Heavy metals in Cassava

Quick read

Lead is the metal to check in cassava: 18 of 50 European root results exceeded the 100 µg/kg limit, while roots from the Americas were far lower; cadmium, arsenic and mercury were mostly low.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}^{2European Commission · 2023Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →}

Metals of concern:
Lead decides most cassava questions; mercury, lead and cadmium are highest in cassava from gold-mining districts.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}^{3A. Adjorlolo-Gasokpoh, A. A. Golow, J. Kambo-Dorsa · 2012Mercury in the Surface Soil and Cassava, Manihot esculenta (Flesh, Leaves and Peel) Near Goldmines at Bogoso and Prestea, GhanaBull Environ Contam ToxicolRead source →}^{4Adjei-Mensah R, Ofori H, Tortoe C, Johnson PT, Aryee D, Frimpong SK · 2021Effect of home processing methods on the levels of heavy metal contaminants in four food crops grown in and around two mining towns in GhanaToxicology Reports 8 (2021) 1830–1838Read source →}^{5Romero-Crespo P, Jiménez-Oyola S, Salgado-Almeida B, Zambrano-Anchundia J, Goyburo-Chávez C, González-Valoys A, Higueras P · 2023Trace elements in farmland soils and crops, and probabilistic health risk assessment in areas influenced by mining activity in EcuadorEnvironmental Geochemistry and HealthRead source →}
Main drivers:
Growing soil and nearby mining, the market sampled, the form of cassava and how it is cooked.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}^{4Adjei-Mensah R, Ofori H, Tortoe C, Johnson PT, Aryee D, Frimpong SK · 2021Effect of home processing methods on the levels of heavy metal contaminants in four food crops grown in and around two mining towns in GhanaToxicology Reports 8 (2021) 1830–1838Read source →}^{5Romero-Crespo P, Jiménez-Oyola S, Salgado-Almeida B, Zambrano-Anchundia J, Goyburo-Chávez C, González-Valoys A, Higueras P · 2023Trace elements in farmland soils and crops, and probabilistic health risk assessment in areas influenced by mining activity in EcuadorEnvironmental Geochemistry and HealthRead source →}^{6Cecilia Owusu-Agyemang Fobi, Guy Eshun, Twumasi Ankrah Kwarteng, Afia Sakyiwaa Amponsah, Prince Owusu Adoma · 2025Effect of Boiling on Heavy Metal and Cyanide Concentrations and Associated Health Risks in Cassava and Cocoyam From a Gold Mining Area in Kade, GhanaInternational Journal of Food ScienceRead source →}
Practical framing:
Ask for lot-specific lead results on the form bought, with a quantification limit well below 100 µg/kg.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}^{2European Commission · 2023Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →}

Cassava overview

Lead is the metal that decides most cassava questions. In WHO GEMS/Food, European cassava root results had a median of 46 µg/kg in random sampling and 92 µg/kg where sampling was not described, and 18 of 50 results exceeded 100 µg/kg, the EU and Codex maximum for root and tuber vegetables.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}^{2European Commission · 2023Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →}^{7Codex Alimentarius Commission · 1995General Standard for Contaminants and Toxins in Food and Feed (CXS 193-1995)Codex Alimentarius (Joint FAO/WHO Food Standards Programme)Read source →} Results submitted from the Americas were lower, with a median of 9 µg/kg and a maximum of 82 µg/kg across 26 roots.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}

Cadmium, arsenic and mercury were mostly low or below detection in monitored cassava.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →} No European cassava root result exceeded the EU cadmium limit of 50 µg/kg for tropical roots and tubers, and most total arsenic and mercury results were non-detects.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}^{2European Commission · 2023Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →} The exceptions are gold-mining areas: in Ghana, cassava root carried 120–430 µg/kg mercury, and peeled roots from two mining towns averaged 2,110 µg/kg lead, while a root from an Ecuadorian mining orchard had 60 µg/kg cadmium.^{3A. Adjorlolo-Gasokpoh, A. A. Golow, J. Kambo-Dorsa · 2012Mercury in the Surface Soil and Cassava, Manihot esculenta (Flesh, Leaves and Peel) Near Goldmines at Bogoso and Prestea, GhanaBull Environ Contam ToxicolRead source →}^{4Adjei-Mensah R, Ofori H, Tortoe C, Johnson PT, Aryee D, Frimpong SK · 2021Effect of home processing methods on the levels of heavy metal contaminants in four food crops grown in and around two mining towns in GhanaToxicology Reports 8 (2021) 1830–1838Read source →}^{5Romero-Crespo P, Jiménez-Oyola S, Salgado-Almeida B, Zambrano-Anchundia J, Goyburo-Chávez C, González-Valoys A, Higueras P · 2023Trace elements in farmland soils and crops, and probabilistic health risk assessment in areas influenced by mining activity in EcuadorEnvironmental Geochemistry and HealthRead source →}^{6Cecilia Owusu-Agyemang Fobi, Guy Eshun, Twumasi Ankrah Kwarteng, Afia Sakyiwaa Amponsah, Prince Owusu Adoma · 2025Effect of Boiling on Heavy Metal and Cyanide Concentrations and Associated Health Risks in Cassava and Cocoyam From a Gold Mining Area in Kade, GhanaInternational Journal of Food ScienceRead source →}

The evidence is thin outside monitoring data. Twelve sources measured cassava itself, mostly small studies from Nigerian markets and Ghanaian mining districts, and several report values their own methods cannot support; those values are shown but not compared.^{3A. Adjorlolo-Gasokpoh, A. A. Golow, J. Kambo-Dorsa · 2012Mercury in the Surface Soil and Cassava, Manihot esculenta (Flesh, Leaves and Peel) Near Goldmines at Bogoso and Prestea, GhanaBull Environ Contam ToxicolRead source →}^{4Adjei-Mensah R, Ofori H, Tortoe C, Johnson PT, Aryee D, Frimpong SK · 2021Effect of home processing methods on the levels of heavy metal contaminants in four food crops grown in and around two mining towns in GhanaToxicology Reports 8 (2021) 1830–1838Read source →}^{5Romero-Crespo P, Jiménez-Oyola S, Salgado-Almeida B, Zambrano-Anchundia J, Goyburo-Chávez C, González-Valoys A, Higueras P · 2023Trace elements in farmland soils and crops, and probabilistic health risk assessment in areas influenced by mining activity in EcuadorEnvironmental Geochemistry and HealthRead source →}^{6Cecilia Owusu-Agyemang Fobi, Guy Eshun, Twumasi Ankrah Kwarteng, Afia Sakyiwaa Amponsah, Prince Owusu Adoma · 2025Effect of Boiling on Heavy Metal and Cyanide Concentrations and Associated Health Risks in Cassava and Cocoyam From a Gold Mining Area in Kade, GhanaInternational Journal of Food ScienceRead source →}^{8AuthorS: Kingsley O. Omeje in common food tubers from Nigeria Benjamin O. Ezema, Sabinus O.O. Eze AFFILIAtIoNS: Pesticide residues · 2023Evaluation of pesticide residues and heavy metalsPeer-reviewed articleRead source →}^{9Albishi TS, Adaigho D, Alqahtani NK, Sami R, Uguru H, Helal M, Alaoufi SH, Alkaltham MS, Osiga-Aibangbee D, Okolotu GI, Emumejaye EP, Qari SH · 2026In vitro evaluation of bio-fortification effects on the nutritional quality, toxicological safety, and antioxidant of cassava (Manihot esculenta) flour, and environmental safety of processing water, using natural additivesFrontiers in NutritionRead source →}^{10EDITED BY Arun K. Bhunia, Purdue University, United States REVIEWED BY Marco E. Mng’ong’o, sold in selected markets—taken Mbeya University of Science, Technology · 2023OPEN ACCESS potential toxic levels of cyanidePeer-reviewed articleRead source →}^{11Iyabo OO, Ogundiran MB, Adebayo OL · 2015Toxic and Essential Metals in Staple Foods Commonly Consumed by Students in Ekiti State, South West, NigeriaInternational Journal of ChemistryRead source →}^{12Bortey-Sam N, Nakayama SMM, Akoto O, Ikenaka Y, Fobil JN, Baidoo E, Mizukawa H, Ishizuka M · 2015Accumulation of Heavy Metals and Metalloid in Foodstuffs from Agricultural Soils around Tarkwa Area in Ghana, and Associated Human Health RisksInternational Journal of Environmental Research and Public Health (2015)Read source →}^{13Christian Ebere Enyoh, Qingyue Wang, Mominul Haque Rabin, Rasheed Oluwafemi Bakare, Joseph Longji Dadiel, Wu Shangrong, Senlin Lu, Ifenna Ilechukwu · 2023Preliminary characterization and probabilistic risk assessment of microplastics and potentially toxic elements (PTEs) in garri (cassava flake), a common staple food consumed in West AfricaEnvironmental Analysis Health and ToxicologyRead source →}^{14Orisakwe OE, Nduka JK, Amadi CN, Dike DO, Bede O · 2012Heavy metals health risk assessment for population via consumption of food crops and fruits in Owerri, South Eastern, NigeriaChemistry Central Journal 2012, 6:77Read source →} The supported buyer control is a lot-specific lead result on the form being bought, measured with a method that can see well below 100 µg/kg.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}^{2European Commission · 2023Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →}

What counts as cassava

This page covers cassava (Manihot esculenta) root and the foods made directly from it: raw, boiled, fried or roasted root and its peel, cassava flour including fermented lafun and amala flour, gari, fufu, and cassava starch sold as tapioca.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}^{3A. Adjorlolo-Gasokpoh, A. A. Golow, J. Kambo-Dorsa · 2012Mercury in the Surface Soil and Cassava, Manihot esculenta (Flesh, Leaves and Peel) Near Goldmines at Bogoso and Prestea, GhanaBull Environ Contam ToxicolRead source →}^{4Adjei-Mensah R, Ofori H, Tortoe C, Johnson PT, Aryee D, Frimpong SK · 2021Effect of home processing methods on the levels of heavy metal contaminants in four food crops grown in and around two mining towns in GhanaToxicology Reports 8 (2021) 1830–1838Read source →}^{10EDITED BY Arun K. Bhunia, Purdue University, United States REVIEWED BY Marco E. Mng’ong’o, sold in selected markets—taken Mbeya University of Science, Technology · 2023OPEN ACCESS potential toxic levels of cyanidePeer-reviewed articleRead source →}^{11Iyabo OO, Ogundiran MB, Adebayo OL · 2015Toxic and Essential Metals in Staple Foods Commonly Consumed by Students in Ekiti State, South West, NigeriaInternational Journal of ChemistryRead source →}^{13Christian Ebere Enyoh, Qingyue Wang, Mominul Haque Rabin, Rasheed Oluwafemi Bakare, Joseph Longji Dadiel, Wu Shangrong, Senlin Lu, Ifenna Ilechukwu · 2023Preliminary characterization and probabilistic risk assessment of microplastics and potentially toxic elements (PTEs) in garri (cassava flake), a common staple food consumed in West AfricaEnvironmental Analysis Health and ToxicologyRead source →} Each form is kept separate because drying, fermentation, roasting and starch extraction change both the food and the basis on which a result is reported.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}^{9Albishi TS, Adaigho D, Alqahtani NK, Sami R, Uguru H, Helal M, Alaoufi SH, Alkaltham MS, Osiga-Aibangbee D, Okolotu GI, Emumejaye EP, Qari SH · 2026In vitro evaluation of bio-fortification effects on the nutritional quality, toxicological safety, and antioxidant of cassava (Manihot esculenta) flour, and environmental safety of processing water, using natural additivesFrontiers in NutritionRead source →}

Cassava leaves, eaten as a vegetable in parts of Africa, are a different food and are not covered here; nor are animal feeds, snacks or infant foods that only contain cassava or tapioca.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}^{15Charles A. Barber, Marc Boyer, Jake A. Carter, Patrick J. Gray, Jennifer Fong Sam, Mesay M. Wolle · 2025Toxic elements in baby and young children’s foods in the US and correlation to ingredientsFood Additives & Contaminants: Part BRead source →} Cyanide from cassava’s cyanogenic glycosides is a natural toxicant rather than a heavy metal, so it is outside this page even though the EU sets limits of 50 mg/kg for fresh peeled root and 10 mg/kg for cassava and tapioca flour.^{2European Commission · 2023Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →}

Names need care. GEMS/Food includes results labelled only “Tapioca” or “tapioca, see cassava”, which may describe the root or the starch, so this page reports them as tapioca of unstated form.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →} In Nigerian studies, amala and fufu are named dishes whose raw material and moisture state are not always stated.^{11Iyabo OO, Ogundiran MB, Adebayo OL · 2015Toxic and Essential Metals in Staple Foods Commonly Consumed by Students in Ekiti State, South West, NigeriaInternational Journal of ChemistryRead source →}

