 1.86 | 2.54 |
| Alternative weighting | 1.19 | 1.86 | 2.54 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for chinese cabbage (frozen) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-chinese-cabbage-frozen-fresh-weight:b9070103a68a
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head cabbage tAs reference (fresh weight) · 2 sampling families
Reference and scope
HMI head cabbage tAs reference (fresh weight) · 2 sampling families. Head cabbage; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤3 | ≤3 | 1.03–5.65 (bounded) |
| Alternative weighting | ≤3 | ≤3 | 1.03–5.65 (bounded) |
31 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 30 | 50% |
| FDA TDS FY2018–FY2020 | 1 | 50% |
Methods
HMI pools subgroup medians for head cabbage on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-head-cabbage-fresh-weight:2cadcea5e14f
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head cabbage (cooked) tAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI head cabbage (cooked) tAs reference (fresh weight) · 1 sampling family. Head cabbage (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤1.25 | 0.816–1.25 (bounded) | 1.84–2.16 (bounded) |
| Alternative weighting | ≤1.25 | 0.816–1.25 (bounded) | 1.84–2.16 (bounded) |
16 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 16 | 100% |
Methods
HMI pools subgroup medians for head cabbage (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-head-cabbage-cooked-fresh-weight:a76633b138aa
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head cabbage (dried) tAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI head cabbage (dried) tAs reference (fresh weight) · 1 sampling family. Head cabbage (dried); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 11.7 | 11.7 | 11.7 |
| Alternative weighting | 11.7 | 11.7 | 11.7 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for head cabbage (dried) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-head-cabbage-dried-fresh-weight:4c46c69800be
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head cabbage (fermented) tAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI head cabbage (fermented) tAs reference (fresh weight) · 1 sampling family. Head cabbage (fermented); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤1.1 | ≤1.25 | 0.356–1.25 (bounded) |
| Alternative weighting | ≤1.1 | ≤1.25 | 0.356–1.25 (bounded) |
8 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 8 | 100% |
Methods
HMI pools subgroup medians for head cabbage (fermented) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-head-cabbage-fermented-fresh-weight:0deedcbf52e3
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head lettuce tAs reference (fresh weight) · 3 sampling families
Reference and scope
HMI head lettuce tAs reference (fresh weight) · 3 sampling families. Head lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤0.5 | ≤3 | 0.5–5 (bounded) |
| Alternative weighting | ≤0.5 | ≤3 | 0.5–5 (bounded) |
26 reference units; 3 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 24 | 33% |
| FDA TDS FY2018–FY2020 | 1 | 33% |
| Martin-Leon et al. 2023 | 1 | 33% |
Methods
HMI pools subgroup medians for head lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-head-lettuce-fresh-weight:af658dbb8a81
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI kale tAs reference (fresh weight) · 2 sampling families
Reference and scope
HMI kale tAs reference (fresh weight) · 2 sampling families. Kale; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤10 | ≤10 | 3.12–14 (bounded) |
| Alternative weighting | ≤10 | ≤10 | 10–14 (bounded) |
58 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 20 | 50% |
| Kim et al. 2024 | 38 | 50% |
Methods
HMI pools subgroup medians for kale on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-kale-fresh-weight:47196977033a
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI kale (cooked) tAs reference (fresh weight) · 2 sampling families
Reference and scope
HMI kale (cooked) tAs reference (fresh weight) · 2 sampling families. Kale (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1.51–1.66 (bounded) | 5.54 | 5.9 |
| Alternative weighting | 1.51–1.66 (bounded) | 5.54 | 5.9 |
16 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 15 | 50% |
| FDA TDS FY2018–FY2020 | 1 | 50% |
Methods
HMI pools subgroup medians for kale (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-kale-cooked-fresh-weight:d4d857df376c
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI kale (frozen) tAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI kale (frozen) tAs reference (fresh weight) · 1 sampling family. Kale (frozen); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤20 | ≤20 | ≤20 |
| Alternative weighting | ≤20 | ≤20 | ≤20 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for kale (frozen) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-kale-frozen-fresh-weight:50a9079aba0d
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI lamb's lettuce tAs reference (fresh weight) · 2 sampling families
Reference and scope
HMI lamb's lettuce tAs reference (fresh weight) · 2 sampling families. Lamb's lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 18.9 | 18.9 | 24 |
| Alternative weighting | 18.9 | 18.9 | 24 |
18 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 17 | 50% |
| Martin-Leon et al. 2023 | 1 | 50% |
Methods
HMI pools subgroup medians for lamb's lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-lamb-s-lettuce-fresh-weight:8d0878ecc198
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leaf lettuce tAs reference (fresh weight) · 2 sampling families
Reference and scope
HMI leaf lettuce tAs reference (fresh weight) · 2 sampling families. Leaf lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤3 | ≤3 | ≤9 |
| Alternative weighting | ≤3 | ≤3 | ≤9 |
26 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 25 | 50% |
| FDA TDS FY2018–FY2020 | 1 | 50% |
Methods
HMI pools subgroup medians for leaf lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-leaf-lettuce-fresh-weight:b1ea3a6bb94a
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leafy vegetables, other or unspecified tAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI leafy vegetables, other or unspecified tAs reference (fresh weight) · 1 sampling family. Leafy vegetables, other or unspecified; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤10 | ≤13.2 | 13.2–22.5 (bounded) |
| Alternative weighting | ≤10 | ≤13.2 | 13.2–22.5 (bounded) |
27 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 27 | 100% |
Methods
HMI pools subgroup medians for leafy vegetables, other or unspecified on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-leafy-vegetables-other-or-unspecified-fresh-weight:63efc42b6a94
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leafy vegetables, other or unspecified (cooked) tAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI leafy vegetables, other or unspecified (cooked) tAs reference (fresh weight) · 1 sampling family. Leafy vegetables, other or unspecified (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 3.5 | 3.5 | 3.5 |
| Alternative weighting | 3.5 | 3.5 | 3.5 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for leafy vegetables, other or unspecified (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-leafy-vegetables-other-or-unspecified-cooked-fresh-weight:2039fdf5f87d
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leafy vegetables, other or unspecified (dried) tAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI leafy vegetables, other or unspecified (dried) tAs reference (fresh weight) · 1 sampling family. Leafy vegetables, other or unspecified (dried); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 300 | 1000 | 1700 |
| Alternative weighting | 300 | 1000 | 1700 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for leafy vegetables, other or unspecified (dried) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-leafy-vegetables-other-or-unspecified-dried-fresh-weight:e819bedd0975
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leafy vegetables, other or unspecified (frozen) tAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI leafy vegetables, other or unspecified (frozen) tAs reference (fresh weight) · 1 sampling family. Leafy vegetables, other or unspecified (frozen); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1.92 | 2.94 | 1400 |
| Alternative weighting | 1.92 | 2.94 | 1400 |
3 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 3 | 100% |
Methods
HMI pools subgroup medians for leafy vegetables, other or unspecified (frozen) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-leafy-vegetables-other-or-unspecified-frozen-fresh-weight:29cf89c48c68
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI mustard greens tAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI mustard greens tAs reference (fresh weight) · 1 sampling family. Mustard greens; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 65–85 (bounded) | 65–85 (bounded) | 65–85 (bounded) |
| Alternative weighting | 65–85 (bounded) | 65–85 (bounded) | 65–85 (bounded) |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for mustard greens on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-mustard-greens-fresh-weight:2856f6daf008
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI pak choi tAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI pak choi tAs reference (fresh weight) · 1 sampling family. Pak choi; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤2 | 1.35–4.3 (bounded) | 4.33–7.85 (bounded) |
| Alternative weighting | ≤2 | 1.35–4.3 (bounded) | 4.33–7.85 (bounded) |
18 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 18 | 100% |
Methods
HMI pools subgroup medians for pak choi on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-pak-choi-fresh-weight:8c573b180d69
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI pak choi (cooked) tAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI pak choi (cooked) tAs reference (fresh weight) · 1 sampling family. Pak choi (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤1.25 | ≤1.25 | ≤5 |
| Alternative weighting | ≤1.25 | ≤1.25 | ≤5 |
3 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 3 | 100% |
Methods
HMI pools subgroup medians for pak choi (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-pak-choi-cooked-fresh-weight:24b17fdc30d4
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach tAs reference (fresh weight) · 3 sampling families
Reference and scope
HMI spinach tAs reference (fresh weight) · 3 sampling families. Spinach; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 5–5.9 (bounded) | 5.9–8 (bounded) | 8–10 (bounded) |
| Alternative weighting | 5–5.9 (bounded) | 5.9–8 (bounded) | 8–10 (bounded) |
28 reference units; 3 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 26 | 33% |
| FDA TDS FY2018–FY2020 | 1 | 33% |
| Martin-Leon et al. 2023 | 1 | 33% |
Methods
HMI pools subgroup medians for spinach on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-spinach-fresh-weight:1b2d2edf5ce9
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach (cooked) tAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI spinach (cooked) tAs reference (fresh weight) · 1 sampling family. Spinach (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 6.37 | 8.19 | 9.26 |
| Alternative weighting | 6.37 | 8.19 | 9.26 |
22 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 22 | 100% |
Methods
HMI pools subgroup medians for spinach (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-spinach-cooked-fresh-weight:03d2f1d08c36
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach (frozen) tAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI spinach (frozen) tAs reference (fresh weight) · 1 sampling family. Spinach (frozen); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤6 | ≤10 | ≤20 |
| Alternative weighting | ≤6 | ≤10 | ≤20 |
9 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 9 | 100% |
Methods
HMI pools subgroup medians for spinach (frozen) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-spinach-frozen-fresh-weight:9c89b41d99d4
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI watercress tAs reference (fresh weight) · 2 sampling families
Reference and scope
HMI watercress tAs reference (fresh weight) · 2 sampling families. Watercress; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤6.7 | 3.35–13.3 (bounded) | 6.7–30 (bounded) |
| Alternative weighting | ≤6.7 | 3.35–13.3 (bounded) | 6.7–30 (bounded) |
3 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 50% |
| Martin-Leon et al. 2023 | 1 | 50% |
Methods
HMI pools subgroup medians for watercress on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-watercress-fresh-weight:0ce1753abd49
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI collard greens (cooked) tAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI collard greens (cooked) tAs reference (fresh weight) · 1 sampling family. Collard greens (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤3 | ≤3 | ≤3 |
| Alternative weighting | ≤3 | ≤3 | ≤3 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| FDA TDS FY2018–FY2020 | 1 | 100% |
Methods
HMI pools subgroup medians for collard greens (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-collard-greens-cooked-fresh-weight:a14f4c09147b
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI coriander leaves tAs reference (dry weight) · 1 sampling family
Reference and scope
HMI coriander leaves tAs reference (dry weight) · 1 sampling family. Coriander leaves; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 331 | 827 | 6410 |
| Alternative weighting | 331 | 827 | 6410 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 100% |
Methods
HMI pools subgroup medians for coriander leaves on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-coriander-leaves-dry-weight:e5fe687c21e6
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI lettuce tAs reference (dry weight) · 2 sampling families
Reference and scope
HMI lettuce tAs reference (dry weight) · 2 sampling families. Lettuce; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 222 | 971 | 1660 |
| Alternative weighting | 222 | 971 | 1660 |
5 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 50% |
| Stark et al. 2019 | 1 | 50% |
Methods
HMI pools subgroup medians for lettuce on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-lettuce-dry-weight:d692da526ee9
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI mint tAs reference (dry weight) · 1 sampling family
Reference and scope
HMI mint tAs reference (dry weight) · 1 sampling family. Mint; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 468 | 524 | 629 |
| Alternative weighting | 468 | 524 | 629 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 100% |
Methods
HMI pools subgroup medians for mint on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-mint-dry-weight:ff5570b62853
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spring onion leaves tAs reference (dry weight) · 1 sampling family
Reference and scope
HMI spring onion leaves tAs reference (dry weight) · 1 sampling family. Spring onion leaves; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 314 | 325 | 368 |
| Alternative weighting | 314 | 325 | 368 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 100% |
Methods
HMI pools subgroup medians for spring onion leaves on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-spring-onion-leaves-dry-weight:6cf7e04b26bf
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI swiss chard tAs reference (dry weight) · 1 sampling family
Reference and scope
HMI swiss chard tAs reference (dry weight) · 1 sampling family. Swiss chard; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 319 | 383 | 490 |
| Alternative weighting | 319 | 383 | 490 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 100% |
Methods
HMI pools subgroup medians for swiss chard on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-swiss-chard-dry-weight:c57f05497632
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI lettuce tAs reference (fresh weight) · 2 sampling families
Reference and scope
HMI lettuce tAs reference (fresh weight) · 2 sampling families. Lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 3.9 | 3.95 | 4 |
| Alternative weighting | 3.9 | 3.95 | 4 |
2 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| UK TDS 2014 (Fera 2015) | 1 | 50% |
| Martin-Leon et al. 2023 | 1 | 50% |
Methods
HMI pools subgroup medians for lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-lettuce-fresh-weight:77c3dcc12d2f
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI rocket tAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI rocket tAs reference (fresh weight) · 1 sampling family. Rocket; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 7.2 | 7.2 | 7.2 |
| Alternative weighting | 7.2 | 7.2 | 7.2 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for rocket on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-rocket-fresh-weight:482032dc9ca8
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI romaine lettuce tAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI romaine lettuce tAs reference (fresh weight) · 1 sampling family. Romaine lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1.6 | 1.6 | 1.6 |
| Alternative weighting | 1.6 | 1.6 | 1.6 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for romaine lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-romaine-lettuce-fresh-weight:1ffd51693818
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI swiss chard tAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI swiss chard tAs reference (fresh weight) · 1 sampling family. Swiss chard; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 4.9 | 4.9 | 4.9 |
| Alternative weighting | 4.9 | 4.9 | 4.9 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for swiss chard on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-swiss-chard-fresh-weight:f7de790ecca1
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI coriander leaves tAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI coriander leaves tAs reference (fresh weight) · 1 sampling family. Coriander leaves; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1250 | 1250 | 1250 |
| Alternative weighting | 1250 | 1250 | 1250 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Romero-Crespo et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for coriander leaves on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-coriander-leaves-fresh-weight:53c64d5be2a7
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI mallow tAs reference (dry weight) · 1 sampling family
Reference and scope
HMI mallow tAs reference (dry weight) · 1 sampling family. Mallow; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1710 | 1710 | 1710 |
| Alternative weighting | 1710 | 1710 | 1710 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 1 | 100% |
Methods
HMI pools subgroup medians for mallow on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-mallow-dry-weight:21bfe0ff3aca
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI bristly ox tongue tAs reference (dry weight) · 1 sampling family
Reference and scope
HMI bristly ox tongue tAs reference (dry weight) · 1 sampling family. Bristly ox tongue; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1660 | 1680 | 1690 |
| Alternative weighting | 1660 | 1680 | 1690 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 2 | 100% |
Methods
HMI pools subgroup medians for bristly ox tongue on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-bristly-ox-tongue-dry-weight:20765a763f54
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI cat's ear tAs reference (dry weight) · 1 sampling family
Reference and scope
HMI cat's ear tAs reference (dry weight) · 1 sampling family. Cat's ear; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1680 | 1680 | 1680 |
| Alternative weighting | 1680 | 1680 | 1680 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 1 | 100% |
Methods
HMI pools subgroup medians for cat's ear on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-cat-s-ear-dry-weight:73a675eacb26
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI dandelion greens tAs reference (dry weight) · 1 sampling family
Reference and scope
HMI dandelion greens tAs reference (dry weight) · 1 sampling family. Dandelion greens; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1690 | 1690 | 1690 |
| Alternative weighting | 1690 | 1690 | 1690 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 1 | 100% |
Methods
HMI pools subgroup medians for dandelion greens on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-dandelion-greens-dry-weight:2db45cb9a363
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI english plantain tAs reference (dry weight) · 1 sampling family
Reference and scope
HMI english plantain tAs reference (dry weight) · 1 sampling family. English plantain; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1520 | 1570 | 1610 |
| Alternative weighting | 1520 | 1570 | 1610 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 2 | 100% |
Methods
HMI pools subgroup medians for english plantain on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-english-plantain-dry-weight:61e265c65cf6
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI nasturtium (tropaeolum) tAs reference (dry weight) · 1 sampling family
Reference and scope
HMI nasturtium (tropaeolum) tAs reference (dry weight) · 1 sampling family. Nasturtium (Tropaeolum); Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1620 | 1620 | 1620 |
| Alternative weighting | 1620 | 1620 | 1620 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 1 | 100% |
Methods
HMI pools subgroup medians for nasturtium (tropaeolum) on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-nasturtium-tropaeolum-dry-weight:92085edb40e5
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI sweet fennel tAs reference (dry weight) · 1 sampling family
Reference and scope
HMI sweet fennel tAs reference (dry weight) · 1 sampling family. Sweet fennel; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1570 | 1580 | 1590 |
| Alternative weighting | 1570 | 1580 | 1590 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 2 | 100% |
Methods
HMI pools subgroup medians for sweet fennel on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-tas-sweet-fennel-dry-weight:b505577e3ff5
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI amaranth leaves (cooked) iAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI amaranth leaves (cooked) iAs reference (fresh weight) · 1 sampling family. Amaranth leaves (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤9.5 | 4.75–14.8 (bounded) | 9.5–20 (bounded) |
| Alternative weighting | ≤9.5 | 4.75–14.8 (bounded) | 9.5–20 (bounded) |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for amaranth leaves (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-amaranth-leaves-cooked-fresh-weight:80e5bfa919ed
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI chinese cabbage (cooked) iAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI chinese cabbage (cooked) iAs reference (fresh weight) · 1 sampling family. Chinese cabbage (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤3 | ≤3 | ≤3 |
| Alternative weighting | ≤3 | ≤3 | ≤3 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for chinese cabbage (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-chinese-cabbage-cooked-fresh-weight:3e235193c957
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head cabbage iAs reference (dry weight) · 1 sampling family
Reference and scope
HMI head cabbage iAs reference (dry weight) · 1 sampling family. Head cabbage; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 39 | 39 | 39 |
| Alternative weighting | 39 | 39 | 39 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for head cabbage on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-head-cabbage-dry-weight:c5cb3a0fce16
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head cabbage iAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI head cabbage iAs reference (fresh weight) · 1 sampling family. Head cabbage; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤10 | ≤39 | 39–40 (bounded) |
| Alternative weighting | ≤10 | ≤39 | 39–40 (bounded) |
3 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 3 | 100% |
Methods
HMI pools subgroup medians for head cabbage on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-head-cabbage-fresh-weight:aadebf70434b
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head cabbage (canned) iAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI head cabbage (canned) iAs reference (fresh weight) · 1 sampling family. Head cabbage (canned); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 30 | 30 | 30 |
| Alternative weighting | 30 | 30 | 30 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for head cabbage (canned) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-head-cabbage-canned-fresh-weight:fd92759de508
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head cabbage (cooked) iAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI head cabbage (cooked) iAs reference (fresh weight) · 1 sampling family. Head cabbage (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤3 | ≤3 | ≤3 |
| Alternative weighting | ≤3 | ≤3 | ≤3 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for head cabbage (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-head-cabbage-cooked-fresh-weight:605967ffbbd6
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head lettuce (cooked) iAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI head lettuce (cooked) iAs reference (fresh weight) · 1 sampling family. Head lettuce (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤3 | ≤3 | ≤3 |
| Alternative weighting | ≤3 | ≤3 | ≤3 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for head lettuce (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-head-lettuce-cooked-fresh-weight:cc67fff359cf
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI kale iAs reference (dry weight) · 1 sampling family
Reference and scope
HMI kale iAs reference (dry weight) · 1 sampling family. Kale; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 103 | 103 | 103 |
