Heavy metals in Soy

Quick read

Soy in the HMI library has 613 reported results for lead, cadmium, total arsenic, inorganic arsenic, total mercury, nickel, chromium, aluminium, tin, antimony, uranium, Co, Cu, Fe, Mn, Zn, V, Se, arsenic (As), mercury (Hg), I, Li, Mo and Sr from 32 sources.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}^{2Vanda Lopes Andrade, Iolanda Ribeiro, A. P. Marreilha dos Santos, Michael Aschner, Maria Luisa Mateus · 2023Metals in Cow Milk and Soy Beverages: Is There a Concern?ToxicsRead source →}^{3Food Standards Agency, Hallmark Meat Hygiene · 2014Survey of metals and other elements in commercial infant foods, infant formula and non-infant specific foodsFood Standards Agency reportRead source →}^{4Marín S, Pardo O, Sánchez A, Sanchis Y, Vélez D, Devesa V, Font G, Yusà V · 2018Assessment of metal levels in foodstuffs from the Region of Valencia (Spain)Toxicology Reports 5 (2018) 654-670Read source →}^{5Marques M, Correig E, Capdevila E, Gargallo E, Gonzalez N, Nadal M, Domingo JL · 2021Essential and Non-essential Trace Elements in Milks and Plant-Based DrinksBiological Trace Element ResearchRead source →}^{6Milani RF, Mauri AA, Sanches VL, Morgano MA, Cadore S · 2023Trace Elements in Soy-Based Beverages: A Comprehensive Study of Total Content and In Vitro BioaccessibilityInternational Journal of Environmental Research and Public HealthRead source →}

Metals measured:
Lead, cadmium, total arsenic, inorganic arsenic, total mercury and nickel and others, in 613 results from 32 sources.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}^{2Vanda Lopes Andrade, Iolanda Ribeiro, A. P. Marreilha dos Santos, Michael Aschner, Maria Luisa Mateus · 2023Metals in Cow Milk and Soy Beverages: Is There a Concern?ToxicsRead source →}^{3Food Standards Agency, Hallmark Meat Hygiene · 2014Survey of metals and other elements in commercial infant foods, infant formula and non-infant specific foodsFood Standards Agency reportRead source →}^{4Marín S, Pardo O, Sánchez A, Sanchis Y, Vélez D, Devesa V, Font G, Yusà V · 2018Assessment of metal levels in foodstuffs from the Region of Valencia (Spain)Toxicology Reports 5 (2018) 654-670Read source →}
Main drivers:
Soy, as sold from all 25 cultivars, field soil 1.22 mg cadmium (Cd)/kg (experiment) had cadmium higher at 470 µg/kg, against 180 µg/kg for soybean cultivar Hefeng 40, control soil in the same study.^{7Zhi Y, Sun T, Zhou Q, Leng X · 2020Screening of safe soybean cultivars for cadmium contaminated fieldsScientific Reports 10:12965Read source →}
Practical framing:
Ask for lot results with a quantification limit well below the limit that applies in the destination market.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}^{2Vanda Lopes Andrade, Iolanda Ribeiro, A. P. Marreilha dos Santos, Michael Aschner, Maria Luisa Mateus · 2023Metals in Cow Milk and Soy Beverages: Is There a Concern?ToxicsRead source →}^{3Food Standards Agency, Hallmark Meat Hygiene · 2014Survey of metals and other elements in commercial infant foods, infant formula and non-infant specific foodsFood Standards Agency reportRead source →}^{4Marín S, Pardo O, Sánchez A, Sanchis Y, Vélez D, Devesa V, Font G, Yusà V · 2018Assessment of metal levels in foodstuffs from the Region of Valencia (Spain)Toxicology Reports 5 (2018) 654-670Read source →}

Soy overview

Soy in the HMI library has 613 reported results for lead, cadmium, total arsenic, inorganic arsenic, total mercury, nickel, chromium, aluminium, tin, antimony, uranium, Co, Cu, Fe, Mn, Zn, V, Se, As, Hg, I, Li, Mo and Sr from 32 sources.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}^{2Vanda Lopes Andrade, Iolanda Ribeiro, A. P. Marreilha dos Santos, Michael Aschner, Maria Luisa Mateus · 2023Metals in Cow Milk and Soy Beverages: Is There a Concern?ToxicsRead source →}^{3Food Standards Agency, Hallmark Meat Hygiene · 2014Survey of metals and other elements in commercial infant foods, infant formula and non-infant specific foodsFood Standards Agency reportRead source →}^{4Marín S, Pardo O, Sánchez A, Sanchis Y, Vélez D, Devesa V, Font G, Yusà V · 2018Assessment of metal levels in foodstuffs from the Region of Valencia (Spain)Toxicology Reports 5 (2018) 654-670Read source →}^{5Marques M, Correig E, Capdevila E, Gargallo E, Gonzalez N, Nadal M, Domingo JL · 2021Essential and Non-essential Trace Elements in Milks and Plant-Based DrinksBiological Trace Element ResearchRead source →}^{6Milani RF, Mauri AA, Sanches VL, Morgano MA, Cadore S · 2023Trace Elements in Soy-Based Beverages: A Comprehensive Study of Total Content and In Vitro BioaccessibilityInternational Journal of Environmental Research and Public HealthRead source →} WHO GEMS/Food submissions reported lead in soy, as sold from 2,031 results in 5 reporting regions, with subgroup medians from 2.69 to 50 µg/kg.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →} For cadmium, GEMS/Food gave subgroup medians from 1.17 to 130 µg/kg from 3,905 results.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}

Across 2 other studies of ordinary farms and markets, reported lead means and medians on an as-sold or fresh-weight basis ranged from 9.8 µg/kg (Marín et al. 2018) to 15 µg/kg (Zhu et al. 2011).^{4Marín S, Pardo O, Sánchez A, Sanchis Y, Vélez D, Devesa V, Font G, Yusà V · 2018Assessment of metal levels in foodstuffs from the Region of Valencia (Spain)Toxicology Reports 5 (2018) 654-670Read source →}^{8Zhu F, Fan W, Wang X, Qu L, Yao S · 2011Health risk assessment of eight heavy metals in nine varieties of edible vegetable oils consumed in ChinaFood and Chemical ToxicologyRead source →} For cadmium, 5 studies on an as-sold or fresh-weight basis reported means and medians from 4.12 µg/kg (Zhu et al. 2011) to 69 µg/kg (EFSA Panel on Contaminants in the Food Chain (CONTAM) 2009).^{8Zhu F, Fan W, Wang X, Qu L, Yao S · 2011Health risk assessment of eight heavy metals in nine varieties of edible vegetable oils consumed in ChinaFood and Chemical ToxicologyRead source →}^{9EFSA Panel on Contaminants in the Food Chain (CONTAM) · 2009Scientific Opinion of the Panel on Contaminants in the Food Chain on a request from the European Commission on cadmium in foodThe EFSA JournalRead source →} Higher results came from contaminated or exposed settings rather than ordinary supply. Soy, as sold from soybean cultivar Hefeng 40, pot soil spiked to 2.5 mg Cd/kg (T2) (experiment) had cadmium higher at 2,060 µg/kg, against 180 µg/kg for soybean cultivar Hefeng 40, control soil in the same study.^{7Zhi Y, Sun T, Zhou Q, Leng X · 2020Screening of safe soybean cultivars for cadmium contaminated fieldsScientific Reports 10:12965Read source →}

For a buyer the practical points are the form bought, the basis a result is reported on, where the soy came from and the laboratory’s reporting limit. Testing finds contamination but does not remove it.

What counts as soy

This page covers soy in 4 forms kept separate in every comparison: soy, as sold (575 results), soy, dried (17 results), soy, powder or ground (11 results) and soy, oil (10 results).^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}^{2Vanda Lopes Andrade, Iolanda Ribeiro, A. P. Marreilha dos Santos, Michael Aschner, Maria Luisa Mateus · 2023Metals in Cow Milk and Soy Beverages: Is There a Concern?ToxicsRead source →}^{3Food Standards Agency, Hallmark Meat Hygiene · 2014Survey of metals and other elements in commercial infant foods, infant formula and non-infant specific foodsFood Standards Agency reportRead source →}^{4Marín S, Pardo O, Sánchez A, Sanchis Y, Vélez D, Devesa V, Font G, Yusà V · 2018Assessment of metal levels in foodstuffs from the Region of Valencia (Spain)Toxicology Reports 5 (2018) 654-670Read source →}^{5Marques M, Correig E, Capdevila E, Gargallo E, Gonzalez N, Nadal M, Domingo JL · 2021Essential and Non-essential Trace Elements in Milks and Plant-Based DrinksBiological Trace Element ResearchRead source →}^{6Milani RF, Mauri AA, Sanches VL, Morgano MA, Cadore S · 2023Trace Elements in Soy-Based Beverages: A Comprehensive Study of Total Content and In Vitro BioaccessibilityInternational Journal of Environmental Research and Public HealthRead source →} Foods made from soy that have their own pages, such as soy-based infant formula, soy milk, soy protein isolate and soy sauce, are covered there.

Results stay on the basis each source reports: as sold or fresh weight, dry weight, or unstated. Dry-weight, as-sold and unstated results are never pooled, and no moisture conversion is applied.

Heavy metal contamination profile of soy

The table holds every extracted soy result: one row for each statistic a source reports for a form, group and metal, 613 rows in all, grouped by metal and form; percentiles, ranges and maxima for the same samples as a mean or median are shown on that row.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}^{2Vanda Lopes Andrade, Iolanda Ribeiro, A. P. Marreilha dos Santos, Michael Aschner, Maria Luisa Mateus · 2023Metals in Cow Milk and Soy Beverages: Is There a Concern?ToxicsRead source →}^{3Food Standards Agency, Hallmark Meat Hygiene · 2014Survey of metals and other elements in commercial infant foods, infant formula and non-infant specific foodsFood Standards Agency reportRead source →}^{4Marín S, Pardo O, Sánchez A, Sanchis Y, Vélez D, Devesa V, Font G, Yusà V · 2018Assessment of metal levels in foodstuffs from the Region of Valencia (Spain)Toxicology Reports 5 (2018) 654-670Read source →}^{5Marques M, Correig E, Capdevila E, Gargallo E, Gonzalez N, Nadal M, Domingo JL · 2021Essential and Non-essential Trace Elements in Milks and Plant-Based DrinksBiological Trace Element ResearchRead source →}^{6Milani RF, Mauri AA, Sanches VL, Morgano MA, Cadore S · 2023Trace Elements in Soy-Based Beverages: A Comprehensive Study of Total Content and In Vitro BioaccessibilityInternational Journal of Environmental Research and Public HealthRead source →} Monitoring medians are HMI calculations from individual submissions; non-detects remain intervals between zero and the reporting limit and are never read as zero.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}

54 rows can be rated against an HMI reference for the same form and basis, 79 rows have a direct same-study comparison, and the rest stay unrated with the reason recorded. 16 values from 3 sources failed a check against their own source and are shown but not compared.^{7Zhi Y, Sun T, Zhou Q, Leng X · 2020Screening of safe soybean cultivars for cadmium contaminated fieldsScientific Reports 10:12965Read source →}^{10Adam AA, Sackey LNA, Ofori LA · 2022Risk assessment of heavy metals concentration in cereals and legumes sold in the Tamale Aboabo market, GhanaHeliyon 8 (2022) e10162Read source →}^{11Yu S, Pan XD, Han JL · 2023Toxic Elements in Beans from Zhejiang, Southeast China: Distribution and Probabilistic Health Risk AssessmentFoodsRead source →}

Comparison methods and scope

Each row is one statistic a source reports for one form of soy, group and metal; percentiles, ranges and maxima reported for the same samples as a mean or median are shown on that row. Different forms and preparation states are kept apart, and dry-weight results never share a reference with as-sold or fresh-weight results. Ratings place a result against an HMI reference built from monitoring results and ordinary-setting medians for the same form and basis. Same-study contrasts compare a contaminated or treated site with its control, targeted with random sampling, a prepared form with the unprocessed one, or one group of a study with the median of its other groups. Held rows failed a check against their own source and are shown but not compared.