Heavy metal contamination profile of cassava

The table holds every cassava result extracted from WHO GEMS/Food and eleven field, market, mining-area and processing studies: one row for each statistic a source reports for a form of cassava, group and metal.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}^{3A. Adjorlolo-Gasokpoh, A. A. Golow, J. Kambo-Dorsa · 2012Mercury in the Surface Soil and Cassava, Manihot esculenta (Flesh, Leaves and Peel) Near Goldmines at Bogoso and Prestea, GhanaBull Environ Contam ToxicolRead source →}^{4Adjei-Mensah R, Ofori H, Tortoe C, Johnson PT, Aryee D, Frimpong SK · 2021Effect of home processing methods on the levels of heavy metal contaminants in four food crops grown in and around two mining towns in GhanaToxicology Reports 8 (2021) 1830–1838Read source →}^{5Romero-Crespo P, Jiménez-Oyola S, Salgado-Almeida B, Zambrano-Anchundia J, Goyburo-Chávez C, González-Valoys A, Higueras P · 2023Trace elements in farmland soils and crops, and probabilistic health risk assessment in areas influenced by mining activity in EcuadorEnvironmental Geochemistry and HealthRead source →}^{6Cecilia Owusu-Agyemang Fobi, Guy Eshun, Twumasi Ankrah Kwarteng, Afia Sakyiwaa Amponsah, Prince Owusu Adoma · 2025Effect of Boiling on Heavy Metal and Cyanide Concentrations and Associated Health Risks in Cassava and Cocoyam From a Gold Mining Area in Kade, GhanaInternational Journal of Food ScienceRead source →}^{8AuthorS: Kingsley O. Omeje in common food tubers from Nigeria Benjamin O. Ezema, Sabinus O.O. Eze AFFILIAtIoNS: Pesticide residues · 2023Evaluation of pesticide residues and heavy metalsPeer-reviewed articleRead source →}^{9Albishi TS, Adaigho D, Alqahtani NK, Sami R, Uguru H, Helal M, Alaoufi SH, Alkaltham MS, Osiga-Aibangbee D, Okolotu GI, Emumejaye EP, Qari SH · 2026In vitro evaluation of bio-fortification effects on the nutritional quality, toxicological safety, and antioxidant of cassava (Manihot esculenta) flour, and environmental safety of processing water, using natural additivesFrontiers in NutritionRead source →}^{10EDITED BY Arun K. Bhunia, Purdue University, United States REVIEWED BY Marco E. Mng’ong’o, sold in selected markets—taken Mbeya University of Science, Technology · 2023OPEN ACCESS potential toxic levels of cyanidePeer-reviewed articleRead source →}^{11Iyabo OO, Ogundiran MB, Adebayo OL · 2015Toxic and Essential Metals in Staple Foods Commonly Consumed by Students in Ekiti State, South West, NigeriaInternational Journal of ChemistryRead source →}^{12Bortey-Sam N, Nakayama SMM, Akoto O, Ikenaka Y, Fobil JN, Baidoo E, Mizukawa H, Ishizuka M · 2015Accumulation of Heavy Metals and Metalloid in Foodstuffs from Agricultural Soils around Tarkwa Area in Ghana, and Associated Human Health RisksInternational Journal of Environmental Research and Public Health (2015)Read source →}^{13Christian Ebere Enyoh, Qingyue Wang, Mominul Haque Rabin, Rasheed Oluwafemi Bakare, Joseph Longji Dadiel, Wu Shangrong, Senlin Lu, Ifenna Ilechukwu · 2023Preliminary characterization and probabilistic risk assessment of microplastics and potentially toxic elements (PTEs) in garri (cassava flake), a common staple food consumed in West AfricaEnvironmental Analysis Health and ToxicologyRead source →}^{14Orisakwe OE, Nduka JK, Amadi CN, Dike DO, Bede O · 2012Heavy metals health risk assessment for population via consumption of food crops and fruits in Owerri, South Eastern, NigeriaChemistry Central Journal 2012, 6:77Read source →} GEMS/Food rows are HMI-calculated medians of individual national submissions by WHO region and sampling type; 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 →} Ratings place a result against an HMI reference for the same form and basis, built from GEMS/Food region-by-year medians and, for flour, the five market values in a Nigerian lafun survey.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}^{10EDITED BY Arun K. Bhunia, Purdue University, United States REVIEWED BY Marco E. Mng’ong’o, sold in selected markets—taken Mbeya University of Science, Technology · 2023OPEN ACCESS potential toxic levels of cyanidePeer-reviewed articleRead source →}

Dry-weight laboratory results and results with no stated basis are never pooled with as-sold monitoring results.^{8AuthorS: Kingsley O. Omeje in common food tubers from Nigeria Benjamin O. Ezema, Sabinus O.O. Eze AFFILIAtIoNS: Pesticide residues · 2023Evaluation of pesticide residues and heavy metalsPeer-reviewed articleRead source →}^{9Albishi TS, Adaigho D, Alqahtani NK, Sami R, Uguru H, Helal M, Alaoufi SH, Alkaltham MS, Osiga-Aibangbee D, Okolotu GI, Emumejaye EP, Qari SH · 2026In vitro evaluation of bio-fortification effects on the nutritional quality, toxicological safety, and antioxidant of cassava (Manihot esculenta) flour, and environmental safety of processing water, using natural additivesFrontiers in NutritionRead source →}^{11Iyabo OO, Ogundiran MB, Adebayo OL · 2015Toxic and Essential Metals in Staple Foods Commonly Consumed by Students in Ekiti State, South West, NigeriaInternational Journal of ChemistryRead source →} Values that fail a check against their own paper are held and stay visible with the reason: low arsenic measured by atomic absorption without hydride generation in three studies, implausible chromium and nickel in garri, and cadmium and roasted-cassava lead cells that contradict their own tables.^{4Adjei-Mensah R, Ofori H, Tortoe C, Johnson PT, Aryee D, Frimpong SK · 2021Effect of home processing methods on the levels of heavy metal contaminants in four food crops grown in and around two mining towns in GhanaToxicology Reports 8 (2021) 1830–1838Read source →}^{6Cecilia Owusu-Agyemang Fobi, Guy Eshun, Twumasi Ankrah Kwarteng, Afia Sakyiwaa Amponsah, Prince Owusu Adoma · 2025Effect of Boiling on Heavy Metal and Cyanide Concentrations and Associated Health Risks in Cassava and Cocoyam From a Gold Mining Area in Kade, GhanaInternational Journal of Food ScienceRead source →}^{9Albishi TS, Adaigho D, Alqahtani NK, Sami R, Uguru H, Helal M, Alaoufi SH, Alkaltham MS, Osiga-Aibangbee D, Okolotu GI, Emumejaye EP, Qari SH · 2026In vitro evaluation of bio-fortification effects on the nutritional quality, toxicological safety, and antioxidant of cassava (Manihot esculenta) flour, and environmental safety of processing water, using natural additivesFrontiers in NutritionRead source →}^{10EDITED BY Arun K. Bhunia, Purdue University, United States REVIEWED BY Marco E. Mng’ong’o, sold in selected markets—taken Mbeya University of Science, Technology · 2023OPEN ACCESS potential toxic levels of cyanidePeer-reviewed articleRead source →}^{11Iyabo OO, Ogundiran MB, Adebayo OL · 2015Toxic and Essential Metals in Staple Foods Commonly Consumed by Students in Ekiti State, South West, NigeriaInternational Journal of ChemistryRead source →}^{13Christian Ebere Enyoh, Qingyue Wang, Mominul Haque Rabin, Rasheed Oluwafemi Bakare, Joseph Longji Dadiel, Wu Shangrong, Senlin Lu, Ifenna Ilechukwu · 2023Preliminary characterization and probabilistic risk assessment of microplastics and potentially toxic elements (PTEs) in garri (cassava flake), a common staple food consumed in West AfricaEnvironmental Analysis Health and ToxicologyRead source →}

Comparison methods and scope

Each row is one statistic a source reports for one form of cassava, group and metal. Raw root, cooked cassava, flour, gari, fufu and starch are kept apart, and dry-weight results never share a reference with fresh or as-sold results. Ratings place a result against an HMI reference built from monitoring results for the same form of cassava. Same-study contrasts compare a treated flour with its untreated control, or cooked dried with cooked fresh cassava. Held rows failed a check against their own paper and are shown but not compared.

Filter findings

AnalyteFood formCollection regionCollection periodSourceCompare withComparable findings only

226 of 226 findings · 8 analytes

About this table

Food formCollection scopeFindingComparisonEvidence

Cassava findings from WHO GEMS/Food monitoring and every extracted field, market and processing study, in µg/kg with each source’s form of cassava, basis, statistic and censoring retained.
Food form	Collection scope	Finding (µg/kg)

11 findings · 6 studies · 2 spread figures folded in
Cassava rootWHO GEMS/Food	EURO, random2008–2019As sold / fresh	median.
Cassava rootWHO GEMS/Food	EURO, sampling unstated2011–2015As sold / fresh	median.
Cassava rootWHO GEMS/Food	PAHO, random2014–2016As sold / fresh	median.
Cassava rootAdjei-Mensah et al. 2021	Akwatia mining town2019Basis not stated	mean.
Cassava rootAdjei-Mensah et al. 2021	Obuasi mining town2019Basis not stated	mean.
Cassava rootBortey-Sam et al. 2015	Tarkwa farms, samples above 300 µg/kgCollection date unreportedBasis unresolved	range.
Cassava rootBortey-Sam et al. 2015	Tarkwa farms, 11 communitiesCollection date unreportedBasis unresolved	median.Same samples: 17 µg/kg (minimum); 450 µg/kg (maximum)
Cassava rootBortey-Sam et al. 2015	Tarkwa farms, 11 communitiesCollection date unreportedBasis unresolved	geometric mean.
Cassava rootBortey-Sam et al. 2015	Tarkwa farms, 11 communitiesCollection date unreportedBasis unresolved	min.
Cassava rootBortey-Sam et al. 2015	Tarkwa farms, 11 communitiesCollection date unreportedBasis unresolved	max.
Cassava rootOmeje et al. 2024	Nsukka market, Nigeria2022Basis not stated	not detected.
Cassava rootOrisakwe et al. 2012	Owerri area, NigeriaCollection date unreportedBasis not stated	mean.
Cassava rootRomero-Crespo et al. 2023	Mining-area orchard, Ecuador2021As sold / fresh	below LOD.
2 findings · 1 study
Boiled cassavaAdjei-Mensah et al. 2021	Akwatia mining town2019Basis not stated	mean.
Boiled cassavaAdjei-Mensah et al. 2021	Obuasi mining town2019Basis not stated	mean.
2 findings · 1 study
Fried cassavaAdjei-Mensah et al. 2021	Akwatia mining town2019Basis not stated	mean.
Fried cassavaAdjei-Mensah et al. 2021	Obuasi mining town2019Basis not stated	mean.
2 findings · 1 study
Roasted cassavaAdjei-Mensah et al. 2021	Akwatia mining town2019Basis not stated	mean.
Roasted cassavaAdjei-Mensah et al. 2021	Obuasi mining town2019Basis not stated	mean.
1 finding · 1 study
Fresh cassava, cookedWHO GEMS/Food	AFRO, random2018As sold / fresh	median.
1 finding · 1 study
Dried cassava, cookedWHO GEMS/Food	AFRO, random2018As sold / fresh	median.
10 findings · 4 studies · 1 spread figure folded in
Cassava flourWHO GEMS/Food	PAHO, random2013–2014As sold / fresh	median.
Cassava flourAlbishi et al. 2026	Control A (unsoaked)2024–2025Dry weight	mean.
Cassava flourAlbishi et al. 2026	Control B (water-soaked)2024–2025Dry weight	mean.
Cassava flourIyabo et al. 2015	Amala (Cassava Flour)Collection date unreportedBasis not stated	sample value.
Fermented cassava flour (lafun)Jacob et al. 2023	All five Oke Ogun marketsCollection date unreportedAs sold / fresh	mean.Same samples: 28–53 µg/kg (range)
Fermented cassava flour (lafun)Jacob et al. 2023	All five Oke Ogun marketsCollection date unreportedAs sold / fresh	range.
Fermented cassava flour (lafun)Jacob et al. 2023	K1 IgbohoCollection date unreportedAs sold / fresh	sample value.
Fermented cassava flour (lafun)Jacob et al. 2023	K2 IgbetiCollection date unreportedAs sold / fresh	sample value.
Fermented cassava flour (lafun)Jacob et al. 2023	K3 IseyinCollection date unreportedAs sold / fresh	sample value.
Fermented cassava flour (lafun)Jacob et al. 2023	K4 KishiCollection date unreportedAs sold / fresh	sample value.
Fermented cassava flour (lafun)Jacob et al. 2023	K5 ShakiCollection date unreportedAs sold / fresh	sample value.
1 finding · 1 study
GariIyabo et al. 2015	Fried Cassava (Garri)Collection date unreportedBasis not stated	sample value.
1 finding · 1 study
FufuIyabo et al. 2015	FufuCollection date unreportedBasis not stated	sample value.
2 findings · 1 study
Cassava starch / tapioca flourWHO GEMS/Food	PAHO, random2013–2017As sold / fresh	median.
Cassava starch / tapioca flourWHO GEMS/Food	WPRO, random2013As sold / fresh	median.
2 findings · 1 study
Tapioca, form not statedWHO GEMS/Food	EURO, random2003–2019As sold / fresh	median.
Tapioca, form not statedWHO GEMS/Food	WPRO, random2005As sold / fresh	median.
11 findings · 1 study
Treated cassava flourAlbishi et al. 2026	F1 (1% CSP + 2% CSE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F2 (1% CSP + 4% CSE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F3 (1% TP + 2% TE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F4 (1% TP + 4%TE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F5 (1% MLP + 2% MLE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F6 (1% MLP 4% MLE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F7 (1% CSP + 2% CSE + 2%TE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F8 (1% CSP + 2% CSE + 2% MLE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F9 (1% TP + 2% TE + 2% MLE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F10 (1% TP + 2% CSE + 2% TE + 2% MLE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F11 (1% TP + 4% CSE + 4% TE + 4% MLE)2024–2025Dry weight	mean.