| Alternative weighting | 103 | 103 | 103 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for kale on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-kale-dry-weight:92226493f5d9
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI kale iAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI kale iAs reference (fresh weight) · 1 sampling family. Kale; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤10 | 51.6–56.6 (bounded) | 103 |
| Alternative weighting | ≤10 | 51.6–56.6 (bounded) | 103 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for kale on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-kale-fresh-weight:9709994b12f3
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI kale (cooked) iAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI kale (cooked) iAs reference (fresh weight) · 1 sampling family. Kale (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤3 | 1.5–3 (bounded) | 3 |
| Alternative weighting | ≤3 | 1.5–3 (bounded) | 3 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for kale (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-kale-cooked-fresh-weight:8002c6931761
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leaf lettuce iAs reference (dry weight) · 1 sampling family
Reference and scope
HMI leaf lettuce iAs reference (dry weight) · 1 sampling family. Leaf lettuce; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 104 | 104 | 104 |
| Alternative weighting | 104 | 104 | 104 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for leaf lettuce on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-leaf-lettuce-dry-weight:6277fb09ba8c
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leaf lettuce iAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI leaf lettuce iAs reference (fresh weight) · 1 sampling family. Leaf lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 0.8–10.8 (bounded) | 18.8–28 (bounded) | 69.9 |
| Alternative weighting | 0.8–10.8 (bounded) | 18.8–28 (bounded) | 69.9 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 4 | 100% |
Methods
HMI pools subgroup medians for leaf lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-leaf-lettuce-fresh-weight:1b1c7c7c4b94
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leaf lettuce (cooked) iAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI leaf lettuce (cooked) iAs reference (fresh weight) · 1 sampling family. Leaf lettuce (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤3 | 1.5–3 (bounded) | 3 |
| Alternative weighting | ≤3 | 1.5–3 (bounded) | 3 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for leaf lettuce (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-leaf-lettuce-cooked-fresh-weight:1c0442e980e2
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leafy vegetables, other or unspecified iAs reference (dry weight) · 1 sampling family
Reference and scope
HMI leafy vegetables, other or unspecified iAs reference (dry weight) · 1 sampling family. Leafy vegetables, other or unspecified; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 257 | 257 | 257 |
| Alternative weighting | 257 | 257 | 257 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for leafy vegetables, other or unspecified on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-leafy-vegetables-other-or-unspecified-dry-weight:eb9cccb30a58
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leafy vegetables, other or unspecified iAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI leafy vegetables, other or unspecified iAs reference (fresh weight) · 1 sampling family. Leafy vegetables, other or unspecified; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤10 | ≤17.1 | 99.8–115 (bounded) |
| Alternative weighting | ≤10 | ≤17.1 | 99.8–115 (bounded) |
8 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 8 | 100% |
Methods
HMI pools subgroup medians for leafy vegetables, other or unspecified on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-leafy-vegetables-other-or-unspecified-fresh-weight:248de661c4e6
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leafy vegetables, other or unspecified (cooked) iAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI leafy vegetables, other or unspecified (cooked) iAs reference (fresh weight) · 1 sampling family. Leafy vegetables, other or unspecified (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 6 | 10.5 | 15 |
| Alternative weighting | 6 | 10.5 | 15 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for leafy vegetables, other or unspecified (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-leafy-vegetables-other-or-unspecified-cooked-fresh-weight:78039ba9f875
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leafy vegetables, other or unspecified (dried) iAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI leafy vegetables, other or unspecified (dried) iAs reference (fresh weight) · 1 sampling family. Leafy vegetables, other or unspecified (dried); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤110 | ≤110 | ≤110 |
| Alternative weighting | ≤110 | ≤110 | ≤110 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for leafy vegetables, other or unspecified (dried) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-leafy-vegetables-other-or-unspecified-dried-fresh-weight:3a17c41eaed2
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leafy vegetables, other or unspecified (frozen) iAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI leafy vegetables, other or unspecified (frozen) iAs reference (fresh weight) · 1 sampling family. Leafy vegetables, other or unspecified (frozen); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 310 | 310 | 310 |
| Alternative weighting | 310 | 310 | 310 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for leafy vegetables, other or unspecified (frozen) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-leafy-vegetables-other-or-unspecified-frozen-fresh-weight:55970c42953f
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI mustard greens (cooked) iAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI mustard greens (cooked) iAs reference (fresh weight) · 1 sampling family. Mustard greens (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤3 | 3.5–5 (bounded) | 7 |
| Alternative weighting | ≤3 | 3.5–5 (bounded) | 7 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for mustard greens (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-mustard-greens-cooked-fresh-weight:7aa6a5c5b0c9
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI pak choi iAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI pak choi iAs reference (fresh weight) · 1 sampling family. Pak choi; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤20 | ≤20 | ≤20 |
| Alternative weighting | ≤20 | ≤20 | ≤20 |
3 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 3 | 100% |
Methods
HMI pools subgroup medians for pak choi on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-pak-choi-fresh-weight:41b3afac2912
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI pak choi (cooked) iAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI pak choi (cooked) iAs reference (fresh weight) · 1 sampling family. Pak choi (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 4.5 | 4.5 | 4.5 |
| Alternative weighting | 4.5 | 4.5 | 4.5 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for pak choi (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-pak-choi-cooked-fresh-weight:14c22bb6f1c9
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach iAs reference (dry weight) · 1 sampling family
Reference and scope
HMI spinach iAs reference (dry weight) · 1 sampling family. Spinach; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 291 | 291 | 291 |
| Alternative weighting | 291 | 291 | 291 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for spinach on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-spinach-dry-weight:e433bb8e877d
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach iAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI spinach iAs reference (fresh weight) · 1 sampling family. Spinach; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤13 | 145–155 (bounded) | 530 |
| Alternative weighting | ≤13 | 145–155 (bounded) | 530 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 4 | 100% |
Methods
HMI pools subgroup medians for spinach on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-spinach-fresh-weight:aa95cf502cf4
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach (cooked) iAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI spinach (cooked) iAs reference (fresh weight) · 1 sampling family. Spinach (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤5 | 5 | 5–20 (bounded) |
| Alternative weighting | ≤5 | 5 | 5–20 (bounded) |
3 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 3 | 100% |
Methods
HMI pools subgroup medians for spinach (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-spinach-cooked-fresh-weight:822a0883926c
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach (frozen) iAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI spinach (frozen) iAs reference (fresh weight) · 1 sampling family. Spinach (frozen); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤6 | ≤6 | ≤6 |
| Alternative weighting | ≤6 | ≤6 | ≤6 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for spinach (frozen) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-spinach-frozen-fresh-weight:ca6250b0b6ba
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI water spinach iAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI water spinach iAs reference (fresh weight) · 1 sampling family. Water spinach; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 50 | 50 | 50 |
| Alternative weighting | 50 | 50 | 50 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for water spinach on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-water-spinach-fresh-weight:23114dbb6949
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI water spinach (cooked) iAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI water spinach (cooked) iAs reference (fresh weight) · 1 sampling family. Water spinach (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤20 | 35.3–45.3 (bounded) | 70.5 |
| Alternative weighting | ≤20 | 35.3–45.3 (bounded) | 70.5 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for water spinach (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-water-spinach-cooked-fresh-weight:aeeebe5d284b
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI watercress (cooked) iAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI watercress (cooked) iAs reference (fresh weight) · 1 sampling family. Watercress (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 9 | 21.5 | 34 |
| Alternative weighting | 9 | 21.5 | 34 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for watercress (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-watercress-cooked-fresh-weight:877ceb379bd7
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI lettuce iAs reference (fresh weight) · 1 sampling family
Reference and scope
HMI lettuce iAs reference (fresh weight) · 1 sampling family. Lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤12 | ≤12 | ≤12 |
| Alternative weighting | ≤12 | ≤12 | ≤12 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| UK TDS 2014 (Fera 2015) | 1 | 100% |
Methods
HMI pools subgroup medians for lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ias-lettuce-fresh-weight:08aadab63717
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leaf lettuce DMA reference (fresh weight) · 1 sampling family
Reference and scope
HMI leaf lettuce DMA reference (fresh weight) · 1 sampling family. Leaf lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤20 | ≤20 | ≤20 |
| Alternative weighting | ≤20 | ≤20 | ≤20 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for leaf lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-dma-leaf-lettuce-fresh-weight:adb5f8bfca03
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head cabbage tHg reference (fresh weight) · 2 sampling families
Reference and scope
HMI head cabbage tHg reference (fresh weight) · 2 sampling families. Head cabbage; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤0.3 | ≤1 | ≤1 |
| Alternative weighting | ≤0.3 | ≤1 | ≤1 |
19 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 18 | 50% |
| FDA TDS FY2018–FY2020 | 1 | 50% |
Methods
HMI pools subgroup medians for head cabbage on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-head-cabbage-fresh-weight:ce0a4031f8a9
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head cabbage (cooked) tHg reference (fresh weight) · 1 sampling family
Reference and scope
HMI head cabbage (cooked) tHg reference (fresh weight) · 1 sampling family. Head cabbage (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤4.2 | ≤4.2 | ≤4.2 |
| Alternative weighting | ≤4.2 | ≤4.2 | ≤4.2 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for head cabbage (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-head-cabbage-cooked-fresh-weight:c515c97d6ade
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head cabbage (fermented) tHg reference (fresh weight) · 1 sampling family
Reference and scope
HMI head cabbage (fermented) tHg reference (fresh weight) · 1 sampling family. Head cabbage (fermented); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤0.1 | ≤0.105 | 0.16–0.65 (bounded) |
| Alternative weighting | ≤0.1 | ≤0.105 | 0.16–0.65 (bounded) |
15 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 15 | 100% |
Methods
HMI pools subgroup medians for head cabbage (fermented) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-head-cabbage-fermented-fresh-weight:cf449374e58b
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head lettuce tHg reference (fresh weight) · 3 sampling families
Reference and scope
HMI head lettuce tHg reference (fresh weight) · 3 sampling families. Head lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤1 | ≤2.23 | 2.3 |
| Alternative weighting | ≤1 | ≤2.23 | 2.3 |
10 reference units; 3 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 8 | 33% |
| FDA TDS FY2018–FY2020 | 1 | 33% |
| Martin-Leon et al. 2023 | 1 | 33% |
Methods
HMI pools subgroup medians for head lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-head-lettuce-fresh-weight:464f2310cc47
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI kale tHg reference (fresh weight) · 1 sampling family
Reference and scope
HMI kale tHg reference (fresh weight) · 1 sampling family. Kale; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤0.3 | 0.05–2 (bounded) | 0.87–4 (bounded) |
| Alternative weighting | ≤0.3 | 0.05–2 (bounded) | 0.87–4 (bounded) |
13 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 13 | 100% |
Methods
HMI pools subgroup medians for kale on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-kale-fresh-weight:72698f6dc232
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI lamb's lettuce tHg reference (fresh weight) · 2 sampling families
Reference and scope
HMI lamb's lettuce tHg reference (fresh weight) · 2 sampling families. Lamb's lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤1.9 | 1.1–2.2 (bounded) | 2.2–2.6 (bounded) |
| Alternative weighting | ≤1.9 | 1.1–2.2 (bounded) | 2.2–2.6 (bounded) |
9 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 8 | 50% |
| Martin-Leon et al. 2023 | 1 | 50% |
Methods
HMI pools subgroup medians for lamb's lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-lamb-s-lettuce-fresh-weight:837e3e027eef
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leaf lettuce tHg reference (fresh weight) · 2 sampling families
Reference and scope
HMI leaf lettuce tHg reference (fresh weight) · 2 sampling families. Leaf lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤1 | ≤1 | ≤1.2 |
| Alternative weighting | ≤1 | ≤1 | ≤1.2 |
18 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 17 | 50% |
| FDA TDS FY2018–FY2020 | 1 | 50% |
Methods
HMI pools subgroup medians for leaf lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-leaf-lettuce-fresh-weight:7ba09346ed6a
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leafy vegetables, other or unspecified tHg reference (fresh weight) · 1 sampling family
Reference and scope
HMI leafy vegetables, other or unspecified tHg reference (fresh weight) · 1 sampling family. Leafy vegetables, other or unspecified; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤5 | ≤6 | ≤6.13 |
| Alternative weighting | ≤5 | ≤6 | ≤6.13 |
13 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 13 | 100% |
Methods
HMI pools subgroup medians for leafy vegetables, other or unspecified on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-leafy-vegetables-other-or-unspecified-fresh-weight:63c2a284e688
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leafy vegetables, other or unspecified (cooked) tHg reference (fresh weight) · 1 sampling family
Reference and scope
HMI leafy vegetables, other or unspecified (cooked) tHg reference (fresh weight) · 1 sampling family. Leafy vegetables, other or unspecified (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤4.2 | ≤4.2 | ≤4.2 |
| Alternative weighting | ≤4.2 | ≤4.2 | ≤4.2 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for leafy vegetables, other or unspecified (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-leafy-vegetables-other-or-unspecified-cooked-fresh-weight:8b863fb0bdcd
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI mustard greens tHg reference (fresh weight) · 1 sampling family
Reference and scope
HMI mustard greens tHg reference (fresh weight) · 1 sampling family. Mustard greens; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤10 | 5–30 (bounded) | 10–50 (bounded) |
| Alternative weighting | ≤10 | 5–30 (bounded) | 10–50 (bounded) |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for mustard greens on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-mustard-greens-fresh-weight:0f65fe7e2dad
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI pak choi tHg reference (fresh weight) · 1 sampling family
Reference and scope
HMI pak choi tHg reference (fresh weight) · 1 sampling family. Pak choi; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤0.1 | ≤0.2 | ≤2 |
| Alternative weighting | ≤0.1 | ≤0.2 | ≤2 |
14 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 14 | 100% |
Methods
HMI pools subgroup medians for pak choi on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-pak-choi-fresh-weight:e306382f35df
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI pak choi (cooked) tHg reference (fresh weight) · 1 sampling family
Reference and scope
HMI pak choi (cooked) tHg reference (fresh weight) · 1 sampling family. Pak choi (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤5 | ≤5 | ≤5 |
| Alternative weighting | ≤5 | ≤5 | ≤5 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for pak choi (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-pak-choi-cooked-fresh-weight:ede833d34fdf
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach tHg reference (fresh weight) · 3 sampling families
Reference and scope
HMI spinach tHg reference (fresh weight) · 3 sampling families. Spinach; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤1 | 0.217–1.5 (bounded) | 1.7 |
| Alternative weighting | ≤1 | 0.217–1.5 (bounded) | 1.7 |
18 reference units; 3 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 16 | 33% |
| FDA TDS FY2018–FY2020 | 1 | 33% |
| Martin-Leon et al. 2023 | 1 | 33% |
Methods
HMI pools subgroup medians for spinach on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-spinach-fresh-weight:cedf06558716
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach (cooked) tHg reference (fresh weight) · 1 sampling family
Reference and scope
HMI spinach (cooked) tHg reference (fresh weight) · 1 sampling family. Spinach (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤5 | ≤5 | ≤5 |
| Alternative weighting | ≤5 | ≤5 | ≤5 |
3 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 3 | 100% |
Methods
HMI pools subgroup medians for spinach (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-spinach-cooked-fresh-weight:d32228320f62
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI watercress tHg reference (fresh weight) · 2 sampling families
Reference and scope
HMI watercress tHg reference (fresh weight) · 2 sampling families. Watercress; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤1 | 0.5–1 (bounded) | 1 |
| Alternative weighting | ≤1 | 0.5–1 (bounded) | 1 |
2 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 50% |
| Martin-Leon et al. 2023 | 1 | 50% |
Methods
HMI pools subgroup medians for watercress on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-watercress-fresh-weight:cbde7a19a7a6
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI collard greens (cooked) tHg reference (fresh weight) · 1 sampling family
Reference and scope
HMI collard greens (cooked) tHg reference (fresh weight) · 1 sampling family. Collard greens (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤1 | ≤1 | ≤1 |
| Alternative weighting | ≤1 | ≤1 | ≤1 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| FDA TDS FY2018–FY2020 | 1 | 100% |
Methods
HMI pools subgroup medians for collard greens (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-collard-greens-cooked-fresh-weight:1e12554ef31c
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI kale (cooked) tHg reference (fresh weight) · 1 sampling family
Reference and scope
HMI kale (cooked) tHg reference (fresh weight) · 1 sampling family. Kale (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1.05 | 1.05 | 1.05 |
| Alternative weighting | 1.05 | 1.05 | 1.05 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| FDA TDS FY2018–FY2020 | 1 | 100% |
Methods
HMI pools subgroup medians for kale (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-kale-cooked-fresh-weight:ee50a7c1e8fe
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI coriander leaves tHg reference (dry weight) · 1 sampling family
Reference and scope
HMI coriander leaves tHg reference (dry weight) · 1 sampling family. Coriander leaves; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 130 | 186 | 5000 |
| Alternative weighting | 130 | 186 | 5000 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 100% |
Methods
HMI pools subgroup medians for coriander leaves on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-coriander-leaves-dry-weight:591b5448ef0b
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI lettuce tHg reference (dry weight) · 2 sampling families
Reference and scope
HMI lettuce tHg reference (dry weight) · 2 sampling families. Lettuce; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤39 | 50.3–69.8 (bounded) | 154 |
| Alternative weighting | ≤39 | 50.3–69.8 (bounded) | 154 |
5 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 50% |
| Stark et al. 2019 | 1 | 50% |
Methods
HMI pools subgroup medians for lettuce on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-lettuce-dry-weight:9cadc136a50b
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI mint tHg reference (dry weight) · 1 sampling family
Reference and scope
HMI mint tHg reference (dry weight) · 1 sampling family. Mint; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 158 | 205 | 374 |
| Alternative weighting | 158 | 205 | 374 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 100% |
Methods
HMI pools subgroup medians for mint on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-mint-dry-weight:242b090517b2
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spring onion leaves tHg reference (dry weight) · 1 sampling family
Reference and scope
HMI spring onion leaves tHg reference (dry weight) · 1 sampling family. Spring onion leaves; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 110 | 146 | 194 |
| Alternative weighting | 110 | 146 | 194 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 100% |
Methods
HMI pools subgroup medians for spring onion leaves on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-spring-onion-leaves-dry-weight:f6e334998373
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI swiss chard tHg reference (dry weight) · 1 sampling family
Reference and scope
HMI swiss chard tHg reference (dry weight) · 1 sampling family. Swiss chard; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 111 | 124 | 167 |
| Alternative weighting | 111 | 124 | 167 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 100% |
Methods
HMI pools subgroup medians for swiss chard on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-swiss-chard-dry-weight:ddabb90571f4
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI lettuce tHg reference (fresh weight) · 2 sampling families
Reference and scope
HMI lettuce tHg reference (fresh weight) · 2 sampling families. Lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤0.5 | 1.35–1.6 (bounded) | 2.7 |
| Alternative weighting | ≤0.5 | 1.35–1.6 (bounded) | 2.7 |
2 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| UK TDS 2014 (Fera 2015) | 1 | 50% |
| Martin-Leon et al. 2023 | 1 | 50% |
Methods
HMI pools subgroup medians for lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-lettuce-fresh-weight:5dd111c0998a
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI rocket tHg reference (fresh weight) · 1 sampling family
Reference and scope
HMI rocket tHg reference (fresh weight) · 1 sampling family. Rocket; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 2.3 | 2.3 | 2.3 |
| Alternative weighting | 2.3 | 2.3 | 2.3 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for rocket on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-rocket-fresh-weight:2d1f8897c272
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI romaine lettuce tHg reference (fresh weight) · 1 sampling family
Reference and scope
HMI romaine lettuce tHg reference (fresh weight) · 1 sampling family. Romaine lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 2.7 | 2.7 | 2.7 |
| Alternative weighting | 2.7 | 2.7 | 2.7 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for romaine lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-romaine-lettuce-fresh-weight:251e8c93a224