Filter findings

AnalyteFood formCollection regionCollection periodSourceCompare withComparable findings only

570 of 570 findings · 24 analytes

About this table

Food formCollection scopeFindingComparisonEvidence

Soy findings from monitoring programmes, total diet studies and every extracted field, market and contaminated-site study, in µg/kg with each source’s form, basis, statistic and censoring retained.
Food form	Collection scope	Finding (µg/kg)

83 findings · 15 studies · 2 spread figures folded in
Soy, as soldWHO GEMS/Food	EURO2015	median.
Soy, as soldWHO GEMS/Food	EURO2004–2019	median.
Soy, as soldWHO GEMS/Food	EURO2012	median.
Soy, as soldWHO GEMS/Food	EURO2004–2019	median.
Soy, as soldWHO GEMS/Food	EURO2007–2016	median.
Soy, as soldWHO GEMS/Food	EURO2007–2019	median.
Soy, as soldWHO GEMS/Food	EURO2004–2018	median.
Soy, as soldWHO GEMS/Food	EURO2003–2017	median.
Soy, as soldWHO GEMS/Food	EURO2007–2008	median.
Soy, as soldWHO GEMS/Food	EURO2001–2005	median.
Soy, as soldWHO GEMS/Food	PAHO2015	median.
Soy, as soldWHO GEMS/Food	PAHO2015	median.
Soy, as soldWHO GEMS/Food	PAHO2015	median.
Soy, as soldWHO GEMS/Food	PAHO2003–2012	median.
Soy, as soldWHO GEMS/Food	PAHO2015	median.
Soy, as soldWHO GEMS/Food	PAHO2015	median.
Soy, as soldWHO GEMS/Food	PAHO2015	median.
Soy, as soldWHO GEMS/Food	PAHO2011–2015	median.
Soy, as soldWHO GEMS/Food	PAHO2013–2015	median.
Soy, as soldWHO GEMS/Food	PAHO2015	median.
Soy, as soldWHO GEMS/Food	PAHO2011–2012	median.
Soy, as soldWHO GEMS/Food	PAHO1995–2012	median.
Soy, as soldWHO GEMS/Food	PAHO2014	median.
Soy, as soldWHO GEMS/Food	SEARO2007–2016	median.
Soy, as soldWHO GEMS/Food	SEARO2009–2012	median.
Soy, as soldWHO GEMS/Food	WPRO2008	median.
Soy, as soldWHO GEMS/Food	WPRO2013	median.
Soy, as soldWHO GEMS/Food	WPRO2010	median.
Soy, as soldWHO GEMS/Food	WPRO2014	median.
Soy, as soldWHO GEMS/Food	WPRO2008	median.
Soy, as soldWHO GEMS/Food	WPRO2010–2012	median.
Soy, as soldWHO GEMS/Food	WPRO2015	median.
Soy, as soldWHO GEMS/Food	WPRO2003–2013	median.
Soy, as soldWHO GEMS/Food	WPRO2001–2002	median.
Soy, as soldWHO GEMS/Food	WPRO2005–2014	median.
Soy, as soldAndrade et al. 2023	all 14 products2019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 12019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 22019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 32019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 42019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 52019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 62019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 72019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 82019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 92019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 102019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 112019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 122019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 132019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 142019	mean.
Soy, as soldFood Standards Agency 2014	dry powder formula category2013–2014	mean.
Soy, as soldMarín et al. 2018	Milk and dairy products group, Region of Valencia TDS 2010-20112010–2011	mean.Same samples: 8.3 µg/kg (minimum); 13.1 µg/kg (maximum)
Soy, as soldMarín et al. 2018	Milk and dairy products group, Region of Valencia TDS 2010-20112010–2011	min.
Soy, as soldMarín et al. 2018	Milk and dairy products group, Region of Valencia TDS 2010-20112010–2011	max.
Soy, as soldMarques et al. 2021	all samples2021	not detected.
Soy, as soldMarques et al. 2021	all samples2021	not detected.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	max.
Soy, as soldWHO GEMS/Food	AFRO2012Dry weight	median.
Soy, as soldWHO GEMS/Food	PAHO2015Dry weight	median.
Soy, as soldWHO GEMS/Food	PAHO2013–2015Dry weight	median.
Soy, as soldAdam et al. 2022	Tamale Aboabo market composite2017Dry weight	mean.
Soy, as soldAdam et al. 2022	Tamale Aboabo market composite2017Dry weight	min.
Soy, as soldAdam et al. 2022	Tamale Aboabo market composite2017Dry weight	max.
Soy, as soldOrisakwe et al. 2012	all samples (Ohaji, Umuagwo, Owerri sites)Collection date unreportedDry weight	mean.
Soy, as soldZhang et al. 2026	Xiushan Mn mining and smelting area2020Dry weight	mean.
Soy, as soldauthors 2016	raw oilseed, fraction 0, one production lotCollection date unreportedBasis not stated	mean.
Soy, as soldCantoral et al. 2024	Legumes (retail, Mexico City)2022–2023Basis not stated	mean.
Soy, as soldCentre for Food Safety (Hong Kong) 2013	all samples2010–2011Basis not stated	mean.
Soy, as soldCentre for Food Safety (Hong Kong) 2013	all samples2010–2011Basis not stated	mean.
Soy, as soldEFSA Panel on Contaminants in the Food Chain (CONTAM) 2010	Milk and dairy-based products2000–2009Basis not stated	max.
Soy, as soldEFSA Panel on Contaminants in the Food Chain (CONTAM) 2010	Milk and dairy-based products2000–2009Basis not stated	median.
Soy, as soldEFSA Panel on Contaminants in the Food Chain (CONTAM) 2010	Milk and dairy-based products2000–2009Basis not stated	mean.
Soy, as soldEFSA Panel on Contaminants in the Food Chain (CONTAM) 2010	Milk and dairy-based products2000–2009Basis not stated	median.
Soy, as soldEFSA Panel on Contaminants in the Food Chain (CONTAM) 2010	Milk and dairy-based products2000–2009Basis not stated	mean.
Soy, as soldWang et al. 2020	all samples (both areas)2015–2016Basis not stated	mean.
Soy, as soldWang et al. 2020	all samples (both areas)2015–2016Basis not stated	mean.
Soy, as soldWang et al. 2020	Other (non-Songhua) area2015–2016Basis not stated	mean.
Soy, as soldWang et al. 2020	Other (non-Songhua) area2015–2016Basis not stated	mean.
Soy, as soldWang et al. 2020	Songhua River basin area2015–2016Basis not stated	mean.
Soy, as soldWang et al. 2020	Songhua River basin area2015–2016Basis not stated	mean.
Soy, as soldZhu et al. 2024	all samples2021–2022Basis not stated	mean.
1 finding · 1 study · 2 spread figures folded in
Soy, driedYu et al. 2023	all samples of this bean type, 11 Zhejiang monitoring areasCollection date unreported	mean.Same samples: 196 µg/kg (maximum); 138 µg/kg (percentile)
Soy, driedYu et al. 2023	all samples of this bean type, 11 Zhejiang monitoring areasCollection date unreported	max.
Soy, driedYu et al. 2023	all samples of this bean type, 11 Zhejiang monitoring areasCollection date unreported	percentile.
4 findings · 1 study
Soy, powder or groundWHO GEMS/Food	EURO2007–2018	median.
Soy, powder or groundWHO GEMS/Food	EURO2003–2013	median.
Soy, powder or groundWHO GEMS/Food	WPRO2012–2013	median.
Soy, powder or groundWHO GEMS/Food	WPRO2016	median.
2 findings · 2 studies
Soy, oilWHO GEMS/Food	PAHO2015	median.
Soy, oilZhu et al. 2011	all samples (nationwide supermarkets)2009–2010	mean.