15 findings · 7 studies · 2 spread figures folded in
Cassava rootWHO GEMS/Food	EURO, random2008–2019As sold / fresh	median.
Cassava rootWHO GEMS/Food	EURO, sampling unstated2011–2015As sold / fresh	median.
Cassava rootAdjei-Mensah et al. 2021	Akwatia mining town2019Basis not stated	below LOD.
Cassava rootAdjei-Mensah et al. 2021	Obuasi mining town2019Basis not stated	below LOD.
Cassava rootBortey-Sam et al. 2015	Tarkwa farms, 11 communitiesCollection date unreportedBasis unresolved	median.Same samples: 2 µg/kg (minimum); 16 µg/kg (maximum)
Cassava rootBortey-Sam et al. 2015	Tarkwa farms, 11 communitiesCollection date unreportedBasis unresolved	geometric mean.
Cassava rootBortey-Sam et al. 2015	Tarkwa farms, 11 communitiesCollection date unreportedBasis unresolved	min.
Cassava rootBortey-Sam et al. 2015	Tarkwa farms, 11 communitiesCollection date unreportedBasis unresolved	max.
Cassava rootFobi et al. 2025	Kade mining area, P12024Dry weight	mean.
Cassava rootFobi et al. 2025	Kade mining area, P22024Dry weight	mean.
Cassava rootFobi et al. 2025	Kade mining area, P32024Dry weight	mean.
Cassava rootFobi et al. 2025	Kade mining area, P42024Dry weight	mean.
Cassava rootFobi et al. 2025	Kade mining area, P52024Dry weight	mean.
Cassava rootFobi et al. 2025	Kade mining area, P62024Dry weight	mean.
Cassava rootOmeje et al. 2024	Nsukka market, Nigeria2022Basis not stated	not detected.
Cassava rootOrisakwe et al. 2012	Owerri area, NigeriaCollection date unreportedBasis not stated	mean.
Cassava rootRomero-Crespo et al. 2023	Mining-area orchard, Ecuador2021As sold / fresh	sample value.
8 findings · 2 studies
Boiled cassavaAdjei-Mensah et al. 2021	Akwatia mining town2019Basis not stated	not detected.
Boiled cassavaAdjei-Mensah et al. 2021	Obuasi mining town2019Basis not stated	below LOD.
Boiled cassavaFobi et al. 2025	Kade mining area, P12024Basis not stated	mean.
Boiled cassavaFobi et al. 2025	Kade mining area, P22024Basis not stated	mean.
Boiled cassavaFobi et al. 2025	Kade mining area, P32024Basis not stated	mean.
Boiled cassavaFobi et al. 2025	Kade mining area, P42024Basis not stated	mean.
Boiled cassavaFobi et al. 2025	Kade mining area, P52024Basis not stated	not detected.
Boiled cassavaFobi et al. 2025	Kade mining area, P62024Basis not stated	not detected.
2 findings · 1 study
Fried cassavaAdjei-Mensah et al. 2021	Akwatia mining town2019Basis not stated	not detected.
Fried cassavaAdjei-Mensah et al. 2021	Obuasi mining town2019Basis not stated	not detected.
2 findings · 1 study
Roasted cassavaAdjei-Mensah et al. 2021	Akwatia mining town2019Basis not stated	below LOD.
Roasted cassavaAdjei-Mensah et al. 2021	Obuasi mining town2019Basis not stated	not detected.
1 finding · 1 study
Fresh cassava, cookedWHO GEMS/Food	AFRO, random2018As sold / fresh	median.
1 finding · 1 study
Dried cassava, cookedWHO GEMS/Food	AFRO, random2018As sold / fresh	median.
9 findings · 3 studies · 1 spread figure folded in
Cassava flourAlbishi et al. 2026	Control A (unsoaked)2024–2025Dry weight	mean.
Cassava flourAlbishi et al. 2026	Control B (water-soaked)2024–2025Dry weight	mean.
Cassava flourIyabo et al. 2015	Amala (Cassava Flour)Collection date unreportedBasis not stated	sample value.
Fermented cassava flour (lafun)Jacob et al. 2023	All five Oke Ogun marketsCollection date unreportedAs sold / fresh	mean.Same samples: 34–65 µg/kg (range)
Fermented cassava flour (lafun)Jacob et al. 2023	All five Oke Ogun marketsCollection date unreportedAs sold / fresh	range.
Fermented cassava flour (lafun)Jacob et al. 2023	K1 IgbohoCollection date unreportedAs sold / fresh	sample value.
Fermented cassava flour (lafun)Jacob et al. 2023	K2 IgbetiCollection date unreportedAs sold / fresh	sample value.
Fermented cassava flour (lafun)Jacob et al. 2023	K3 IseyinCollection date unreportedAs sold / fresh	sample value.
Fermented cassava flour (lafun)Jacob et al. 2023	K4 KishiCollection date unreportedAs sold / fresh	sample value.
Fermented cassava flour (lafun)Jacob et al. 2023	K5 ShakiCollection date unreportedAs sold / fresh	sample value.
1 finding · 1 study
GariIyabo et al. 2015	Fried Cassava (Garri)Collection date unreportedBasis not stated	below LOD.
2 findings · 1 study
Tapioca, form not statedWHO GEMS/Food	EURO, random2003–2019As sold / fresh	median.
Tapioca, form not statedWHO GEMS/Food	WPRO, random2005As sold / fresh	median.
11 findings · 1 study
Treated cassava flourAlbishi et al. 2026	F1 (1% CSP + 2% CSE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F2 (1% CSP + 4% CSE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F3 (1% TP + 2% TE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F4 (1% TP + 4%TE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F5 (1% MLP + 2% MLE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F6 (1% MLP 4% MLE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F7 (1% CSP + 2% CSE + 2%TE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F8 (1% CSP + 2% CSE + 2% MLE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F9 (1% TP + 2% TE + 2% MLE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F10 (1% TP + 2% CSE + 2% TE + 2% MLE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F11 (1% TP + 4% CSE + 4% TE + 4% MLE)2024–2025Dry weight	mean.

16 findings · 6 studies · 2 spread figures folded in
Cassava rootWHO GEMS/Food	EURO, random2010–2023As sold / fresh	median.
Cassava rootWHO GEMS/Food	EURO, sampling unstated2011–2012As sold / fresh	median.
Cassava rootWHO GEMS/Food	European Union, random2013As sold / fresh	median.
Cassava rootWHO GEMS/Food	European Union, sampling unstated2012As sold / fresh	median.
Cassava rootAdjei-Mensah et al. 2021	Akwatia mining town2019Basis not stated	mean.
Cassava rootAdjei-Mensah et al. 2021	Obuasi mining town2019Basis not stated	mean.
Cassava rootBortey-Sam et al. 2015	Tarkwa farms, 11 communitiesCollection date unreportedBasis unresolved	median.Same samples: 3 µg/kg (minimum); 17 µg/kg (maximum)
Cassava rootBortey-Sam et al. 2015	Tarkwa farms, 11 communitiesCollection date unreportedBasis unresolved	geometric mean.
Cassava rootBortey-Sam et al. 2015	Tarkwa farms, 11 communitiesCollection date unreportedBasis unresolved	min.
Cassava rootBortey-Sam et al. 2015	Tarkwa farms, 11 communitiesCollection date unreportedBasis unresolved	max.
Cassava rootFobi et al. 2025	Kade mining area, P12024Dry weight	mean.
Cassava rootFobi et al. 2025	Kade mining area, P22024Dry weight	mean.
Cassava rootFobi et al. 2025	Kade mining area, P32024Dry weight	mean.
Cassava rootFobi et al. 2025	Kade mining area, P42024Dry weight	mean.
Cassava rootFobi et al. 2025	Kade mining area, P52024Dry weight	mean.
Cassava rootFobi et al. 2025	Kade mining area, P62024Dry weight	mean.
Cassava rootOmeje et al. 2024	Nsukka market, Nigeria2022Basis not stated	mean.
Cassava rootRomero-Crespo et al. 2023	Mining-area orchard, Ecuador2021As sold / fresh	below LOD.
8 findings · 2 studies
Boiled cassavaAdjei-Mensah et al. 2021	Akwatia mining town2019Basis not stated	mean.
Boiled cassavaAdjei-Mensah et al. 2021	Obuasi mining town2019Basis not stated	mean.
Boiled cassavaFobi et al. 2025	Kade mining area, P12024Basis not stated	not detected.
Boiled cassavaFobi et al. 2025	Kade mining area, P22024Basis not stated	mean.
Boiled cassavaFobi et al. 2025	Kade mining area, P32024Basis not stated	not detected.
Boiled cassavaFobi et al. 2025	Kade mining area, P42024Basis not stated	not detected.
Boiled cassavaFobi et al. 2025	Kade mining area, P52024Basis not stated	mean.
Boiled cassavaFobi et al. 2025	Kade mining area, P62024Basis not stated	not detected.
2 findings · 1 study
Fried cassavaAdjei-Mensah et al. 2021	Akwatia mining town2019Basis not stated	below LOD.
Fried cassavaAdjei-Mensah et al. 2021	Obuasi mining town2019Basis not stated	below LOD.
2 findings · 1 study
Roasted cassavaAdjei-Mensah et al. 2021	Akwatia mining town2019Basis not stated	mean.
Roasted cassavaAdjei-Mensah et al. 2021	Obuasi mining town2019Basis not stated	mean.
1 finding · 1 study
Fresh cassava, cookedWHO GEMS/Food	AFRO, random2018As sold / fresh	median.
1 finding · 1 study
Dried cassava, cookedWHO GEMS/Food	AFRO, random2018As sold / fresh	median.
10 findings · 3 studies
Cassava flourWHO GEMS/Food	EURO, random2018–2023As sold / fresh	median.
Cassava flourAlbishi et al. 2026	Control A (unsoaked)2024–2025Dry weight	mean.
Cassava flourAlbishi et al. 2026	Control B (water-soaked)2024–2025Dry weight	mean.
Fermented cassava flour (lafun)Jacob et al. 2023	All five Oke Ogun marketsCollection date unreportedAs sold / fresh	mean.
Fermented cassava flour (lafun)Jacob et al. 2023	All five Oke Ogun marketsCollection date unreportedAs sold / fresh	range.
Fermented cassava flour (lafun)Jacob et al. 2023	K1 IgbohoCollection date unreportedAs sold / fresh	sample value.
Fermented cassava flour (lafun)Jacob et al. 2023	K2 IgbetiCollection date unreportedAs sold / fresh	sample value.
Fermented cassava flour (lafun)Jacob et al. 2023	K3 IseyinCollection date unreportedAs sold / fresh	sample value.
Fermented cassava flour (lafun)Jacob et al. 2023	K4 KishiCollection date unreportedAs sold / fresh	sample value.
Fermented cassava flour (lafun)Jacob et al. 2023	K5 ShakiCollection date unreportedAs sold / fresh	sample value.
1 finding · 1 study
Cassava starch / tapioca flourWHO GEMS/Food	EURO, random2014–2023As sold / fresh	median.
2 findings · 1 study
Tapioca, form not statedWHO GEMS/Food	EURO, random2013–2014As sold / fresh	median.
Tapioca, form not statedWHO GEMS/Food	WPRO, random2005As sold / fresh	median.
11 findings · 1 study
Treated cassava flourAlbishi et al. 2026	F1 (1% CSP + 2% CSE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F2 (1% CSP + 4% CSE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F3 (1% TP + 2% TE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F4 (1% TP + 4%TE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F5 (1% MLP + 2% MLE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F6 (1% MLP 4% MLE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F7 (1% CSP + 2% CSE + 2%TE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F8 (1% CSP + 2% CSE + 2% MLE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F9 (1% TP + 2% TE + 2% MLE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F10 (1% TP + 2% CSE + 2% TE + 2% MLE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F11 (1% TP + 4% CSE + 4% TE + 4% MLE)2024–2025Dry weight	mean.

2 findings · 1 study
Cassava starch / tapioca flourWHO GEMS/Food	PAHO, random2016As sold / fresh	median.
Cassava starch / tapioca flourWHO GEMS/Food	WPRO, random2021As sold / fresh	median.