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI swiss chard tHg reference (fresh weight) · 1 sampling family
Reference and scope
HMI swiss chard tHg reference (fresh weight) · 1 sampling family. Swiss chard; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1.4 | 1.4 | 1.4 |
| Alternative weighting | 1.4 | 1.4 | 1.4 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for swiss chard on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-swiss-chard-fresh-weight:84dde29a0b9d
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI bristly ox tongue tHg reference (dry weight) · 1 sampling family
Reference and scope
HMI bristly ox tongue tHg reference (dry weight) · 1 sampling family. Bristly ox tongue; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤38.3 | ≤38.3 | ≤38.3 |
| Alternative weighting | ≤38.3 | ≤38.3 | ≤38.3 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 2 | 100% |
Methods
HMI pools subgroup medians for bristly ox tongue on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-bristly-ox-tongue-dry-weight:7de0c0be5f91
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI cat's ear tHg reference (dry weight) · 1 sampling family
Reference and scope
HMI cat's ear tHg reference (dry weight) · 1 sampling family. Cat's ear; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤39 | ≤39 | ≤39 |
| Alternative weighting | ≤39 | ≤39 | ≤39 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 1 | 100% |
Methods
HMI pools subgroup medians for cat's ear on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-cat-s-ear-dry-weight:c0b686a94560
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI dandelion greens tHg reference (dry weight) · 1 sampling family
Reference and scope
HMI dandelion greens tHg reference (dry weight) · 1 sampling family. Dandelion greens; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤39.8 | ≤39.8 | ≤39.8 |
| Alternative weighting | ≤39.8 | ≤39.8 | ≤39.8 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 1 | 100% |
Methods
HMI pools subgroup medians for dandelion greens on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-dandelion-greens-dry-weight:c4b1f9b71bfe
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI english plantain tHg reference (dry weight) · 1 sampling family
Reference and scope
HMI english plantain tHg reference (dry weight) · 1 sampling family. English plantain; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤38.3 | ≤38.3 | ≤38.4 |
| Alternative weighting | ≤38.3 | ≤38.3 | ≤38.4 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 2 | 100% |
Methods
HMI pools subgroup medians for english plantain on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-english-plantain-dry-weight:d0f427e35d6a
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI mallow tHg reference (dry weight) · 1 sampling family
Reference and scope
HMI mallow tHg reference (dry weight) · 1 sampling family. Mallow; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤39.3 | ≤39.3 | ≤39.3 |
| Alternative weighting | ≤39.3 | ≤39.3 | ≤39.3 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 1 | 100% |
Methods
HMI pools subgroup medians for mallow on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-mallow-dry-weight:d43ac9de9d9c
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI nasturtium (tropaeolum) tHg reference (dry weight) · 1 sampling family
Reference and scope
HMI nasturtium (tropaeolum) tHg reference (dry weight) · 1 sampling family. Nasturtium (Tropaeolum); Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤38.2 | ≤38.2 | ≤38.2 |
| Alternative weighting | ≤38.2 | ≤38.2 | ≤38.2 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 1 | 100% |
Methods
HMI pools subgroup medians for nasturtium (tropaeolum) on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-nasturtium-tropaeolum-dry-weight:ddc81526ebcf
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI sweet fennel tHg reference (dry weight) · 1 sampling family
Reference and scope
HMI sweet fennel tHg reference (dry weight) · 1 sampling family. Sweet fennel; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤39.1 | ≤39.5 | ≤39.9 |
| Alternative weighting | ≤39.1 | ≤39.5 | ≤39.9 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 2 | 100% |
Methods
HMI pools subgroup medians for sweet fennel on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-thg-sweet-fennel-dry-weight:80f378b732f6
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head cabbage Ni reference (fresh weight) · 2 sampling families
Reference and scope
HMI head cabbage Ni reference (fresh weight) · 2 sampling families. Head cabbage; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤40 | 291–311 (bounded) | 640 |
| Alternative weighting | ≤40 | 291–311 (bounded) | 640 |
11 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| FDA TDS FY2018–FY2020 | 1 | 50% |
| Armand et al. 2026 | 10 | 50% |
Methods
HMI pools subgroup medians for head cabbage on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-head-cabbage-fresh-weight:0f35c7e52dc9
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach Ni reference (fresh weight) · 2 sampling families
Reference and scope
HMI spinach Ni reference (fresh weight) · 2 sampling families. Spinach; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 72 | 78.8 | 85.7 |
| Alternative weighting | 72 | 78.8 | 85.7 |
2 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| FDA TDS FY2018–FY2020 | 1 | 50% |
| Martin-Leon et al. 2023 | 1 | 50% |
Methods
HMI pools subgroup medians for spinach on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-spinach-fresh-weight:a91229145f30
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI collard greens (cooked) Ni reference (fresh weight) · 1 sampling family
Reference and scope
HMI collard greens (cooked) Ni reference (fresh weight) · 1 sampling family. Collard greens (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤40 | ≤40 | ≤40 |
| Alternative weighting | ≤40 | ≤40 | ≤40 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| FDA TDS FY2018–FY2020 | 1 | 100% |
Methods
HMI pools subgroup medians for collard greens (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-collard-greens-cooked-fresh-weight:d850bd629f71
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head lettuce Ni reference (fresh weight) · 2 sampling families
Reference and scope
HMI head lettuce Ni reference (fresh weight) · 2 sampling families. Head lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 60.4 | 75.2 | 90 |
| Alternative weighting | 60.4 | 75.2 | 90 |
2 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| FDA TDS FY2018–FY2020 | 1 | 50% |
| Martin-Leon et al. 2023 | 1 | 50% |
Methods
HMI pools subgroup medians for head lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-head-lettuce-fresh-weight:4dda4dcb7373
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI kale (cooked) Ni reference (fresh weight) · 1 sampling family
Reference and scope
HMI kale (cooked) Ni reference (fresh weight) · 1 sampling family. Kale (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 60.5 | 60.5 | 60.5 |
| Alternative weighting | 60.5 | 60.5 | 60.5 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| FDA TDS FY2018–FY2020 | 1 | 100% |
Methods
HMI pools subgroup medians for kale (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-kale-cooked-fresh-weight:364306ab9e54
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leaf lettuce Ni reference (fresh weight) · 1 sampling family
Reference and scope
HMI leaf lettuce Ni reference (fresh weight) · 1 sampling family. Leaf lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 56 | 56 | 56 |
| Alternative weighting | 56 | 56 | 56 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| FDA TDS FY2018–FY2020 | 1 | 100% |
Methods
HMI pools subgroup medians for leaf lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-leaf-lettuce-fresh-weight:00252bf7c7bb
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI coriander leaves Ni reference (dry weight) · 2 sampling families
Reference and scope
HMI coriander leaves Ni reference (dry weight) · 2 sampling families. Coriander leaves; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 3000 | 7500 | 18100 |
| Alternative weighting | 3000 | 7500 | 18100 |
8 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 50% |
| Sultana et al. 2022 | 4 | 50% |
Methods
HMI pools subgroup medians for coriander leaves on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-coriander-leaves-dry-weight:d00b40b5ba7d
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI lettuce Ni reference (dry weight) · 2 sampling families
Reference and scope
HMI lettuce Ni reference (dry weight) · 2 sampling families. Lettuce; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤781 | 2260–2660 (bounded) | 41500 |
| Alternative weighting | ≤781 | 2260–2660 (bounded) | 41500 |
5 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 50% |
| Stark et al. 2019 | 1 | 50% |
Methods
HMI pools subgroup medians for lettuce on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-lettuce-dry-weight:6606af2b314d
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI mint Ni reference (dry weight) · 2 sampling families
Reference and scope
HMI mint Ni reference (dry weight) · 2 sampling families. Mint; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1050 | 7610 | 20800 |
| Alternative weighting | 1050 | 7610 | 20800 |
8 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 50% |
| Sultana et al. 2022 | 4 | 50% |
Methods
HMI pools subgroup medians for mint on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-mint-dry-weight:13d4ae2d9745
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spring onion leaves Ni reference (dry weight) · 1 sampling family
Reference and scope
HMI spring onion leaves Ni reference (dry weight) · 1 sampling family. Spring onion leaves; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 6490 | 7550 | 10200 |
| Alternative weighting | 6490 | 7550 | 10200 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 100% |
Methods
HMI pools subgroup medians for spring onion leaves on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-spring-onion-leaves-dry-weight:7647c3a29aa3
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI swiss chard Ni reference (dry weight) · 1 sampling family
Reference and scope
HMI swiss chard Ni reference (dry weight) · 1 sampling family. Swiss chard; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 6940 | 7340 | 8990 |
| Alternative weighting | 6940 | 7340 | 8990 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 100% |
Methods
HMI pools subgroup medians for swiss chard on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-swiss-chard-dry-weight:3376350b9c9a
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI lettuce Ni reference (fresh weight) · 5 sampling families
Reference and scope
HMI lettuce Ni reference (fresh weight) · 5 sampling families. Lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 31 | 49.1 | 336 |
| Alternative weighting | 30 | 49.1 | 336 |
21 reference units; 5 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Armand et al. 2026 | 10 | 20% |
| Dearing et al. 2025 | 8 | 20% |
| Romero-Crespo et al. 2023 | 1 | 20% |
| UK TDS 2014 (Fera 2015) | 1 | 20% |
| Martin-Leon et al. 2023 | 1 | 20% |
Methods
HMI pools subgroup medians for lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
- Armand et al. 2026
- Dearing et al. 2025
- Romero-Crespo et al. 2023
- UK TDS 2014 (Fera 2015)
- Martin-Leon et al. 2023
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-lettuce-fresh-weight:8b728cf8520f
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI parsley Ni reference (fresh weight) · 1 sampling family
Reference and scope
HMI parsley Ni reference (fresh weight) · 1 sampling family. Parsley; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 115 | 155 | 280 |
| Alternative weighting | 120 | 155 | 370 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Dearing et al. 2025 | 4 | 100% |
Methods
HMI pools subgroup medians for parsley on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-parsley-fresh-weight:80497c09d312
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI water spinach Ni reference (fresh weight) · 1 sampling family
Reference and scope
HMI water spinach Ni reference (fresh weight) · 1 sampling family. Water spinach; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 40 | 40 | 40 |
| Alternative weighting | 40 | 40 | 40 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Dearing et al. 2025 | 1 | 100% |
Methods
HMI pools subgroup medians for water spinach on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-water-spinach-fresh-weight:606cc4ee45ca
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach Ni reference (dry weight) · 1 sampling family
Reference and scope
HMI spinach Ni reference (dry weight) · 1 sampling family. Spinach; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 150–327 (bounded) | 600–627 (bounded) | 1300 |
| Alternative weighting | 150–327 (bounded) | 600–627 (bounded) | 1300 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Sultana et al. 2022 | 4 | 100% |
Methods
HMI pools subgroup medians for spinach on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-spinach-dry-weight:d11ba43cbe6f
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head cabbage Ni reference (dry weight) · 1 sampling family
Reference and scope
HMI head cabbage Ni reference (dry weight) · 1 sampling family. Head cabbage; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤355 | ≤355 | 1350–1530 (bounded) |
| Alternative weighting | ≤355 | ≤355 | 1350–1530 (bounded) |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Sultana et al. 2022 | 4 | 100% |
Methods
HMI pools subgroup medians for head cabbage on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-head-cabbage-dry-weight:96af87a28139
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI kale Ni reference (fresh weight) · 1 sampling family
Reference and scope
HMI kale Ni reference (fresh weight) · 1 sampling family. Kale; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤50 | 63.1 | 104 |
| Alternative weighting | ≤50 | 50.9 | 104 |
38 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Kim et al. 2024 | 38 | 100% |
Methods
HMI pools subgroup medians for kale on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-kale-fresh-weight:9fd9fee4e011
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI mustard greens Ni reference (dry weight) · 1 sampling family
Reference and scope
HMI mustard greens Ni reference (dry weight) · 1 sampling family. Mustard greens; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1600 | 2450 | 4850 |
| Alternative weighting | 1600 | 2450 | 4850 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Sultana et al. 2022 | 4 | 100% |
Methods
HMI pools subgroup medians for mustard greens on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-mustard-greens-dry-weight:5e8c7581a9a3
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI lamb's lettuce Ni reference (fresh weight) · 1 sampling family
Reference and scope
HMI lamb's lettuce Ni reference (fresh weight) · 1 sampling family. Lamb's lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 40.4 | 40.4 | 40.4 |
| Alternative weighting | 40.4 | 40.4 | 40.4 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for lamb's lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-lamb-s-lettuce-fresh-weight:c6f89d595e74
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI rocket Ni reference (fresh weight) · 1 sampling family
Reference and scope
HMI rocket Ni reference (fresh weight) · 1 sampling family. Rocket; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 59.3 | 59.3 | 59.3 |
| Alternative weighting | 59.3 | 59.3 | 59.3 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for rocket on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-rocket-fresh-weight:0bbf1a6c5637
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI romaine lettuce Ni reference (fresh weight) · 1 sampling family
Reference and scope
HMI romaine lettuce Ni reference (fresh weight) · 1 sampling family. Romaine lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 39.4 | 39.4 | 39.4 |
| Alternative weighting | 39.4 | 39.4 | 39.4 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for romaine lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-romaine-lettuce-fresh-weight:5a191e662796
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI swiss chard Ni reference (fresh weight) · 1 sampling family
Reference and scope
HMI swiss chard Ni reference (fresh weight) · 1 sampling family. Swiss chard; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 56.5 | 56.5 | 56.5 |
| Alternative weighting | 56.5 | 56.5 | 56.5 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for swiss chard on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-swiss-chard-fresh-weight:82300a164d93
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI watercress Ni reference (fresh weight) · 1 sampling family
Reference and scope
HMI watercress Ni reference (fresh weight) · 1 sampling family. Watercress; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 33.3 | 33.3 | 33.3 |
| Alternative weighting | 33.3 | 33.3 | 33.3 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for watercress on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-watercress-fresh-weight:8239acf63f56
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI chives Ni reference (fresh weight) · 1 sampling family
Reference and scope
HMI chives Ni reference (fresh weight) · 1 sampling family. Chives; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 2060 | 2060 | 2060 |
| Alternative weighting | 2060 | 2060 | 2060 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Romero-Crespo et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for chives on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-chives-fresh-weight:09cdecc4be8d
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI coriander leaves Ni reference (fresh weight) · 1 sampling family
Reference and scope
HMI coriander leaves Ni reference (fresh weight) · 1 sampling family. Coriander leaves; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 9100 | 9100 | 9100 |
| Alternative weighting | 9100 | 9100 | 9100 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Romero-Crespo et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for coriander leaves on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-coriander-leaves-fresh-weight:056fe1fda6ba
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI bristly ox tongue Ni reference (dry weight) · 1 sampling family
Reference and scope
HMI bristly ox tongue Ni reference (dry weight) · 1 sampling family. Bristly ox tongue; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤767 | ≤767 | ≤767 |
| Alternative weighting | ≤767 | ≤767 | ≤767 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 2 | 100% |
Methods
HMI pools subgroup medians for bristly ox tongue on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-bristly-ox-tongue-dry-weight:ad07000d3744
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI cat's ear Ni reference (dry weight) · 1 sampling family
Reference and scope
HMI cat's ear Ni reference (dry weight) · 1 sampling family. Cat's ear; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤781 | ≤781 | ≤781 |
| Alternative weighting | ≤781 | ≤781 | ≤781 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 1 | 100% |
Methods
HMI pools subgroup medians for cat's ear on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-cat-s-ear-dry-weight:055eef09bdf1
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI dandelion greens Ni reference (dry weight) · 1 sampling family
Reference and scope
HMI dandelion greens Ni reference (dry weight) · 1 sampling family. Dandelion greens; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤797 | ≤797 | ≤797 |
| Alternative weighting | ≤797 | ≤797 | ≤797 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 1 | 100% |
Methods
HMI pools subgroup medians for dandelion greens on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-dandelion-greens-dry-weight:c93d8d16a7cb
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI english plantain Ni reference (dry weight) · 1 sampling family
Reference and scope
HMI english plantain Ni reference (dry weight) · 1 sampling family. English plantain; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤767 | ≤768 | ≤769 |
| Alternative weighting | ≤767 | ≤768 | ≤769 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 2 | 100% |
Methods
HMI pools subgroup medians for english plantain on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-english-plantain-dry-weight:4be687e4686a
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI mallow Ni reference (dry weight) · 1 sampling family
Reference and scope
HMI mallow Ni reference (dry weight) · 1 sampling family. Mallow; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤787 | ≤787 | ≤787 |
| Alternative weighting | ≤787 | ≤787 | ≤787 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 1 | 100% |
Methods
HMI pools subgroup medians for mallow on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-mallow-dry-weight:99adb1c042d8
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI nasturtium (tropaeolum) Ni reference (dry weight) · 1 sampling family
Reference and scope
HMI nasturtium (tropaeolum) Ni reference (dry weight) · 1 sampling family. Nasturtium (Tropaeolum); Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤766 | ≤766 | ≤766 |
| Alternative weighting | ≤766 | ≤766 | ≤766 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 1 | 100% |
Methods
HMI pools subgroup medians for nasturtium (tropaeolum) on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-nasturtium-tropaeolum-dry-weight:dfc100e53657
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI sweet fennel Ni reference (dry weight) · 1 sampling family
Reference and scope
HMI sweet fennel Ni reference (dry weight) · 1 sampling family. Sweet fennel; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤784 | ≤792 | ≤799 |
| Alternative weighting | ≤784 | ≤792 | ≤799 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 2 | 100% |
Methods
HMI pools subgroup medians for sweet fennel on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-ni-sweet-fennel-dry-weight:394922868ffe
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head cabbage Cr reference (fresh weight) · 2 sampling families
Reference and scope
HMI head cabbage Cr reference (fresh weight) · 2 sampling families. Head cabbage; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤50 | 331–356 (bounded) | 698 |
| Alternative weighting | ≤50 | 331–356 (bounded) | 698 |
11 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| FDA TDS FY2018–FY2020 | 1 | 50% |
| Armand et al. 2026 | 10 | 50% |
Methods
HMI pools subgroup medians for head cabbage on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-head-cabbage-fresh-weight:924ea663b95e
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach Cr reference (fresh weight) · 2 sampling families
Reference and scope
HMI spinach Cr reference (fresh weight) · 2 sampling families. Spinach; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 72.4 | 86.2 | 100 |
| Alternative weighting | 72.4 | 86.2 | 100 |
2 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| FDA TDS FY2018–FY2020 | 1 | 50% |
| Martin-Leon et al. 2023 | 1 | 50% |
Methods
HMI pools subgroup medians for spinach on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-spinach-fresh-weight:6fb3433b9154
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI collard greens (cooked) Cr reference (fresh weight) · 1 sampling family
Reference and scope
HMI collard greens (cooked) Cr reference (fresh weight) · 1 sampling family. Collard greens (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤50 | ≤50 | ≤50 |
| Alternative weighting | ≤50 | ≤50 | ≤50 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| FDA TDS FY2018–FY2020 | 1 | 100% |
Methods
HMI pools subgroup medians for collard greens (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-collard-greens-cooked-fresh-weight:81139e1cb912
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head lettuce Cr reference (fresh weight) · 2 sampling families
Reference and scope
HMI head lettuce Cr reference (fresh weight) · 2 sampling families. Head lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤8.9 | 4.45–29.4 (bounded) | 8.9–50 (bounded) |
| Alternative weighting | ≤8.9 | 4.45–29.4 (bounded) | 8.9–50 (bounded) |
2 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| FDA TDS FY2018–FY2020 | 1 | 50% |
| Martin-Leon et al. 2023 | 1 | 50% |
Methods
HMI pools subgroup medians for head lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-head-lettuce-fresh-weight:1a57f83462f1
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI kale (cooked) Cr reference (fresh weight) · 1 sampling family
Reference and scope
HMI kale (cooked) Cr reference (fresh weight) · 1 sampling family. Kale (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤50 | ≤50 | ≤50 |
| Alternative weighting | ≤50 | ≤50 | ≤50 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| FDA TDS FY2018–FY2020 | 1 | 100% |
Methods
HMI pools subgroup medians for kale (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-kale-cooked-fresh-weight:47f452768109
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leaf lettuce Cr reference (fresh weight) · 1 sampling family
Reference and scope
HMI leaf lettuce Cr reference (fresh weight) · 1 sampling family. Leaf lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤50 | ≤50 | ≤50 |
| Alternative weighting | ≤50 | ≤50 | ≤50 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| FDA TDS FY2018–FY2020 | 1 | 100% |
Methods
HMI pools subgroup medians for leaf lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-leaf-lettuce-fresh-weight:4bcb55bd1144
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI amaranth leaves Cr reference (dry weight) · 1 sampling family
Reference and scope
HMI amaranth leaves Cr reference (dry weight) · 1 sampling family. Amaranth leaves; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 246 | 288 | 3590 |
| Alternative weighting | 246 | 288 | 3590 |
3 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Hossain et al. 2016 | 3 | 100% |