169 findings · 18 studies · 17 spread figures folded in
Soy, as soldWHO GEMS/Food	EMRO2019	median.
Soy, as soldWHO GEMS/Food	EURO2015	median.
Soy, as soldWHO GEMS/Food	EURO2003–2019	median.
Soy, as soldWHO GEMS/Food	EURO2003–2012	median.
Soy, as soldWHO GEMS/Food	EURO2004–2019	median.
Soy, as soldWHO GEMS/Food	EURO2003–2016	median.
Soy, as soldWHO GEMS/Food	EURO2016	median.
Soy, as soldWHO GEMS/Food	EURO2005	median.
Soy, as soldWHO GEMS/Food	EURO2005–2006	median.
Soy, as soldWHO GEMS/Food	EURO2006	median.
Soy, as soldWHO GEMS/Food	EURO2003–2006	median.
Soy, as soldWHO GEMS/Food	EURO2005	median.
Soy, as soldWHO GEMS/Food	EURO2006	median.
Soy, as soldWHO GEMS/Food	EURO2004	median.
Soy, as soldWHO GEMS/Food	EURO2005–2006	median.
Soy, as soldWHO GEMS/Food	EURO2004	median.
Soy, as soldWHO GEMS/Food	EURO2003–2019	median.
Soy, as soldWHO GEMS/Food	EURO2007–2019	median.
Soy, as soldWHO GEMS/Food	EURO2003–2006	median.
Soy, as soldWHO GEMS/Food	EURO2003–2017	median.
Soy, as soldWHO GEMS/Food	EURO2007–2008	median.
Soy, as soldWHO GEMS/Food	EURO2001–2005	median.
Soy, as soldWHO GEMS/Food	PAHO2019	median.
Soy, as soldWHO GEMS/Food	PAHO1999–2018	median.
Soy, as soldWHO GEMS/Food	PAHO2006–2018	median.
Soy, as soldWHO GEMS/Food	PAHO2000–2018	median.
Soy, as soldWHO GEMS/Food	PAHO2019	median.
Soy, as soldWHO GEMS/Food	WPRO2015	median.
Soy, as soldWHO GEMS/Food	WPRO2010	median.
Soy, as soldWHO GEMS/Food	WPRO2014	median.
Soy, as soldWHO GEMS/Food	WPRO2008	median.
Soy, as soldWHO GEMS/Food	WPRO2010	median.
Soy, as soldWHO GEMS/Food	WPRO2002–2005	median.
Soy, as soldWHO GEMS/Food	WPRO2000–2002	median.
Soy, as soldWHO GEMS/Food	WPRO2002–2005	median.
Soy, as soldWHO GEMS/Food	WPRO2011–2025	median.
Soy, as soldAndrade et al. 2023	all 14 products2019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 12019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 22019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 32019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 42019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 52019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 62019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 72019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 82019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 92019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 102019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 112019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 122019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 132019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 142019	mean.
Soy, as soldEFSA Panel on Contaminants in the Food Chain (CONTAM) 2009	all samples (EU Member States and others, 2003-2007)2003–2007	median.
Soy, as soldEFSA Panel on Contaminants in the Food Chain (CONTAM) 2009	all samples (EU Member States and others, 2003-2007)2003–2007	median.
Soy, as soldEFSA Panel on Contaminants in the Food Chain (CONTAM) 2009	all samples (EU Member States and others, 2003-2007)2003–2007	median.
Soy, as soldEFSA Panel on Contaminants in the Food Chain (CONTAM) 2009	all samples (EU Member States and others, 2003-2007)2003–2007	mean.Same samples: 124 µg/kg (maximum); 61 µg/kg (maximum); 45 µg/kg (maximum); 0.5 µg/kg (percentile); 64 µg/kg (percentile); 2 µg/kg (percentile); 61 µg/kg (percentile); 4 µg/kg (percentile); 45 µg/kg (percentile)
Soy, as soldEFSA Panel on Contaminants in the Food Chain (CONTAM) 2009	all samples (EU Member States and others, 2003-2007)2003–2007	mean.
Soy, as soldEFSA Panel on Contaminants in the Food Chain (CONTAM) 2009	all samples (EU Member States and others, 2003-2007)2003–2007	mean.
Soy, as soldEFSA Panel on Contaminants in the Food Chain (CONTAM) 2009	all samples (EU Member States and others, 2003-2007)2003–2007	mean.
Soy, as soldEFSA Panel on Contaminants in the Food Chain (CONTAM) 2009	all samples (EU Member States and others, 2003-2007)2003–2007	max.
Soy, as soldEFSA Panel on Contaminants in the Food Chain (CONTAM) 2009	all samples (EU Member States and others, 2003-2007)2003–2007	max.
Soy, as soldEFSA Panel on Contaminants in the Food Chain (CONTAM) 2009	all samples (EU Member States and others, 2003-2007)2003–2007	max.
Soy, as soldEFSA Panel on Contaminants in the Food Chain (CONTAM) 2009	all samples (EU Member States and others, 2003-2007)2003–2007	percentile.
Soy, as soldEFSA Panel on Contaminants in the Food Chain (CONTAM) 2009	all samples (EU Member States and others, 2003-2007)2003–2007	percentile.
Soy, as soldEFSA Panel on Contaminants in the Food Chain (CONTAM) 2009	all samples (EU Member States and others, 2003-2007)2003–2007	percentile.
Soy, as soldEFSA Panel on Contaminants in the Food Chain (CONTAM) 2009	all samples (EU Member States and others, 2003-2007)2003–2007	percentile.
Soy, as soldEFSA Panel on Contaminants in the Food Chain (CONTAM) 2009	all samples (EU Member States and others, 2003-2007)2003–2007	percentile.
Soy, as soldEFSA Panel on Contaminants in the Food Chain (CONTAM) 2009	all samples (EU Member States and others, 2003-2007)2003–2007	percentile.
Soy, as soldFood Standards Agency 2014	dry powder formula category2013–2014	mean.
Soy, as soldMarín et al. 2018	Milk and dairy products group, Region of Valencia TDS 2010-20112010–2011	mean.Same samples: 7.7 µg/kg (minimum); 13.4 µg/kg (maximum)
Soy, as soldMarín et al. 2018	Milk and dairy products group, Region of Valencia TDS 2010-20112010–2011	min.
Soy, as soldMarín et al. 2018	Milk and dairy products group, Region of Valencia TDS 2010-20112010–2011	max.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	mean.
Soy, as soldShindoh et al. 2010	Soybean A (Cd 0.040 µg/g), miso trialCollection date unreported	mean.
Soy, as soldShindoh et al. 2010	Soybean A (Cd 0.040 µg/g), miso trialCollection date unreported	mean.
Soy, as soldShindoh et al. 2010	Soybean A (Cd 0.040 µg/g), soy sauce trialCollection date unreported	mean.
Soy, as soldShindoh et al. 2010	Soybean B (Cd 0.320 µg/g), miso trialCollection date unreported	mean.
Soy, as soldShindoh et al. 2010	Soybean B (Cd 0.320 µg/g), miso trialCollection date unreported	mean.
Soy, as soldShindoh et al. 2010	Soybean B (Cd 0.320 µg/g), soy sauce trialCollection date unreported	mean.
Soy, as soldWHO GEMS/Food	AFRO2012Dry weight	median.
Soy, as soldAdam et al. 2022	Tamale Aboabo market composite2017Dry weight	not detected.
Soy, as soldHajara et al. 2026	Alagarno ward markets, Gamawa LGACollection date unreportedDry weight	mean.
Soy, as soldHajara et al. 2026	all wards (Kafiromi, Alagarno, Tumbi), market samplesCollection date unreportedDry weight	mean.
Soy, as soldHajara et al. 2026	Kafiromi ward markets, Gamawa LGACollection date unreportedDry weight	mean.
Soy, as soldHajara et al. 2026	Tumbi ward markets, Gamawa LGACollection date unreportedDry weight	mean.
Soy, as soldOrisakwe et al. 2012	all samples (Ohaji, Umuagwo, Owerri sites)Collection date unreportedDry weight	mean.
Soy, as soldZhang et al. 2026	Xiushan Mn mining and smelting area2020Dry weight	mean.
Soy, as soldZhi et al. 2020	all 25 cultivars, field soil 1.22 mg Cd/kgCollection date unreportedDry weight	mean.Same samples: 110 µg/kg (minimum); 990 µg/kg (maximum)
Soy, as soldZhi et al. 2020	all 25 cultivars, field soil 1.22 mg Cd/kgCollection date unreportedDry weight	min.
Soy, as soldZhi et al. 2020	all 25 cultivars, field soil 1.22 mg Cd/kgCollection date unreportedDry weight	max.
Soy, as soldZhi et al. 2020	all 25 cultivars, pot soil spiked to 1.0 mg Cd/kg (T1)Collection date unreportedDry weight	mean.Same samples: 90 µg/kg (minimum); 980 µg/kg (maximum)
Soy, as soldZhi et al. 2020	all 25 cultivars, pot soil spiked to 1.0 mg Cd/kg (T1)Collection date unreportedDry weight	min.
Soy, as soldZhi et al. 2020	all 25 cultivars, pot soil spiked to 1.0 mg Cd/kg (T1)Collection date unreportedDry weight	max.
Soy, as soldZhi et al. 2020	all 25 cultivars, pot soil spiked to 2.5 mg Cd/kg (T2)Collection date unreportedDry weight	mean.Same samples: 290 µg/kg (minimum); 2,060 µg/kg (maximum)
Soy, as soldZhi et al. 2020	all 25 cultivars, pot soil spiked to 2.5 mg Cd/kg (T2)Collection date unreportedDry weight	min.
Soy, as soldZhi et al. 2020	all 25 cultivars, pot soil spiked to 2.5 mg Cd/kg (T2)Collection date unreportedDry weight	max.
Soy, as soldZhi et al. 2020	soybean cultivar Beijiang 1, control soilCollection date unreportedDry weight	not detected.
Soy, as soldZhi et al. 2020	soybean cultivar Beijiang 1, pot soil spiked to 1.0 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Beijiang 1, pot soil spiked to 2.5 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Hefeng 40, control soilCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Hefeng 40, pot soil spiked to 1.0 mg Cd/kg (T1)Collection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Hefeng 40, pot soil spiked to 2.5 mg Cd/kg (T2)Collection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Heinong 48, control soilCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Heinong 48, pot soil spiked to 1.0 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Heinong 48, pot soil spiked to 2.5 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Ji 1005, control soilCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Ji 1005, pot soil spiked to 1.0 mg Cd/kg (T1)Collection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Ji 1005, pot soil spiked to 2.5 mg Cd/kg (T2)Collection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Kaijiao 8157, control soilCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Kaijiao 8157, pot soil spiked to 1.0 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Kaijiao 8157, pot soil spiked to 2.5 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Kaiyu 11, control soilCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Kaiyu 11, pot soil spiked to 1.0 mg Cd/kg (T1)Collection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Kaiyu 11, pot soil spiked to 2.5 mg Cd/kg (T2)Collection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Kaiyu 12, control soilCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Kaiyu 12, pot soil spiked to 1.0 mg Cd/kg (T1)Collection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Kaiyu 12, pot soil spiked to 2.5 mg Cd/kg (T2)Collection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Kaiyu 13, control soilCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Kaiyu 13, pot soil spiked to 1.0 mg Cd/kg (T1)Collection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Kaiyu 13, pot soil spiked to 2.5 mg Cd/kg (T2)Collection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Kenqianbei 4, control soilCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Kenqianbei 4, pot soil spiked to 1.0 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Kenqianbei 4, pot soil spiked to 2.5 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Liaodou 15, control soilCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Liaodou 15, pot soil spiked to 1.0 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Liaodou 15, pot soil spiked to 2.5 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Liaodou 21, control soilCollection date unreportedDry weight	not detected.
Soy, as soldZhi et al. 2020	soybean cultivar Liaodou 21, pot soil spiked to 1.0 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Liaodou 21, pot soil spiked to 2.5 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Liaoxian 1, control soilCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Liaoxian 1, pot soil spiked to 1.0 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Liaoxian 1, pot soil spiked to 2.5 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Shennong 6, control soilCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Shennong 6, pot soil spiked to 1.0 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Shennong 6, pot soil spiked to 2.5 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Shennong 8, control soilCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Shennong 8, pot soil spiked to 1.0 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Shennong 8, pot soil spiked to 2.5 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Shennong 10, control soilCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Shennong 10, pot soil spiked to 1.0 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Shennong 10, pot soil spiked to 2.5 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Suke 1, control soilCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Suke 1, pot soil spiked to 1.0 mg Cd/kg (T1)Collection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Suke 1, pot soil spiked to 2.5 mg Cd/kg (T2)Collection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Taiwan 292, control soilCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Taiwan 292, pot soil spiked to 1.0 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Taiwan 292, pot soil spiked to 2.5 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Tiedou 36, control soilCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Tiedou 36, pot soil spiked to 1.0 mg Cd/kg (T1)Collection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Tiedou 36, pot soil spiked to 2.5 mg Cd/kg (T2)Collection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Tiefeng 29, control soilCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Tiefeng 29, pot soil spiked to 1.0 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Tiefeng 29, pot soil spiked to 2.5 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Tiefeng 30, control soilCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Tiefeng 30, pot soil spiked to 1.0 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Tiefeng 30, pot soil spiked to 2.5 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Tiefeng 31, control soilCollection date unreportedDry weight	not detected.
Soy, as soldZhi et al. 2020	soybean cultivar Tiefeng 31, pot soil spiked to 1.0 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Tiefeng 31, pot soil spiked to 2.5 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Tiefeng 33, control soilCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Tiefeng 33, pot soil spiked to 1.0 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Tiefeng 33, pot soil spiked to 2.5 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Tiefeng 35, control soilCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Tiefeng 35, pot soil spiked to 1.0 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Tiefeng 35, pot soil spiked to 2.5 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Tiefeng 37, control soilCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Tiefeng 37, pot soil spiked to 1.0 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Tiefeng 37, pot soil spiked to 2.5 mg Cd/kgCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Xinfeng 1, control soilCollection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Xinfeng 1, pot soil spiked to 1.0 mg Cd/kg (T1)Collection date unreportedDry weight	mean.
Soy, as soldZhi et al. 2020	soybean cultivar Xinfeng 1, pot soil spiked to 2.5 mg Cd/kg (T2)Collection date unreportedDry weight	mean.
Soy, as soldauthors 2016	raw oilseed, fraction 0, one production lotCollection date unreportedBasis not stated	mean.
Soy, as soldCantoral et al. 2024	all samples (single foodstuff, duplicate mean)2022–2023Basis not stated	sample value.
Soy, as soldCentre for Food Safety (Hong Kong) 2013	all samples2010–2011Basis not stated	mean.
Soy, as soldCentre for Food Safety (Hong Kong) 2013	all samples2010–2011Basis not stated	mean.
Soy, as soldJoint FAO/WHO Expert Committee on Food Additives 2022	GEMS/Food contaminants database2011Basis not stated	mean.
Soy, as soldJoint FAO/WHO Expert Committee on Food Additives 2022	GEMS/Food contaminants database2011Basis not stated	mean.
Soy, as soldJoint FAO/WHO Expert Committee on Food Additives 2022	GEMS/Food contaminants database2011Basis not stated	mean.
Soy, as soldWang et al. 2020	all samples (both areas)2015–2016Basis not stated	mean.
Soy, as soldWang et al. 2020	all samples (both areas)2015–2016Basis not stated	mean.
Soy, as soldWang et al. 2020	Other (non-Songhua) area2015–2016Basis not stated	mean.
Soy, as soldWang et al. 2020	Other (non-Songhua) area2015–2016Basis not stated	mean.
Soy, as soldWang et al. 2020	Songhua River basin area2015–2016Basis not stated	mean.
Soy, as soldWang et al. 2020	Songhua River basin area2015–2016Basis not stated	mean.
Soy, as soldZhu et al. 2024	all samples2021–2022Basis not stated	mean.
2 findings · 1 study · 2 spread figures folded in
Soy, driedYu et al. 2023	all samples of this bean type, 11 Zhejiang monitoring areasCollection date unreported	median.
Soy, driedYu et al. 2023	all samples of this bean type, 11 Zhejiang monitoring areasCollection date unreported	mean.Same samples: 489 µg/kg (maximum); 169 µg/kg (percentile)
Soy, driedYu et al. 2023	all samples of this bean type, 11 Zhejiang monitoring areasCollection date unreported	max.
Soy, driedYu et al. 2023	all samples of this bean type, 11 Zhejiang monitoring areasCollection date unreported	percentile.
3 findings · 1 study
Soy, powder or groundWHO GEMS/Food	EURO2004–2015	median.
Soy, powder or groundWHO GEMS/Food	PAHO2017	median.
Soy, powder or groundWHO GEMS/Food	PAHO2017	median.
2 findings · 2 studies
Soy, oilWHO GEMS/Food	EURO2003–2004	median.
Soy, oilZhu et al. 2011	all samples (nationwide supermarkets)2009–2010	mean.

36 findings · 9 studies · 4 spread figures folded in
Soy, as soldWHO GEMS/Food	EURO2015	median.
Soy, as soldWHO GEMS/Food	EURO2015	median.
Soy, as soldWHO GEMS/Food	EURO2012–2015	median.
Soy, as soldWHO GEMS/Food	EURO2007–2012	median.
Soy, as soldWHO GEMS/Food	EURO2012–2015	median.
Soy, as soldWHO GEMS/Food	EURO2007–2012	median.
Soy, as soldWHO GEMS/Food	EURO2015–2022	median.
Soy, as soldWHO GEMS/Food	EURO2015	median.
Soy, as soldWHO GEMS/Food	EURO2019–2023	median.
Soy, as soldWHO GEMS/Food	EURO2023	median.
Soy, as soldWHO GEMS/Food	EURO2003–2023	median.
Soy, as soldWHO GEMS/Food	EURO2007–2015	median.
Soy, as soldWHO GEMS/Food	EURO2012	median.
Soy, as soldWHO GEMS/Food	EURO2004–2009	median.
Soy, as soldWHO GEMS/Food	EURO2001–2021	median.
Soy, as soldWHO GEMS/Food	EURO2007–2008	median.
Soy, as soldWHO GEMS/Food	WPRO2014	median.
Soy, as soldWHO GEMS/Food	WPRO2008	median.
Soy, as soldWHO GEMS/Food	WPRO2015	median.
Soy, as soldWHO GEMS/Food	WPRO2005	median.
Soy, as soldWHO GEMS/Food	WPRO2000–2002	median.
Soy, as soldWHO GEMS/Food	WPRO2005–2014	median.
Soy, as soldFood Standards Agency 2014	dry powder formula category2013–2014	mean.
Soy, as soldMarín et al. 2018	Milk and dairy products group, Region of Valencia TDS 2010-20112010–2011	mean.Same samples: 3.4 µg/kg (minimum); 4.8 µg/kg (maximum)
Soy, as soldMarín et al. 2018	Milk and dairy products group, Region of Valencia TDS 2010-20112010–2011	min.
Soy, as soldMarín et al. 2018	Milk and dairy products group, Region of Valencia TDS 2010-20112010–2011	max.
Soy, as soldRuzik et al. 2022	Polish marketsCollection date unreported	mean.
Soy, as soldVentura et al. 2025	pooled TDS sample2014–2016	sample value.
Soy, as soldAdam et al. 2022	Tamale Aboabo market composite2017Dry weight	mean.Same samples: 49 µg/kg (minimum); 60 µg/kg (maximum)
Soy, as soldAdam et al. 2022	Tamale Aboabo market composite2017Dry weight	min.
Soy, as soldAdam et al. 2022	Tamale Aboabo market composite2017Dry weight	max.
Soy, as soldZhang et al. 2026	Xiushan Mn mining and smelting area2020Dry weight	mean.
Soy, as soldRuzik et al. 2022	Polish marketsCollection date unreportedBasis not stated	mean.
Soy, as soldWang et al. 2020	all samples (both areas)2015–2016Basis not stated	mean.
Soy, as soldWang et al. 2020	all samples (both areas)2015–2016Basis not stated	mean.
Soy, as soldWang et al. 2020	Other (non-Songhua) area2015–2016Basis not stated	mean.
Soy, as soldWang et al. 2020	Other (non-Songhua) area2015–2016Basis not stated	mean.
Soy, as soldWang et al. 2020	Songhua River basin area2015–2016Basis not stated	mean.
Soy, as soldWang et al. 2020	Songhua River basin area2015–2016Basis not stated	mean.
Soy, as soldZhu et al. 2024	all samples2021–2022Basis not stated	mean.
1 finding · 1 study · 2 spread figures folded in
Soy, driedYu et al. 2023	all samples of this bean type, 11 Zhejiang monitoring areasCollection date unreported	mean.Same samples: 170 µg/kg (maximum); 50 µg/kg (percentile)
Soy, driedYu et al. 2023	all samples of this bean type, 11 Zhejiang monitoring areasCollection date unreported	max.
Soy, driedYu et al. 2023	all samples of this bean type, 11 Zhejiang monitoring areasCollection date unreported	percentile.
3 findings · 1 study
Soy, powder or groundWHO GEMS/Food	EURO2015	median.
Soy, powder or groundWHO GEMS/Food	EURO2015	median.
Soy, powder or groundWHO GEMS/Food	EURO2019	median.