16 findings · 6 studies · 2 spread figures folded in
Cassava rootWHO GEMS/Food	EURO, random2010–2016As sold / fresh	median.
Cassava rootAdjei-Mensah et al. 2021	Akwatia mining town2019Basis not stated	mean.
Cassava rootAdjei-Mensah et al. 2021	Obuasi mining town2019Basis not stated	mean.
Cassava rootAdjorlolo-Gasokpoh et al. 2012	Dumasi, 5.5 km from mineCollection date unreportedBasis not stated	mean.
Cassava rootAdjorlolo-Gasokpoh et al. 2012	Bogoso mine entrance, 1.5 kmCollection date unreportedBasis not stated	mean.
Cassava rootAdjorlolo-Gasokpoh et al. 2012	Prestea, Ankobra, 14 kmCollection date unreportedBasis not stated	mean.
Cassava rootAdjorlolo-Gasokpoh et al. 2012	River Bogo bank, 8.4 kmCollection date unreportedBasis not stated	mean.
Cassava rootBortey-Sam et al. 2015	Tarkwa farms, 11 communitiesCollection date unreportedBasis unresolved	median.Same samples: 1 µg/kg (minimum); 10 µg/kg (maximum)
Cassava rootBortey-Sam et al. 2015	Tarkwa farms, 11 communitiesCollection date unreportedBasis unresolved	geometric mean.
Cassava rootBortey-Sam et al. 2015	Tarkwa farms, 11 communitiesCollection date unreportedBasis unresolved	min.
Cassava rootBortey-Sam et al. 2015	Tarkwa farms, 11 communitiesCollection date unreportedBasis unresolved	max.
Cassava rootFobi et al. 2025	Kade mining area, P12024Dry weight	mean.
Cassava rootFobi et al. 2025	Kade mining area, P22024Dry weight	mean.
Cassava rootFobi et al. 2025	Kade mining area, P32024Dry weight	mean.
Cassava rootFobi et al. 2025	Kade mining area, P42024Dry weight	mean.
Cassava rootFobi et al. 2025	Kade mining area, P52024Dry weight	mean.
Cassava rootFobi et al. 2025	Kade mining area, P62024Dry weight	mean.
Cassava rootOmeje et al. 2024	Nsukka market, Nigeria2022Basis not stated	not detected.
8 findings · 1 study
Cassava peel (outer skin)Adjorlolo-Gasokpoh et al. 2012	Dumasi, 5.5 km from mineCollection date unreportedBasis not stated	mean.
Cassava peel (inner cortex)Adjorlolo-Gasokpoh et al. 2012	Dumasi, 5.5 km from mineCollection date unreportedBasis not stated	mean.
Cassava peel (outer skin)Adjorlolo-Gasokpoh et al. 2012	Bogoso mine entrance, 1.5 kmCollection date unreportedBasis not stated	mean.
Cassava peel (inner cortex)Adjorlolo-Gasokpoh et al. 2012	Bogoso mine entrance, 1.5 kmCollection date unreportedBasis not stated	mean.
Cassava peel (outer skin)Adjorlolo-Gasokpoh et al. 2012	Prestea, Ankobra, 14 kmCollection date unreportedBasis not stated	mean.
Cassava peel (inner cortex)Adjorlolo-Gasokpoh et al. 2012	Prestea, Ankobra, 14 kmCollection date unreportedBasis not stated	mean.
Cassava peel (outer skin)Adjorlolo-Gasokpoh et al. 2012	River Bogo bank, 8.4 kmCollection date unreportedBasis not stated	mean.
Cassava peel (inner cortex)Adjorlolo-Gasokpoh et al. 2012	River Bogo bank, 8.4 kmCollection date unreportedBasis not stated	mean.
8 findings · 2 studies
Boiled cassavaAdjei-Mensah et al. 2021	Akwatia mining town2019Basis not stated	mean.
Boiled cassavaAdjei-Mensah et al. 2021	Obuasi mining town2019Basis not stated	mean.
Boiled cassavaFobi et al. 2025	Kade mining area, P12024Basis not stated	mean.
Boiled cassavaFobi et al. 2025	Kade mining area, P22024Basis not stated	mean.
Boiled cassavaFobi et al. 2025	Kade mining area, P32024Basis not stated	mean.
Boiled cassavaFobi et al. 2025	Kade mining area, P42024Basis not stated	mean.
Boiled cassavaFobi et al. 2025	Kade mining area, P52024Basis not stated	mean.
Boiled cassavaFobi et al. 2025	Kade mining area, P62024Basis not stated	mean.
2 findings · 1 study
Fried cassavaAdjei-Mensah et al. 2021	Akwatia mining town2019Basis not stated	mean.
Fried cassavaAdjei-Mensah et al. 2021	Obuasi mining town2019Basis not stated	mean.
2 findings · 1 study
Roasted cassavaAdjei-Mensah et al. 2021	Akwatia mining town2019Basis not stated	mean.
Roasted cassavaAdjei-Mensah et al. 2021	Obuasi mining town2019Basis not stated	mean.
1 finding · 1 study
Fresh cassava, cookedWHO GEMS/Food	AFRO, random2018As sold / fresh	median.
1 finding · 1 study
Dried cassava, cookedWHO GEMS/Food	AFRO, random2018As sold / fresh	median.
1 finding · 1 study
Tapioca, form not statedWHO GEMS/Food	EURO, random2007–2013As sold / fresh	median.

5 findings · 4 studies · 2 spread figures folded in
Cassava rootBortey-Sam et al. 2015	Tarkwa farms, 11 communitiesCollection date unreportedBasis unresolved	median.Same samples: 2,700 µg/kg (minimum); 5,400 µg/kg (maximum)
Cassava rootBortey-Sam et al. 2015	Tarkwa farms, 11 communitiesCollection date unreportedBasis unresolved	geometric mean.
Cassava rootBortey-Sam et al. 2015	Tarkwa farms, 11 communitiesCollection date unreportedBasis unresolved	min.
Cassava rootBortey-Sam et al. 2015	Tarkwa farms, 11 communitiesCollection date unreportedBasis unresolved	max.
Cassava rootOmeje et al. 2024	Nsukka market, Nigeria2022Basis not stated	mean.
Cassava rootOrisakwe et al. 2012	Owerri area, NigeriaCollection date unreportedBasis not stated	mean.
Cassava rootRomero-Crespo et al. 2023	Mining-area orchard, Ecuador2021As sold / fresh	sample value.
5 findings · 1 study
GariEnyoh et al. 2023	S1 open market, NigeriaCollection date unreportedDry weight	mean.
GariEnyoh et al. 2023	S2 supermarket, NigeriaCollection date unreportedDry weight	mean.
GariEnyoh et al. 2023	S3 shop, JapanCollection date unreportedDry weight	mean.
GariEnyoh et al. 2023	S4 shop, JapanCollection date unreportedDry weight	mean.
GariEnyoh et al. 2023	S5 yellow garri, JapanCollection date unreportedDry weight	mean.

4 findings · 3 studies · 2 spread figures folded in
Cassava rootBortey-Sam et al. 2015	Tarkwa farms, 11 communitiesCollection date unreportedBasis unresolved	median.Same samples: 10 µg/kg (minimum); 200 µg/kg (maximum)
Cassava rootBortey-Sam et al. 2015	Tarkwa farms, 11 communitiesCollection date unreportedBasis unresolved	geometric mean.
Cassava rootBortey-Sam et al. 2015	Tarkwa farms, 11 communitiesCollection date unreportedBasis unresolved	min.
Cassava rootBortey-Sam et al. 2015	Tarkwa farms, 11 communitiesCollection date unreportedBasis unresolved	max.
Cassava rootOmeje et al. 2024	Nsukka market, Nigeria2022Basis not stated	mean.
Cassava rootRomero-Crespo et al. 2023	Mining-area orchard, Ecuador2021As sold / fresh	sample value.
2 findings · 1 study
Cassava flourAlbishi et al. 2026	Control A (unsoaked)2024–2025Dry weight	mean.
Cassava flourAlbishi et al. 2026	Control B (water-soaked)2024–2025Dry weight	mean.
5 findings · 1 study
GariEnyoh et al. 2023	S1 open market, NigeriaCollection date unreportedDry weight	mean.
GariEnyoh et al. 2023	S2 supermarket, NigeriaCollection date unreportedDry weight	mean.
GariEnyoh et al. 2023	S3 shop, JapanCollection date unreportedDry weight	mean.
GariEnyoh et al. 2023	S4 shop, JapanCollection date unreportedDry weight	mean.
GariEnyoh et al. 2023	S5 yellow garri, JapanCollection date unreportedDry weight	not detected.
11 findings · 1 study
Treated cassava flourAlbishi et al. 2026	F1 (1% CSP + 2% CSE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F2 (1% CSP + 4% CSE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F3 (1% TP + 2% TE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F4 (1% TP + 4%TE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F5 (1% MLP + 2% MLE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F6 (1% MLP 4% MLE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F7 (1% CSP + 2% CSE + 2%TE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F8 (1% CSP + 2% CSE + 2% MLE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F9 (1% TP + 2% TE + 2% MLE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F10 (1% TP + 2% CSE + 2% TE + 2% MLE)2024–2025Dry weight	mean.
Treated cassava flourAlbishi et al. 2026	F11 (1% TP + 4% CSE + 4% TE + 4% MLE)2024–2025Dry weight	mean.

1 finding · 1 study
Tapioca, form not statedWHO GEMS/Food	WPRO, random2005As sold / fresh	median.

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 cassava Pb reference · Cassava root

Reference and scope

HMI cassava Pb reference · Cassava root. Cassava root; 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	7.5–9 (bounded)	31.4–36 (bounded)	80
Alternative weighting	7.5–9 (bounded)	31.4–36 (bounded)	80

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

Contributing family	Reference units	Weight
WHO GEMS/Food	13	100%

Methods

HMI pools subgroup medians of individual WHO GEMS/Food cassava root results, one unit per WHO region, sampling type and collection year, with per-market values from market surveys of the same form where they exist, in the result basis as reported (as sold or as eaten). 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 of cassava, dry-weight results and results without a stated basis 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–Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

GEMS/Food is the only compatible sampling family for cassava root in the completed cassava ledger, so this single-family reference cannot represent other markets. WHO regions are reporting regions, not growing areas. Many results are non-detects with differing limits, which widens 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:cassava-root-fw-Pb:c8751e17edba

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

HMI cassava Pb reference · Fresh cassava, cooked

Reference and scope

HMI cassava Pb reference · Fresh cassava, cooked. Fresh cassava, cooked; 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	35	35	35
Alternative weighting	35	35	35

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 of individual WHO GEMS/Food fresh cassava, cooked results, one unit per WHO region, sampling type and collection year, with per-market values from market surveys of the same form where they exist, in the result basis as reported (as sold or as eaten). 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 of cassava, dry-weight results and results without a stated basis 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–Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

GEMS/Food is the only compatible sampling family for fresh cassava, cooked in the completed cassava ledger, so this single-family reference cannot represent other markets. WHO regions are reporting regions, not growing areas. Many results are non-detects with differing limits, which widens 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:cassava-cooked_root-fw-Pb:d7afad9b460a

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

HMI cassava Pb reference · Dried cassava, cooked

Reference and scope

HMI cassava Pb reference · Dried cassava, cooked. Dried cassava, cooked; 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	82	82	82
Alternative weighting	82	82	82

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 of individual WHO GEMS/Food dried cassava, cooked results, one unit per WHO region, sampling type and collection year, with per-market values from market surveys of the same form where they exist, in the result basis as reported (as sold or as eaten). 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 of cassava, dry-weight results and results without a stated basis 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–Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

GEMS/Food is the only compatible sampling family for dried cassava, cooked in the completed cassava ledger, so this single-family reference cannot represent other markets. WHO regions are reporting regions, not growing areas. Many results are non-detects with differing limits, which widens 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:cassava-cooked_dry-fw-Pb:73cd7d37e4a8

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

HMI cassava Pb reference · Cassava flour

Reference and scope

HMI cassava Pb reference · Cassava flour. Cassava flour; 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	19	36
Alternative weighting	10	19	36

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

Contributing family	Reference units	Weight
WHO GEMS/Food	2	50%
Jacob et al. 2023	5	50%

Methods

HMI pools subgroup medians of individual WHO GEMS/Food cassava flour results, one unit per WHO region, sampling type and collection year, with per-market values from market surveys of the same form where they exist, in the result basis as reported (as sold or as eaten). 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 of cassava, dry-weight results and results without a stated basis 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–Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

Two sampling families with different designs; neither is a probability sample of world supply. WHO regions are reporting regions, not growing areas. Many results are non-detects with differing limits, which widens 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:cassava-flour-fw-Pb:61637ee00917

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

HMI cassava Pb reference · Cassava starch / tapioca flour

Reference and scope

HMI cassava Pb reference · Cassava starch / tapioca flour. Cassava starch / tapioca flour; 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	10–11.3 (bounded)	12.7–25.5 (bounded)
Alternative weighting	10	10–11.3 (bounded)	12.7–25.5 (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 of individual WHO GEMS/Food cassava starch / tapioca flour results, one unit per WHO region, sampling type and collection year, with per-market values from market surveys of the same form where they exist, in the result basis as reported (as sold or as eaten). 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 of cassava, dry-weight results and results without a stated basis 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–Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