Methods
HMI pools subgroup medians for amaranth leaves on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-amaranth-leaves-dry-weight:839abdb99236
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI coriander leaves Cr reference (dry weight) · 2 sampling families
Reference and scope
HMI coriander leaves Cr reference (dry weight) · 2 sampling families. Coriander leaves; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 3750 | 5810 | 13200 |
| Alternative weighting | 3750 | 5810 | 13200 |
8 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 50% |
| Sultana et al. 2022 | 4 | 50% |
Methods
HMI pools subgroup medians for coriander leaves on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-coriander-leaves-dry-weight:3773fa1fffc6
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leafy vegetables, other or unspecified Cr reference (dry weight) · 1 sampling family
Reference and scope
HMI leafy vegetables, other or unspecified Cr reference (dry weight) · 1 sampling family. Leafy vegetables, other or unspecified; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 4170 | 4170 | 4170 |
| Alternative weighting | 4170 | 4170 | 4170 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Cherfi et al. 2014 | 1 | 100% |
Methods
HMI pools subgroup medians for leafy vegetables, other or unspecified on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-leafy-vegetables-other-or-unspecified-dry-weight:dea98a21d89d
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI lettuce Cr reference (dry weight) · 2 sampling families
Reference and scope
HMI lettuce Cr reference (dry weight) · 2 sampling families. Lettuce; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤781 | 1960–2350 (bounded) | 5020 |
| Alternative weighting | ≤781 | 1960–2350 (bounded) | 5020 |
5 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 50% |
| Stark et al. 2019 | 1 | 50% |
Methods
HMI pools subgroup medians for lettuce on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-lettuce-dry-weight:c12302b7e80e
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI mint Cr reference (dry weight) · 2 sampling families
Reference and scope
HMI mint Cr reference (dry weight) · 2 sampling families. Mint; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 3300 | 9140 | 14600 |
| Alternative weighting | 3300 | 9140 | 14600 |
8 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 50% |
| Sultana et al. 2022 | 4 | 50% |
Methods
HMI pools subgroup medians for mint on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-mint-dry-weight:441a679ee86e
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spring onion leaves Cr reference (dry weight) · 1 sampling family
Reference and scope
HMI spring onion leaves Cr reference (dry weight) · 1 sampling family. Spring onion leaves; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 6070 | 8860 | 10500 |
| Alternative weighting | 6070 | 8860 | 10500 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 100% |
Methods
HMI pools subgroup medians for spring onion leaves on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-spring-onion-leaves-dry-weight:6c319e722506
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI swiss chard Cr reference (dry weight) · 1 sampling family
Reference and scope
HMI swiss chard Cr reference (dry weight) · 1 sampling family. Swiss chard; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 6300 | 8030 | 16400 |
| Alternative weighting | 6300 | 8030 | 16400 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 100% |
Methods
HMI pools subgroup medians for swiss chard on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-swiss-chard-dry-weight:2978ee20e5d5
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head cabbage Cr reference (dry weight) · 2 sampling families
Reference and scope
HMI head cabbage Cr reference (dry weight) · 2 sampling families. Head cabbage; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 526 | 1340 | 2060 |
| Alternative weighting | 526 | 1340 | 2060 |
12 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Hossain et al. 2016 | 8 | 50% |
| Sultana et al. 2022 | 4 | 50% |
Methods
HMI pools subgroup medians for head cabbage on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-head-cabbage-dry-weight:bc9f3240de4c
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI lettuce Cr reference (fresh weight) · 4 sampling families
Reference and scope
HMI lettuce Cr reference (fresh weight) · 4 sampling families. Lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 26.4 | 152 | 1690 |
| Alternative weighting | 26.4 | 152 | 1690 |
13 reference units; 4 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Armand et al. 2026 | 10 | 25% |
| Romero-Crespo et al. 2023 | 1 | 25% |
| UK TDS 2014 (Fera 2015) | 1 | 25% |
| Martin-Leon et al. 2023 | 1 | 25% |
Methods
HMI pools subgroup medians for lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-lettuce-fresh-weight:e5b13cd84e2b
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach Cr reference (dry weight) · 2 sampling families
Reference and scope
HMI spinach Cr reference (dry weight) · 2 sampling families. Spinach; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 2100 | 5820 | 8330 |
| Alternative weighting | 2100 | 5820 | 8330 |
5 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Cherfi et al. 2014 | 1 | 50% |
| Sultana et al. 2022 | 4 | 50% |
Methods
HMI pools subgroup medians for spinach on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-spinach-dry-weight:798ec676164b
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI kale Cr reference (fresh weight) · 1 sampling family
Reference and scope
HMI kale Cr reference (fresh weight) · 1 sampling family. Kale; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 31.3 | 39.8 | 72 |
| Alternative weighting | 31.3 | 37.3 | 71.5 |
38 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Kim et al. 2024 | 38 | 100% |
Methods
HMI pools subgroup medians for kale on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-kale-fresh-weight:1be4c5c5349b
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI mustard greens Cr reference (dry weight) · 1 sampling family
Reference and scope
HMI mustard greens Cr reference (dry weight) · 1 sampling family. Mustard greens; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 2800 | 4200 | 6600 |
| Alternative weighting | 2800 | 4200 | 6600 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Sultana et al. 2022 | 4 | 100% |
Methods
HMI pools subgroup medians for mustard greens on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-mustard-greens-dry-weight:dad73d21b459
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI lamb's lettuce Cr reference (fresh weight) · 1 sampling family
Reference and scope
HMI lamb's lettuce Cr reference (fresh weight) · 1 sampling family. Lamb's lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 38 | 38 | 38 |
| Alternative weighting | 38 | 38 | 38 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for lamb's lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-lamb-s-lettuce-fresh-weight:f6b12e8fd23f
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI rocket Cr reference (fresh weight) · 1 sampling family
Reference and scope
HMI rocket Cr reference (fresh weight) · 1 sampling family. Rocket; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 52.3 | 52.3 | 52.3 |
| Alternative weighting | 52.3 | 52.3 | 52.3 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for rocket on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-rocket-fresh-weight:a4f79bcf1a0c
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI romaine lettuce Cr reference (fresh weight) · 1 sampling family
Reference and scope
HMI romaine lettuce Cr reference (fresh weight) · 1 sampling family. Romaine lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 21.4 | 21.4 | 21.4 |
| Alternative weighting | 21.4 | 21.4 | 21.4 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for romaine lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-romaine-lettuce-fresh-weight:301e58ffc898
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI swiss chard Cr reference (fresh weight) · 1 sampling family
Reference and scope
HMI swiss chard Cr reference (fresh weight) · 1 sampling family. Swiss chard; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 53.7 | 53.7 | 53.7 |
| Alternative weighting | 53.7 | 53.7 | 53.7 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for swiss chard on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-swiss-chard-fresh-weight:1fafe6c80751
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI watercress Cr reference (fresh weight) · 1 sampling family
Reference and scope
HMI watercress Cr reference (fresh weight) · 1 sampling family. Watercress; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 31.8 | 31.8 | 31.8 |
| Alternative weighting | 31.8 | 31.8 | 31.8 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for watercress on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-watercress-fresh-weight:2c14640e79c4
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI chives Cr reference (fresh weight) · 1 sampling family
Reference and scope
HMI chives Cr reference (fresh weight) · 1 sampling family. Chives; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 380 | 380 | 380 |
| Alternative weighting | 380 | 380 | 380 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Romero-Crespo et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for chives on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-chives-fresh-weight:ea58f302bdfb
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI coriander leaves Cr reference (fresh weight) · 1 sampling family
Reference and scope
HMI coriander leaves Cr reference (fresh weight) · 1 sampling family. Coriander leaves; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 490 | 490 | 490 |
| Alternative weighting | 490 | 490 | 490 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Romero-Crespo et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for coriander leaves on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-coriander-leaves-fresh-weight:9df0866b7c4d
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI mallow Cr reference (dry weight) · 1 sampling family
Reference and scope
HMI mallow Cr reference (dry weight) · 1 sampling family. Mallow; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤787 | ≤787 | ≤787 |
| Alternative weighting | ≤787 | ≤787 | ≤787 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 1 | 100% |
Methods
HMI pools subgroup medians for mallow on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-mallow-dry-weight:07ebadaa5f26
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI bristly ox tongue Cr reference (dry weight) · 1 sampling family
Reference and scope
HMI bristly ox tongue Cr reference (dry weight) · 1 sampling family. Bristly ox tongue; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤767 | ≤767 | ≤767 |
| Alternative weighting | ≤767 | ≤767 | ≤767 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 2 | 100% |
Methods
HMI pools subgroup medians for bristly ox tongue on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-bristly-ox-tongue-dry-weight:0dd03a176b27
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI cat's ear Cr reference (dry weight) · 1 sampling family
Reference and scope
HMI cat's ear Cr reference (dry weight) · 1 sampling family. Cat's ear; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤781 | ≤781 | ≤781 |
| Alternative weighting | ≤781 | ≤781 | ≤781 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 1 | 100% |
Methods
HMI pools subgroup medians for cat's ear on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-cat-s-ear-dry-weight:9bf17ea31bd1
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI dandelion greens Cr reference (dry weight) · 1 sampling family
Reference and scope
HMI dandelion greens Cr reference (dry weight) · 1 sampling family. Dandelion greens; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤797 | ≤797 | ≤797 |
| Alternative weighting | ≤797 | ≤797 | ≤797 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 1 | 100% |
Methods
HMI pools subgroup medians for dandelion greens on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-dandelion-greens-dry-weight:bd34cd533bc0
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI english plantain Cr reference (dry weight) · 1 sampling family
Reference and scope
HMI english plantain Cr reference (dry weight) · 1 sampling family. English plantain; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤767 | ≤768 | ≤769 |
| Alternative weighting | ≤767 | ≤768 | ≤769 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 2 | 100% |
Methods
HMI pools subgroup medians for english plantain on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-english-plantain-dry-weight:c9ed5d6dbe51
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI nasturtium (tropaeolum) Cr reference (dry weight) · 1 sampling family
Reference and scope
HMI nasturtium (tropaeolum) Cr reference (dry weight) · 1 sampling family. Nasturtium (Tropaeolum); Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤766 | ≤766 | ≤766 |
| Alternative weighting | ≤766 | ≤766 | ≤766 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 1 | 100% |
Methods
HMI pools subgroup medians for nasturtium (tropaeolum) on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-nasturtium-tropaeolum-dry-weight:84479d0cc64d
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI sweet fennel Cr reference (dry weight) · 1 sampling family
Reference and scope
HMI sweet fennel Cr reference (dry weight) · 1 sampling family. Sweet fennel; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤784 | ≤792 | ≤799 |
| Alternative weighting | ≤784 | ≤792 | ≤799 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Stark et al. 2019 | 2 | 100% |
Methods
HMI pools subgroup medians for sweet fennel on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-cr-sweet-fennel-dry-weight:26a505e12000
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leafy vegetables, other or unspecified Al reference (fresh weight) · 1 sampling family
Reference and scope
HMI leafy vegetables, other or unspecified Al reference (fresh weight) · 1 sampling family. Leafy vegetables, other or unspecified; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 4210 | 5560 | 12600 |
| Alternative weighting | 4210 | 5560 | 12600 |
3 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 3 | 100% |
Methods
HMI pools subgroup medians for leafy vegetables, other or unspecified on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-al-leafy-vegetables-other-or-unspecified-fresh-weight:6c0389546b21
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach Al reference (fresh weight) · 1 sampling family
Reference and scope
HMI spinach Al reference (fresh weight) · 1 sampling family. Spinach; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 41900 | 41900 | 41900 |
| Alternative weighting | 41900 | 41900 | 41900 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for spinach on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-al-spinach-fresh-weight:b3b935acd1d5
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach (cooked) Al reference (fresh weight) · 1 sampling family
Reference and scope
HMI spinach (cooked) Al reference (fresh weight) · 1 sampling family. Spinach (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 10500 | 13900 | 52100 |
| Alternative weighting | 10500 | 13900 | 52100 |
3 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 3 | 100% |
Methods
HMI pools subgroup medians for spinach (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-al-spinach-cooked-fresh-weight:8410f9f1ebdf
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI coriander leaves Al reference (dry weight) · 1 sampling family
Reference and scope
HMI coriander leaves Al reference (dry weight) · 1 sampling family. Coriander leaves; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 501000 | 1940000 | 3360000 |
| Alternative weighting | 501000 | 1940000 | 3360000 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 100% |
Methods
HMI pools subgroup medians for coriander leaves on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-al-coriander-leaves-dry-weight:9ffb405bfe96
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI lettuce Al reference (dry weight) · 1 sampling family
Reference and scope
HMI lettuce Al reference (dry weight) · 1 sampling family. Lettuce; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 52500 | 106000 | 276000 |
| Alternative weighting | 52500 | 106000 | 276000 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 100% |
Methods
HMI pools subgroup medians for lettuce on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-al-lettuce-dry-weight:eb11639d5fa4
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI mint Al reference (dry weight) · 1 sampling family
Reference and scope
HMI mint Al reference (dry weight) · 1 sampling family. Mint; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 371000 | 671000 | 1010000 |
| Alternative weighting | 371000 | 671000 | 1010000 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 100% |
Methods
HMI pools subgroup medians for mint on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-al-mint-dry-weight:0c606326c1ae
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spring onion leaves Al reference (dry weight) · 1 sampling family
Reference and scope
HMI spring onion leaves Al reference (dry weight) · 1 sampling family. Spring onion leaves; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 193000 | 317000 | 459000 |
| Alternative weighting | 193000 | 317000 | 459000 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 100% |
Methods
HMI pools subgroup medians for spring onion leaves on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-al-spring-onion-leaves-dry-weight:4c4a2fd27fce
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI swiss chard Al reference (dry weight) · 1 sampling family
Reference and scope
HMI swiss chard Al reference (dry weight) · 1 sampling family. Swiss chard; Dry weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 282000 | 401000 | 1100000 |
| Alternative weighting | 282000 | 401000 | 1100000 |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Adhikari et al. 2024 | 4 | 100% |
Methods
HMI pools subgroup medians for swiss chard on a dry weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-al-swiss-chard-dry-weight:1646e91086a8
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head lettuce Al reference (fresh weight) · 1 sampling family
Reference and scope
HMI head lettuce Al reference (fresh weight) · 1 sampling family. Head lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1580 | 1580 | 1580 |
| Alternative weighting | 1580 | 1580 | 1580 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for head lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-al-head-lettuce-fresh-weight:ec310e80d7a0
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI lamb's lettuce Al reference (fresh weight) · 1 sampling family
Reference and scope
HMI lamb's lettuce Al reference (fresh weight) · 1 sampling family. Lamb's lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 14200 | 14200 | 14200 |
| Alternative weighting | 14200 | 14200 | 14200 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for lamb's lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-al-lamb-s-lettuce-fresh-weight:06f69dc8cf6f
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI lettuce Al reference (fresh weight) · 2 sampling families
Reference and scope
HMI lettuce Al reference (fresh weight) · 2 sampling families. Lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1640 | 7660 | 13700 |
| Alternative weighting | 1640 | 7660 | 13700 |
2 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| UK TDS 2014 (Fera 2015) | 1 | 50% |
| Martin-Leon et al. 2023 | 1 | 50% |
Methods
HMI pools subgroup medians for lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-al-lettuce-fresh-weight:6735d7bfcff0
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI rocket Al reference (fresh weight) · 1 sampling family
Reference and scope
HMI rocket Al reference (fresh weight) · 1 sampling family. Rocket; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 32100 | 32100 | 32100 |
| Alternative weighting | 32100 | 32100 | 32100 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for rocket on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-al-rocket-fresh-weight:46fd60f6f405
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI romaine lettuce Al reference (fresh weight) · 1 sampling family
Reference and scope
HMI romaine lettuce Al reference (fresh weight) · 1 sampling family. Romaine lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 8830 | 8830 | 8830 |
| Alternative weighting | 8830 | 8830 | 8830 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for romaine lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-al-romaine-lettuce-fresh-weight:07e82a7cec64
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI swiss chard Al reference (fresh weight) · 1 sampling family
Reference and scope
HMI swiss chard Al reference (fresh weight) · 1 sampling family. Swiss chard; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 17900 | 17900 | 17900 |
| Alternative weighting | 17900 | 17900 | 17900 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for swiss chard on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-al-swiss-chard-fresh-weight:177d7ece2474
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI watercress Al reference (fresh weight) · 1 sampling family
Reference and scope
HMI watercress Al reference (fresh weight) · 1 sampling family. Watercress; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 11800 | 11800 | 11800 |
| Alternative weighting | 11800 | 11800 | 11800 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for watercress on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-al-watercress-fresh-weight:aab16b19050d
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI amaranth leaves (cooked) Sn reference (fresh weight) · 1 sampling family
Reference and scope
HMI amaranth leaves (cooked) Sn reference (fresh weight) · 1 sampling family. Amaranth leaves (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 15 | 15 | 15 |
| Alternative weighting | 15 | 15 | 15 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for amaranth leaves (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-sn-amaranth-leaves-cooked-fresh-weight:1abbe1ecbdc4
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI chinese cabbage (cooked) Sn reference (fresh weight) · 1 sampling family
Reference and scope
HMI chinese cabbage (cooked) Sn reference (fresh weight) · 1 sampling family. Chinese cabbage (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤10 | ≤10 | ≤10 |
| Alternative weighting | ≤10 | ≤10 | ≤10 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for chinese cabbage (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-sn-chinese-cabbage-cooked-fresh-weight:9c9144c33809
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head cabbage (cooked) Sn reference (fresh weight) · 1 sampling family
Reference and scope
HMI head cabbage (cooked) Sn reference (fresh weight) · 1 sampling family. Head cabbage (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤10 | ≤10 | ≤10 |
| Alternative weighting | ≤10 | ≤10 | ≤10 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for head cabbage (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-sn-head-cabbage-cooked-fresh-weight:0209cdb98946
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head lettuce Sn reference (fresh weight) · 1 sampling family
Reference and scope
HMI head lettuce Sn reference (fresh weight) · 1 sampling family. Head lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1.1 | 1.1 | 1.1 |
| Alternative weighting | 1.1 | 1.1 | 1.1 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for head lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-sn-head-lettuce-fresh-weight:6e663ae54c17
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head lettuce (cooked) Sn reference (fresh weight) · 1 sampling family
Reference and scope
HMI head lettuce (cooked) Sn reference (fresh weight) · 1 sampling family. Head lettuce (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤10 | ≤10 | ≤10 |
| Alternative weighting | ≤10 | ≤10 | ≤10 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for head lettuce (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-sn-head-lettuce-cooked-fresh-weight:978c8d9bfb66
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI kale (cooked) Sn reference (fresh weight) · 1 sampling family
Reference and scope
HMI kale (cooked) Sn reference (fresh weight) · 1 sampling family. Kale (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤10 | 5.5–10.5 (bounded) | 11 |
| Alternative weighting | ≤10 | 5.5–10.5 (bounded) | 11 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for kale (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-sn-kale-cooked-fresh-weight:776f15ab678f
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leaf lettuce (cooked) Sn reference (fresh weight) · 1 sampling family
Reference and scope
HMI leaf lettuce (cooked) Sn reference (fresh weight) · 1 sampling family. Leaf lettuce (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤10 | 6–11 (bounded) | 12 |
| Alternative weighting | ≤10 | 6–11 (bounded) | 12 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for leaf lettuce (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-sn-leaf-lettuce-cooked-fresh-weight:b84f24a2805a
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leafy vegetables, other or unspecified Sn reference (fresh weight) · 1 sampling family
Reference and scope
HMI leafy vegetables, other or unspecified Sn reference (fresh weight) · 1 sampling family. Leafy vegetables, other or unspecified; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 6.15 | 9.14 | 55.4–70.4 (bounded) |
| Alternative weighting | 6.15 | 9.14 | 55.4–70.4 (bounded) |
4 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 4 | 100% |
Methods
HMI pools subgroup medians for leafy vegetables, other or unspecified on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-sn-leafy-vegetables-other-or-unspecified-fresh-weight:c5bfadaf921b