14 findings · 5 studies · 2 spread figures folded in
Soy, as soldWHO GEMS/Food	EURO2018	median.
Soy, as soldWHO GEMS/Food	EURO2021	median.
Soy, as soldWHO GEMS/Food	EURO2018	median.
Soy, as soldWHO GEMS/Food	EURO2015–2016	median.
Soy, as soldWHO GEMS/Food	WPRO2010	median.
Soy, as soldWHO GEMS/Food	WPRO2010	median.
Soy, as soldFood Standards Agency 2014	dry powder formula category2013–2014	mean.
Soy, as soldMarín et al. 2018	Milk and dairy products group, Region of Valencia TDS 2010-20112010–2011	mean.Same samples: 1.2 µg/kg (minimum); 8.2 µg/kg (maximum)
Soy, as soldMarín et al. 2018	Milk and dairy products group, Region of Valencia TDS 2010-20112010–2011	min.
Soy, as soldMarín et al. 2018	Milk and dairy products group, Region of Valencia TDS 2010-20112010–2011	max.
Soy, as soldCentre for Food Safety FEHD HKSAR 2012	all samples (4 sampling occasions, composites)2010–2011Basis not stated	mean.
Soy, as soldCentre for Food Safety FEHD HKSAR 2012	all samples (4 sampling occasions, composites)2010–2011Basis not stated	not detected.
Soy, as soldEuropean Food Safety Authority (EFSA) 2014	European dietary-survey occurrence data2003–2013Basis not stated	mean.
Soy, as soldEuropean Food Safety Authority (EFSA) 2014	European dietary-survey occurrence data2003–2013Basis not stated	mean.
Soy, as soldEuropean Food Safety Authority (EFSA) 2014	European dietary-survey occurrence data2003–2013Basis not stated	mean.
Soy, as soldEuropean Food Safety Authority (EFSA) 2014	European dietary-survey occurrence data2003–2013Basis not stated	mean.

20 findings · 5 studies
Soy, as soldWHO GEMS/Food	EURO2015	median.
Soy, as soldWHO GEMS/Food	EURO2004–2015	median.
Soy, as soldWHO GEMS/Food	EURO2012	median.
Soy, as soldWHO GEMS/Food	EURO2004–2017	median.
Soy, as soldWHO GEMS/Food	EURO2012	median.
Soy, as soldWHO GEMS/Food	EURO2007–2008	median.
Soy, as soldWHO GEMS/Food	EURO2004–2019	median.
Soy, as soldWHO GEMS/Food	EURO2002	median.
Soy, as soldWHO GEMS/Food	EURO2007–2008	median.
Soy, as soldWHO GEMS/Food	EURO2003–2005	median.
Soy, as soldWHO GEMS/Food	WPRO2014	median.
Soy, as soldWHO GEMS/Food	WPRO2008	median.
Soy, as soldWHO GEMS/Food	WPRO2005	median.
Soy, as soldWHO GEMS/Food	WPRO2001–2002	median.
Soy, as soldWHO GEMS/Food	WPRO2005	median.
Soy, as soldDabeka et al. 2003	Ottawa composite (retail Oct-Nov 2000)1998–2000	sample value.
Soy, as soldDabeka et al. 2003	Whitehorse composite (retail Jan-Feb 1998)1998–2000	not detected.
Soy, as soldFood Standards Agency 2014	dry powder formula category2013–2014	mean.
Soy, as soldKurniawati et al. 2021	all samplesCollection date unreported	mean.
Soy, as soldZhu et al. 2024	all samples2021–2022Basis not stated	not detected.
1 finding · 1 study · 2 spread figures folded in
Soy, driedYu et al. 2023	all samples of this bean type, 11 Zhejiang monitoring areasCollection date unreported	mean.Same samples: 37.1 µg/kg (maximum); 8.8 µg/kg (percentile)
Soy, driedYu et al. 2023	all samples of this bean type, 11 Zhejiang monitoring areasCollection date unreported	max.
Soy, driedYu et al. 2023	all samples of this bean type, 11 Zhejiang monitoring areasCollection date unreported	percentile.
1 finding · 1 study
Soy, powder or groundWHO GEMS/Food	EURO2007–2015	median.

16 findings · 7 studies · 2 spread figures folded in
Soy, as soldFood Standards Agency 2014	dry powder formula category2013–2014	mean.
Soy, as soldMarques et al. 2021	all samples2021	sample value.
Soy, as soldMarques et al. 2021	all samples2021	sample value.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	max.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	max.
Soy, as soldOrisakwe et al. 2012	all samples (Ohaji, Umuagwo, Owerri sites)Collection date unreportedDry weight	mean.
Soy, as soldZhang et al. 2026	Xiushan Mn mining and smelting area2020Dry weight	mean.
Soy, as soldCentre for Food Safety (Hong Kong) 2013	all samples2010–2011Basis not stated	mean.
Soy, as soldCentre for Food Safety (Hong Kong) 2013	all samples2010–2011Basis not stated	mean.
Soy, as soldHan et al. 2024	Beans (Soybean)2018–2019Basis not stated	median.
Soy, as soldHan et al. 2024	Beans (Soybean)2018–2019Basis not stated	mean.Same samples: 9,210–9,210 µg/kg (range); 9,210 µg/kg (percentile)
Soy, as soldHan et al. 2024	Beans (Soybean)2018–2019Basis not stated	range.
Soy, as soldHan et al. 2024	Beans (Soybean)2018–2019Basis not stated	percentile.
1 finding · 1 study
Soy, oilZhu et al. 2011	all samples (nationwide supermarkets)2009–2010	mean.

17 findings · 9 studies · 2 spread figures folded in
Soy, as soldEFSA Panel on Contaminants in the Food Chain (CONTAM) 2014	all samples, EFSA FoodEx subgroup, LB2000–2013	mean.
Soy, as soldEFSA Panel on Contaminants in the Food Chain (CONTAM) 2014	all samples, EFSA FoodEx subgroup, UB2000–2013	mean.
Soy, as soldFood Standards Agency 2014	dry powder formula category2013–2014	mean.
Soy, as soldKurniawati et al. 2021	all samplesCollection date unreported	mean.
Soy, as soldMarín et al. 2018	Milk and dairy products group, Region of Valencia TDS 2010-20112010–2011	mean.Same samples: 9.2 µg/kg (minimum); 28.9 µg/kg (maximum)
Soy, as soldMarín et al. 2018	Milk and dairy products group, Region of Valencia TDS 2010-20112010–2011	min.
Soy, as soldMarín et al. 2018	Milk and dairy products group, Region of Valencia TDS 2010-20112010–2011	max.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	max.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	mean.
Soy, as soldWu 2024	soybean samples, Jiaozuo marketsCollection date unreported	mean.
Soy, as soldHajara et al. 2026	Alagarno ward markets, Gamawa LGACollection date unreportedDry weight	mean.
Soy, as soldHajara et al. 2026	all wards (Kafiromi, Alagarno, Tumbi), market samplesCollection date unreportedDry weight	mean.
Soy, as soldHajara et al. 2026	Kafiromi ward markets, Gamawa LGACollection date unreportedDry weight	mean.
Soy, as soldHajara et al. 2026	Tumbi ward markets, Gamawa LGACollection date unreportedDry weight	mean.
Soy, as soldZhang et al. 2026	Xiushan Mn mining and smelting area2020Dry weight	mean.
Soy, as soldZhu et al. 2024	all samples2021–2022Basis not stated	mean.
2 findings · 1 study · 2 spread figures folded in
Soy, driedYu et al. 2023	all samples of this bean type, 11 Zhejiang monitoring areasCollection date unreported	median.
Soy, driedYu et al. 2023	all samples of this bean type, 11 Zhejiang monitoring areasCollection date unreported	mean.Same samples: 1,510 µg/kg (maximum); 720 µg/kg (percentile)
Soy, driedYu et al. 2023	all samples of this bean type, 11 Zhejiang monitoring areasCollection date unreported	max.
Soy, driedYu et al. 2023	all samples of this bean type, 11 Zhejiang monitoring areasCollection date unreported	percentile.

20 findings · 5 studies
Soy, as soldWHO GEMS/Food	EURO2007–2008	median.
Soy, as soldWHO GEMS/Food	EURO2007–2008	median.
Soy, as soldFood Standards Agency 2014	dry powder formula category2013–2014	mean.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	max.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	max.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	max.
Soy, as soldCentre for Food Safety (Hong Kong) 2013	all samples2010–2011Basis not stated	mean.
Soy, as soldCentre for Food Safety (Hong Kong) 2013	all samples2010–2011Basis not stated	mean.
Soy, as soldWang et al. 2020	all samples (both areas)2015–2016Basis not stated	mean.
Soy, as soldWang et al. 2020	all samples (both areas)2015–2016Basis not stated	mean.
Soy, as soldWang et al. 2020	Other (non-Songhua) area2015–2016Basis not stated	mean.
Soy, as soldWang et al. 2020	Other (non-Songhua) area2015–2016Basis not stated	mean.
Soy, as soldWang et al. 2020	Songhua River basin area2015–2016Basis not stated	mean.
Soy, as soldWang et al. 2020	Songhua River basin area2015–2016Basis not stated	mean.

16 findings · 5 studies · 2 spread figures folded in
Soy, as soldWHO GEMS/Food	EURO2007–2008	median.
Soy, as soldWHO GEMS/Food	EURO2007–2008	median.
Soy, as soldWHO GEMS/Food	WPRO2010	median.
Soy, as soldWHO GEMS/Food	WPRO2010	median.
Soy, as soldWHO GEMS/Food	WPRO2005	median.
Soy, as soldWHO GEMS/Food	WPRO2002	median.
Soy, as soldWHO GEMS/Food	WPRO2005	median.
Soy, as soldFood Standards Agency 2014	dry powder formula category2013–2014	mean.
Soy, as soldMarín et al. 2018	Milk and dairy products group, Region of Valencia TDS 2010-20112010–2011	mean.Same samples: 2.1 µg/kg (minimum); 556 µg/kg (maximum)
Soy, as soldMarín et al. 2018	Milk and dairy products group, Region of Valencia TDS 2010-20112010–2011	min.
Soy, as soldMarín et al. 2018	Milk and dairy products group, Region of Valencia TDS 2010-20112010–2011	max.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	max.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	mean.
Soy, as soldCentre for Food Safety (Hong Kong) 2013	all samples2010–2011Basis not stated	not detected.
Soy, as soldCentre for Food Safety (Hong Kong) 2013	all samples2010–2011Basis not stated	not detected.

13 findings · 4 studies
Soy, as soldFood Standards Agency 2014	dry powder formula category2013–2014	mean.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	max.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	max.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	max.
Soy, as soldZhang et al. 2026	Xiushan Mn mining and smelting area2020Dry weight	mean.
Soy, as soldCentre for Food Safety (Hong Kong) 2013	all samples2010–2011Basis not stated	mean.
Soy, as soldCentre for Food Safety (Hong Kong) 2013	all samples2010–2011Basis not stated	not detected.

2 findings · 1 study
Soy, as soldMarques et al. 2021	all samples2021	not detected.
Soy, as soldMarques et al. 2021	all samples2021	not detected.

9 findings · 4 studies
Soy, as soldKurniawati et al. 2021	all samplesCollection date unreported	mean.
Soy, as soldMarques et al. 2021	all samples2021	not detected.
Soy, as soldMarques et al. 2021	all samples2021	not detected.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	max.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	mean.
Soy, as soldVentura et al. 2025	pooled TDS sample2014–2016	sample value.