GEMS/Food is the only compatible sampling family for cassava starch / tapioca flour in the completed cassava ledger, so this single-family reference cannot represent other markets. WHO regions are reporting regions, not growing areas. Many results are non-detects with differing limits, which widens 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:cassava-starch-fw-Pb:e7370a7a19ea

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

HMI cassava Pb reference · Tapioca, form not stated

Reference and scope

HMI cassava Pb reference · Tapioca, form not stated. Tapioca, form not stated; 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	≤90	55–105 (bounded)	230
Alternative weighting	≤90	55–105 (bounded)	230

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 of individual WHO GEMS/Food tapioca, form not stated results, one unit per WHO region, sampling type and collection year, with per-market values from market surveys of the same form where they exist, in the result basis as reported (as sold or as eaten). 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 of cassava, dry-weight results and results without a stated basis 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–Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

GEMS/Food is the only compatible sampling family for tapioca, form not stated in the completed cassava ledger, so this single-family reference cannot represent other markets. WHO regions are reporting regions, not growing areas. Many results are non-detects with differing limits, which widens 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:cassava-tapioca-fw-Pb:5f35b58cf7e6

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

HMI cassava Cd reference · Cassava root

Reference and scope

HMI cassava Cd reference · Cassava root. Cassava root; 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	1.8–3 (bounded)	3.5–5 (bounded)
Alternative weighting	≤2	1.8–3 (bounded)	3.5–5 (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 of individual WHO GEMS/Food cassava root results, one unit per WHO region, sampling type and collection year, with per-market values from market surveys of the same form where they exist, in the result basis as reported (as sold or as eaten). 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 of cassava, dry-weight results and results without a stated basis 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–Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

GEMS/Food is the only compatible sampling family for cassava root in the completed cassava ledger, so this single-family reference cannot represent other markets. WHO regions are reporting regions, not growing areas. Many results are non-detects with differing limits, which widens 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:cassava-root-fw-Cd:93aab0028c50

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

HMI cassava Cd reference · Fresh cassava, cooked

Reference and scope

HMI cassava Cd reference · Fresh cassava, cooked. Fresh cassava, cooked; 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	0.5–0.65 (bounded)	0.5–0.65 (bounded)	0.5–0.65 (bounded)
Alternative weighting	0.5–0.65 (bounded)	0.5–0.65 (bounded)	0.5–0.65 (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 of individual WHO GEMS/Food fresh cassava, cooked results, one unit per WHO region, sampling type and collection year, with per-market values from market surveys of the same form where they exist, in the result basis as reported (as sold or as eaten). 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 of cassava, dry-weight results and results without a stated basis 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–Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

GEMS/Food is the only compatible sampling family for fresh cassava, cooked in the completed cassava ledger, so this single-family reference cannot represent other markets. WHO regions are reporting regions, not growing areas. Many results are non-detects with differing limits, which widens 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:cassava-cooked_root-fw-Cd:35bae1d133d6

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

HMI cassava Cd reference · Dried cassava, cooked

Reference and scope

HMI cassava Cd reference · Dried cassava, cooked. Dried cassava, cooked; 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	1	1	1
Alternative weighting	1	1	1

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

Contributing family	Reference units	Weight
WHO GEMS/Food	1	100%

Methods

HMI pools subgroup medians of individual WHO GEMS/Food dried cassava, cooked results, one unit per WHO region, sampling type and collection year, with per-market values from market surveys of the same form where they exist, in the result basis as reported (as sold or as eaten). 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 of cassava, dry-weight results and results without a stated basis 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–Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

GEMS/Food is the only compatible sampling family for dried cassava, cooked in the completed cassava ledger, so this single-family reference cannot represent other markets. WHO regions are reporting regions, not growing areas. Many results are non-detects with differing limits, which widens 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:cassava-cooked_dry-fw-Cd:b29ab604a44e

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

HMI cassava Cd reference · Cassava flour

Reference and scope

HMI cassava Cd reference · Cassava flour. Cassava flour; 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	36	45	61
Alternative weighting	36	45	61

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

Contributing family	Reference units	Weight
Jacob et al. 2023	5	100%

Methods

HMI pools subgroup medians of individual WHO GEMS/Food cassava flour results, one unit per WHO region, sampling type and collection year, with per-market values from market surveys of the same form where they exist, in the result basis as reported (as sold or as eaten). 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 of cassava, dry-weight results and results without a stated basis 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–Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

One market survey is the only compatible sampling family for cassava flour in the completed cassava ledger, so this single-family reference cannot represent other markets. WHO regions are reporting regions, not growing areas. Many results are non-detects with differing limits, which widens the bounds. A position is descriptive concentration, not a safety or compliance judgment.

Admitted source families

Jacob et al. 2023

Reference version and interpretation

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

hmi-pooled-reference-v2:cassava-flour-fw-Cd:8b611dd8a864

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

HMI cassava Cd reference · Tapioca, form not stated

Reference and scope

HMI cassava Cd reference · Tapioca, form not stated. Tapioca, form not stated; 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	≤4.3	≤8.33	≤39
Alternative weighting	≤4.3	≤8.33	≤39

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 of individual WHO GEMS/Food tapioca, form not stated results, one unit per WHO region, sampling type and collection year, with per-market values from market surveys of the same form where they exist, in the result basis as reported (as sold or as eaten). 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 of cassava, dry-weight results and results without a stated basis 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–Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

GEMS/Food is the only compatible sampling family for tapioca, form not stated in the completed cassava ledger, so this single-family reference cannot represent other markets. WHO regions are reporting regions, not growing areas. Many results are non-detects with differing limits, which widens 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:cassava-tapioca-fw-Cd:2dcc06e4530e

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

HMI cassava tAs reference · Cassava root

Reference and scope

HMI cassava tAs reference · Cassava root. Cassava root; 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	≤30.5	≤60	≤90
Alternative weighting	≤30.5	≤60	≤90

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

Contributing family	Reference units	Weight
WHO GEMS/Food	12	100%

Methods

HMI pools subgroup medians of individual WHO GEMS/Food cassava root results, one unit per WHO region, sampling type and collection year, with per-market values from market surveys of the same form where they exist, in the result basis as reported (as sold or as eaten). 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 of cassava, dry-weight results and results without a stated basis 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–Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

GEMS/Food is the only compatible sampling family for cassava root in the completed cassava ledger, so this single-family reference cannot represent other markets. WHO regions are reporting regions, not growing areas. Many results are non-detects with differing limits, which widens 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:cassava-root-fw-tAs:048ba32bb61b

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

HMI cassava tAs reference · Fresh cassava, cooked

Reference and scope

HMI cassava tAs reference · Fresh cassava, cooked. Fresh cassava, cooked; 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	≤1	≤1	≤1
Alternative weighting	≤1	≤1	≤1

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

Contributing family	Reference units	Weight
WHO GEMS/Food	1	100%

Methods

HMI pools subgroup medians of individual WHO GEMS/Food fresh cassava, cooked results, one unit per WHO region, sampling type and collection year, with per-market values from market surveys of the same form where they exist, in the result basis as reported (as sold or as eaten). 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 of cassava, dry-weight results and results without a stated basis 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–Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

GEMS/Food is the only compatible sampling family for fresh cassava, cooked in the completed cassava ledger, so this single-family reference cannot represent other markets. WHO regions are reporting regions, not growing areas. Many results are non-detects with differing limits, which widens 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:cassava-cooked_root-fw-tAs:de0d6fca1b6e

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

HMI cassava tAs reference · Dried cassava, cooked

Reference and scope

HMI cassava tAs reference · Dried cassava, cooked. Dried cassava, cooked; 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	≤1	≤1	≤1
Alternative weighting	≤1	≤1	≤1

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

Contributing family	Reference units	Weight
WHO GEMS/Food	1	100%

Methods

HMI pools subgroup medians of individual WHO GEMS/Food dried cassava, cooked results, one unit per WHO region, sampling type and collection year, with per-market values from market surveys of the same form where they exist, in the result basis as reported (as sold or as eaten). 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 of cassava, dry-weight results and results without a stated basis 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–Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

GEMS/Food is the only compatible sampling family for dried cassava, cooked in the completed cassava ledger, so this single-family reference cannot represent other markets. WHO regions are reporting regions, not growing areas. Many results are non-detects with differing limits, which widens 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:cassava-cooked_dry-fw-tAs:cb3f05fc88b1

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

HMI cassava tAs reference · Cassava flour

Reference and scope

HMI cassava tAs reference · Cassava flour. Cassava flour; 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	≤17	≤32.5	≤37.5
Alternative weighting	≤17	≤32.5	≤37.5

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

Contributing family	Reference units	Weight
WHO GEMS/Food	4	100%

Methods

HMI pools subgroup medians of individual WHO GEMS/Food cassava flour results, one unit per WHO region, sampling type and collection year, with per-market values from market surveys of the same form where they exist, in the result basis as reported (as sold or as eaten). 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 of cassava, dry-weight results and results without a stated basis 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–Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

GEMS/Food is the only compatible sampling family for cassava flour in the completed cassava ledger, so this single-family reference cannot represent other markets. WHO regions are reporting regions, not growing areas. Many results are non-detects with differing limits, which widens 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:cassava-flour-fw-tAs:08acc9751fa9

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

HMI cassava tAs reference · Cassava starch / tapioca flour

Reference and scope

HMI cassava tAs reference · Cassava starch / tapioca flour. Cassava starch / tapioca flour; 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	≤21.5	≤44.5	≤70
Alternative weighting	≤21.5	≤44.5	≤70

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 of individual WHO GEMS/Food cassava starch / tapioca flour results, one unit per WHO region, sampling type and collection year, with per-market values from market surveys of the same form where they exist, in the result basis as reported (as sold or as eaten). 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 of cassava, dry-weight results and results without a stated basis 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–Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

GEMS/Food is the only compatible sampling family for cassava starch / tapioca flour in the completed cassava ledger, so this single-family reference cannot represent other markets. WHO regions are reporting regions, not growing areas. Many results are non-detects with differing limits, which widens 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:cassava-starch-fw-tAs:d7bccd4b5355

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

HMI cassava tAs reference · Tapioca, form not stated

Reference and scope

HMI cassava tAs reference · Tapioca, form not stated. Tapioca, form not stated; 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	≤40	≤80	≤90
Alternative weighting	≤40	≤80	≤90

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

Contributing family	Reference units	Weight
WHO GEMS/Food	3	100%

Methods

HMI pools subgroup medians of individual WHO GEMS/Food tapioca, form not stated results, one unit per WHO region, sampling type and collection year, with per-market values from market surveys of the same form where they exist, in the result basis as reported (as sold or as eaten). 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 of cassava, dry-weight results and results without a stated basis 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–Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

GEMS/Food is the only compatible sampling family for tapioca, form not stated in the completed cassava ledger, so this single-family reference cannot represent other markets. WHO regions are reporting regions, not growing areas. Many results are non-detects with differing limits, which widens 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:cassava-tapioca-fw-tAs:173e80285f54

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

HMI cassava iAs reference · Cassava starch / tapioca flour

Reference and scope

HMI cassava iAs reference · Cassava starch / tapioca flour. Cassava starch / tapioca flour; 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	≤0.03	≤10	≤20
Alternative weighting	≤0.03	≤10	≤20

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 of individual WHO GEMS/Food cassava starch / tapioca flour results, one unit per WHO region, sampling type and collection year, with per-market values from market surveys of the same form where they exist, in the result basis as reported (as sold or as eaten). 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 of cassava, dry-weight results and results without a stated basis 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–Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

GEMS/Food is the only compatible sampling family for cassava starch / tapioca flour in the completed cassava ledger, so this single-family reference cannot represent other markets. WHO regions are reporting regions, not growing areas. Many results are non-detects with differing limits, which widens 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:cassava-starch-fw-iAs:72becbdb0e3d

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

HMI cassava tHg reference · Cassava root

Reference and scope

HMI cassava tHg reference · Cassava root. Cassava root; 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	≤11	≤30	6–50 (bounded)
Alternative weighting	≤11	≤30	6–50 (bounded)

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

Contributing family	Reference units	Weight
WHO GEMS/Food	3	100%

Methods

HMI pools subgroup medians of individual WHO GEMS/Food cassava root results, one unit per WHO region, sampling type and collection year, with per-market values from market surveys of the same form where they exist, in the result basis as reported (as sold or as eaten). 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 of cassava, dry-weight results and results without a stated basis 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–Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

GEMS/Food is the only compatible sampling family for cassava root in the completed cassava ledger, so this single-family reference cannot represent other markets. WHO regions are reporting regions, not growing areas. Many results are non-detects with differing limits, which widens 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:cassava-root-fw-tHg:caaea4a4c0ef

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

HMI cassava tHg reference · Fresh cassava, cooked

Reference and scope

HMI cassava tHg reference · Fresh cassava, cooked. Fresh cassava, cooked; 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	≤4.2	≤4.2	≤4.2
Alternative weighting	≤4.2	≤4.2	≤4.2

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

Contributing family	Reference units	Weight
WHO GEMS/Food	1	100%

Methods

HMI pools subgroup medians of individual WHO GEMS/Food fresh cassava, cooked results, one unit per WHO region, sampling type and collection year, with per-market values from market surveys of the same form where they exist, in the result basis as reported (as sold or as eaten). 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 of cassava, dry-weight results and results without a stated basis 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–Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