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leafy vegetables, other or unspecified (cooked) Sn reference (fresh weight) · 1 sampling family
Reference and scope
HMI leafy vegetables, other or unspecified (cooked) Sn reference (fresh weight) · 1 sampling family. Leafy vegetables, other or unspecified (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤10 | ≤10 | ≤10 |
| Alternative weighting | ≤10 | ≤10 | ≤10 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for leafy vegetables, other or unspecified (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-sn-leafy-vegetables-other-or-unspecified-cooked-fresh-weight:7b89c07adcde
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI mustard greens Sn reference (fresh weight) · 1 sampling family
Reference and scope
HMI mustard greens Sn reference (fresh weight) · 1 sampling family. Mustard greens; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 100–130 (bounded) | 100–130 (bounded) | 100–130 (bounded) |
| Alternative weighting | 100–130 (bounded) | 100–130 (bounded) | 100–130 (bounded) |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for mustard greens on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-sn-mustard-greens-fresh-weight:9a953c62edbf
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI mustard greens (cooked) Sn reference (fresh weight) · 1 sampling family
Reference and scope
HMI mustard greens (cooked) Sn reference (fresh weight) · 1 sampling family. Mustard greens (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤10 | 9–14 (bounded) | 18 |
| Alternative weighting | ≤10 | 9–14 (bounded) | 18 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for mustard greens (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-sn-mustard-greens-cooked-fresh-weight:7df394ca21b5
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI pak choi (cooked) Sn reference (fresh weight) · 1 sampling family
Reference and scope
HMI pak choi (cooked) Sn reference (fresh weight) · 1 sampling family. Pak choi (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤10 | ≤10 | ≤10 |
| Alternative weighting | ≤10 | ≤10 | ≤10 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for pak choi (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-sn-pak-choi-cooked-fresh-weight:94fe3cfe9df2
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach (cooked) Sn reference (fresh weight) · 1 sampling family
Reference and scope
HMI spinach (cooked) Sn reference (fresh weight) · 1 sampling family. Spinach (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤10 | 12.3 | 33.3 |
| Alternative weighting | ≤10 | 12.3 | 33.3 |
5 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 5 | 100% |
Methods
HMI pools subgroup medians for spinach (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-sn-spinach-cooked-fresh-weight:f0264e641b99
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI water spinach (cooked) Sn reference (fresh weight) · 1 sampling family
Reference and scope
HMI water spinach (cooked) Sn reference (fresh weight) · 1 sampling family. Water spinach (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 27 | 27 | 27 |
| Alternative weighting | 27 | 27 | 27 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 1 | 100% |
Methods
HMI pools subgroup medians for water spinach (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-sn-water-spinach-cooked-fresh-weight:6066b18821fd
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI watercress (cooked) Sn reference (fresh weight) · 1 sampling family
Reference and scope
HMI watercress (cooked) Sn reference (fresh weight) · 1 sampling family. Watercress (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 12 | 32.5 | 53 |
| Alternative weighting | 12 | 32.5 | 53 |
2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| WHO GEMS/Food | 2 | 100% |
Methods
HMI pools subgroup medians for watercress (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-sn-watercress-cooked-fresh-weight:5cb1a19b07ab
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI lamb's lettuce Sn reference (fresh weight) · 1 sampling family
Reference and scope
HMI lamb's lettuce Sn reference (fresh weight) · 1 sampling family. Lamb's lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1.7 | 1.7 | 1.7 |
| Alternative weighting | 1.7 | 1.7 | 1.7 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for lamb's lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-sn-lamb-s-lettuce-fresh-weight:3f32bd532f13
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI lettuce Sn reference (fresh weight) · 2 sampling families
Reference and scope
HMI lettuce Sn reference (fresh weight) · 2 sampling families. Lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤3.3 | 1.65–6.65 (bounded) | 3.3–10 (bounded) |
| Alternative weighting | ≤3.3 | 1.65–6.65 (bounded) | 3.3–10 (bounded) |
2 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| UK TDS 2014 (Fera 2015) | 1 | 50% |
| Martin-Leon et al. 2023 | 1 | 50% |
Methods
HMI pools subgroup medians for lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-sn-lettuce-fresh-weight:138b2639aa4f
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI rocket Sn reference (fresh weight) · 1 sampling family
Reference and scope
HMI rocket Sn reference (fresh weight) · 1 sampling family. Rocket; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 27.9 | 27.9 | 27.9 |
| Alternative weighting | 27.9 | 27.9 | 27.9 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for rocket on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-sn-rocket-fresh-weight:45aadab99da7
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI romaine lettuce Sn reference (fresh weight) · 1 sampling family
Reference and scope
HMI romaine lettuce Sn reference (fresh weight) · 1 sampling family. Romaine lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 4.4 | 4.4 | 4.4 |
| Alternative weighting | 4.4 | 4.4 | 4.4 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for romaine lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-sn-romaine-lettuce-fresh-weight:d6431da5360e
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach Sn reference (fresh weight) · 1 sampling family
Reference and scope
HMI spinach Sn reference (fresh weight) · 1 sampling family. Spinach; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 13.1 | 13.1 | 13.1 |
| Alternative weighting | 13.1 | 13.1 | 13.1 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for spinach on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-sn-spinach-fresh-weight:dec165fff8dc
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI swiss chard Sn reference (fresh weight) · 1 sampling family
Reference and scope
HMI swiss chard Sn reference (fresh weight) · 1 sampling family. Swiss chard; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 3.4 | 3.4 | 3.4 |
| Alternative weighting | 3.4 | 3.4 | 3.4 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for swiss chard on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-sn-swiss-chard-fresh-weight:6ed3ac05f81e
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI watercress Sn reference (fresh weight) · 1 sampling family
Reference and scope
HMI watercress Sn reference (fresh weight) · 1 sampling family. Watercress; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 2.7 | 2.7 | 2.7 |
| Alternative weighting | 2.7 | 2.7 | 2.7 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for watercress on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-sn-watercress-fresh-weight:19b731231fb2
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI collard greens (cooked) U reference (fresh weight) · 1 sampling family
Reference and scope
HMI collard greens (cooked) U reference (fresh weight) · 1 sampling family. Collard greens (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤1 | ≤1 | ≤1 |
| Alternative weighting | ≤1 | ≤1 | ≤1 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| FDA TDS FY2018–FY2020 | 1 | 100% |
Methods
HMI pools subgroup medians for collard greens (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-u-collard-greens-cooked-fresh-weight:cdbb8f826cd7
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head cabbage U reference (fresh weight) · 1 sampling family
Reference and scope
HMI head cabbage U reference (fresh weight) · 1 sampling family. Head cabbage; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤1 | ≤1 | ≤1 |
| Alternative weighting | ≤1 | ≤1 | ≤1 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| FDA TDS FY2018–FY2020 | 1 | 100% |
Methods
HMI pools subgroup medians for head cabbage on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-u-head-cabbage-fresh-weight:6e53bc575270
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI head lettuce U reference (fresh weight) · 2 sampling families
Reference and scope
HMI head lettuce U reference (fresh weight) · 2 sampling families. Head lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤0.1 | 0.05–0.55 (bounded) | 0.1–1 (bounded) |
| Alternative weighting | ≤0.1 | 0.05–0.55 (bounded) | 0.1–1 (bounded) |
2 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| FDA TDS FY2018–FY2020 | 1 | 50% |
| Martin-Leon et al. 2023 | 1 | 50% |
Methods
HMI pools subgroup medians for head lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-u-head-lettuce-fresh-weight:20b8f5e33d63
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI kale (cooked) U reference (fresh weight) · 1 sampling family
Reference and scope
HMI kale (cooked) U reference (fresh weight) · 1 sampling family. Kale (cooked); Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤1 | ≤1 | ≤1 |
| Alternative weighting | ≤1 | ≤1 | ≤1 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| FDA TDS FY2018–FY2020 | 1 | 100% |
Methods
HMI pools subgroup medians for kale (cooked) on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-u-kale-cooked-fresh-weight:b2d3c4eee2ed
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI leaf lettuce U reference (fresh weight) · 1 sampling family
Reference and scope
HMI leaf lettuce U reference (fresh weight) · 1 sampling family. Leaf lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤1 | ≤1 | ≤1 |
| Alternative weighting | ≤1 | ≤1 | ≤1 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| FDA TDS FY2018–FY2020 | 1 | 100% |
Methods
HMI pools subgroup medians for leaf lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-u-leaf-lettuce-fresh-weight:09b2e35790c5
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI spinach U reference (fresh weight) · 2 sampling families
Reference and scope
HMI spinach U reference (fresh weight) · 2 sampling families. Spinach; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 2.1 | 3 | 3.9 |
| Alternative weighting | 2.1 | 3 | 3.9 |
2 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| FDA TDS FY2018–FY2020 | 1 | 50% |
| Martin-Leon et al. 2023 | 1 | 50% |
Methods
HMI pools subgroup medians for spinach on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-u-spinach-fresh-weight:54ec47f371be
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI kale U reference (fresh weight) · 1 sampling family
Reference and scope
HMI kale U reference (fresh weight) · 1 sampling family. Kale; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | ≤5 | ≤5 | ≤5 |
| Alternative weighting | ≤5 | ≤5 | 5 |
38 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Kim et al. 2024 | 38 | 100% |
Methods
HMI pools subgroup medians for kale on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-u-kale-fresh-weight:64056e5421e5
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI lamb's lettuce U reference (fresh weight) · 1 sampling family
Reference and scope
HMI lamb's lettuce U reference (fresh weight) · 1 sampling family. Lamb's lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 7.9 | 7.9 | 7.9 |
| Alternative weighting | 7.9 | 7.9 | 7.9 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for lamb's lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-u-lamb-s-lettuce-fresh-weight:94aa368a3b69
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI lettuce U reference (fresh weight) · 1 sampling family
Reference and scope
HMI lettuce U reference (fresh weight) · 1 sampling family. Lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 0.9 | 0.9 | 0.9 |
| Alternative weighting | 0.9 | 0.9 | 0.9 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-u-lettuce-fresh-weight:34667b79cea4
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI rocket U reference (fresh weight) · 1 sampling family
Reference and scope
HMI rocket U reference (fresh weight) · 1 sampling family. Rocket; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1.8 | 1.8 | 1.8 |
| Alternative weighting | 1.8 | 1.8 | 1.8 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for rocket on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-u-rocket-fresh-weight:58c2ef4ed7f6
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI romaine lettuce U reference (fresh weight) · 1 sampling family
Reference and scope
HMI romaine lettuce U reference (fresh weight) · 1 sampling family. Romaine lettuce; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 0.3 | 0.3 | 0.3 |
| Alternative weighting | 0.3 | 0.3 | 0.3 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for romaine lettuce on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-u-romaine-lettuce-fresh-weight:9afbe37620fb
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI swiss chard U reference (fresh weight) · 1 sampling family
Reference and scope
HMI swiss chard U reference (fresh weight) · 1 sampling family. Swiss chard; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1.6 | 1.6 | 1.6 |
| Alternative weighting | 1.6 | 1.6 | 1.6 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for swiss chard on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-u-swiss-chard-fresh-weight:f13f0ff4e97c
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
HMI watercress U reference (fresh weight) · 1 sampling family
Reference and scope
HMI watercress U reference (fresh weight) · 1 sampling family. Watercress; Fresh weight, µg/kg; no moisture or cooking conversion. Distribution of subgroup median concentration values.
| Reference (µg/kg) | Q1 | Median | Q3 |
|---|---|---|---|
| Primary HMI reference | 1.1 | 1.1 | 1.1 |
| Alternative weighting | 1.1 | 1.1 | 1.1 |
1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.
| Contributing family | Reference units | Weight |
|---|---|---|
| Martin-Leon et al. 2023 | 1 | 100% |
Methods
HMI pools subgroup medians for watercress on a fresh weight basis from every compatible sampling family in the evidence ledger: WHO GEMS/Food region-by-year medians (random sampling only), FDA Total Diet Study composites, per-sample results printed by individual surveys and published subgroup medians. Each family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations, and fresh-weight and dry-weight results are never converted into each other.
Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1–Q3 spread are required. This is a display eligibility rule, not statistical validation.
Limitations
An evidence-library reference for one leaf and basis, not a probability sample of any market. This is the only compatible family in the completed leafy-greens ledger for this leaf, analyte and basis; it cannot represent other markets. GEMS/Food regions are reporting regions, not growing origins. A position is descriptive concentration, not a safety or compliance judgment.
Admitted source families
Reference version and interpretation
Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.
hmi-pooled-reference-v2:leafy-u-watercress-fresh-weight:84ff4bb1d713
The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.
The source reports only a percentile, range or maximum for these samples, with no mean or median to place against the reference. Spread figures that do have a mean or median are shown on that row instead.
The value could not be reconciled with its own paper, so it is shown but not compared.
The result is below a reporting limit, and the range from zero to that limit crosses more than one band.
No medians or sample values exist for this form and metal, so no HMI reference could be built.
The HMI reference for this form has fewer than four units or no measurable spread, below the display minimum.
The result is on a basis (dry weight, or not stated) that no HMI reference for this form uses, so it cannot be placed.
Reporting limits overlap the study comparison, so no direction can be shown.
No findings match these filters. Change the selection or reset to show all source findings.
How contamination reaches leafy greens
Leaves take metals up from soil and irrigation water, and cadmium moves into leaves more readily than lead.3Heavy metal accumulation in leafy vegetables grown in industrial areas under varying levels of pollutionBangladesh Journal of Agricultural ResearchRead source →14Uptake of Arsenic by Irrigated Vegetables and Cooked Food Products in Burkina FasoFrontiers in WaterRead source → In Gazipur, Bangladesh, spinach leaves from a highly polluted industrial area carried 5,970 µg/kg lead and 1,020 µg/kg cadmium dry weight, against 1,160 and 430 µg/kg on a low-pollution research field.3Heavy metal accumulation in leafy vegetables grown in industrial areas under varying levels of pollutionBangladesh Journal of Agricultural ResearchRead source → In a greenhouse trial, irrigating with water containing 500 µg/L arsenic raised spinach leaf arsenic from 1.05 to 6.57 µg/g dry matter.14Uptake of Arsenic by Irrigated Vegetables and Cooked Food Products in Burkina FasoFrontiers in WaterRead source →
Dust and air deposition add a surface load. Near coal-fired power plants in China, water spinach carried 86.7 µg/kg mercury fresh weight at 1 km and 23.9 µg/kg at 10 km, against 0.85 µg/kg in grocery-store leaves, and rinsing removed 19–63% of the mercury.6Mercury pollution in vegetables, grains and soils from areas surrounding coal-fired power plantsScientific ReportsRead source → Kale grown on urban farms in Baltimore averaged 130 µg/kg lead fresh weight against 10 µg/kg on rural farms nearby.17Nutrients and non-essential metals in darkibor kale grown at urban and rural farms: a pilot studyPLOS ONERead source → Unwashed coriander from a Johannesburg roadside market carried 191.9 mg/kg cadmium dry weight against 6.6 mg/kg once washed.7Concentrations and health risks of selected elements in leafy vegetables: a comparison between roadside open-air markets and large stores in Johannesburg, South AfricaNot reportedRead source →
Inputs and land use also shape levels. Phosphate fertilisers raised lettuce cadmium by up to 91% in a Chinese greenhouse field trial, while farmyard manure lowered cadmium in cabbage and lettuce compared with mineral fertiliser in a long-term English trial.19Effect of Climatic Conditions, and Agronomic Practices Used in Organic and Conventional Crop Production on Yield and Nutritional Composition Parameters in Potato, Cabbage, Lettuce and Onion; Results from the Long-Term NFSC-TrialsAgronomyRead source →20Impact of Physical Interventions, Phosphorus Fertilization, and the Utilization of Soil Amendments on the Absorption of Cadmium by Lettuce Grown in a Solar-Powered GreenhouseBiologyRead source → Manures raised spinach lead from 0.11 to 0.50–0.75 mg/kg in a Turkish field trial.21compared to the chemical fertilizer. In organic fertilizer applications, heavy metal contents such as arsenic, chrome, cobalt, and nickel were found toPeer-reviewed articleRead source → Mining areas raised mercury in Houttuynia leaves to 260 µg/kg dry weight against 40 µg/kg elsewhere, and lead in chard from Spanish mining-town gardens to 698 µg/kg against 118 µg/kg from a city market.10Is it healthy urban agriculture? Human exposure to potentially toxic elements in urban gardens from Andalusia, SpainEnvironmental Science and Pollution ResearchRead source →22Mercury accumulation in vegetable Houttuynia cordata Thunb. from two different geological areas in southwest China and implications for human consumptionScientific Reports 11:1470Read source →
Lower measured levels by source and type
For cadmium the leaf type gives the most consistent lower-contamination choice.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →2FY2018-FY2020 TDS Elements Analytical ResultsFDA Total Diet StudyRead source →15Evaluation of essential, toxic and potentially toxic elements in leafy vegetables grown in the Canary IslandsToxicsRead source → Head cabbage was among the lowest in every large dataset: GEMS/Food median 2–4 µg/kg fresh weight (only lamb’s lettuce lower), the lowest of the five FDA foods at 5.1 µg/kg, 6 µg/kg against 25 µg/kg for lettuce in the UK total diet study, and 4.6–5.4 µg/kg in an English field trial.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →2FY2018-FY2020 TDS Elements Analytical ResultsFDA Total Diet StudyRead source →19Effect of Climatic Conditions, and Agronomic Practices Used in Organic and Conventional Crop Production on Yield and Nutritional Composition Parameters in Potato, Cabbage, Lettuce and Onion; Results from the Long-Term NFSC-TrialsAgronomyRead source →18Total Diet Study of metals and other elements in foodFood and Environment Research Agency report for the UK Food Standards Agency, Fera report 15/06, project FS102081Read source → Head and iceberg lettuce were lower than leaf and romaine lettuce (GEMS/Food 11–12 against 19–20 µg/kg; FDA 27 against 57 µg/kg; Gran Canaria iceberg 4.0 against romaine 11.9 µg/kg), and spinach was highest in all of them.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →2FY2018-FY2020 TDS Elements Analytical ResultsFDA Total Diet StudyRead source →15Evaluation of essential, toxic and potentially toxic elements in leafy vegetables grown in the Canary IslandsToxicsRead source → Lamb’s lettuce and watercress were among the lowest-cadmium leaves where they were measured (GEMS/Food lamb’s lettuce median 0–2 µg/kg; Gran Canaria 1.1 and 1.9 µg/kg).1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →15Evaluation of essential, toxic and potentially toxic elements in leafy vegetables grown in the Canary IslandsToxicsRead source → Because the order repeats across independent European, US and Spanish datasets, this is a supported pattern for cadmium, not a single-study observation.
Lead and arsenic do not follow the cadmium order, so the lowest-cadmium leaf is not automatically lowest in everything.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →15Evaluation of essential, toxic and potentially toxic elements in leafy vegetables grown in the Canary IslandsToxicsRead source → Lamb’s lettuce had the highest total-arsenic median among the leaves in GEMS/Food (22–23 µg/kg) and in Gran Canaria (18.9 µg/kg), and a higher lead median (23–25 µg/kg) than lettuce or cabbage.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →15Evaluation of essential, toxic and potentially toxic elements in leafy vegetables grown in the Canary IslandsToxicsRead source → Iceberg lettuce was lowest for lead (2.0 µg/kg), arsenic (0.5 µg/kg) and aluminium (1,581 µg/kg) in Gran Canaria, and FDA detected lead in none of 27 iceberg or cabbage composites.2FY2018-FY2020 TDS Elements Analytical ResultsFDA Total Diet StudyRead source →15Evaluation of essential, toxic and potentially toxic elements in leafy vegetables grown in the Canary IslandsToxicsRead source → In Hong Kong, water spinach carried 74 µg/kg inorganic arsenic as eaten, against non-detects for European cabbage and lettuce, and watercress had the highest lead (96 µg/kg).5The First Hong Kong Total Diet Study: Inorganic ArsenicCentre for Food Safety, Food and Environmental Hygiene Department, Government of the Hong Kong Special Administrative RegionRead source →16The First Hong Kong Total Diet Study: Metallic ContaminantsCentre for Food Safety, Food and Environmental Hygiene Department, Government of the Hong Kong Special Administrative RegionRead source →
Origin comparisons are weaker, because GEMS/Food records the reporting region rather than where leaves were grown and the regions cover different years.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source → Lead was higher in Western Pacific submissions from 2001–2009 (median about 107 µg/kg for leafy vegetables, 47–50 µg/kg for pak choi) than in European ones (4–21 and 0–10 µg/kg), while cadmium in spinach and leaf lettuce was higher in the Americas (125 and 40 µg/kg) than in Europe (40 and 18–19 µg/kg).1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source → These are tentative regional differences, not a supported country ranking; within single studies, the clearest lower-contamination origin is simply a site away from industry, mining, roads and contaminated water.6Mercury pollution in vegetables, grains and soils from areas surrounding coal-fired power plantsScientific ReportsRead source →10Is it healthy urban agriculture? Human exposure to potentially toxic elements in urban gardens from Andalusia, SpainEnvironmental Science and Pollution ResearchRead source →17Nutrients and non-essential metals in darkibor kale grown at urban and rural farms: a pilot studyPLOS ONERead source →23BIO-ACCUMULATION OF HEAVY METALS IN SOME NORTHPeer-reviewed articleRead source →
Which leaves measured lowest in arsenic?
In Gran Canaria supermarkets, iceberg (0.5 µg/kg) and romaine lettuce (1.6 µg/kg) had the lowest total arsenic and lamb’s lettuce the highest (18.9 µg/kg).15Evaluation of essential, toxic and potentially toxic elements in leafy vegetables grown in the Canary IslandsToxicsRead source → In Hong Kong, inorganic arsenic was not detected in European cabbage or lettuce, while water spinach averaged 74 µg/kg.5The First Hong Kong Total Diet Study: Inorganic ArsenicCentre for Food Safety, Food and Environmental Hygiene Department, Government of the Hong Kong Special Administrative RegionRead source → Near contaminated irrigation in Gazipur, the aquatic leaf helencha carried 1.85–2.02 mg/kg arsenic dry weight.24Arsenic Contamination of Water-Soil-Crop System in an Industrial Area of BangladeshInternational Journal of EnvironmentRead source →
Which leaves were lowest in nickel and aluminium?