20 findings · 7 studies · 2 spread figures folded in
Soy, as soldFood Standards Agency 2014	dry powder formula category2013–2014	mean.
Soy, as soldMarín et al. 2018	Milk and dairy products group, Region of Valencia TDS 2010-20112010–2011	mean.Same samples: 1,310 µg/kg (minimum); 2,030 µg/kg (maximum)
Soy, as soldMarín et al. 2018	Milk and dairy products group, Region of Valencia TDS 2010-20112010–2011	min.
Soy, as soldMarín et al. 2018	Milk and dairy products group, Region of Valencia TDS 2010-20112010–2011	max.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	max.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	max.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	max.
Soy, as soldAdam et al. 2022	Tamale Aboabo market composite2017Dry weight	mean.
Soy, as soldAdam et al. 2022	Tamale Aboabo market composite2017Dry weight	min.
Soy, as soldAdam et al. 2022	Tamale Aboabo market composite2017Dry weight	max.
Soy, as soldHajara et al. 2026	Alagarno ward markets, Gamawa LGACollection date unreportedDry weight	mean.
Soy, as soldHajara et al. 2026	all wards (Kafiromi, Alagarno, Tumbi), market samplesCollection date unreportedDry weight	mean.
Soy, as soldHajara et al. 2026	Kafiromi ward markets, Gamawa LGACollection date unreportedDry weight	mean.
Soy, as soldHajara et al. 2026	Tumbi ward markets, Gamawa LGACollection date unreportedDry weight	mean.
Soy, as soldZhang et al. 2026	Xiushan Mn mining and smelting area2020Dry weight	mean.
Soy, as soldauthors 2016	raw oilseed, fraction 0, one production lotCollection date unreportedBasis not stated	mean.
1 finding · 1 study
Soy, oilZhu et al. 2011	all samples (nationwide supermarkets)2009–2010	mean.

11 findings · 3 studies
Soy, as soldFood Standards Agency 2014	dry powder formula category2013–2014	mean.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	max.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	max.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	max.
Soy, as soldauthors 2016	raw oilseed, fraction 0, one production lotCollection date unreportedBasis not stated	mean.
1 finding · 1 study
Soy, oilZhu et al. 2011	all samples (nationwide supermarkets)2009–2010	mean.

34 findings · 5 studies
Soy, as soldAndrade et al. 2023	all 14 products2019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 12019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 22019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 32019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 42019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 52019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 62019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 72019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 82019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 92019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 102019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 112019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 122019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 132019	mean.
Soy, as soldAndrade et al. 2023	retail soy beverage product 142019	mean.
Soy, as soldFood Standards Agency 2014	dry powder formula category2013–2014	mean.
Soy, as soldMarques et al. 2021	all samples2021	sample value.
Soy, as soldMarques et al. 2021	all samples2021	sample value.
Soy, as soldMilani et al. 2023	Soy drinkCollection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drinkCollection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	max.
Soy, as soldMilani et al. 2023	Soy drinkCollection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drinkCollection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	max.
Soy, as soldMilani et al. 2023	Soy drinkCollection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drinkCollection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	max.
Soy, as soldauthors 2016	raw oilseed, fraction 0, one production lotCollection date unreportedBasis not stated	mean.
1 finding · 1 study
Soy, oilZhu et al. 2011	all samples (nationwide supermarkets)2009–2010	mean.

20 findings · 7 studies
Soy, as soldFood Standards Agency 2014	dry powder formula category2013–2014	mean.
Soy, as soldKurniawati et al. 2021	all samplesCollection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	max.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	max.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	max.
Soy, as soldAdam et al. 2022	Tamale Aboabo market composite2017Dry weight	mean.
Soy, as soldAdam et al. 2022	Tamale Aboabo market composite2017Dry weight	min.
Soy, as soldAdam et al. 2022	Tamale Aboabo market composite2017Dry weight	max.
Soy, as soldHajara et al. 2026	Alagarno ward markets, Gamawa LGACollection date unreportedDry weight	mean.
Soy, as soldHajara et al. 2026	all wards (Kafiromi, Alagarno, Tumbi), market samplesCollection date unreportedDry weight	mean.
Soy, as soldHajara et al. 2026	Kafiromi ward markets, Gamawa LGACollection date unreportedDry weight	mean.
Soy, as soldHajara et al. 2026	Tumbi ward markets, Gamawa LGACollection date unreportedDry weight	mean.
Soy, as soldZhang et al. 2026	Xiushan Mn mining and smelting area2020Dry weight	mean.
Soy, as soldauthors 2016	raw oilseed, fraction 0, one production lotCollection date unreportedBasis not stated	mean.
1 finding · 1 study
Soy, oilZhu et al. 2011	all samples (nationwide supermarkets)2009–2010	mean.

4 findings · 2 studies
Soy, as soldMarques et al. 2021	all samples2021	not detected.
Soy, as soldMarques et al. 2021	all samples2021	not detected.
Soy, as soldCentre for Food Safety (Hong Kong) 2013	all samples2010–2011Basis not stated	mean.
Soy, as soldCentre for Food Safety (Hong Kong) 2013	all samples2010–2011Basis not stated	not detected.

11 findings · 3 studies
Soy, as soldFood Standards Agency 2014	dry powder formula category2013–2014	mean.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	max.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	max.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	max.
Soy, as soldVentura et al. 2025	pooled TDS sample2014–2016	sample value.

3 findings · 1 study
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	mean.
1 finding · 1 study
Soy, oilZhu et al. 2011	all samples (nationwide supermarkets)2009–2010	mean.

2 findings · 1 study
Soy, as soldMarques et al. 2021	all samples2021	not detected.
Soy, as soldMarques et al. 2021	all samples2021	not detected.

2 findings · 2 studies
Soy, as soldFood Standards Agency 2014	dry powder formula category2013–2014	mean.
Soy, as soldVentura et al. 2025	pooled TDS sample2014–2016	sample value.

3 findings · 1 study
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	mean.

1 finding · 1 study
Soy, as soldVentura et al. 2025	pooled TDS sample2014–2016	sample value.

16 findings · 2 studies
Soy, as soldMilani et al. 2023	Soy drinkCollection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drinkCollection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (isolate soy protein), total content, n=6Collection date unreported	max.
Soy, as soldMilani et al. 2023	Soy drinkCollection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drinkCollection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (soy hydrosoluble extract), total content, n=6Collection date unreported	max.
Soy, as soldMilani et al. 2023	Soy drinkCollection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drinkCollection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	mean.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	min.
Soy, as soldMilani et al. 2023	Soy drink (soybeans), total content, n=6Collection date unreported	max.
Soy, as soldVentura et al. 2025	pooled TDS sample2014–2016	sample value.

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 soy Pb reference · Soy, as sold

Reference and scope

HMI soy Pb reference · Soy, as sold. Soy, as sold; As sold, as eaten or fresh weight, µg/kg; no moisture conversion. Distribution of subgroup median concentration values.

Reference (µg/kg)	Q1	Median	Q3
Primary HMI reference	≤19	≤50	≤50
Alternative weighting	≤22.5	≤50	≤50

128 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.

Contributing family	Reference units	Weight
WHO GEMS/Food	126	50%
Marques et al. 2021	2	50%

Methods

HMI pools subgroup medians for soy, as sold on a as sold, as eaten or fresh weight basis: monitoring results summarized as one median per reporting region, sampling type and collection year, total diet composites and printed sample values as single observations, and printed subgroup medians from ordinary farms or markets. Each sampling family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations. Other forms, other bases, targeted sampling and results from mining, industrial, wastewater-affected, experimental, fertiliser or manufacturer settings are kept out of this reference.

Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1 to Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

2 sampling families with different designs; none is a probability sample of world supply. Reporting regions and purchase markets are not growing areas. Non-detects with differing limits widen the bounds. A position is descriptive concentration, not a safety or compliance judgment.

Admitted source families

Reference version and interpretation

Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.

hmi-pooled-reference-v2:soy-9ee274a26bfb:fe5ad72873de

The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.

HMI soy Pb reference · Soy, as sold, dry weight

Reference and scope

HMI soy Pb reference · Soy, as sold, dry weight. Soy, as sold; Dry weight, µg/kg; no moisture conversion. Distribution of subgroup median concentration values.

Reference (µg/kg)	Q1	Median	Q3
Primary HMI reference	≤5	≤5	50
Alternative weighting	≤3	25–27.5 (bounded)	50

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 soy, as sold on a dry weight basis: monitoring results summarized as one median per reporting region, sampling type and collection year, total diet composites and printed sample values as single observations, and printed subgroup medians from ordinary farms or markets. Each sampling family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations. Other forms, other bases, targeted sampling and results from mining, industrial, wastewater-affected, experimental, fertiliser or manufacturer settings are kept out of this reference.

Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1 to Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

WHO GEMS/Food is the only compatible sampling family for soy, as sold on this basis in the completed soy ledger, so this single-family reference cannot represent other markets. Reporting regions and purchase markets are not growing areas. Non-detects with differing limits widen the bounds. A position is descriptive concentration, not a safety or compliance judgment.

Admitted source families

WHO GEMS/Food

Reference version and interpretation

Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.

hmi-pooled-reference-v2:soy-ca9318cebdc8:775c2b5211bf

The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.

HMI soy Pb reference · Soy, powder or ground

Reference and scope

HMI soy Pb reference · Soy, powder or ground. Soy, powder or ground; As sold, as eaten or fresh weight, µg/kg; no moisture conversion. Distribution of subgroup median concentration values.

Reference (µg/kg)	Q1	Median	Q3
Primary HMI reference	≤5.25	≤20	2.2–90 (bounded)
Alternative weighting	≤5.25	≤20	2.2–90 (bounded)

10 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.

Contributing family	Reference units	Weight
WHO GEMS/Food	10	100%

Methods

HMI pools subgroup medians for soy, powder or ground on a as sold, as eaten or fresh weight basis: monitoring results summarized as one median per reporting region, sampling type and collection year, total diet composites and printed sample values as single observations, and printed subgroup medians from ordinary farms or markets. Each sampling family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations. Other forms, other bases, targeted sampling and results from mining, industrial, wastewater-affected, experimental, fertiliser or manufacturer settings are kept out of this reference.

Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1 to Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

WHO GEMS/Food is the only compatible sampling family for soy, powder or ground on this basis in the completed soy ledger, so this single-family reference cannot represent other markets. Reporting regions and purchase markets are not growing areas. Non-detects with differing limits widen the bounds. A position is descriptive concentration, not a safety or compliance judgment.

Admitted source families

WHO GEMS/Food

Reference version and interpretation

Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.

hmi-pooled-reference-v2:soy-981cf097d13c:71a000282089

The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.

HMI soy Pb reference · Soy, oil

Reference and scope

HMI soy Pb reference · Soy, oil. Soy, oil; As sold, as eaten or fresh weight, µg/kg; no moisture conversion. Distribution of subgroup median concentration values.

Reference (µg/kg)	Q1	Median	Q3
Primary HMI reference	10–15 (bounded)	10–15 (bounded)	10–15 (bounded)
Alternative weighting	10–15 (bounded)	10–15 (bounded)	10–15 (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 soy, oil on a as sold, as eaten or fresh weight basis: monitoring results summarized as one median per reporting region, sampling type and collection year, total diet composites and printed sample values as single observations, and printed subgroup medians from ordinary farms or markets. Each sampling family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations. Other forms, other bases, targeted sampling and results from mining, industrial, wastewater-affected, experimental, fertiliser or manufacturer settings are kept out of this reference.

Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1 to Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

WHO GEMS/Food is the only compatible sampling family for soy, oil on this basis in the completed soy ledger, so this single-family reference cannot represent other markets. Reporting regions and purchase markets are not growing areas. Non-detects with differing limits widen the bounds. A position is descriptive concentration, not a safety or compliance judgment.

Admitted source families

WHO GEMS/Food

Reference version and interpretation

Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.

hmi-pooled-reference-v2:soy-0b4e394766a0:cbb0db639651

The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.

HMI soy Cd reference · Soy, as sold

Reference and scope

HMI soy Cd reference · Soy, as sold. Soy, as sold; As sold, as eaten or fresh weight, µg/kg; no moisture conversion. Distribution of subgroup median concentration values.

Reference (µg/kg)	Q1	Median	Q3
Primary HMI reference	4–4.5 (bounded)	12	17.2
Alternative weighting	3.79–4.5 (bounded)	12	17.2

149 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.

Contributing family	Reference units	Weight
WHO GEMS/Food	146	50%
EFSA Panel on Contaminants in the Food Chain (CONTAM) 2009	3	50%

Methods

HMI pools subgroup medians for soy, as sold on a as sold, as eaten or fresh weight basis: monitoring results summarized as one median per reporting region, sampling type and collection year, total diet composites and printed sample values as single observations, and printed subgroup medians from ordinary farms or markets. Each sampling family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations. Other forms, other bases, targeted sampling and results from mining, industrial, wastewater-affected, experimental, fertiliser or manufacturer settings are kept out of this reference.

Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1 to Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

2 sampling families with different designs; none is a probability sample of world supply. Reporting regions and purchase markets are not growing areas. Non-detects with differing limits widen the bounds. A position is descriptive concentration, not a safety or compliance judgment.

Admitted source families

Reference version and interpretation

Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.

hmi-pooled-reference-v2:soy-30cc150efab3:ecc35ba60028

The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.

HMI soy Cd reference · Soy, as sold, dry weight

Reference and scope

HMI soy Cd reference · Soy, as sold, dry weight. Soy, as sold; Dry weight, µg/kg; no moisture conversion. Distribution of subgroup median concentration values.

Reference (µg/kg)	Q1	Median	Q3
Primary HMI reference	70	70	70
Alternative weighting	70	70	70

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 soy, as sold on a dry weight basis: monitoring results summarized as one median per reporting region, sampling type and collection year, total diet composites and printed sample values as single observations, and printed subgroup medians from ordinary farms or markets. Each sampling family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations. Other forms, other bases, targeted sampling and results from mining, industrial, wastewater-affected, experimental, fertiliser or manufacturer settings are kept out of this reference.

Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1 to Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

WHO GEMS/Food is the only compatible sampling family for soy, as sold on this basis in the completed soy ledger, so this single-family reference cannot represent other markets. Reporting regions and purchase markets are not growing areas. Non-detects with differing limits widen the bounds. A position is descriptive concentration, not a safety or compliance judgment.

Admitted source families

WHO GEMS/Food

Reference version and interpretation

Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.

hmi-pooled-reference-v2:soy-de4625814d05:219bf86ee798

The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.

HMI soy Cd reference · Soy, dried

Reference and scope

HMI soy Cd reference · Soy, dried. Soy, dried; As sold, as eaten or fresh weight, µg/kg; no moisture conversion. Distribution of subgroup median concentration values.

Reference (µg/kg)	Q1	Median	Q3
Primary HMI reference	22.6	22.6	22.6
Alternative weighting	22.6	22.6	22.6

1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.

Contributing family	Reference units	Weight
Yu et al. 2023	1	100%

Methods

HMI pools subgroup medians for soy, dried on a as sold, as eaten or fresh weight basis: monitoring results summarized as one median per reporting region, sampling type and collection year, total diet composites and printed sample values as single observations, and printed subgroup medians from ordinary farms or markets. Each sampling family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations. Other forms, other bases, targeted sampling and results from mining, industrial, wastewater-affected, experimental, fertiliser or manufacturer settings are kept out of this reference.

Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1 to Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

Yu et al. 2023 is the only compatible sampling family for soy, dried on this basis in the completed soy ledger, so this single-family reference cannot represent other markets. Reporting regions and purchase markets are not growing areas. Non-detects with differing limits widen the bounds. A position is descriptive concentration, not a safety or compliance judgment.

Admitted source families

Yu et al. 2023

Reference version and interpretation

Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.

hmi-pooled-reference-v2:soy-0961a3f408a5:b425ef0217a9

The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.

HMI soy Cd reference · Soy, powder or ground

Reference and scope

HMI soy Cd reference · Soy, powder or ground. Soy, powder or ground; As sold, as eaten or fresh weight, µg/kg; no moisture conversion. Distribution of subgroup median concentration values.

Reference (µg/kg)	Q1	Median	Q3
Primary HMI reference	≤2.71	0.45–5 (bounded)	2.71–10 (bounded)
Alternative weighting	≤1.42	0.9–5 (bounded)	4–15 (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 soy, powder or ground on a as sold, as eaten or fresh weight basis: monitoring results summarized as one median per reporting region, sampling type and collection year, total diet composites and printed sample values as single observations, and printed subgroup medians from ordinary farms or markets. Each sampling family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations. Other forms, other bases, targeted sampling and results from mining, industrial, wastewater-affected, experimental, fertiliser or manufacturer settings are kept out of this reference.

Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1 to Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

WHO GEMS/Food is the only compatible sampling family for soy, powder or ground on this basis in the completed soy ledger, so this single-family reference cannot represent other markets. Reporting regions and purchase markets are not growing areas. Non-detects with differing limits widen the bounds. A position is descriptive concentration, not a safety or compliance judgment.

Admitted source families

WHO GEMS/Food

Reference version and interpretation

Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.

hmi-pooled-reference-v2:soy-6c2d05e926a3:f6b7a0b07f81

The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.

HMI soy tAs reference · Soy, as sold

Reference and scope

HMI soy tAs reference · Soy, as sold. Soy, as sold; As sold, as eaten or fresh weight, µg/kg; no moisture conversion. Distribution of subgroup median concentration values.

Reference (µg/kg)	Q1	Median	Q3
Primary HMI reference	≤8.1	8.1	8.1–13.7 (bounded)
Alternative weighting	1.15–8.1 (bounded)	8.1	8.1–14 (bounded)

56 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.

Contributing family	Reference units	Weight
WHO GEMS/Food	55	50%
Ventura et al. 2025	1	50%

Methods

HMI pools subgroup medians for soy, as sold on a as sold, as eaten or fresh weight basis: monitoring results summarized as one median per reporting region, sampling type and collection year, total diet composites and printed sample values as single observations, and printed subgroup medians from ordinary farms or markets. Each sampling family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations. Other forms, other bases, targeted sampling and results from mining, industrial, wastewater-affected, experimental, fertiliser or manufacturer settings are kept out of this reference.

Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1 to Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

2 sampling families with different designs; none is a probability sample of world supply. Reporting regions and purchase markets are not growing areas. Non-detects with differing limits widen the bounds. A position is descriptive concentration, not a safety or compliance judgment.

Admitted source families

Reference version and interpretation

Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.

hmi-pooled-reference-v2:soy-97db0f917409:44b7f5752f3d

The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.

HMI soy tAs reference · Soy, powder or ground

Reference and scope

HMI soy tAs reference · Soy, powder or ground. Soy, powder or ground; As sold, as eaten or fresh weight, µg/kg; no moisture conversion. Distribution of subgroup median concentration values.

Reference (µg/kg)	Q1	Median	Q3
Primary HMI reference	0.2–2.25 (bounded)	0.2–3.92 (bounded)	0.2–5.6 (bounded)
Alternative weighting	0.2–2.25 (bounded)	0.2–3.92 (bounded)	0.2–5.6 (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 soy, powder or ground on a as sold, as eaten or fresh weight basis: monitoring results summarized as one median per reporting region, sampling type and collection year, total diet composites and printed sample values as single observations, and printed subgroup medians from ordinary farms or markets. Each sampling family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations. Other forms, other bases, targeted sampling and results from mining, industrial, wastewater-affected, experimental, fertiliser or manufacturer settings are kept out of this reference.

Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1 to Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

WHO GEMS/Food is the only compatible sampling family for soy, powder or ground on this basis in the completed soy ledger, so this single-family reference cannot represent other markets. Reporting regions and purchase markets are not growing areas. Non-detects with differing limits widen the bounds. A position is descriptive concentration, not a safety or compliance judgment.

Admitted source families

WHO GEMS/Food

Reference version and interpretation

Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.

hmi-pooled-reference-v2:soy-dc8145c36053:56d2c46781c7

The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.

HMI soy iAs reference · Soy, as sold

Reference and scope

HMI soy iAs reference · Soy, as sold. Soy, as sold; As sold, as eaten or fresh weight, µg/kg; no moisture conversion. Distribution of subgroup median concentration values.

Reference (µg/kg)	Q1	Median	Q3
Primary HMI reference	≤3.9	≤5.5	7–10 (bounded)
Alternative weighting	≤3.9	≤5.5	7–10 (bounded)

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 soy, as sold on a as sold, as eaten or fresh weight basis: monitoring results summarized as one median per reporting region, sampling type and collection year, total diet composites and printed sample values as single observations, and printed subgroup medians from ordinary farms or markets. Each sampling family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations. Other forms, other bases, targeted sampling and results from mining, industrial, wastewater-affected, experimental, fertiliser or manufacturer settings are kept out of this reference.

Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1 to Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

WHO GEMS/Food is the only compatible sampling family for soy, as sold on this basis in the completed soy ledger, so this single-family reference cannot represent other markets. Reporting regions and purchase markets are not growing areas. Non-detects with differing limits widen the bounds. A position is descriptive concentration, not a safety or compliance judgment.

Admitted source families

WHO GEMS/Food

Reference version and interpretation

Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.

hmi-pooled-reference-v2:soy-11e6e7b0485c:79fec1cd9079

The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.

HMI soy tHg reference · Soy, as sold

Reference and scope

HMI soy tHg reference · Soy, as sold. Soy, as sold; As sold, as eaten or fresh weight, µg/kg; no moisture conversion. Distribution of subgroup median concentration values.

Reference (µg/kg)	Q1	Median	Q3
Primary HMI reference	≤0.048	≤0.23	0.23–2 (bounded)
Alternative weighting	≤0.048	≤0.23	0.23–2 (bounded)

41 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.

Contributing family	Reference units	Weight
WHO GEMS/Food	39	50%
Dabeka et al. 2003	2	50%

Methods

HMI pools subgroup medians for soy, as sold on a as sold, as eaten or fresh weight basis: monitoring results summarized as one median per reporting region, sampling type and collection year, total diet composites and printed sample values as single observations, and printed subgroup medians from ordinary farms or markets. Each sampling family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations. Other forms, other bases, targeted sampling and results from mining, industrial, wastewater-affected, experimental, fertiliser or manufacturer settings are kept out of this reference.

Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1 to Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

2 sampling families with different designs; none is a probability sample of world supply. Reporting regions and purchase markets are not growing areas. Non-detects with differing limits widen the bounds. A position is descriptive concentration, not a safety or compliance judgment.

Admitted source families

Reference version and interpretation

Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.

hmi-pooled-reference-v2:soy-e26b6be98bc0:1164e9e732cc

The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.

HMI soy tHg reference · Soy, powder or ground

Reference and scope

HMI soy tHg reference · Soy, powder or ground. Soy, powder or ground; As sold, as eaten or fresh weight, µg/kg; no moisture conversion. Distribution of subgroup median concentration values.

Reference (µg/kg)	Q1	Median	Q3
Primary HMI reference	≤0.747	≤1.6	0.147–26 (bounded)
Alternative weighting	≤0.747	≤1.6	0.147–26 (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 soy, powder or ground on a as sold, as eaten or fresh weight basis: monitoring results summarized as one median per reporting region, sampling type and collection year, total diet composites and printed sample values as single observations, and printed subgroup medians from ordinary farms or markets. Each sampling family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations. Other forms, other bases, targeted sampling and results from mining, industrial, wastewater-affected, experimental, fertiliser or manufacturer settings are kept out of this reference.

Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1 to Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

WHO GEMS/Food is the only compatible sampling family for soy, powder or ground on this basis in the completed soy ledger, so this single-family reference cannot represent other markets. Reporting regions and purchase markets are not growing areas. Non-detects with differing limits widen the bounds. A position is descriptive concentration, not a safety or compliance judgment.

Admitted source families

WHO GEMS/Food

Reference version and interpretation

Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.

hmi-pooled-reference-v2:soy-27069bdca31c:f2db164163c8

The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.

HMI soy Ni reference · Soy, as sold

Reference and scope

HMI soy Ni reference · Soy, as sold. Soy, as sold; As sold, as eaten or fresh weight, µg/kg; no moisture conversion. Distribution of subgroup median concentration values.

Reference (µg/kg)	Q1	Median	Q3
Primary HMI reference	283	307	330
Alternative weighting	283	307	330

2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.

Contributing family	Reference units	Weight
Marques et al. 2021	2	100%

Methods

HMI pools subgroup medians for soy, as sold on a as sold, as eaten or fresh weight basis: monitoring results summarized as one median per reporting region, sampling type and collection year, total diet composites and printed sample values as single observations, and printed subgroup medians from ordinary farms or markets. Each sampling family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations. Other forms, other bases, targeted sampling and results from mining, industrial, wastewater-affected, experimental, fertiliser or manufacturer settings are kept out of this reference.

Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1 to Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

Marques et al. 2021 is the only compatible sampling family for soy, as sold on this basis in the completed soy ledger, so this single-family reference cannot represent other markets. Reporting regions and purchase markets are not growing areas. Non-detects with differing limits widen the bounds. A position is descriptive concentration, not a safety or compliance judgment.

Admitted source families

Marques et al. 2021

Reference version and interpretation

Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.

hmi-pooled-reference-v2:soy-43a428224377:6365ccfaa4fd

The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.

HMI soy Cr reference · Soy, dried

Reference and scope

HMI soy Cr reference · Soy, dried. Soy, dried; As sold, as eaten or fresh weight, µg/kg; no moisture conversion. Distribution of subgroup median concentration values.

Reference (µg/kg)	Q1	Median	Q3
Primary HMI reference	186	186	186
Alternative weighting	186	186	186

1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.

Contributing family	Reference units	Weight
Yu et al. 2023	1	100%

Methods

HMI pools subgroup medians for soy, dried on a as sold, as eaten or fresh weight basis: monitoring results summarized as one median per reporting region, sampling type and collection year, total diet composites and printed sample values as single observations, and printed subgroup medians from ordinary farms or markets. Each sampling family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations. Other forms, other bases, targeted sampling and results from mining, industrial, wastewater-affected, experimental, fertiliser or manufacturer settings are kept out of this reference.

Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1 to Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

Yu et al. 2023 is the only compatible sampling family for soy, dried on this basis in the completed soy ledger, so this single-family reference cannot represent other markets. Reporting regions and purchase markets are not growing areas. Non-detects with differing limits widen the bounds. A position is descriptive concentration, not a safety or compliance judgment.

Admitted source families

Yu et al. 2023

Reference version and interpretation

Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.

hmi-pooled-reference-v2:soy-a25737be9985:4e5878108cb6

The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.

HMI soy Al reference · Soy, as sold

Reference and scope

HMI soy Al reference · Soy, as sold. Soy, as sold; As sold, as eaten or fresh weight, µg/kg; no moisture conversion. Distribution of subgroup median concentration values.

Reference (µg/kg)	Q1	Median	Q3
Primary HMI reference	407–525 (bounded)	931	1900
Alternative weighting	407–525 (bounded)	931	1900

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 soy, as sold on a as sold, as eaten or fresh weight basis: monitoring results summarized as one median per reporting region, sampling type and collection year, total diet composites and printed sample values as single observations, and printed subgroup medians from ordinary farms or markets. Each sampling family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations. Other forms, other bases, targeted sampling and results from mining, industrial, wastewater-affected, experimental, fertiliser or manufacturer settings are kept out of this reference.

Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1 to Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

WHO GEMS/Food is the only compatible sampling family for soy, as sold on this basis in the completed soy ledger, so this single-family reference cannot represent other markets. Reporting regions and purchase markets are not growing areas. Non-detects with differing limits widen the bounds. A position is descriptive concentration, not a safety or compliance judgment.

Admitted source families

WHO GEMS/Food

Reference version and interpretation

Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.

hmi-pooled-reference-v2:soy-e825bb66395a:50b4f84fd848

The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.

HMI soy Sn reference · Soy, as sold

Reference and scope

HMI soy Sn reference · Soy, as sold. Soy, as sold; As sold, as eaten or fresh weight, µg/kg; no moisture conversion. Distribution of subgroup median concentration values.

Reference (µg/kg)	Q1	Median	Q3
Primary HMI reference	≤10	47.4	100–130 (bounded)
Alternative weighting	≤10	47.4	100–130 (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 soy, as sold on a as sold, as eaten or fresh weight basis: monitoring results summarized as one median per reporting region, sampling type and collection year, total diet composites and printed sample values as single observations, and printed subgroup medians from ordinary farms or markets. Each sampling family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations. Other forms, other bases, targeted sampling and results from mining, industrial, wastewater-affected, experimental, fertiliser or manufacturer settings are kept out of this reference.

Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1 to Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

WHO GEMS/Food is the only compatible sampling family for soy, as sold on this basis in the completed soy ledger, so this single-family reference cannot represent other markets. Reporting regions and purchase markets are not growing areas. Non-detects with differing limits widen the bounds. A position is descriptive concentration, not a safety or compliance judgment.

Admitted source families

WHO GEMS/Food

Reference version and interpretation

Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.

hmi-pooled-reference-v2:soy-c179b537864f:8a266af3b956

The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.

HMI soy U reference · Soy, as sold

Reference and scope

HMI soy U reference · Soy, as sold. Soy, as sold; As sold, as eaten or fresh weight, µg/kg; no moisture conversion. Distribution of subgroup median concentration values.

Reference (µg/kg)	Q1	Median	Q3
Primary HMI reference	≤25	≤25	≤25
Alternative weighting	≤25	≤25	≤25

2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.

Contributing family	Reference units	Weight
Marques et al. 2021	2	100%

Methods

HMI pools subgroup medians for soy, as sold on a as sold, as eaten or fresh weight basis: monitoring results summarized as one median per reporting region, sampling type and collection year, total diet composites and printed sample values as single observations, and printed subgroup medians from ordinary farms or markets. Each sampling family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations. Other forms, other bases, targeted sampling and results from mining, industrial, wastewater-affected, experimental, fertiliser or manufacturer settings are kept out of this reference.

Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1 to Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

Marques et al. 2021 is the only compatible sampling family for soy, as sold on this basis in the completed soy ledger, so this single-family reference cannot represent other markets. Reporting regions and purchase markets are not growing areas. Non-detects with differing limits widen the bounds. A position is descriptive concentration, not a safety or compliance judgment.

Admitted source families

Marques et al. 2021

Reference version and interpretation

Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.

hmi-pooled-reference-v2:soy-0386d6631753:720e06b4da4f

The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.

HMI soy Co reference · Soy, as sold

Reference and scope

HMI soy Co reference · Soy, as sold. Soy, as sold; As sold, as eaten or fresh weight, µg/kg; no moisture conversion. Distribution of subgroup median concentration values.

Reference (µg/kg)	Q1	Median	Q3
Primary HMI reference	≤12	6–18.5 (bounded)	12–25 (bounded)
Alternative weighting	≤12	6–18.5 (bounded)	12–25 (bounded)

3 reference units; 2 sampling families. Counts describe evidence units, not independent studies or certainty.

Contributing family	Reference units	Weight
Marques et al. 2021	2	50%
Ventura et al. 2025	1	50%

Methods

HMI pools subgroup medians for soy, as sold on a as sold, as eaten or fresh weight basis: monitoring results summarized as one median per reporting region, sampling type and collection year, total diet composites and printed sample values as single observations, and printed subgroup medians from ordinary farms or markets. Each sampling family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations. Other forms, other bases, targeted sampling and results from mining, industrial, wastewater-affected, experimental, fertiliser or manufacturer settings are kept out of this reference.

Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1 to Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

2 sampling families with different designs; none is a probability sample of world supply. Reporting regions and purchase markets are not growing areas. Non-detects with differing limits widen the bounds. A position is descriptive concentration, not a safety or compliance judgment.

Admitted source families

Reference version and interpretation

Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.

hmi-pooled-reference-v2:soy-473c6062babb:1b9e7a735ccb

The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.

HMI soy Mn reference · Soy, as sold

Reference and scope

HMI soy Mn reference · Soy, as sold. Soy, as sold; As sold, as eaten or fresh weight, µg/kg; no moisture conversion. Distribution of subgroup median concentration values.

Reference (µg/kg)	Q1	Median	Q3
Primary HMI reference	1440	1750	2050
Alternative weighting	1440	1750	2050

2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.

Contributing family	Reference units	Weight
Marques et al. 2021	2	100%

Methods

HMI pools subgroup medians for soy, as sold on a as sold, as eaten or fresh weight basis: monitoring results summarized as one median per reporting region, sampling type and collection year, total diet composites and printed sample values as single observations, and printed subgroup medians from ordinary farms or markets. Each sampling family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations. Other forms, other bases, targeted sampling and results from mining, industrial, wastewater-affected, experimental, fertiliser or manufacturer settings are kept out of this reference.

Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1 to Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

Marques et al. 2021 is the only compatible sampling family for soy, as sold on this basis in the completed soy ledger, so this single-family reference cannot represent other markets. Reporting regions and purchase markets are not growing areas. Non-detects with differing limits widen the bounds. A position is descriptive concentration, not a safety or compliance judgment.

Admitted source families

Marques et al. 2021

Reference version and interpretation

Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.

hmi-pooled-reference-v2:soy-7cffcbac58a3:7ad81b44ba4a

The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.

HMI soy V reference · Soy, as sold

Reference and scope

HMI soy V reference · Soy, as sold. Soy, as sold; As sold, as eaten or fresh weight, µg/kg; no moisture conversion. Distribution of subgroup median concentration values.

Reference (µg/kg)	Q1	Median	Q3
Primary HMI reference	≤50	≤50	≤50
Alternative weighting	≤50	≤50	≤50

2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.

Contributing family	Reference units	Weight
Marques et al. 2021	2	100%

Methods

HMI pools subgroup medians for soy, as sold on a as sold, as eaten or fresh weight basis: monitoring results summarized as one median per reporting region, sampling type and collection year, total diet composites and printed sample values as single observations, and printed subgroup medians from ordinary farms or markets. Each sampling family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations. Other forms, other bases, targeted sampling and results from mining, industrial, wastewater-affected, experimental, fertiliser or manufacturer settings are kept out of this reference.

Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1 to Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

Marques et al. 2021 is the only compatible sampling family for soy, as sold on this basis in the completed soy ledger, so this single-family reference cannot represent other markets. Reporting regions and purchase markets are not growing areas. Non-detects with differing limits widen the bounds. A position is descriptive concentration, not a safety or compliance judgment.

Admitted source families

Marques et al. 2021

Reference version and interpretation

Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.

hmi-pooled-reference-v2:soy-c82de1da2573:1ebf0ab8b18f

The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.

HMI soy Se reference · Soy, as sold

Reference and scope

HMI soy Se reference · Soy, as sold. Soy, as sold; As sold, as eaten or fresh weight, µg/kg; no moisture conversion. Distribution of subgroup median concentration values.

Reference (µg/kg)	Q1	Median	Q3
Primary HMI reference	14	14	14
Alternative weighting	14	14	14

1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.

Contributing family	Reference units	Weight
Ventura et al. 2025	1	100%

Methods

HMI pools subgroup medians for soy, as sold on a as sold, as eaten or fresh weight basis: monitoring results summarized as one median per reporting region, sampling type and collection year, total diet composites and printed sample values as single observations, and printed subgroup medians from ordinary farms or markets. Each sampling family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations. Other forms, other bases, targeted sampling and results from mining, industrial, wastewater-affected, experimental, fertiliser or manufacturer settings are kept out of this reference.

Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1 to Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

Ventura et al. 2025 is the only compatible sampling family for soy, as sold on this basis in the completed soy ledger, so this single-family reference cannot represent other markets. Reporting regions and purchase markets are not growing areas. Non-detects with differing limits widen the bounds. A position is descriptive concentration, not a safety or compliance judgment.

Admitted source families

Ventura et al. 2025

Reference version and interpretation

Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.

hmi-pooled-reference-v2:soy-d8bb629d836e:01e7abe6f847

The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.

HMI soy Hg reference · Soy, as sold

Reference and scope

HMI soy Hg reference · Soy, as sold. Soy, as sold; As sold, as eaten or fresh weight, µg/kg; no moisture conversion. Distribution of subgroup median concentration values.

Reference (µg/kg)	Q1	Median	Q3
Primary HMI reference	≤50	≤50	≤50
Alternative weighting	≤50	≤50	≤50

2 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.

Contributing family	Reference units	Weight
Marques et al. 2021	2	100%

Methods

HMI pools subgroup medians for soy, as sold on a as sold, as eaten or fresh weight basis: monitoring results summarized as one median per reporting region, sampling type and collection year, total diet composites and printed sample values as single observations, and printed subgroup medians from ordinary farms or markets. Each sampling family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations. Other forms, other bases, targeted sampling and results from mining, industrial, wastewater-affected, experimental, fertiliser or manufacturer settings are kept out of this reference.

Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1 to Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

Marques et al. 2021 is the only compatible sampling family for soy, as sold on this basis in the completed soy ledger, so this single-family reference cannot represent other markets. Reporting regions and purchase markets are not growing areas. Non-detects with differing limits widen the bounds. A position is descriptive concentration, not a safety or compliance judgment.

Admitted source families

Marques et al. 2021

Reference version and interpretation

Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.

hmi-pooled-reference-v2:soy-f0caecdd198b:1c666d1910c3

The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.

HMI soy I reference · Soy, as sold

Reference and scope

HMI soy I reference · Soy, as sold. Soy, as sold; As sold, as eaten or fresh weight, µg/kg; no moisture conversion. Distribution of subgroup median concentration values.

Reference (µg/kg)	Q1	Median	Q3
Primary HMI reference	5.9	5.9	5.9
Alternative weighting	5.9	5.9	5.9

1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.

Contributing family	Reference units	Weight
Ventura et al. 2025	1	100%

Methods

HMI pools subgroup medians for soy, as sold on a as sold, as eaten or fresh weight basis: monitoring results summarized as one median per reporting region, sampling type and collection year, total diet composites and printed sample values as single observations, and printed subgroup medians from ordinary farms or markets. Each sampling family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations. Other forms, other bases, targeted sampling and results from mining, industrial, wastewater-affected, experimental, fertiliser or manufacturer settings are kept out of this reference.

Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1 to Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

Ventura et al. 2025 is the only compatible sampling family for soy, as sold on this basis in the completed soy ledger, so this single-family reference cannot represent other markets. Reporting regions and purchase markets are not growing areas. Non-detects with differing limits widen the bounds. A position is descriptive concentration, not a safety or compliance judgment.

Admitted source families

Ventura et al. 2025

Reference version and interpretation

Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.

hmi-pooled-reference-v2:soy-38b4acc4cf6e:6dfb9838005b

The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.

HMI soy Mo reference · Soy, as sold

Reference and scope

HMI soy Mo reference · Soy, as sold. Soy, as sold; As sold, as eaten or fresh weight, µg/kg; no moisture conversion. Distribution of subgroup median concentration values.

Reference (µg/kg)	Q1	Median	Q3
Primary HMI reference	197	197	197
Alternative weighting	197	197	197

1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.

Contributing family	Reference units	Weight
Ventura et al. 2025	1	100%

Methods

HMI pools subgroup medians for soy, as sold on a as sold, as eaten or fresh weight basis: monitoring results summarized as one median per reporting region, sampling type and collection year, total diet composites and printed sample values as single observations, and printed subgroup medians from ordinary farms or markets. Each sampling family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations. Other forms, other bases, targeted sampling and results from mining, industrial, wastewater-affected, experimental, fertiliser or manufacturer settings are kept out of this reference.

Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1 to Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

Ventura et al. 2025 is the only compatible sampling family for soy, as sold on this basis in the completed soy ledger, so this single-family reference cannot represent other markets. Reporting regions and purchase markets are not growing areas. Non-detects with differing limits widen the bounds. A position is descriptive concentration, not a safety or compliance judgment.

Admitted source families

Ventura et al. 2025

Reference version and interpretation

Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.

hmi-pooled-reference-v2:soy-1be13756993d:c0e89d391bd5

The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.

HMI soy Sr reference · Soy, as sold

Reference and scope

HMI soy Sr reference · Soy, as sold. Soy, as sold; As sold, as eaten or fresh weight, µg/kg; no moisture conversion. Distribution of subgroup median concentration values.

Reference (µg/kg)	Q1	Median	Q3
Primary HMI reference	829	829	829
Alternative weighting	829	829	829

1 reference units; 1 sampling family. Counts describe evidence units, not independent studies or certainty.