GEMS/Food is the only compatible sampling family for fresh cassava, cooked in the completed cassava ledger, so this single-family reference cannot represent other markets. WHO regions are reporting regions, not growing areas. Many results are non-detects with differing limits, which widens 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:cassava-cooked_root-fw-tHg:c985cc3817a3

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

HMI cassava tHg reference · Dried cassava, cooked

Reference and scope

HMI cassava tHg reference · Dried cassava, cooked. Dried cassava, cooked; 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	≤4.2	≤4.2	≤4.2
Alternative weighting	≤4.2	≤4.2	≤4.2

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

Contributing family	Reference units	Weight
WHO GEMS/Food	1	100%

Methods

HMI pools subgroup medians of individual WHO GEMS/Food dried cassava, cooked results, one unit per WHO region, sampling type and collection year, with per-market values from market surveys of the same form where they exist, in the result basis as reported (as sold or as eaten). 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 of cassava, dry-weight results and results without a stated basis 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–Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

GEMS/Food is the only compatible sampling family for dried cassava, cooked in the completed cassava ledger, so this single-family reference cannot represent other markets. WHO regions are reporting regions, not growing areas. Many results are non-detects with differing limits, which widens 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:cassava-cooked_dry-fw-tHg:a752d9472c51

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

HMI cassava tHg reference · Tapioca, form not stated

Reference and scope

HMI cassava tHg reference · Tapioca, form not stated. Tapioca, form not stated; 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	≤20	≤35	≤50
Alternative weighting	≤20	≤35	≤50

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 of individual WHO GEMS/Food tapioca, form not stated results, one unit per WHO region, sampling type and collection year, with per-market values from market surveys of the same form where they exist, in the result basis as reported (as sold or as eaten). 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 of cassava, dry-weight results and results without a stated basis 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–Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

GEMS/Food is the only compatible sampling family for tapioca, form not stated in the completed cassava ledger, so this single-family reference cannot represent other markets. WHO regions are reporting regions, not growing areas. Many results are non-detects with differing limits, which widens 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:cassava-tapioca-fw-tHg:298217b9cf88

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

HMI cassava Sn reference · Tapioca, form not stated

Reference and scope

HMI cassava Sn reference · Tapioca, form not stated. Tapioca, form not stated; 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	≤60	≤60	≤60
Alternative weighting	≤60	≤60	≤60

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 of individual WHO GEMS/Food tapioca, form not stated results, one unit per WHO region, sampling type and collection year, with per-market values from market surveys of the same form where they exist, in the result basis as reported (as sold or as eaten). 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 of cassava, dry-weight results and results without a stated basis 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–Q3 spread are required. This is a display eligibility rule, not statistical validation.

Limitations

GEMS/Food is the only compatible sampling family for tapioca, form not stated in the completed cassava ledger, so this single-family reference cannot represent other markets. WHO regions are reporting regions, not growing areas. Many results are non-detects with differing limits, which widens 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:cassava-tapioca-fw-Sn:7906f1eaaf1c

The reference is computed by HMI from admitted evidence, not copied from a study baseline. It describes the evidence currently in this scope; it is 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 cassava

Cassava takes up metals from the soil it grows in.^{5Romero-Crespo P, Jiménez-Oyola S, Salgado-Almeida B, Zambrano-Anchundia J, Goyburo-Chávez C, González-Valoys A, Higueras P · 2023Trace elements in farmland soils and crops, and probabilistic health risk assessment in areas influenced by mining activity in EcuadorEnvironmental Geochemistry and HealthRead source →} In an orchard on gold-mining waste in Ponce Enríquez, Ecuador, where soils held 1.06–1.94 mg/kg cadmium and up to 422 mg/kg arsenic, the cassava root carried 60 µg/kg cadmium and 2,350 µg/kg nickel but arsenic and lead below 200 and 100 µg/kg.^{5Romero-Crespo P, Jiménez-Oyola S, Salgado-Almeida B, Zambrano-Anchundia J, Goyburo-Chávez C, González-Valoys A, Higueras P · 2023Trace elements in farmland soils and crops, and probabilistic health risk assessment in areas influenced by mining activity in EcuadorEnvironmental Geochemistry and HealthRead source →} In the same orchards cassava was one of three crops with arsenic below detection and among the lowest in cadmium of nine crops, while lettuce reached 12,500 µg/kg arsenic.^{5Romero-Crespo P, Jiménez-Oyola S, Salgado-Almeida B, Zambrano-Anchundia J, Goyburo-Chávez C, González-Valoys A, Higueras P · 2023Trace elements in farmland soils and crops, and probabilistic health risk assessment in areas influenced by mining activity in EcuadorEnvironmental Geochemistry and HealthRead source →}

Most of the lead found in monitoring is likely surface or soil-related contamination of the root or post-harvest contact rather than a property of the plant, but no source in the ledger tested that directly. The EU limit for root and tuber vegetables applies after washing and separating the edible part, which shows where regulators expect soil to be removed.^{2European Commission · 2023Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →} A Nigerian lafun survey attributes its low cadmium to reduced use of cadmium-containing fertilisers, without data on fertiliser use.^{10EDITED BY Arun K. Bhunia, Purdue University, United States REVIEWED BY Marco E. Mng’ong’o, sold in selected markets—taken Mbeya University of Science, Technology · 2023OPEN ACCESS potential toxic levels of cyanidePeer-reviewed articleRead source →}

Gold mining is the clearest source in the evidence. Near goldmines at Bogoso and Prestea in Ghana, peeled cassava flesh carried 122–168 µg/kg mercury at sites 1.5 to 14 km from the mines, with no clear fall-off with distance.^{3A. Adjorlolo-Gasokpoh, A. A. Golow, J. Kambo-Dorsa · 2012Mercury in the Surface Soil and Cassava, Manihot esculenta (Flesh, Leaves and Peel) Near Goldmines at Bogoso and Prestea, GhanaBull Environ Contam ToxicolRead source →} In the Kade mining area, raw cassava held 301–426 µg/kg mercury and 24–30 µg/kg cadmium on a dry-weight basis, and peeled roots bought in the mining towns of Obuasi and Akwatia averaged 2,110 µg/kg lead and 120–160 µg/kg mercury, on a basis the paper does not state.^{4Adjei-Mensah R, Ofori H, Tortoe C, Johnson PT, Aryee D, Frimpong SK · 2021Effect of home processing methods on the levels of heavy metal contaminants in four food crops grown in and around two mining towns in GhanaToxicology Reports 8 (2021) 1830–1838Read source →}^{6Cecilia Owusu-Agyemang Fobi, Guy Eshun, Twumasi Ankrah Kwarteng, Afia Sakyiwaa Amponsah, Prince Owusu Adoma · 2025Effect of Boiling on Heavy Metal and Cyanide Concentrations and Associated Health Risks in Cassava and Cocoyam From a Gold Mining Area in Kade, GhanaInternational Journal of Food ScienceRead source →} Around Tarkwa’s gold mines, 33 cassava roots from 11 farming communities had a geometric mean of 180 µg/kg lead, with about 30% at 370–450 µg/kg, but only 4 µg/kg mercury; the paper labels its results as both wet and dry weight, so the basis is unresolved.^{12Bortey-Sam N, Nakayama SMM, Akoto O, Ikenaka Y, Fobil JN, Baidoo E, Mizukawa H, Ishizuka M · 2015Accumulation of Heavy Metals and Metalloid in Foodstuffs from Agricultural Soils around Tarkwa Area in Ghana, and Associated Human Health RisksInternational Journal of Environmental Research and Public Health (2015)Read source →} Mercury sat in the edible flesh as well as the peel: at all four Ghanaian sites the outer skin held less mercury than the flesh beneath it, and the inner cortex held less at three.^{3A. Adjorlolo-Gasokpoh, A. A. Golow, J. Kambo-Dorsa · 2012Mercury in the Surface Soil and Cassava, Manihot esculenta (Flesh, Leaves and Peel) Near Goldmines at Bogoso and Prestea, GhanaBull Environ Contam ToxicolRead source →}

Processing water is another route. When peeled roots were soaked for flour, the soaking water itself gained lead, cadmium and chromium, showing that metals move between the root and the water in both directions depending on its quality.^{9Albishi TS, Adaigho D, Alqahtani NK, Sami R, Uguru H, Helal M, Alaoufi SH, Alkaltham MS, Osiga-Aibangbee D, Okolotu GI, Emumejaye EP, Qari SH · 2026In vitro evaluation of bio-fortification effects on the nutritional quality, toxicological safety, and antioxidant of cassava (Manihot esculenta) flour, and environmental safety of processing water, using natural additivesFrontiers in NutritionRead source →}

Which forms and markets measured lower

For lead in raw cassava root, the lowest results in the evidence are those submitted to GEMS/Food from the Americas: a median of 9 µg/kg across 26 imported roots in 2014–2016, with a maximum of 82 µg/kg.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →} European submissions were higher, with medians of 46 µg/kg under random sampling and 92 µg/kg where sampling was not described.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →} WHO regions are reporting regions, not growing areas, so this is a lowest observed pattern by market, not a ranking of producing countries.

Within Europe, lead fell over time and differed by origin label.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →} 17 of 39 root results from 2008–2012 exceeded 100 µg/kg against 1 of 11 from 2013–2019, and roots labelled imported had a median of 140 µg/kg against 64 µg/kg for those labelled domestic.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →} These are national monitoring submissions, not a fixed panel of suppliers, and the export does not name the country of growth.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}

Among processed forms, cassava starch and flour submitted from the Americas were lowest for lead, at a median of 10 µg/kg (15 starch and 3 flour results).^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →} Results labelled “Tapioca” in Europe were among the highest, with a median of 115 µg/kg and 12 of 22 above 100 µg/kg, but their form is not stated.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →} In Nigeria, market lafun ranged from 28 to 53 µg/kg lead across five towns, while single vendor samples of amala flour, fufu and garri reported 140, 150 and 300 µg/kg, and cassava from around Owerri 330 µg/kg lead and 100 µg/kg cadmium, all on an unstated basis.^{10EDITED BY Arun K. Bhunia, Purdue University, United States REVIEWED BY Marco E. Mng’ong’o, sold in selected markets—taken Mbeya University of Science, Technology · 2023OPEN ACCESS potential toxic levels of cyanidePeer-reviewed articleRead source →}^{11Iyabo OO, Ogundiran MB, Adebayo OL · 2015Toxic and Essential Metals in Staple Foods Commonly Consumed by Students in Ekiti State, South West, NigeriaInternational Journal of ChemistryRead source →}^{14Orisakwe OE, Nduka JK, Amadi CN, Dike DO, Bede O · 2012Heavy metals health risk assessment for population via consumption of food crops and fruits in Owerri, South Eastern, NigeriaChemistry Central Journal 2012, 6:77Read source →}

The highest values in the evidence come from cassava grown in gold-mining districts, not from any market.^{3A. Adjorlolo-Gasokpoh, A. A. Golow, J. Kambo-Dorsa · 2012Mercury in the Surface Soil and Cassava, Manihot esculenta (Flesh, Leaves and Peel) Near Goldmines at Bogoso and Prestea, GhanaBull Environ Contam ToxicolRead source →}^{4Adjei-Mensah R, Ofori H, Tortoe C, Johnson PT, Aryee D, Frimpong SK · 2021Effect of home processing methods on the levels of heavy metal contaminants in four food crops grown in and around two mining towns in GhanaToxicology Reports 8 (2021) 1830–1838Read source →}^{6Cecilia Owusu-Agyemang Fobi, Guy Eshun, Twumasi Ankrah Kwarteng, Afia Sakyiwaa Amponsah, Prince Owusu Adoma · 2025Effect of Boiling on Heavy Metal and Cyanide Concentrations and Associated Health Risks in Cassava and Cocoyam From a Gold Mining Area in Kade, GhanaInternational Journal of Food ScienceRead source →} Ghanaian mining-area roots carried mercury of 120–430 µg/kg, far above the European monitoring results, where most mercury results were below detection.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}^{3A. Adjorlolo-Gasokpoh, A. A. Golow, J. Kambo-Dorsa · 2012Mercury in the Surface Soil and Cassava, Manihot esculenta (Flesh, Leaves and Peel) Near Goldmines at Bogoso and Prestea, GhanaBull Environ Contam ToxicolRead source →}^{6Cecilia Owusu-Agyemang Fobi, Guy Eshun, Twumasi Ankrah Kwarteng, Afia Sakyiwaa Amponsah, Prince Owusu Adoma · 2025Effect of Boiling on Heavy Metal and Cyanide Concentrations and Associated Health Risks in Cassava and Cocoyam From a Gold Mining Area in Kade, GhanaInternational Journal of Food ScienceRead source →} Because the mining studies report dry-weight or unstated bases while monitoring reports fresh weight, these are not rated against the market references.^{3A. Adjorlolo-Gasokpoh, A. A. Golow, J. Kambo-Dorsa · 2012Mercury in the Surface Soil and Cassava, Manihot esculenta (Flesh, Leaves and Peel) Near Goldmines at Bogoso and Prestea, GhanaBull Environ Contam ToxicolRead source →}^{4Adjei-Mensah R, Ofori H, Tortoe C, Johnson PT, Aryee D, Frimpong SK · 2021Effect of home processing methods on the levels of heavy metal contaminants in four food crops grown in and around two mining towns in GhanaToxicology Reports 8 (2021) 1830–1838Read source →}^{6Cecilia Owusu-Agyemang Fobi, Guy Eshun, Twumasi Ankrah Kwarteng, Afia Sakyiwaa Amponsah, Prince Owusu Adoma · 2025Effect of Boiling on Heavy Metal and Cyanide Concentrations and Associated Health Risks in Cassava and Cocoyam From a Gold Mining Area in Kade, GhanaInternational Journal of Food ScienceRead source →}

For cadmium, monitored cassava root is consistently low, with a European median below 4 µg/kg and a maximum of 22 µg/kg.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →} Nigerian flours measured higher, 34–65 µg/kg in market lafun and about 60 µg/kg dry weight in laboratory flour from one farm, so the cadmium evidence for flour comes from few studies in one country.^{9Albishi TS, Adaigho D, Alqahtani NK, Sami R, Uguru H, Helal M, Alaoufi SH, Alkaltham MS, Osiga-Aibangbee D, Okolotu GI, Emumejaye EP, Qari SH · 2026In vitro evaluation of bio-fortification effects on the nutritional quality, toxicological safety, and antioxidant of cassava (Manihot esculenta) flour, and environmental safety of processing water, using natural additivesFrontiers in NutritionRead source →}^{10EDITED BY Arun K. Bhunia, Purdue University, United States REVIEWED BY Marco E. Mng’ong’o, sold in selected markets—taken Mbeya University of Science, Technology · 2023OPEN ACCESS potential toxic levels of cyanidePeer-reviewed articleRead source →}

Which form of cassava measured lowest in lead?