In Gran Canaria, watercress (33 µg/kg) and romaine lettuce (39 µg/kg) had the lowest nickel and spinach the highest (86 µg/kg), while aluminium ranged from 1,581 µg/kg in iceberg lettuce to 41,947 µg/kg in spinach.15Evaluation of essential, toxic and potentially toxic elements in leafy vegetables grown in the Canary IslandsToxicsRead source → In Hong Kong, Chinese cabbage (200 µg/kg) and European cabbage and lettuce (320 µg/kg) were lowest in aluminium and Chinese spinach highest (35,000 µg/kg).16The First Hong Kong Total Diet Study: Metallic ContaminantsCentre for Food Safety, Food and Environmental Hygiene Department, Government of the Hong Kong Special Administrative RegionRead source →
Washing, preparation and derivatives
Washing removes surface deposits but not metal taken up into the leaf.6Mercury pollution in vegetables, grains and soils from areas surrounding coal-fired power plantsScientific ReportsRead source →7Concentrations and health risks of selected elements in leafy vegetables: a comparison between roadside open-air markets and large stores in Johannesburg, South AfricaNot reportedRead source →25Quantification and Reduction in Heavy Metal Residues in Some Fruits and Vegetables: A Case Study Galați County, RomaniaHorticulturaeRead source → Washing halves of the same Johannesburg roadside purchases lowered the group mean cadmium from 51.5 to 6.8 mg/kg and nickel from 199 to 8 mg/kg dry weight, yet the washed leaves stayed above supermarket leaves for cadmium.7Concentrations and health risks of selected elements in leafy vegetables: a comparison between roadside open-air markets and large stores in Johannesburg, South AfricaNot reportedRead source → A 10% vinegar soak lowered lead in market lettuce from 155 to 95 µg/kg fresh weight but left farm lettuce unchanged (73.3 against 73.1 µg/kg), consistent with removal of surface lead only.25Quantification and Reduction in Heavy Metal Residues in Some Fruits and Vegetables: A Case Study Galați County, RomaniaHorticulturaeRead source →
Cooking acts differently on different metals and leaves. Boiling in clean water removed 55–69% of the arsenic from amaranth but only 3–31% from spinach, while boiling market spinach in water containing 1,000 µg/L arsenic raised it from 0.01 to 0.75 µg/g fresh mass.14Uptake of Arsenic by Irrigated Vegetables and Cooked Food Products in Burkina FasoFrontiers in WaterRead source → Cooking with coconut milk or oil generally lowered the in-vitro bioaccessible fraction of nickel, cadmium, chromium and lead in Sri Lankan leaves, but not in every case, and bioaccessibility is not the total concentration.26Bioaccessibility of Heavy Metals in Raw and Processed Alternanthera sessilis and Centella asiatica: An In vitro StudyNature Environment and Pollution Technology25(1):D1830Read source → Frozen, canned and fermented leaves are reported separately in GEMS/Food; sauerkraut, for example, had a lead median of 13–14 µg/kg, which is not a measure of what fermentation does to a given cabbage.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →
Controls for leafy greens
Controls for leafy greens act mainly where the leaf grows: choosing the site, the irrigation water and the soil inputs.6Mercury pollution in vegetables, grains and soils from areas surrounding coal-fired power plantsScientific ReportsRead source →14Uptake of Arsenic by Irrigated Vegetables and Cooked Food Products in Burkina FasoFrontiers in WaterRead source →20Impact of Physical Interventions, Phosphorus Fertilization, and the Utilization of Soil Amendments on the Absorption of Cadmium by Lettuce Grown in a Solar-Powered GreenhouseBiologyRead source → Covering contaminated greenhouse soil with 15 cm of clean soil cut lettuce cadmium by 62% (2.57 to 0.98 mg/kg dry weight), and a biochar–alginate amendment cut it by 63% in pots and 77% in the field.20Impact of Physical Interventions, Phosphorus Fertilization, and the Utilization of Soil Amendments on the Absorption of Cadmium by Lettuce Grown in a Solar-Powered GreenhouseBiologyRead source →27Development and Application of a Modified Biochar-Calcium Alginate Composite (MB-CA) for In Situ Remediation of Cadmium-Contaminated SoilGelsRead source → Arsenic-free irrigation water is the control that the irrigation trials support for arsenic.14Uptake of Arsenic by Irrigated Vegetables and Cooked Food Products in Burkina FasoFrontiers in WaterRead source →28Arsenic uptake and accumulation in bean and lettuce plants at different developmental stagesEnvironmental Science and Pollution ResearchRead source →
After harvest, washing and testing are the levers. Washing lowers surface lead, mercury and cadmium, and testing finds high lots without removing anything; the EU limits apply to the edible part after washing.6Mercury pollution in vegetables, grains and soils from areas surrounding coal-fired power plantsScientific ReportsRead source →7Concentrations and health risks of selected elements in leafy vegetables: a comparison between roadside open-air markets and large stores in Johannesburg, South AfricaNot reportedRead source →8Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source → Most amendment evidence comes from pot or greenhouse experiments with spiked soils, so effect sizes do not transfer directly to commercial fields.29Humic + Fulvic acid mitigated Cd adverse effects on plant growth, physiology and biochemical properties of garden cressScientific Reports 11: 8040Read source →30The Cd:Zn ratio in a soil affects Cd toxicity in spinach (Spinacea oleracea L.)Pakistan Journal of Agricultural SciencesRead source →31Minimization of cadmium toxicity and improvement in growth and biochemical attributes of spinach by using acidified biocharScientific ReportsRead source →
Filter findings
14 control comparisons
About this table
| Control | Material | Effects | Evidence | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Grow leaves away from coal-fired power plantscultivation · As reported | Leafy greens | tHg · Decreased tHg · Decreased | 1Li et al. 2017 HMI source record · Original source One dataset family. Linked analytes, repeated seasons and reports are not independent studies. Effects and comparators Total mercuryDecreasedLower mercury farther from the plant (observational)Water spinach mean total mercury1 km versus 10 kmFresh weightHigh · HMI calculates the relative change from the retained source means: (23.88 − 86.69) / 86.69 × 100 = -72.5%. This is a descriptive contrast of means; no confidence interval or significance test is inferred. Total mercuryDecreasedLower mercury farther from the plant (observational)Lettuce mean total mercury1 km versus 10 kmFresh weightHigh · HMI calculates the relative change from the retained source means: (7.23 − 39.04) / 39.04 × 100 = -81.5%. This is a descriptive contrast of means; no confidence interval or significance test is inferred. Farms at 1, 3, 5 and 10 km from the plants and grocery-store produce. How magnitude is classified HMI editorial magnitude bands: Low is a change greater than 0% and below 25%; Moderate is 25% to below 50%; High is 50% or more, relative to that control’s own comparator. Direction is shown separately. These are descriptive display bands, not validated efficacy cutoffs, occurrence-baseline positions, statistical significance, evidence certainty, safety or net benefit. A range spanning bands retains both labels; a qualitative label has no implied percentage. Control in context Leaf mercury fell with distance from two coal-fired power plants, from 86.7 µg/kg at 1 km to 23.9 µg/kg at 10 km in water spinach; grocery-store leaves from more than 55 km away were below 1 µg/kg.
Design, applicability and operating conditions Observational field or market comparison. Areas with point-source air emissions. One study family. As reported by the source Reduces metal reaching or remaining in the edible leaf, or detects it before release. Source measurements and linked outcomes
Observational; distance is confounded with soil mercury. No percentage calculated unless shown in outcomes. Basis as stated in each outcome; no conversion. Trade-offs and implementation As reported in outcomes. Not assessed unless stated. Not assessed. Not quantified. Not assessed. Applicability and verification Areas with point-source air emissions. Lot testing. See source. Extraction limits and next actions
Provenance and assessment As disclosed by the source. Source facts and HMI calculations are distinguished above. These selected source findings do not constitute a systematic review; outcome-level appraisal remains incomplete. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Irrigate with arsenic-free watercultivation · As reported | Leafy greens | tAs · Increased tAs · Increased | 1Clair-Caliot et al. 2021 HMI source record · Original source One dataset family. Linked analytes, repeated seasons and reports are not independent studies. Effects and comparators Total arsenicIncreasedHigher arsenic with contaminated irrigation waterSpinach leaf total arsenic0 versus 500 µg/L irrigation arsenicDry weightHigh · HMI calculates the relative change from the retained source means: (6570 − 1050) / 1050 × 100 = 525.7%. This is a descriptive contrast of means; no confidence interval or significance test is inferred. Total arsenicIncreasedHigher arsenic with contaminated irrigation waterAmaranth leaf total arsenic0 versus 500 µg/L irrigation arsenicDry weightHigh · HMI calculates the relative change from the retained source means: (2230 − 120) / 120 × 100 = 1758.3%. This is a descriptive contrast of means; no confidence interval or significance test is inferred. Arsenic-free irrigation versus water spiked to 100, 500 and 1,000 µg/L arsenic(V), same soil. How magnitude is classified HMI editorial magnitude bands: Low is a change greater than 0% and below 25%; Moderate is 25% to below 50%; High is 50% or more, relative to that control’s own comparator. Direction is shown separately. These are descriptive display bands, not validated efficacy cutoffs, occurrence-baseline positions, statistical significance, evidence certainty, safety or net benefit. A range spanning bands retains both labels; a qualitative label has no implied percentage. Control in context Leaf arsenic rose roughly in proportion to irrigation-water arsenic; at 500 µg/L spinach reached 6.57 µg/g dry matter against 1.05 µg/g with arsenic-free water.
Design, applicability and operating conditions Measured paired comparison or trial. Areas with arsenic-bearing groundwater. One study family. As reported by the source Reduces metal reaching or remaining in the edible leaf, or detects it before release. Source measurements and linked outcomes
Greenhouse experiment; control-plant values differ between the paper’s tables. No percentage calculated unless shown in outcomes. Basis as stated in each outcome; no conversion. Trade-offs and implementation As reported in outcomes. Not assessed unless stated. Not assessed. Not quantified. Not assessed. Applicability and verification Areas with arsenic-bearing groundwater. Lot testing. See source. Extraction limits and next actions
Provenance and assessment As disclosed by the source. Source facts and HMI calculations are distinguished above. These selected source findings do not constitute a systematic review; outcome-level appraisal remains incomplete. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Cover contaminated soil with clean soilsoil remediation · As reported | Leafy greens | Cd · Decreased Cd · Decreased | 1Zhang et al. 2024 HMI source record · Original source One dataset family. Linked analytes, repeated seasons and reports are not independent studies. Effects and comparators CadmiumDecreasedLower cadmium with a clean-soil coverLettuce cadmiumUntreated versus 15 cm coverDry weightHigh · HMI calculates the relative change from the retained source means: (980 − 2570) / 2570 × 100 = -61.9%. This is a descriptive contrast of means; no confidence interval or significance test is inferred. CadmiumDecreasedLower cadmium with deep ploughingLettuce cadmiumUntreated versus 40 cm ploughingDry weightModerate · HMI calculates the relative change from the retained source means: (1840 − 2570) / 2570 × 100 = -28.4%. This is a descriptive contrast of means; no confidence interval or significance test is inferred. Covered or ploughed plots versus untreated plots in the same greenhouse. How magnitude is classified HMI editorial magnitude bands: Low is a change greater than 0% and below 25%; Moderate is 25% to below 50%; High is 50% or more, relative to that control’s own comparator. Direction is shown separately. These are descriptive display bands, not validated efficacy cutoffs, occurrence-baseline positions, statistical significance, evidence certainty, safety or net benefit. A range spanning bands retains both labels; a qualitative label has no implied percentage. Control in context A 15 cm layer of uncontaminated soil cut lettuce cadmium by 62%; deep ploughing to 30–50 cm cut it by 11–28%.
Design, applicability and operating conditions Measured paired comparison or trial. Protected cultivation on naturally cadmium-enriched soil. One study family. As reported by the source Reduces metal reaching or remaining in the edible leaf, or detects it before release. Source measurements and linked outcomes
One greenhouse; the abstract and figure give different values (figure used). No percentage calculated unless shown in outcomes. Basis as stated in each outcome; no conversion. Trade-offs and implementation As reported in outcomes. Not assessed unless stated. Not assessed. Not quantified. Not assessed. Applicability and verification Protected cultivation on naturally cadmium-enriched soil. Lot testing. See source. Extraction limits and next actions
Provenance and assessment As disclosed by the source. Source facts and HMI calculations are distinguished above. These selected source findings do not constitute a systematic review; outcome-level appraisal remains incomplete. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Limit phosphate fertiliser on cadmium-prone soilcultivation · As reported | Leafy greens | Cd · Increased | 1Zhang et al. 2024 HMI source record · Original source One dataset family. Linked analytes, repeated seasons and reports are not independent studies. Effects and comparators CadmiumIncreasedHigher cadmium with superphosphateLettuce cadmiumNo phosphate versus 225 kg/ha P2O5 superphosphateDry weightHigh · HMI calculates the relative change from the retained source means: (1890 − 990) / 990 × 100 = 90.9%. This is a descriptive contrast of means; no confidence interval or significance test is inferred. Fertilised versus unfertilised plots. How magnitude is classified HMI editorial magnitude bands: Low is a change greater than 0% and below 25%; Moderate is 25% to below 50%; High is 50% or more, relative to that control’s own comparator. Direction is shown separately. These are descriptive display bands, not validated efficacy cutoffs, occurrence-baseline positions, statistical significance, evidence certainty, safety or net benefit. A range spanning bands retains both labels; a qualitative label has no implied percentage. Control in context All three phosphate fertilisers raised lettuce cadmium, by up to 91% at the highest superphosphate rate.
Design, applicability and operating conditions Measured paired comparison or trial. Phosphate-fertilised leafy crops. One study family. As reported by the source Reduces metal reaching or remaining in the edible leaf, or detects it before release. Source measurements and linked outcomes
One site; untreated means differ between experiments. No percentage calculated unless shown in outcomes. Basis as stated in each outcome; no conversion. Trade-offs and implementation As reported in outcomes. Not assessed unless stated. Not assessed. Not quantified. Not assessed. Applicability and verification Phosphate-fertilised leafy crops. Lot testing. See source. Extraction limits and next actions
Provenance and assessment As disclosed by the source. Source facts and HMI calculations are distinguished above. These selected source findings do not constitute a systematic review; outcome-level appraisal remains incomplete. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Use farmyard manure rather than mineral NPKcultivation · As reported | Leafy greens | Cd · Decreased Cd · Decreased Ni · Decreased Pb · Increased | 1Rempelos et al. 2023 HMI source record · Original source One dataset family. Linked analytes, repeated seasons and reports are not independent studies. Effects and comparators CadmiumDecreasedLower cadmium with manureLettuce cadmiumNPK versus manureFresh weightLow · HMI calculates the relative change from the retained source means: (11.4 − 14.2) / 14.2 × 100 = -19.7%. This is a descriptive contrast of means; no confidence interval or significance test is inferred. CadmiumDecreasedLower cadmium with manureCabbage cadmiumNPK versus manureFresh weightLow · HMI calculates the relative change from the retained source means: (4.6 − 5.4) / 5.4 × 100 = -14.8%. This is a descriptive contrast of means; no confidence interval or significance test is inferred. NickelDecreasedLower nickel with manureCabbage nickelNPK versus manureFresh weightModerate · HMI calculates the relative change from the retained source means: (17 − 23) / 23 × 100 = -26.1%. This is a descriptive contrast of means; no confidence interval or significance test is inferred. LeadIncreasedSlightly higher, not significantLettuce leadNPK versus manureFresh weightLow · HMI calculates the relative change from the retained source means: (9.3 − 8) / 8 × 100 = 16.3%. This is a descriptive contrast of means; no confidence interval or significance test is inferred. Mineral NPK versus farmyard manure, main-effect means over six seasons. How magnitude is classified HMI editorial magnitude bands: Low is a change greater than 0% and below 25%; Moderate is 25% to below 50%; High is 50% or more, relative to that control’s own comparator. Direction is shown separately. These are descriptive display bands, not validated efficacy cutoffs, occurrence-baseline positions, statistical significance, evidence certainty, safety or net benefit. A range spanning bands retains both labels; a qualitative label has no implied percentage. Control in context Farmyard manure gave lower cadmium in cabbage and lettuce and lower nickel in cabbage than mineral fertiliser at the same nitrogen input; lead was unaffected.
Design, applicability and operating conditions Measured paired comparison or trial. Temperate field vegetables. One study family. As reported by the source Reduces metal reaching or remaining in the edible leaf, or detects it before release. Source measurements and linked outcomes
One trial site; per-mean sample counts not stated. No percentage calculated unless shown in outcomes. Basis as stated in each outcome; no conversion. Trade-offs and implementation As reported in outcomes. Not assessed unless stated. Not assessed. Not quantified. Not assessed. Applicability and verification Temperate field vegetables. Lot testing. See source. Extraction limits and next actions
Provenance and assessment As disclosed by the source. Source facts and HMI calculations are distinguished above. These selected source findings do not constitute a systematic review; outcome-level appraisal remains incomplete. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Check manures for lead before usecultivation · As reported | Leafy greens | Pb · Increased tAs · Decreased | 1Yaman Türkkan and Kibar 2022 HMI source record · Original source One dataset family. Linked analytes, repeated seasons and reports are not independent studies. Effects and comparators LeadIncreasedHigher lead with chicken manureSpinach leadControl versus chicken manureBasis not statedHigh · HMI calculates the relative change from the retained source means: (750 − 110) / 110 × 100 = 581.8%. This is a descriptive contrast of means; no confidence interval or significance test is inferred. Total arsenicDecreasedLower arsenic with chicken manureSpinach arsenicControl versus chicken manureBasis not statedLow · HMI calculates the relative change from the retained source means: (350 − 410) / 410 × 100 = -14.6%. This is a descriptive contrast of means; no confidence interval or significance test is inferred. Unfertilised control plots versus each fertiliser. How magnitude is classified HMI editorial magnitude bands: Low is a change greater than 0% and below 25%; Moderate is 25% to below 50%; High is 50% or more, relative to that control’s own comparator. Direction is shown separately. These are descriptive display bands, not validated efficacy cutoffs, occurrence-baseline positions, statistical significance, evidence certainty, safety or net benefit. A range spanning bands retains both labels; a qualitative label has no implied percentage. Control in context Chicken, sheep and cattle manure and vermicompost raised spinach lead four- to seven-fold while lowering arsenic, chromium and nickel.
Design, applicability and operating conditions Measured paired comparison or trial. Manure-fertilised leafy crops. One study family. As reported by the source Reduces metal reaching or remaining in the edible leaf, or detects it before release. Source measurements and linked outcomes
One season; basis not stated; no detection limits reported. No percentage calculated unless shown in outcomes. Basis as stated in each outcome; no conversion. Trade-offs and implementation As reported in outcomes. Not assessed unless stated. Not assessed. Not quantified. Not assessed. Applicability and verification Manure-fertilised leafy crops. Lot testing. See source. Extraction limits and next actions
Provenance and assessment As disclosed by the source. Source facts and HMI calculations are distinguished above. These selected source findings do not constitute a systematic review; outcome-level appraisal remains incomplete. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Immobilise soil cadmium with a biochar–alginate compositesoil remediation · As reported | Leafy greens | Cd · Decreased | 1Sun et al. 2025 HMI source record · Original source One dataset family. Linked analytes, repeated seasons and reports are not independent studies. Effects and comparators CadmiumDecreasedLower cadmium (63–77%)Lettuce cadmiumUnamended versus amendedBasis not statedNot quantified · No supported numerical treatment contrast is retained for this finding. A direction, market association, source-selection strategy or guidance statement alone does not establish a percentage change. Amended versus unamended soil. How magnitude is classified HMI editorial magnitude bands: Low is a change greater than 0% and below 25%; Moderate is 25% to below 50%; High is 50% or more, relative to that control’s own comparator. Direction is shown separately. These are descriptive display bands, not validated efficacy cutoffs, occurrence-baseline positions, statistical significance, evidence certainty, safety or net benefit. A range spanning bands retains both labels; a qualitative label has no implied percentage. Control in context The amendment cut lettuce cadmium by 63% in pots and 77% in the field.
Design, applicability and operating conditions Measured paired comparison or trial. Cadmium-polluted soils. One study family. As reported by the source Reduces metal reaching or remaining in the edible leaf, or detects it before release. Source measurements and linked outcomes
Values shown only in a figure; percentages from the abstract. No percentage calculated unless shown in outcomes. Basis as stated in each outcome; no conversion. Trade-offs and implementation As reported in outcomes. Not assessed unless stated. Not assessed. Not quantified. Not assessed. Applicability and verification Cadmium-polluted soils. Lot testing. See source. Extraction limits and next actions
Provenance and assessment As disclosed by the source. Source facts and HMI calculations are distinguished above. These selected source findings do not constitute a systematic review; outcome-level appraisal remains incomplete. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Test immobilising amendments before relying on themsoil remediation · As reported | Leafy greens | Cd · Increased | 1Guo et al. 2024 HMI source record · Original source One dataset family. Linked analytes, repeated seasons and reports are not independent studies. Effects and comparators CadmiumIncreasedHigher cadmium with wollastonitePak choi cadmiumUntreated versus wollastonite, moderate soilDry weightHigh · HMI calculates the relative change from the retained source means: (16620 − 7630) / 7630 × 100 = 117.8%. This is a descriptive contrast of means; no confidence interval or significance test is inferred. Amended versus untreated soil at the same cadmium level. How magnitude is classified HMI editorial magnitude bands: Low is a change greater than 0% and below 25%; Moderate is 25% to below 50%; High is 50% or more, relative to that control’s own comparator. Direction is shown separately. These are descriptive display bands, not validated efficacy cutoffs, occurrence-baseline positions, statistical significance, evidence certainty, safety or net benefit. A range spanning bands retains both labels; a qualitative label has no implied percentage. Control in context Wollastonite-based amendments raised soil pH but increased pak choi cadmium in most soils.
Design, applicability and operating conditions Measured paired comparison or trial. Any soil amendment programme. One study family. As reported by the source Reduces metal reaching or remaining in the edible leaf, or detects it before release. Source measurements and linked outcomes
Pot experiment. No percentage calculated unless shown in outcomes. Basis as stated in each outcome; no conversion. Trade-offs and implementation As reported in outcomes. Not assessed unless stated. Not assessed. Not quantified. Not assessed. Applicability and verification Any soil amendment programme. Lot testing. See source. Extraction limits and next actions
Provenance and assessment As disclosed by the source. Source facts and HMI calculations are distinguished above. These selected source findings do not constitute a systematic review; outcome-level appraisal remains incomplete. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Wash leaves thoroughly under running waterfood preparation · As reported | Leafy greens | Cd · Decreased Ni · Decreased | 1Adhikari et al. 2024 HMI source record · Original source One dataset family. Linked analytes, repeated seasons and reports are not independent studies. Effects and comparators CadmiumDecreasedLower cadmium after washingGroup mean cadmiumUnwashed versus washedDry weightHigh · HMI calculates the relative change from the retained source means: (6840 − 51470) / 51470 × 100 = -86.7%. This is a descriptive contrast of means; no confidence interval or significance test is inferred. NickelDecreasedLower nickel after washingGroup mean nickelUnwashed versus washedDry weightHigh · HMI calculates the relative change from the retained source means: (8000 − 198710) / 198710 × 100 = -96.0%. This is a descriptive contrast of means; no confidence interval or significance test is inferred. Unwashed versus washed halves of one composite. How magnitude is classified HMI editorial magnitude bands: Low is a change greater than 0% and below 25%; Moderate is 25% to below 50%; High is 50% or more, relative to that control’s own comparator. Direction is shown separately. These are descriptive display bands, not validated efficacy cutoffs, occurrence-baseline positions, statistical significance, evidence certainty, safety or net benefit. A range spanning bands retains both labels; a qualitative label has no implied percentage. Control in context Washing halves of the same roadside purchases lowered the group mean cadmium from 51.5 to 6.8 mg/kg and nickel from 199 to 8 mg/kg dry weight; washed roadside leaves stayed above supermarket leaves for cadmium.