Contributing family	Reference units	Weight
Ventura et al. 2025	1	100%

Methods

HMI pools subgroup medians for soy, as sold on a as sold, as eaten or fresh weight basis: monitoring results summarized as one median per reporting region, sampling type and collection year, total diet composites and printed sample values as single observations, and printed subgroup medians from ordinary farms or markets. Each sampling family has equal total weight. Non-detects stay as zero-to-limit intervals, so quartiles carry lower and upper bounds. Means, ranges and maxima never enter as observations. Other forms, other bases, targeted sampling and results from mining, industrial, wastewater-affected, experimental, fertiliser or manufacturer settings are kept out of this reference.

Inverse weighted empirical CDF; midpoint at exact cumulative-weight boundaries. At least four reference units and a definitely positive Q1 to Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

Ventura et al. 2025 is the only compatible sampling family for soy, as sold on this basis in the completed soy ledger, so this single-family reference cannot represent other markets. Reporting regions and purchase markets are not growing areas. Non-detects with differing limits widen the bounds. A position is descriptive concentration, not a safety or compliance judgment.

Admitted source families

Ventura et al. 2025

Reference version and interpretation

Frozen evidence selection: 2026-10-22. Method: hmi-pooled-reference-v2.

hmi-pooled-reference-v2:soy-1510e3c69bc1:aa7ebacb19f9

The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is not a worldwide or current-lot benchmark. Source, year and region filters do not change it. Different HMI scopes are not one common ranking.

How metals reach soy

The evidence on how metals reach soy comes from studies that compared exposed and ordinary settings. Soy, as sold from soybean cultivar Hefeng 40, pot soil spiked to 2.5 mg Cd/kg (T2) (experiment) had cadmium higher at 2,060 µg/kg, against 180 µg/kg for soybean cultivar Hefeng 40, control soil in the same study.^{7Zhi Y, Sun T, Zhou Q, Leng X · 2020Screening of safe soybean cultivars for cadmium contaminated fieldsScientific Reports 10:12965Read source →}

These are observational comparisons from single studies; the size of the difference depends on the site, metal and distance.

Which forms and regions measured lower

The HMI reference for lead in soy, as sold pools 128 subgroup units from 2 sampling families, with a median of below 50 µg/kg and a middle half spanning below 19 (lower quartile) and below 50 µg/kg (upper quartile).^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}^{5Marques M, Correig E, Capdevila E, Gargallo E, Gonzalez N, Nadal M, Domingo JL · 2021Essential and Non-essential Trace Elements in Milks and Plant-Based DrinksBiological Trace Element ResearchRead source →} For lead in soy, powder or ground, the reference median was below 20 µg/kg, with a middle half from below 5.25 to between 2.2 and 90 µg/kg across 10 units.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →} For lead in soy, as sold, the reference median was below 5 µg/kg, with a middle half from below 5 to 50 µg/kg across 5 units.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}

Among GEMS/Food reporting regions, soy, as sold had the lowest lead subgroup median in WPRO (4 µg/kg) and the highest in PAHO (50 µg/kg); reporting regions are markets, not growing areas.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →} For cadmium, the lowest regional median was in PAHO (1.17 µg/kg) and the highest in WPRO (130 µg/kg).^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}

Targeted soy, as sold samples in WPRO had cadmium lower at 2.05 µg/kg, against 20 µg/kg for WPRO in the same study.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →} In another study, targeted soy, as sold samples in WPRO had cadmium higher at 50 µg/kg, against 20 µg/kg for WPRO in the same study.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}

Processing, cooking and preparation

No processing study for metals in this food is in the library yet.

Mitigation for soy

Lot testing against the applicable limit is the check that can be verified.

No measured way to lower metals in this food is in the library yet.

Trade-offs and limits of controls

Avoiding exposed growing areas lowered the metal measured in the studies available, but no study measured the cost or supply effect of doing so.^{7Zhi Y, Sun T, Zhou Q, Leng X · 2020Screening of safe soybean cultivars for cadmium contaminated fieldsScientific Reports 10:12965Read source →} Metals do not move together, so a lot that passes a lead limit says little about cadmium, arsenic or nickel unless those are tested as well.

Testing soy for metals

Results depend on the method and its reporting limit. Reporting limits for lead ranged from 0.04 µg/kg (WHO GEMS/Food) to 25,000 µg/kg (WHO GEMS/Food), so a non-detect from a less sensitive method says much less.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →} Reporting limits for cadmium ranged from 0.4 µg/kg (WHO GEMS/Food) to 100 µg/kg (WHO GEMS/Food), so a non-detect from a less sensitive method says much less.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →} Reporting limits for total arsenic ranged from 2.5 µg/kg (WHO GEMS/Food) to 997 µg/kg (WHO GEMS/Food), so a non-detect from a less sensitive method says much less.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}

A useful lot result names the form, the basis and a quantification limit well below the limit that applies.

Buying and verifying soy

No maximum level for metals in soy was verified from primary text for this page; buyers should confirm the limit that applies in the destination market.

Exposure from soy

No source in the library converted soy concentrations into a verified estimate of intake for a defined population. Concentrations describe the food, not individual risk.

Evidence review and remaining gaps

Every source record that names soy was decided on its content: 32 occurrence extracted, 32 records in all. The occurrence table renders every extracted result; held values remain visible with their reason.

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.

Soy FAQs

What heavy metals are found in soy?

The library reports lead, cadmium, total arsenic, inorganic arsenic, total mercury, nickel, chromium, aluminium, tin, antimony, uranium, Co, Cu, Fe, Mn, Zn, V, Se, As, Hg, I, Li, Mo and Sr in soy, most often lead, cadmium and total arsenic.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}^{2Vanda Lopes Andrade, Iolanda Ribeiro, A. P. Marreilha dos Santos, Michael Aschner, Maria Luisa Mateus · 2023Metals in Cow Milk and Soy Beverages: Is There a Concern?ToxicsRead source →}^{3Food Standards Agency, Hallmark Meat Hygiene · 2014Survey of metals and other elements in commercial infant foods, infant formula and non-infant specific foodsFood Standards Agency reportRead source →}^{4Marín S, Pardo O, Sánchez A, Sanchis Y, Vélez D, Devesa V, Font G, Yusà V · 2018Assessment of metal levels in foodstuffs from the Region of Valencia (Spain)Toxicology Reports 5 (2018) 654-670Read source →} WHO GEMS/Food submissions reported lead in soy, as sold from 2,031 results in 5 reporting regions, with subgroup medians from 2.69 to 50 µg/kg.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}

How do heavy metals get into soy?

In the studies available, soy from contaminated or exposed settings carried more metal than soy from the same study’s control.^{7Zhi Y, Sun T, Zhou Q, Leng X · 2020Screening of safe soybean cultivars for cadmium contaminated fieldsScientific Reports 10:12965Read source →} Soy, as sold from all 25 cultivars, field soil 1.22 mg Cd/kg (experiment) had cadmium higher at 470 µg/kg, against 180 µg/kg for soybean cultivar Hefeng 40, control soil in the same study.^{7Zhi Y, Sun T, Zhou Q, Leng X · 2020Screening of safe soybean cultivars for cadmium contaminated fieldsScientific Reports 10:12965Read source →}

Which regions report the lowest levels in soy?

Among GEMS/Food reporting regions, soy, as sold had the lowest lead subgroup median in WPRO (4 µg/kg) and the highest in PAHO (50 µg/kg); reporting regions are markets, not growing areas.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →} For cadmium, the lowest regional median was in PAHO (1.17 µg/kg) and the highest in WPRO (130 µg/kg).^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}

How much of the arsenic in soy is inorganic?

Inorganic arsenic, the more toxic form, was measured in 16 results from 5 sources, far fewer than total arsenic.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}^{3Food Standards Agency, Hallmark Meat Hygiene · 2014Survey of metals and other elements in commercial infant foods, infant formula and non-infant specific foodsFood Standards Agency reportRead source →}^{4Marín S, Pardo O, Sánchez A, Sanchis Y, Vélez D, Devesa V, Font G, Yusà V · 2018Assessment of metal levels in foodstuffs from the Region of Valencia (Spain)Toxicology Reports 5 (2018) 654-670Read source →}^{12Centre for Food Safety FEHD HKSAR · 2012The 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 →}^{13European Food Safety Authority (EFSA) · 2014Dietary exposure to inorganic arsenic in the European populationEFSA Journal 2014;12(3):3597Read source →} Across 2 other studies of ordinary farms and markets, reported inorganic arsenic means and medians on an as-sold or fresh-weight basis ranged from 3.3 µg/kg (Marín et al. 2018) to 4.6 µg/kg (Food Standards Agency 2014).^{3Food Standards Agency, Hallmark Meat Hygiene · 2014Survey of metals and other elements in commercial infant foods, infant formula and non-infant specific foodsFood Standards Agency reportRead source →}^{4Marín S, Pardo O, Sánchez A, Sanchis Y, Vélez D, Devesa V, Font G, Yusà V · 2018Assessment of metal levels in foodstuffs from the Region of Valencia (Spain)Toxicology Reports 5 (2018) 654-670Read 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 exportsWorld Health Organization · GEMS/Food Contamination Monitoring and Assessment Programme · 2026 · extranet.who.intDataset
Metals in Cow Milk and Soy Beverages: Is There a Concern?Vanda Lopes Andrade, Iolanda Ribeiro, A. P. Marreilha dos Santos, Michael Aschner, and Maria Luisa Mateus · Toxics · 2023 · doi.org/10.3390/toxics11121013Peer-reviewed
Survey of metals and other elements in commercial infant foods, infant formula and non-infant specific foodsFood Standards Agency and Hallmark Meat Hygiene · Food Standards Agency report · 2014Government
Assessment of metal levels in foodstuffs from the Region of Valencia (Spain)Marín S, Pardo O, Sánchez A, Sanchis Y, Vélez D, Devesa V, et al. · Toxicology Reports 5 (2018) 654-670 · 2018 · doi.org/10.1016/j.toxrep.2018.05.005Peer-reviewed
Essential and Non-essential Trace Elements in Milks and Plant-Based DrinksMarques M, Correig E, Capdevila E, Gargallo E, Gonzalez N, Nadal M, et al. · Biological Trace Element Research · 2021 · doi.org/10.1007/s12011-021-03021-5Peer-reviewed
Trace Elements in Soy-Based Beverages: A Comprehensive Study of Total Content and In Vitro BioaccessibilityMilani RF, Mauri AA, Sanches VL, Morgano MA, and Cadore S · International Journal of Environmental Research and Public Health · 2023 · doi.org/10.3390/ijerph20064986Peer-reviewed
Screening of safe soybean cultivars for cadmium contaminated fieldsZhi Y, Sun T, Zhou Q, and Leng X · Scientific Reports 10:12965 · 2020 · doi.org/10.1038/s41598-020-69803-4Peer-reviewed
Health risk assessment of eight heavy metals in nine varieties of edible vegetable oils consumed in ChinaZhu F, Fan W, Wang X, Qu L, and Yao S · Food and Chemical Toxicology · 2011 · doi.org/10.1016/j.fct.2011.09.019Peer-reviewed
Scientific Opinion of the Panel on Contaminants in the Food Chain on a request from the European Commission on cadmium in foodEFSA Panel on Contaminants in the Food Chain (CONTAM) · The EFSA Journal · 2009 · doi.org/10.2903/j.efsa.2009.980Government
Risk assessment of heavy metals concentration in cereals and legumes sold in the Tamale Aboabo market, GhanaAdam AA, Sackey LNA, and Ofori LA · Heliyon 8 (2022) e10162 · 2022 · doi.org/10.1016/j.heliyon.2022.e10162Peer-reviewed
Toxic Elements in Beans from Zhejiang, Southeast China: Distribution and Probabilistic Health Risk AssessmentYu S, Pan XD, and Han JL · Foods · 2023 · doi.org/10.3390/foods12173300Peer-reviewed
The First Hong Kong Total Diet Study: Inorganic ArsenicCentre for Food Safety FEHD HKSAR · Centre for Food Safety, Food and Environmental Hygiene Department, Government of the Hong Kong Special Administrative Region · 2012 · www.cfs.gov.hkGovernment
Dietary exposure to inorganic arsenic in the European populationEuropean Food Safety Authority (EFSA) · EFSA Journal 2014;12(3):3597 · 2014 · doi.org/10.2903/j.efsa.2014.3597Government

Sources

Browse Soy 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 soy record in the library: 32 occurrence sources, references by form and basis, same-study contrasts.

1

2

3

4

5

6

7

8

9

10

11

12

13

European Food Safety Authority

Heavy metals in Soy

Soy overview

What counts as soy

Heavy metal contamination profile of soy

How metals reach soy

Which forms and regions measured lower

Processing, cooking and preparation

Mitigation for soy

Trade-offs and limits of controls

Testing soy for metals

Buying and verifying soy

Exposure from soy

Evidence review and remaining gaps

Soy FAQs

References

Sources

Update history

Referenced by 101

Graph View

Heavy metals in Soy

Soy overview

What counts as soy

Heavy metal contamination profile of soy

How metals reach soy

Which forms and regions measured lower

Processing, cooking and preparation

Mitigation for soy

Trade-offs and limits of controls

Testing soy for metals

Buying and verifying soy

Exposure from soy

Evidence review and remaining gaps

Soy FAQs

References

Sources

Update history

Related ingredients

Referenced by 101

Graph View