Starch and flour submitted to GEMS/Food from the Americas were lowest, at a median of 10 µg/kg, followed by raw root from the same region at 9 µg/kg median.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →} European “tapioca” of unstated form and European root were higher, and Nigerian vendor samples of garri, fufu and amala reported 140–300 µg/kg on an unstated basis.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}^{11Iyabo OO, Ogundiran MB, Adebayo OL · 2015Toxic and Essential Metals in Staple Foods Commonly Consumed by Students in Ekiti State, South West, NigeriaInternational Journal of ChemistryRead source →} These forms come from different markets and laboratories, so the pattern is a lowest observed comparison, not a ranking of processing methods.

Is imported cassava higher in lead than locally grown cassava?

In European GEMS/Food submissions, roots labelled imported had a median of 140 µg/kg lead against 64 µg/kg for those labelled domestic, and 12 of 21 imported roots exceeded 100 µg/kg against 6 of 28 domestic ones.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →} The export does not name the country of growth, so this cannot identify which producing countries are higher.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}

Soaking, drying and cooking

Two kinds of processing evidence exist, and neither is large. In a laboratory experiment with one cultivar from one farm, soaking peeled roots in tap water for 48 hours before drying and milling gave flour about 11% lower in lead (61 against 69 µg/kg dry weight), 8% lower in cadmium and 4% lower in chromium than unsoaked roots.^{9Albishi TS, Adaigho D, Alqahtani NK, Sami R, Uguru H, Helal M, Alaoufi SH, Alkaltham MS, Osiga-Aibangbee D, Okolotu GI, Emumejaye EP, Qari SH · 2026In vitro evaluation of bio-fortification effects on the nutritional quality, toxicological safety, and antioxidant of cassava (Manihot esculenta) flour, and environmental safety of processing water, using natural additivesFrontiers in NutritionRead source →} Soaking in plant-extract solutions with 1% plant powder added to the flour gave larger reductions, down to 21 µg/kg lead and 27 µg/kg cadmium, but the extracts and powders were never tested for metals and the results are repeat analyses of a single lot per treatment.^{9Albishi TS, Adaigho D, Alqahtani NK, Sami R, Uguru H, Helal M, Alaoufi SH, Alkaltham MS, Osiga-Aibangbee D, Okolotu GI, Emumejaye EP, Qari SH · 2026In vitro evaluation of bio-fortification effects on the nutritional quality, toxicological safety, and antioxidant of cassava (Manihot esculenta) flour, and environmental safety of processing water, using natural additivesFrontiers in NutritionRead source →}

Cooking lowered lead and mercury per kilogram in the one study that measured the same roots before and after.^{4Adjei-Mensah R, Ofori H, Tortoe C, Johnson PT, Aryee D, Frimpong SK · 2021Effect of home processing methods on the levels of heavy metal contaminants in four food crops grown in and around two mining towns in GhanaToxicology Reports 8 (2021) 1830–1838Read source →} In two Ghanaian mining towns, peeled cassava averaged 2,110 µg/kg lead; boiling for 30 minutes gave 600–610 µg/kg, frying 70–80 µg/kg and roasting 100 µg/kg in Akwatia, while mercury fell from 120–160 to 50–60 µg/kg when boiled and 20–30 µg/kg when fried, and fell least when roasted (110 µg/kg in both towns).^{4Adjei-Mensah R, Ofori H, Tortoe C, Johnson PT, Aryee D, Frimpong SK · 2021Effect of home processing methods on the levels of heavy metal contaminants in four food crops grown in and around two mining towns in GhanaToxicology Reports 8 (2021) 1830–1838Read source →} The results were not corrected for water lost or gained or oil taken up during cooking, and heat does not destroy metals, so part of the drop may reflect transfer to cooking water or dilution rather than removal.^{4Adjei-Mensah R, Ofori H, Tortoe C, Johnson PT, Aryee D, Frimpong SK · 2021Effect of home processing methods on the levels of heavy metal contaminants in four food crops grown in and around two mining towns in GhanaToxicology Reports 8 (2021) 1830–1838Read source →}

A second Ghanaian study reports large reductions in mercury and cadmium after boiling, but it measured raw cassava on a dry-weight basis and does not state the basis for boiled cassava, so its raw-to-boiled differences cannot be read as a cooking effect.^{6Cecilia Owusu-Agyemang Fobi, Guy Eshun, Twumasi Ankrah Kwarteng, Afia Sakyiwaa Amponsah, Prince Owusu Adoma · 2025Effect of Boiling on Heavy Metal and Cyanide Concentrations and Associated Health Risks in Cassava and Cocoyam From a Gold Mining Area in Kade, GhanaInternational Journal of Food ScienceRead source →} African total diet monitoring adds that cooked dried cassava had a median of 82 µg/kg lead against 35 µg/kg for cooked fresh cassava in 2018 submissions; these are different foods sampled as eaten.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}

Does soaking cassava remove heavy metals?

Only partly, and the evidence is one laboratory experiment. Soaking peeled roots in tap water for 48 hours lowered flour lead by about 11% and cadmium by 8%, and the metals appeared in the soaking water.^{9Albishi TS, Adaigho D, Alqahtani NK, Sami R, Uguru H, Helal M, Alaoufi SH, Alkaltham MS, Osiga-Aibangbee D, Okolotu GI, Emumejaye EP, Qari SH · 2026In vitro evaluation of bio-fortification effects on the nutritional quality, toxicological safety, and antioxidant of cassava (Manihot esculenta) flour, and environmental safety of processing water, using natural additivesFrontiers in NutritionRead source →} No study measured the traditional long fermentation used for fufu or lafun against unfermented flour.

Mitigation from field to flour

Controls for metals in cassava act at the field, during flour processing and at lot release. The only measured processing comparison is laboratory soaking, which lowered lead by about a tenth in water alone; the larger plant-extract effects come from one experiment without a check on the added materials.^{9Albishi TS, Adaigho D, Alqahtani NK, Sami R, Uguru H, Helal M, Alaoufi SH, Alkaltham MS, Osiga-Aibangbee D, Okolotu GI, Emumejaye EP, Qari SH · 2026In vitro evaluation of bio-fortification effects on the nutritional quality, toxicological safety, and antioxidant of cassava (Manihot esculenta) flour, and environmental safety of processing water, using natural additivesFrontiers in NutritionRead source →} Field evidence is observational: cassava from a mining-waste orchard stayed below detection for lead and arsenic but exceeded the EU cadmium limit for tropical roots.^{2European Commission · 2023Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →}^{5Romero-Crespo P, Jiménez-Oyola S, Salgado-Almeida B, Zambrano-Anchundia J, Goyburo-Chávez C, González-Valoys A, Higueras P · 2023Trace elements in farmland soils and crops, and probabilistic health risk assessment in areas influenced by mining activity in EcuadorEnvironmental Geochemistry and HealthRead source →}

Boiling or frying peeled roots lowered lead and mercury per kilogram in one laboratory comparison from Ghanaian mining towns, but peeling did not remove mercury, which was present in the flesh.^{3A. Adjorlolo-Gasokpoh, A. A. Golow, J. Kambo-Dorsa · 2012Mercury in the Surface Soil and Cassava, Manihot esculenta (Flesh, Leaves and Peel) Near Goldmines at Bogoso and Prestea, GhanaBull Environ Contam ToxicolRead source →}^{4Adjei-Mensah R, Ofori H, Tortoe C, Johnson PT, Aryee D, Frimpong SK · 2021Effect of home processing methods on the levels of heavy metal contaminants in four food crops grown in and around two mining towns in GhanaToxicology Reports 8 (2021) 1830–1838Read source →} The measured cooking effects come from one study without moisture correction, so they are tentative.^{4Adjei-Mensah R, Ofori H, Tortoe C, Johnson PT, Aryee D, Frimpong SK · 2021Effect of home processing methods on the levels of heavy metal contaminants in four food crops grown in and around two mining towns in GhanaToxicology Reports 8 (2021) 1830–1838Read source →}

Sourcing controls rest on monitoring patterns rather than tested interventions.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →} European lead results were higher for imported roots and before 2013, and testing a lot detects lead but does not remove it.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}^{2European Commission · 2023Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →}

About this table

ControlMaterialEffectsEvidence

Control

Material

Effects

Evidence

Avoid growing cassava on mining waste or contaminated soilcultivation · As reported

Cassava

Cd · Context

tAs · Context

1

Romero-Crespo et al. 2023

HMI source record · Original source

One dataset family. Linked analytes, repeated seasons and reports are not independent studies.

Effects and comparators

CadmiumAbove the EU 50 µg/kg tropical-root limit (one sample)Root cadmium, fresh weightMining-area cassava versus EU limitCassava root, fresh weightContext · No supported numerical treatment contrast is retained for this finding. A direction, market association, source-selection strategy or guidance statement alone does not establish a percentage change.

Total arsenicBelow detection (<200 µg/kg), as in two other cropsRoot arsenic, fresh weightCassava versus other crops on the same landCassava root, fresh weightContext · No supported numerical treatment contrast is retained for this finding. A direction, market association, source-selection strategy or guidance statement alone does not establish a percentage change.

Cassava against eight other crops from the same orchards and against the EU tropical-root cadmium limit; one cassava sample, June 2021.

How magnitude is classified

HMI editorial magnitude bands: Low is a change greater than 0% and below 25%; Moderate is 25% to below 50%; High is 50% or more, relative to that control’s own comparator. Direction is shown separately. These are descriptive display bands, not validated efficacy cutoffs, occurrence-baseline positions, statistical significance, evidence certainty, safety or net benefit. A range spanning bands retains both labels; a qualitative label has no implied percentage.

Control in context

Cassava grown on mining-influenced soil had lead and arsenic below detection but cadmium at 60 µg/kg, above the EU limit for tropical roots, and nickel at 2,350 µg/kg; it was one of three of nine crops grown there with arsenic below detection.

Context	Selected control
Material	Cassava root from a gold-mining-area orchard, Ponce Enríquez, Ecuador
Stage	cultivation
Audience	Ingredient suppliers · Brands
Period	As reported
Conditions	One mining district; one cassava sample.
Tested scale	Field survey
Sample units	Source sampling units; no independent intervention units unless stated.
Trade-offs	Other metals, yield, taste and cost not quantified unless stated.
Verification	Confirm with lot-specific lead and cadmium results on the form sold, using an adequate quantification limit.

Design, applicability and operating conditions

Observational survey. One mining district; one cassava sample.

One study family.

As reported by the source

Reduces metal reaching or remaining in cassava, or detects it before release.

Source measurements and linked outcomes

Analyte	Period	Effect	Converted values	Original means (SD)	Basis	Locator
Cadmium (Cd)	As reported	Above the EU 50 µg/kg tropical-root limit (one sample)	60 → null µg/kg	Absolute values unextracted	Cassava root, fresh weight	Table 2, p. 7
Total arsenic (tAs)	As reported	Below detection (<200 µg/kg), as in two other crops	Numerical baseline not extracted	Absolute values unextracted	Cassava root, fresh weight	Table 2, p. 7

Observational; no siting intervention was tested.

Percentages are calculated from the source means shown in outcomes.

Basis as stated in outcomes; no mass balance reported.

Trade-offs and implementation

Cyanide is outside this table.

Not assessed.

Not quantified.

Not assessed.