Design, applicability and operating conditions Measured paired comparison or trial. Leaves exposed to traffic dust. One study family. As reported by the source Reduces metal reaching or remaining in the edible leaf, or detects it before release. Source measurements and linked outcomes
Single composites; some leaf–metal pairs rose after washing. No percentage calculated unless shown in outcomes. Basis as stated in each outcome; no conversion. Trade-offs and implementation As reported in outcomes. Not assessed unless stated. Not assessed. Not quantified. Not assessed. Applicability and verification Leaves exposed to traffic dust. Lot testing. See source. Extraction limits and next actions
Provenance and assessment As disclosed by the source. Source facts and HMI calculations are distinguished above. These selected source findings do not constitute a systematic review; outcome-level appraisal remains incomplete. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Test lots against leaf-specific limitslot acceptance and release · As reported | Leafy greens | Cd · Context Pb · Context | GuidanceRegulation (EU) 2023/915 HMI source record · Original source Official guidance; not an independent experimental dataset. Reused across actions without adding studies. Effects and comparators CadmiumLimit 0.10–0.20 mg/kgCadmium maximum levelLeaf vegetables; spinachFresh weightContext · No supported numerical treatment contrast is retained for this finding. A direction, market association, source-selection strategy or guidance statement alone does not establish a percentage change. LeadLimit 0.30 mg/kgLead maximum levelLeaf vegetablesFresh weightContext · No supported numerical treatment contrast is retained for this finding. A direction, market association, source-selection strategy or guidance statement alone does not establish a percentage change. Regulatory maximum levels. How magnitude is classified HMI editorial magnitude bands: Low is a change greater than 0% and below 25%; Moderate is 25% to below 50%; High is 50% or more, relative to that control’s own comparator. Direction is shown separately. These are descriptive display bands, not validated efficacy cutoffs, occurrence-baseline positions, statistical significance, evidence certainty, safety or net benefit. A range spanning bands retains both labels; a qualitative label has no implied percentage. Control in context Lead 0.30 mg/kg and cadmium 0.10 mg/kg (0.20 mg/kg for spinach, similar leaves and fresh herbs), wet weight, after washing and separating the edible part.
Design, applicability and operating conditions Regulatory or author guidance. Leaves placed on the EU market. One study family. As reported by the source Reduces metal reaching or remaining in the edible leaf, or detects it before release. Source measurements and linked outcomes
A limit detects non-compliant lots; it does not lower contamination. No percentage calculated unless shown in outcomes. Basis as stated in each outcome; no conversion. Trade-offs and implementation As reported in outcomes. Not assessed unless stated. Not assessed. Not quantified. Not assessed. Applicability and verification Leaves placed on the EU market. Lot testing. See source. Extraction limits and next actions
Provenance and assessment As disclosed by the source. Source facts and HMI calculations are distinguished above. These selected source findings do not constitute a systematic review; outcome-level appraisal remains incomplete. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Buy from controlled supply rather than roadside stallssourcing and procurement · As reported | Leafy greens | Cd · Decreased | 1Adhikari et al. 2024 HMI source record · Original source One dataset family. Linked analytes, repeated seasons and reports are not independent studies. Effects and comparators CadmiumDecreasedLower cadmium in large-store leaves (observational)Group mean cadmiumWashed roadside versus supermarketDry weightHigh · HMI calculates the relative change from the retained source means: (2500 − 6840) / 6840 × 100 = -63.5%. This is a descriptive contrast of means; no confidence interval or significance test is inferred. Roadside versus supermarket and vegetable-market purchases. How magnitude is classified HMI editorial magnitude bands: Low is a change greater than 0% and below 25%; Moderate is 25% to below 50%; High is 50% or more, relative to that control’s own comparator. Direction is shown separately. These are descriptive display bands, not validated efficacy cutoffs, occurrence-baseline positions, statistical significance, evidence certainty, safety or net benefit. A range spanning bands retains both labels; a qualitative label has no implied percentage. Control in context Washed roadside leaves still carried more cadmium (6.84 mg/kg) than leaves from large stores (2.10–2.50 mg/kg dry weight).
Design, applicability and operating conditions Observational field or market comparison. Urban retail sourcing. One study family. As reported by the source Reduces metal reaching or remaining in the edible leaf, or detects it before release. Source measurements and linked outcomes
Observational; growing sites not identified. No percentage calculated unless shown in outcomes. Basis as stated in each outcome; no conversion. Trade-offs and implementation As reported in outcomes. Not assessed unless stated. Not assessed. Not quantified. Not assessed. Applicability and verification Urban retail sourcing. Lot testing. See source. Extraction limits and next actions
Provenance and assessment As disclosed by the source. Source facts and HMI calculations are distinguished above. These selected source findings do not constitute a systematic review; outcome-level appraisal remains incomplete. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Rinse leaves before eatinghome preparation · As reported | Leafy greens | tHg · Decreased | 1Li et al. 2017 HMI source record · Original source One dataset family. Linked analytes, repeated seasons and reports are not independent studies. Effects and comparators Total mercuryDecreasedLower mercury after rinsingLettuce total mercuryUnwashed versus washed, 1 km siteFresh weightModerate · HMI calculates the relative change from the retained source means: (21.03 − 35.7) / 35.7 × 100 = -41.1%. This is a descriptive contrast of means; no confidence interval or significance test is inferred. Unwashed versus rinsed leaves from the same site. How magnitude is classified HMI editorial magnitude bands: Low is a change greater than 0% and below 25%; Moderate is 25% to below 50%; High is 50% or more, relative to that control’s own comparator. Direction is shown separately. These are descriptive display bands, not validated efficacy cutoffs, occurrence-baseline positions, statistical significance, evidence certainty, safety or net benefit. A range spanning bands retains both labels; a qualitative label has no implied percentage. Control in context Rinsing lowered leaf mercury by 19–63%.
Design, applicability and operating conditions Measured paired comparison or trial. Leaves with deposited dust or fly ash. One study family. As reported by the source Reduces metal reaching or remaining in the edible leaf, or detects it before release. Source measurements and linked outcomes
Removes surface mercury only. No percentage calculated unless shown in outcomes. Basis as stated in each outcome; no conversion. Trade-offs and implementation As reported in outcomes. Not assessed unless stated. Not assessed. Not quantified. Not assessed. Applicability and verification Leaves with deposited dust or fly ash. Lot testing. See source. Extraction limits and next actions
Provenance and assessment As disclosed by the source. Source facts and HMI calculations are distinguished above. These selected source findings do not constitute a systematic review; outcome-level appraisal remains incomplete. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Vinegar soak after washinghome preparation · As reported | Leafy greens | Pb · Decreased Pb · Decreased | 1Bora et al. 2022 HMI source record · Original source One dataset family. Linked analytes, repeated seasons and reports are not independent studies. Effects and comparators LeadDecreasedLower lead after vinegar soakMarket lettuce leadTap water versus 10% vinegarFresh weightModerate · HMI calculates the relative change from the retained source means: (94.52 − 155.07) / 155.07 × 100 = -39.0%. This is a descriptive contrast of means; no confidence interval or significance test is inferred. LeadDecreasedEssentially unchangedFarm lettuce leadTap water versus 10% vinegarFresh weightLow · HMI calculates the relative change from the retained source means: (73.14 − 73.25) / 73.25 × 100 = -0.2%. This is a descriptive contrast of means; no confidence interval or significance test is inferred. Tap-water washing versus vinegar soak. How magnitude is classified HMI editorial magnitude bands: Low is a change greater than 0% and below 25%; Moderate is 25% to below 50%; High is 50% or more, relative to that control’s own comparator. Direction is shown separately. These are descriptive display bands, not validated efficacy cutoffs, occurrence-baseline positions, statistical significance, evidence certainty, safety or net benefit. A range spanning bands retains both labels; a qualitative label has no implied percentage. Control in context A 10% vinegar soak lowered market-lettuce lead from 155 to 95 µg/kg and cadmium from 46.9 to 36.2 µg/kg, but left farm lettuce lead unchanged.
Design, applicability and operating conditions Measured paired comparison or trial. Household or commercial washing. One study family. As reported by the source Reduces metal reaching or remaining in the edible leaf, or detects it before release. Source measurements and linked outcomes
Analytical replicates; lot identity across treatments not stated. No percentage calculated unless shown in outcomes. Basis as stated in each outcome; no conversion. Trade-offs and implementation As reported in outcomes. Not assessed unless stated. Not assessed. Not quantified. Not assessed. Applicability and verification Household or commercial washing. Lot testing. See source. Extraction limits and next actions
Provenance and assessment As disclosed by the source. Source facts and HMI calculations are distinguished above. These selected source findings do not constitute a systematic review; outcome-level appraisal remains incomplete. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Boil leaves in arsenic-free waterhome preparation · As reported | Leafy greens | tAs · Decreased tAs · Increased | 1Clair-Caliot et al. 2021 HMI source record · Original source One dataset family. Linked analytes, repeated seasons and reports are not independent studies. Effects and comparators Total arsenicDecreasedLower arsenic after boilingAmaranth total arsenicRaw versus boiledDry weightHigh · HMI calculates the relative change from the retained source means: (1010 − 2230) / 2230 × 100 = -54.7%. This is a descriptive contrast of means; no confidence interval or significance test is inferred. Total arsenicIncreasedHigher arsenic with contaminated cooking waterMarket spinach total arsenicClean versus 1,000 µg/L cooking waterFresh weightHigh · HMI calculates the relative change from the retained source means: (750 − 10) / 10 × 100 = 7400.0%. This is a descriptive contrast of means; no confidence interval or significance test is inferred. Raw versus boiled leaves; clean versus contaminated cooking water. How magnitude is classified HMI editorial magnitude bands: Low is a change greater than 0% and below 25%; Moderate is 25% to below 50%; High is 50% or more, relative to that control’s own comparator. Direction is shown separately. These are descriptive display bands, not validated efficacy cutoffs, occurrence-baseline positions, statistical significance, evidence certainty, safety or net benefit. A range spanning bands retains both labels; a qualitative label has no implied percentage. Control in context Boiling in excess clean water removed 55–69% of arsenic from amaranth but 3–31% from spinach; arsenic-contaminated cooking water added arsenic.
Design, applicability and operating conditions Measured paired comparison or trial. Arsenic-affected areas. One study family. As reported by the source Reduces metal reaching or remaining in the edible leaf, or detects it before release. Source measurements and linked outcomes
Greenhouse-grown leaves. No percentage calculated unless shown in outcomes. Basis as stated in each outcome; no conversion. Trade-offs and implementation As reported in outcomes. Not assessed unless stated. Not assessed. Not quantified. Not assessed. Applicability and verification Arsenic-affected areas. Lot testing. See source. Extraction limits and next actions
Provenance and assessment As disclosed by the source. Source facts and HMI calculations are distinguished above. These selected source findings do not constitute a systematic review; outcome-level appraisal remains incomplete. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
No matching controls in this selected evidence. This is a coverage boundary, not proof that no control exists. Change or reset the filters.
Trade-offs and failure modes
Soil amendments can raise the metal they are meant to lower.20Impact of Physical Interventions, Phosphorus Fertilization, and the Utilization of Soil Amendments on the Absorption of Cadmium by Lettuce Grown in a Solar-Powered GreenhouseBiologyRead source →32Effects of wollastonite and phosphate treatments on cadmium bioaccessibility and health risk in pak choi (Brassica rapa ssp. chinensis) grown in contaminated soilFrontiers in NutritionRead source → In the same greenhouse field trial that found clean-soil covering effective, corn-straw biochar raised lettuce cadmium at most application rates (by up to 44%) and attapulgite clay raised it by 24–56%.20Impact of Physical Interventions, Phosphorus Fertilization, and the Utilization of Soil Amendments on the Absorption of Cadmium by Lettuce Grown in a Solar-Powered GreenhouseBiologyRead source → Wollastonite raised pak choi cadmium in three of four soils, in one case from 7.63 to 16.62 mg/kg.32Effects of wollastonite and phosphate treatments on cadmium bioaccessibility and health risk in pak choi (Brassica rapa ssp. chinensis) grown in contaminated soilFrontiers in NutritionRead source →
Organic or manure-based growing is not uniformly cleaner. Manures that lowered arsenic, chromium and nickel in Turkish spinach raised its lead four- to seven-fold, while farmyard manure lowered cadmium in English cabbage and lettuce without changing lead.19Effect of Climatic Conditions, and Agronomic Practices Used in Organic and Conventional Crop Production on Yield and Nutritional Composition Parameters in Potato, Cabbage, Lettuce and Onion; Results from the Long-Term NFSC-TrialsAgronomyRead source →21compared to the chemical fertilizer. In organic fertilizer applications, heavy metal contents such as arsenic, chrome, cobalt, and nickel were found toPeer-reviewed articleRead source → Organic crop protection changed only cabbage nickel in the English trial, and organic certification was associated with lower cadmium and nickel in New Zealand market-garden vegetables but not consistently across leaves.19Effect of Climatic Conditions, and Agronomic Practices Used in Organic and Conventional Crop Production on Yield and Nutritional Composition Parameters in Potato, Cabbage, Lettuce and Onion; Results from the Long-Term NFSC-TrialsAgronomyRead source →33Assessment of Heavy Metals in Organic and Non-Organic Vegetables Post Severe Tropical Cyclone Gabrielle: A cross-sectional comparative analysisF1000ResearchRead source →
Choosing a leaf for low cadmium can trade off against other metals. Lamb’s lettuce is low in cadmium but higher in arsenic and lead than lettuce, and watercress was low in cadmium in Gran Canaria but highest in lead in Hong Kong.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →15Evaluation of essential, toxic and potentially toxic elements in leafy vegetables grown in the Canary IslandsToxicsRead source →16The First Hong Kong Total Diet Study: Metallic ContaminantsCentre for Food Safety, Food and Environmental Hygiene Department, Government of the Hong Kong Special Administrative RegionRead source →
Sampling and testing leafy greens
Results are only comparable when the basis is stated.17Nutrients and non-essential metals in darkibor kale grown at urban and rural farms: a pilot studyPLOS ONERead source →11Food survey: Levels and potential health risks of chromium, lead, zinc and copper content in fruits and vegetables consumed in AlgeriaFood and Chemical ToxicologyRead source → Many surveys report dry weight and then compare their values with fresh-weight limits, which overstates exceedance about ten-fold for leaves that are around 90% water.11Food survey: Levels and potential health risks of chromium, lead, zinc and copper content in fruits and vegetables consumed in AlgeriaFood and Chemical ToxicologyRead source →12Heavy Metals in Commonly Consumed Root and Leafy Vegetables in Dhaka City, Bangladesh, and Assessment of Associated Public Health RisksEnvironmental Systems ResearchRead source → A useful report gives the leaf, fresh-weight basis, the edible part analysed, whether it was washed, and the quantification limit for each metal.2FY2018-FY2020 TDS Elements Analytical ResultsFDA Total Diet StudyRead source →8Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →
Detection limits must sit well below the limits being checked.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →2FY2018-FY2020 TDS Elements Analytical ResultsFDA Total Diet StudyRead source → FDA reported lead down to 4 µg/kg, but several surveys report values below their own stated detection limits, and those rows are held.2FY2018-FY2020 TDS Elements Analytical ResultsFDA Total Diet StudyRead source →12Heavy Metals in Commonly Consumed Root and Leafy Vegetables in Dhaka City, Bangladesh, and Assessment of Associated Public Health RisksEnvironmental Systems ResearchRead source →13Enrichment Characteristics and Health Risk Assessment of Heavy Metals in Soil-Crop SystemsKSCE Journal of Civil EngineeringRead source → Total arsenic is not inorganic arsenic, and inorganic mercury submitted to GEMS/Food is not total mercury; the page keeps each species separate.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →5The First Hong Kong Total Diet Study: Inorganic ArsenicCentre for Food Safety, Food and Environmental Hygiene Department, Government of the Hong Kong Special Administrative RegionRead source →
Why do some studies report extreme values?
Some extreme values do not survive a check against their own papers. In a Dhaka study every detected lead value was below the method’s own detection limit, and in a Yuyao study every mustard mean was below the crop detection limits printed in the same paper.12Heavy Metals in Commonly Consumed Root and Leafy Vegetables in Dhaka City, Bangladesh, and Assessment of Associated Public Health RisksEnvironmental Systems ResearchRead source →13Enrichment Characteristics and Health Risk Assessment of Heavy Metals in Soil-Crop SystemsKSCE Journal of Civil EngineeringRead source → Other high values are real but site-specific, such as unwashed roadside coriander in Johannesburg.7Concentrations and health risks of selected elements in leafy vegetables: a comparison between roadside open-air markets and large stores in Johannesburg, South AfricaNot reportedRead source →
Sourcing and supplier evidence
Limits differ by market and by leaf.8Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →34General Standard for Contaminants and Toxins in Food and Feed (CXS 193-1995)Codex Alimentarius (Joint FAO/WHO Food Standards Programme)Read source →35China Releases the Standard for Maximum Levels of Contaminants in Foods (USDA FAS GAIN Report CH2023-0040, unofficial translation of GB 2762-2022)USDA Foreign Agricultural Service, Global Agricultural Information Network (GAIN), Report Number CH2023-0040Read source → The EU sets 0.30 mg/kg lead for leaf vegetables (fresh herbs excluded) and 0.10 mg/kg cadmium, or 0.20 mg/kg for spinach, similar leaves and fresh herbs, and from 1 July 2025 0.50 mg/kg nickel for leafy vegetables and 1.2 mg/kg for fresh herbs.8Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →36Commission Regulation (EU) 2024/1987 of 30 July 2024 amending Regulation (EU) 2023/915 as regards maximum levels of nickel in certain foodstuffsOfficial Journal of the European Union, L series, 2024/1987 (31.7.2024)Read source → Codex sets 0.3 mg/kg lead (the edition held in the library excludes spinach from that entry) and 0.2 mg/kg cadmium for leafy vegetables, China 0.3 mg/kg lead and 0.2 mg/kg cadmium, Korea the same values for spinach and Chinese cabbage, and Hong Kong 0.2 mg/kg cadmium.34General Standard for Contaminants and Toxins in Food and Feed (CXS 193-1995)Codex Alimentarius (Joint FAO/WHO Food Standards Programme)Read source →35China Releases the Standard for Maximum Levels of Contaminants in Foods (USDA FAS GAIN Report CH2023-0040, unofficial translation of GB 2762-2022)USDA Foreign Agricultural Service, Global Agricultural Information Network (GAIN), Report Number CH2023-0040Read source →37Article 2. Common Standards and Specifications for General FoodsKorean Food CodeRead source →38Guidelines on the Food Adulteration (Metallic Contamination) (Amendment) Regulation 2018USDA Foreign Agricultural Service GAIN Report HK1922, relaying the Hong Kong Centre for Food Safety Guidelines for the Food Adulteration (Metallic Contamination) (Amendment) Regulation 2018 (Cap. 132V sub. leg.)Read source →
Spinach is the leaf most likely to approach a cadmium limit.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →4Scientific Opinion of the Panel on Contaminants in the Food Chain on a request from the European Commission on cadmium in foodThe EFSA JournalRead source → In GEMS/Food, 105 of 2,262 spinach results were definitely above 0.20 mg/kg cadmium and 193 of 3,456 leaf lettuce results above 0.10 mg/kg, against 11 of 3,523 head cabbage results.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source → Ask for lot results by leaf with the growing site and irrigation source, because most datasets identify only the market and the clearest high-contamination cases are site-specific.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →3Heavy metal accumulation in leafy vegetables grown in industrial areas under varying levels of pollutionBangladesh Journal of Agricultural ResearchRead source →6Mercury pollution in vegetables, grains and soils from areas surrounding coal-fired power plantsScientific ReportsRead source →
Finished products and dietary exposure
Leafy greens are a modest but real share of dietary cadmium and lead.39Cadmium dietary exposure in the European populationEFSA Journal 2012;10(1):2551Read source →40Cadmium: dietary exposure assessmentWHO Food Additives Series, No. 82 (Safety evaluation of certain contaminants in food, prepared by the 91st meeting of JECFA)Read source → In Europe leafy vegetables supplied 3.9% of cadmium exposure overall and 4.8–5.9% for adults and the elderly; JECFA cites national estimates of 11% in China and 20% in the USA.39Cadmium dietary exposure in the European populationEFSA Journal 2012;10(1):2551Read source →40Cadmium: dietary exposure assessmentWHO Food Additives Series, No. 82 (Safety evaluation of certain contaminants in food, prepared by the 91st meeting of JECFA)Read source → In Guangzhou, leafy vegetables supplied 20% of dietary lead, and in Chongqing 4.9–8.9% of arsenic exposure.41Dietary Lead Exposure and Associated Health Risks in Guangzhou, ChinaInternational Journal of Environmental Research and Public HealthRead source →42Assessment of Dietary Arsenic Exposure Levels and the Associated Health Risks in Chongqing City, ChinaChinese Journal of Public HealthRead source →
Exposure depends on how much is eaten as much as on concentration. A Southern California food-frequency survey found a mean intake of 33.6 g of leafy greens a day, and JECFA estimated that leafy vegetables gave less than 5% of the tolerable cadmium intake in every GEMS/Food region it assessed in 2006.43Assessing Dietary Consumption of Toxicant-Laden Foods and Beverages by Age and Ethnicity in California: Implications for Proposition 65NutrientsRead source →44Evaluation of certain food contaminants — Sixty-fourth report of the Joint FAO/WHO Expert Committee on Food AdditivesWHO Technical Report Series 930 (Sixty-fourth meeting of JECFA, Rome, 8-17 February 2005)Read source → No source in the ledger measured metals in finished leafy products together with the leaves that went into them.
Evidence review and remaining questions
This assessment covers every source record in the library tagged to leafy greens or naming a leafy vegetable, and every leafy-vegetable row in WHO GEMS/Food, the FDA Total Diet Study and the other indexed datasets; each item has a recorded decision based on its content.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →2FY2018-FY2020 TDS Elements Analytical ResultsFDA Total Diet StudyRead source → Values from more than 30 sources that could not be reconciled with their own methods or figures are shown and held.7Concentrations and health risks of selected elements in leafy vegetables: a comparison between roadside open-air markets and large stores in Johannesburg, South AfricaNot reportedRead source →12Heavy Metals in Commonly Consumed Root and Leafy Vegetables in Dhaka City, Bangladesh, and Assessment of Associated Public Health RisksEnvironmental Systems ResearchRead source →13Enrichment Characteristics and Health Risk Assessment of Heavy Metals in Soil-Crop SystemsKSCE Journal of Civil EngineeringRead source →
Remaining gaps are specific. The Peruvian mining-region dataset with chard, spinach and lettuce results is held in an external repository not in the library, several papers print leaf values only as charts, and the Korean Food Code was recovered from the library volume rather than the declared file.37Article 2. Common Standards and Specifications for General FoodsKorean Food CodeRead source →45Dataset of metals and metalloids in food crops and soils sampled across the mining region of Moquegua in PeruScientific DataRead source →
References identify the sources used in these answers. The Sources index preserves additional discovery links; inclusion there does not mean a source has been reviewed for this assessment. Corrections, completed searches, new direct evidence and changes in requirements should reopen the relevant answer.
Leafy greens FAQs
What heavy metals have been measured in leafy greens?
Cadmium and lead are measured most often, and cadmium is the one most often found at levels that matter, especially in spinach.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →2FY2018-FY2020 TDS Elements Analytical ResultsFDA Total Diet StudyRead source → Studies also report total and inorganic arsenic, mercury, nickel, chromium, aluminium, tin and uranium; most mercury and arsenic results in monitored retail leaves are near or below detection.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →2FY2018-FY2020 TDS Elements Analytical ResultsFDA Total Diet StudyRead source →15Evaluation of essential, toxic and potentially toxic elements in leafy vegetables grown in the Canary IslandsToxicsRead source →
How do heavy metals get into leafy greens?
Through roots from soil and irrigation water, and on the leaf surface from dust and air deposition.3Heavy metal accumulation in leafy vegetables grown in industrial areas under varying levels of pollutionBangladesh Journal of Agricultural ResearchRead source →6Mercury pollution in vegetables, grains and soils from areas surrounding coal-fired power plantsScientific ReportsRead source →14Uptake of Arsenic by Irrigated Vegetables and Cooked Food Products in Burkina FasoFrontiers in WaterRead source → Near coal power plants, water spinach mercury fell from 86.7 µg/kg at 1 km to 23.9 µg/kg at 10 km.6Mercury pollution in vegetables, grains and soils from areas surrounding coal-fired power plantsScientific ReportsRead source →
Can leafy greens become contaminated after harvest?
Yes, through dust, handling and cooking water.7Concentrations and health risks of selected elements in leafy vegetables: a comparison between roadside open-air markets and large stores in Johannesburg, South AfricaNot reportedRead source →14Uptake of Arsenic by Irrigated Vegetables and Cooked Food Products in Burkina FasoFrontiers in WaterRead source → Leaves sold at a Johannesburg roadside carried far more cadmium unwashed than washed, and boiling spinach in arsenic-contaminated water raised its arsenic from 0.01 to 0.75 µg/g.7Concentrations and health risks of selected elements in leafy vegetables: a comparison between roadside open-air markets and large stores in Johannesburg, South AfricaNot reportedRead source →14Uptake of Arsenic by Irrigated Vegetables and Cooked Food Products in Burkina FasoFrontiers in WaterRead source →
Does this evidence show that organic leafy greens are cleaner?
Not consistently. In a long-term English trial, organic crop protection changed only cabbage nickel, while farmyard manure lowered cadmium; in New Zealand, organic market-garden vegetables were lower in cadmium and nickel overall but not consistently by leaf.19Effect of Climatic Conditions, and Agronomic Practices Used in Organic and Conventional Crop Production on Yield and Nutritional Composition Parameters in Potato, Cabbage, Lettuce and Onion; Results from the Long-Term NFSC-TrialsAgronomyRead source →33Assessment of Heavy Metals in Organic and Non-Organic Vegetables Post Severe Tropical Cyclone Gabrielle: A cross-sectional comparative analysisF1000ResearchRead source → Manures raised lead in Turkish spinach, so an organic label alone does not establish lower metals.21compared to the chemical fertilizer. In organic fertilizer applications, heavy metal contents such as arsenic, chrome, cobalt, and nickel were found toPeer-reviewed articleRead source →
Which leafy greens have the lowest cadmium?
Head cabbage, followed by head and iceberg lettuce; lamb’s lettuce and watercress were also low where measured.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →2FY2018-FY2020 TDS Elements Analytical ResultsFDA Total Diet StudyRead source →15Evaluation of essential, toxic and potentially toxic elements in leafy vegetables grown in the Canary IslandsToxicsRead source → Spinach was highest in every large dataset, with a GEMS/Food median of about 41 µg/kg against 2–4 µg/kg for cabbage.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →
Which leafy greens have the lowest lead?
Iceberg lettuce and cabbage in monitored markets: FDA detected lead in none of 27 composites of either, and iceberg had the lowest lead median (2.0 µg/kg) in Gran Canaria.2FY2018-FY2020 TDS Elements Analytical ResultsFDA Total Diet StudyRead source →15Evaluation of essential, toxic and potentially toxic elements in leafy vegetables grown in the Canary IslandsToxicsRead source → Kale, spinach and lamb’s lettuce had higher lead medians (13–25 µg/kg) in GEMS/Food.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →
Which leafy vegetables had the lowest measured aluminium?
Iceberg lettuce (1,581 µg/kg) in Gran Canaria and Chinese cabbage (200 µg/kg) in Hong Kong; spinach and Chinese spinach were highest.15Evaluation of essential, toxic and potentially toxic elements in leafy vegetables grown in the Canary IslandsToxicsRead source →16The First Hong Kong Total Diet Study: Metallic ContaminantsCentre for Food Safety, Food and Environmental Hygiene Department, Government of the Hong Kong Special Administrative RegionRead source → Aluminium varies more than twenty-fold between leaves in the same survey, so leaf type matters more than origin.15Evaluation of essential, toxic and potentially toxic elements in leafy vegetables grown in the Canary IslandsToxicsRead source →
Which growing country has the lowest heavy metal levels?
No country ranking is supported. GEMS/Food reports reporting regions rather than growing countries: lead was higher in Western Pacific submissions from 2001–2009 and cadmium in spinach higher in the Americas than in Europe, but these regions cover different years.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source → The clearest lower-contamination origin is a growing site away from mining, industry, roads and contaminated water.6Mercury pollution in vegetables, grains and soils from areas surrounding coal-fired power plantsScientific ReportsRead source →10Is it healthy urban agriculture? Human exposure to potentially toxic elements in urban gardens from Andalusia, SpainEnvironmental Science and Pollution ResearchRead source →17Nutrients and non-essential metals in darkibor kale grown at urban and rural farms: a pilot studyPLOS ONERead source →
Do these measurements predict my personal health risk?
No. They describe concentrations; risk depends on how much you eat, your other diet and your health.39Cadmium dietary exposure in the European populationEFSA Journal 2012;10(1):2551Read source →43Assessing Dietary Consumption of Toxicant-Laden Foods and Beverages by Age and Ethnicity in California: Implications for Proposition 65NutrientsRead source → In Europe leafy vegetables supplied about 4–6% of adult cadmium exposure.39Cadmium dietary exposure in the European populationEFSA Journal 2012;10(1):2551Read source →
Does washing remove every metal from leafy greens?
No. Washing removed most surface cadmium, nickel and mercury in roadside and power-plant studies, but a vinegar soak left lead in farm lettuce unchanged, because metal taken up into the leaf stays.6Mercury pollution in vegetables, grains and soils from areas surrounding coal-fired power plantsScientific ReportsRead source →7Concentrations and health risks of selected elements in leafy vegetables: a comparison between roadside open-air markets and large stores in Johannesburg, South AfricaNot reportedRead source →25Quantification and Reduction in Heavy Metal Residues in Some Fruits and Vegetables: A Case Study Galați County, RomaniaHorticulturaeRead source →
Is total arsenic the same as inorganic arsenic?
No. Inorganic arsenic is the more toxic part; Hong Kong measured it directly and found 74 µg/kg in water spinach but none detected in European cabbage or lettuce.5The First Hong Kong Total Diet Study: Inorganic ArsenicCentre for Food Safety, Food and Environmental Hygiene Department, Government of the Hong Kong Special Administrative RegionRead source → Most surveys report total arsenic only.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →15Evaluation of essential, toxic and potentially toxic elements in leafy vegetables grown in the Canary IslandsToxicsRead source →
Does soaking overnight reduce metals?
No study in the library tested an overnight water soak. A 5–10 minute vinegar soak lowered surface lead in market lettuce, and boiling in clean water removed part of the arsenic from amaranth.14Uptake of Arsenic by Irrigated Vegetables and Cooked Food Products in Burkina FasoFrontiers in WaterRead source →25Quantification and Reduction in Heavy Metal Residues in Some Fruits and Vegetables: A Case Study Galați County, RomaniaHorticulturaeRead source →
Can non-detects identify the cleanest type?
Only when the detection limits are low enough.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →2FY2018-FY2020 TDS Elements Analytical ResultsFDA Total Diet StudyRead source → A result below a 4 µg/kg limit says more than one below 200 µg/kg, and non-detects with different limits cannot be ranked against each other.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →2FY2018-FY2020 TDS Elements Analytical ResultsFDA Total Diet StudyRead source →
Does a supermarket guarantee lower levels?
No, though supermarket leaves were lower than roadside leaves in Johannesburg.7Concentrations and health risks of selected elements in leafy vegetables: a comparison between roadside open-air markets and large stores in Johannesburg, South AfricaNot reportedRead source → Supermarket monitoring datasets still include spinach and lettuce results above EU cadmium limits.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →4Scientific Opinion of the Panel on Contaminants in the Food Chain on a request from the European Commission on cadmium in foodThe EFSA JournalRead source →
Why are dry-weight results different from fresh-weight results?
Because fresh leaves are about 90% water.11Food survey: Levels and potential health risks of chromium, lead, zinc and copper content in fruits and vegetables consumed in AlgeriaFood and Chemical ToxicologyRead source → Removing the water concentrates the metal, so a dry-weight result is roughly ten times the fresh-weight value; limits are set on fresh weight.8Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →11Food survey: Levels and potential health risks of chromium, lead, zinc and copper content in fruits and vegetables consumed in AlgeriaFood and Chemical ToxicologyRead source →
Does cooking reduce metals in leafy greens?
Sometimes, and it depends on the metal and leaf.14Uptake of Arsenic by Irrigated Vegetables and Cooked Food Products in Burkina FasoFrontiers in WaterRead source →26Bioaccessibility of Heavy Metals in Raw and Processed Alternanthera sessilis and Centella asiatica: An In vitro StudyNature Environment and Pollution Technology25(1):D1830Read source → Boiling removed 55–69% of arsenic from amaranth but only 3–31% from spinach, and cooking with coconut milk or oil generally lowered the bioaccessible fraction of several metals.14Uptake of Arsenic by Irrigated Vegetables and Cooked Food Products in Burkina FasoFrontiers in WaterRead source →26Bioaccessibility of Heavy Metals in Raw and Processed Alternanthera sessilis and Centella asiatica: An In vitro StudyNature Environment and Pollution Technology25(1):D1830Read source →
Are there legal limits for heavy metals in leafy greens?
Yes: the EU sets 0.30 mg/kg lead and 0.10 mg/kg cadmium (0.20 mg/kg for spinach and fresh herbs), Codex 0.3 mg/kg lead and 0.2 mg/kg cadmium, and China 0.3 and 0.2 mg/kg.8Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →34General Standard for Contaminants and Toxins in Food and Feed (CXS 193-1995)Codex Alimentarius (Joint FAO/WHO Food Standards Programme)Read source →35China Releases the Standard for Maximum Levels of Contaminants in Foods (USDA FAS GAIN Report CH2023-0040, unofficial translation of GB 2762-2022)USDA Foreign Agricultural Service, Global Agricultural Information Network (GAIN), Report Number CH2023-0040Read source → The EU also sets 0.50 mg/kg nickel for leafy vegetables from 1 July 2025.36Commission Regulation (EU) 2024/1987 of 30 July 2024 amending Regulation (EU) 2023/915 as regards maximum levels of nickel in certain foodstuffsOfficial Journal of the European Union, L series, 2024/1987 (31.7.2024)Read source →
Does wastewater irrigation raise metals in leafy greens?
Yes, in the field studies that compared it.3Heavy metal accumulation in leafy vegetables grown in industrial areas under varying levels of pollutionBangladesh Journal of Agricultural ResearchRead source →14Uptake of Arsenic by Irrigated Vegetables and Cooked Food Products in Burkina FasoFrontiers in WaterRead source → Leaves near industrial effluent in Bangladesh carried several times the lead and cadmium of leaves from a low-pollution field, and arsenic in irrigation water raised leaf arsenic in proportion.3Heavy metal accumulation in leafy vegetables grown in industrial areas under varying levels of pollutionBangladesh Journal of Agricultural ResearchRead source →14Uptake of Arsenic by Irrigated Vegetables and Cooked Food Products in Burkina FasoFrontiers in WaterRead source →
What should a supplier laboratory report identify?
The leaf, the edible part analysed, whether it was washed, the fresh-weight basis, the quantification limit for each metal and the growing site.2FY2018-FY2020 TDS Elements Analytical ResultsFDA Total Diet StudyRead source →8Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source → For arsenic and mercury it should say whether totals or species were measured.1GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →5The First Hong Kong Total Diet Study: Inorganic ArsenicCentre for Food Safety, Food and Environmental Hygiene Department, Government of the Hong Kong Special Administrative RegionRead source →
Are leafy greens a major source of cadmium in the diet?
A modest one.39Cadmium dietary exposure in the European populationEFSA Journal 2012;10(1):2551Read source →40Cadmium: dietary exposure assessmentWHO Food Additives Series, No. 82 (Safety evaluation of certain contaminants in food, prepared by the 91st meeting of JECFA)Read source → In Europe they supplied 3.9% of cadmium exposure overall; national estimates cited by JECFA are 11% in China and 20% in the USA.39Cadmium dietary exposure in the European populationEFSA Journal 2012;10(1):2551Read source →40Cadmium: dietary exposure assessmentWHO Food Additives Series, No. 82 (Safety evaluation of certain contaminants in food, prepared by the 91st meeting of JECFA)Read source →
References
Works cited on this page, in citation-number order. See Sources for this page’s source inventory. Each title links to a source record with bibliographic details, reported findings, and document provenance.
- GEMS/Food Contaminants database heavy-metal exportsDataset
- FY2018-FY2020 TDS Elements Analytical ResultsDataset
- Heavy metal accumulation in leafy vegetables grown in industrial areas under varying levels of pollutionPeer-reviewed
- Scientific Opinion of the Panel on Contaminants in the Food Chain on a request from the European Commission on cadmium in foodGovernment
- The First Hong Kong Total Diet Study: Inorganic ArsenicGovernment
- Mercury pollution in vegetables, grains and soils from areas surrounding coal-fired power plantsPeer-reviewed
- Concentrations and health risks of selected elements in leafy vegetables: a comparison between roadside open-air markets and large stores in Johannesburg, South AfricaPeer-reviewed
- Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Regulation
- Scientific Opinion on Lead in FoodGovernment
- Is it healthy urban agriculture? Human exposure to potentially toxic elements in urban gardens from Andalusia, SpainPeer-reviewed
- Food survey: Levels and potential health risks of chromium, lead, zinc and copper content in fruits and vegetables consumed in AlgeriaPeer-reviewed
- Heavy Metals in Commonly Consumed Root and Leafy Vegetables in Dhaka City, Bangladesh, and Assessment of Associated Public Health RisksPeer-reviewed
- Enrichment Characteristics and Health Risk Assessment of Heavy Metals in Soil-Crop SystemsPeer-reviewed
- Uptake of Arsenic by Irrigated Vegetables and Cooked Food Products in Burkina FasoPeer-reviewed
- Evaluation of essential, toxic and potentially toxic elements in leafy vegetables grown in the Canary IslandsPeer-reviewed
- The First Hong Kong Total Diet Study: Metallic ContaminantsGovernment
- Nutrients and non-essential metals in darkibor kale grown at urban and rural farms: a pilot studyPeer-reviewed
- Total Diet Study of metals and other elements in foodGovernment
- Effect of Climatic Conditions, and Agronomic Practices Used in Organic and Conventional Crop Production on Yield and Nutritional Composition Parameters in Potato, Cabbage, Lettuce and Onion; Results from the Long-Term NFSC-TrialsPeer-reviewed
- Impact of Physical Interventions, Phosphorus Fertilization, and the Utilization of Soil Amendments on the Absorption of Cadmium by Lettuce Grown in a Solar-Powered GreenhousePeer-reviewed
- compared to the chemical fertilizer. In organic fertilizer applications, heavy metal contents such as arsenic, chrome, cobalt, and nickel were found toPeer-reviewed
- Mercury accumulation in vegetable Houttuynia cordata Thunb. from two different geological areas in southwest China and implications for human consumptionPeer-reviewed
- BIO-ACCUMULATION OF HEAVY METALS IN SOME NORTHPeer-reviewed
- Arsenic Contamination of Water-Soil-Crop System in an Industrial Area of BangladeshPeer-reviewed
- Quantification and Reduction in Heavy Metal Residues in Some Fruits and Vegetables: A Case Study Galați County, RomaniaPeer-reviewed
- Bioaccessibility of Heavy Metals in Raw and Processed Alternanthera sessilis and Centella asiatica: An In vitro StudyPeer-reviewed
- Development and Application of a Modified Biochar-Calcium Alginate Composite (MB-CA) for In Situ Remediation of Cadmium-Contaminated SoilPeer-reviewed
- Arsenic uptake and accumulation in bean and lettuce plants at different developmental stagesPeer-reviewed
- Humic + Fulvic acid mitigated Cd adverse effects on plant growth, physiology and biochemical properties of garden cressPeer-reviewed
- The Cd:Zn ratio in a soil affects Cd toxicity in spinach (Spinacea oleracea L.)Peer-reviewed
- Minimization of cadmium toxicity and improvement in growth and biochemical attributes of spinach by using acidified biocharPeer-reviewed
- Effects of wollastonite and phosphate treatments on cadmium bioaccessibility and health risk in pak choi (Brassica rapa ssp. chinensis) grown in contaminated soilPeer-reviewed
- Assessment of Heavy Metals in Organic and Non-Organic Vegetables Post Severe Tropical Cyclone Gabrielle: A cross-sectional comparative analysisPeer-reviewed
- General Standard for Contaminants and Toxins in Food and Feed (CXS 193-1995)Government
- China Releases the Standard for Maximum Levels of Contaminants in Foods (USDA FAS GAIN Report CH2023-0040, unofficial translation of GB 2762-2022)Regulation
- Commission Regulation (EU) 2024/1987 of 30 July 2024 amending Regulation (EU) 2023/915 as regards maximum levels of nickel in certain foodstuffsRegulation
- Article 2. Common Standards and Specifications for General FoodsRegulation
- Guidelines on the Food Adulteration (Metallic Contamination) (Amendment) Regulation 2018Government
- Cadmium dietary exposure in the European populationGovernment
- Cadmium: dietary exposure assessmentGovernment
- Dietary Lead Exposure and Associated Health Risks in Guangzhou, ChinaPeer-reviewed
- Assessment of Dietary Arsenic Exposure Levels and the Associated Health Risks in Chongqing City, ChinaPeer-reviewed
- Assessing Dietary Consumption of Toxicant-Laden Foods and Beverages by Age and Ethnicity in California: Implications for Proposition 65Peer-reviewed
- Evaluation of certain food contaminants — Sixty-fourth report of the Joint FAO/WHO Expert Committee on Food AdditivesGovernment
- Dataset of metals and metalloids in food crops and soils sampled across the mining region of Moquegua in PeruPeer-reviewed
Sources
Browse Leafy greens Sources for sources cited on this page and additional discovery candidates. The searchable index distinguishes verified citation use from discovery relationships whose relevance has not been established.
Update history
22 October 2026 — Rebuilt from every leafy-green record in the library: WHO GEMS/Food and FDA monitoring results by leaf, total diet studies, field and market surveys, pooled same-leaf comparisons, verified limits and stage-led controls.