Applicability and verification

One mining district; one cassava sample.

Lot testing.

See source.

Extraction limits and next actions

Unresolved field	Next action
controlled field or commercial-scale intervention effect	No controlled field or commercial cassava intervention study in the library; keep effects unquantified unless measured.

Provenance and assessment

As disclosed by the source.

Source facts and HMI calculations are distinguished above. These selected source findings do not constitute a systematic review; outcome-level appraisal remains incomplete.

Limit cadmium-containing fertiliser on cassava fieldscultivation · As reported

Cassava

Cd · Context

Guidance

Jacob et al. 2023

HMI source record · Original source

Official guidance; not an independent experimental dataset. Reused across actions without adding studies.

Effects and comparators

CadmiumGuidance; effect not measuredFlour cadmiumFertiliser practiceFermented cassava flour, as soldContext · No supported numerical treatment contrast is retained for this finding. A direction, market association, source-selection strategy or guidance statement alone does not establish a percentage change.

Guidance; no measured comparison.

How magnitude is classified

HMI editorial magnitude bands: Low is a change greater than 0% and below 25%; Moderate is 25% to below 50%; High is 50% or more, relative to that control’s own comparator. Direction is shown separately. These are descriptive display bands, not validated efficacy cutoffs, occurrence-baseline positions, statistical significance, evidence certainty, safety or net benefit. A range spanning bands retains both labels; a qualitative label has no implied percentage.

Control in context

The authors recommend prohibiting cadmium-containing fertilisers and attribute the cadmium they found to fertiliser use, without measuring it.

Context	Selected control
Material	Fermented cassava flour (lafun), Oke Ogun markets, Nigeria
Stage	cultivation
Audience	Ingredient suppliers · Brands
Period	As reported
Conditions	Applies where phosphate or cadmium-bearing fertilisers are used.
Tested scale	Author guidance
Sample units	Source sampling units; no independent intervention units unless stated.
Trade-offs	Other metals, yield, taste and cost not quantified unless stated.
Verification	Confirm with lot-specific lead and cadmium results on the form sold, using an adequate quantification limit.

Design, applicability and operating conditions

Regulatory or author guidance. Applies where phosphate or cadmium-bearing fertilisers are used.

One study family.

As reported by the source

Reduces metal reaching or remaining in cassava, or detects it before release.

Source measurements and linked outcomes

Analyte	Period	Effect	Converted values	Original means (SD)	Basis	Locator
Cadmium (Cd)	As reported	Guidance; effect not measured	Numerical baseline not extracted	Absolute values unextracted	Fermented cassava flour, as sold	Section 5 discussion; Section 7 recommendations

No measured effect size.

Percentages are calculated from the source means shown in outcomes.

Basis as stated in outcomes; no mass balance reported.

Trade-offs and implementation

Cyanide is outside this table.

Not assessed.

Not quantified.

Not assessed.

Applicability and verification

Applies where phosphate or cadmium-bearing fertilisers are used.

Lot testing.

See source.

Extraction limits and next actions

Unresolved field	Next action
controlled field or commercial-scale intervention effect	No controlled field or commercial cassava intervention study in the library; keep effects unquantified unless measured.

Provenance and assessment

As disclosed by the source.

Source facts and HMI calculations are distinguished above. These selected source findings do not constitute a systematic review; outcome-level appraisal remains incomplete.

Wash and separate the edible part before testing or processingfood processing · As reported

Cassava

Pb · Context

Cd · Context

Guidance

Commission Regulation (EU) 2023/915

HMI source record · Original source

Official guidance; not an independent experimental dataset. Reused across actions without adding studies.

Effects and comparators

LeadRegulatory definition; effect not measuredLead limit basisEdible part after washingFresh root, wet weightContext · No supported numerical treatment contrast is retained for this finding. A direction, market association, source-selection strategy or guidance statement alone does not establish a percentage change.

CadmiumRegulatory definition; effect not measuredCadmium limit basisEdible part after washingFresh root, wet weightContext · No supported numerical treatment contrast is retained for this finding. A direction, market association, source-selection strategy or guidance statement alone does not establish a percentage change.

Regulatory definition; no measured comparison of peeled and unpeeled cassava.

How magnitude is classified

HMI editorial magnitude bands: Low is a change greater than 0% and below 25%; Moderate is 25% to below 50%; High is 50% or more, relative to that control’s own comparator. Direction is shown separately. These are descriptive display bands, not validated efficacy cutoffs, occurrence-baseline positions, statistical significance, evidence certainty, safety or net benefit. A range spanning bands retains both labels; a qualitative label has no implied percentage.

Control in context

EU limits for lead, cadmium and nickel in root and tuber vegetables apply on a wet-weight basis after washing and separating the edible part, which defines where soil is expected to be removed.

Context	Selected control
Material	Fresh cassava and other tropical roots
Stage	food processing
Audience	Ingredient suppliers · Brands · Consumers
Period	As reported
Conditions	Fresh cassava placed on the EU market.
Tested scale	Regulation
Sample units	Source sampling units; no independent intervention units unless stated.
Trade-offs	Other metals, yield, taste and cost not quantified unless stated.
Verification	Confirm with lot-specific lead and cadmium results on the form sold, using an adequate quantification limit.

Design, applicability and operating conditions

Regulatory or author guidance. Fresh cassava placed on the EU market.

One study family.

As reported by the source

Reduces metal reaching or remaining in cassava, or detects it before release.

Source measurements and linked outcomes

Analyte	Period	Effect	Converted values	Original means (SD)	Basis	Locator
Lead (Pb)	As reported	Regulatory definition; effect not measured	Numerical baseline not extracted	Absolute values unextracted	Fresh root, wet weight	Annex I 3.1.2 remarks
Cadmium (Cd)	As reported	Regulatory definition; effect not measured	Numerical baseline not extracted	Absolute values unextracted	Fresh root, wet weight	Annex I 3.2.2 remarks

No measured effect size for peeling or washing cassava.

Percentages are calculated from the source means shown in outcomes.

Basis as stated in outcomes; no mass balance reported.

Trade-offs and implementation

Cyanide is outside this table.

Not assessed.

Not quantified.

Not assessed.

Applicability and verification

Fresh cassava placed on the EU market.

Lot testing.

See source.

Extraction limits and next actions

Unresolved field	Next action
controlled field or commercial-scale intervention effect	No controlled field or commercial cassava intervention study in the library; keep effects unquantified unless measured.

Provenance and assessment

As disclosed by the source.

Source facts and HMI calculations are distinguished above. These selected source findings do not constitute a systematic review; outcome-level appraisal remains incomplete.

Soak peeled roots in clean water before drying and millingfood processing · As reported

Cassava

Pb · Decreased

Cd · Decreased

Cr · Decreased

1

Albishi et al. 2026

HMI source record · Original source

One dataset family. Linked analytes, repeated seasons and reports are not independent studies.

Effects and comparators

LeadDecreasedAbout 11% lowerFlour lead, dry weightWater-soaked versus unsoakedCassava flour, dry weightLow · HMI calculates the relative change from the retained source means: (61.33 − 68.67) / 68.67 × 100 = -10.7%. This is a descriptive contrast of means; no confidence interval or significance test is inferred.

CadmiumDecreasedAbout 8% lowerFlour cadmium, dry weightWater-soaked versus unsoakedCassava flour, dry weightLow · HMI calculates the relative change from the retained source means: (57.67 − 63) / 63 × 100 = -8.5%. This is a descriptive contrast of means; no confidence interval or significance test is inferred.

Total chromiumDecreasedAbout 4% lowerFlour chromium, dry weightWater-soaked versus unsoakedCassava flour, dry weightLow · HMI calculates the relative change from the retained source means: (65 − 68) / 68 × 100 = -4.4%. This is a descriptive contrast of means; no confidence interval or significance test is inferred.

Control B (water-soaked) against Control A (unsoaked), one flour lot each, three repeat determinations.

How magnitude is classified

HMI editorial magnitude bands: Low is a change greater than 0% and below 25%; Moderate is 25% to below 50%; High is 50% or more, relative to that control’s own comparator. Direction is shown separately. These are descriptive display bands, not validated efficacy cutoffs, occurrence-baseline positions, statistical significance, evidence certainty, safety or net benefit. A range spanning bands retains both labels; a qualitative label has no implied percentage.

Control in context

Soaking peeled roots in tap water for 48 hours gave flour about 11% lower in lead, 8% lower in cadmium and 4% lower in chromium than unsoaked roots; the soaking water gained these metals.

Context	Selected control
Material	Laboratory cassava flour, cultivar TME 419, one farm, Delta State, Nigeria
Stage	food processing
Audience	Ingredient suppliers · Brands · Consumers
Period	As reported
Conditions	Laboratory flour from one cultivar, farm and season.
Tested scale	Laboratory
Sample units	Source sampling units; no independent intervention units unless stated.
Trade-offs	Other metals, yield, taste and cost not quantified unless stated.
Verification	Confirm with lot-specific lead and cadmium results on the form sold, using an adequate quantification limit.

Design, applicability and operating conditions

Laboratory processing experiment. Laboratory flour from one cultivar, farm and season.

One study family.

As reported by the source

Reduces metal reaching or remaining in cassava, or detects it before release.

Source measurements and linked outcomes

Analyte	Period	Effect	Converted values	Original means (SD)	Basis	Locator
Lead (Pb)	As reported	About 11% lower	68.67 → 61.33 µg/kg	Absolute values unextracted	Cassava flour, dry weight	Table 6 Control A/B
Cadmium (Cd)	As reported	About 8% lower	63 → 57.67 µg/kg	Absolute values unextracted	Cassava flour, dry weight	Table 6 Control A/B
Total chromium (Cr)	As reported	About 4% lower	68 → 65 µg/kg	Absolute values unextracted	Cassava flour, dry weight	Table 6 Control A/B

Repeat analyses of one lot per arm; not independent samples or a significance test.

Percentages are calculated from the source means shown in outcomes.

Basis as stated in outcomes; no mass balance reported.

Trade-offs and implementation

Cyanide is outside this table.

Not assessed.

Not quantified.

Not assessed.

Applicability and verification

Laboratory flour from one cultivar, farm and season.

Lot testing.

See source.

Extraction limits and next actions

Unresolved field	Next action
controlled field or commercial-scale intervention effect	No controlled field or commercial cassava intervention study in the library; keep effects unquantified unless measured.

Provenance and assessment

As disclosed by the source.

Source facts and HMI calculations are distinguished above. These selected source findings do not constitute a systematic review; outcome-level appraisal remains incomplete.

Soak roots in plant-extract solutions (experimental)food processing · As reported

Cassava

Pb · Decreased

Cd · Decreased

Cr · Decreased

1

Albishi et al. 2026

HMI source record · Original source

One dataset family. Linked analytes, repeated seasons and reports are not independent studies.

Effects and comparators

LeadDecreasedAbout 69% lower (experimental)Flour lead, dry weightF11 versus unsoakedCassava flour, dry weightHigh · HMI calculates the relative change from the retained source means: (21.33 − 68.67) / 68.67 × 100 = -68.9%. This is a descriptive contrast of means; no confidence interval or significance test is inferred.

CadmiumDecreasedAbout 57% lower (experimental)Flour cadmium, dry weightF11 versus unsoakedCassava flour, dry weightHigh · HMI calculates the relative change from the retained source means: (27.33 − 63) / 63 × 100 = -56.6%. This is a descriptive contrast of means; no confidence interval or significance test is inferred.

Total chromiumDecreasedAbout 19% lower (experimental)Flour chromium, dry weightF11 versus unsoakedCassava flour, dry weightLow · HMI calculates the relative change from the retained source means: (55.33 − 68) / 68 × 100 = -18.6%. This is a descriptive contrast of means; no confidence interval or significance test is inferred.

Treatment F11 against Control A (unsoaked), one flour lot each, three repeat determinations.

How magnitude is classified

HMI editorial magnitude bands: Low is a change greater than 0% and below 25%; Moderate is 25% to below 50%; High is 50% or more, relative to that control’s own comparator. Direction is shown separately. These are descriptive display bands, not validated efficacy cutoffs, occurrence-baseline positions, statistical significance, evidence certainty, safety or net benefit. A range spanning bands retains both labels; a qualitative label has no implied percentage.

Control in context

Soaking in clove, turmeric and moringa extracts with 1% plant powder blended into the flour gave the lowest values, down to 21 µg/kg lead and 27 µg/kg cadmium; the added plant materials were not tested for metals.

Context	Selected control
Material	Laboratory cassava flour, cultivar TME 419, one farm, Delta State, Nigeria
Stage	food processing
Audience	Ingredient suppliers · Brands
Period	As reported
Conditions	Laboratory experiment only; not an established or approved processing method.
Tested scale	Laboratory
Sample units	Source sampling units; no independent intervention units unless stated.
Trade-offs	Other metals, yield, taste and cost not quantified unless stated.
Verification	Confirm with lot-specific lead and cadmium results on the form sold, using an adequate quantification limit.

Design, applicability and operating conditions

Laboratory processing experiment. Laboratory experiment only; not an established or approved processing method.

One study family.

As reported by the source

Reduces metal reaching or remaining in cassava, or detects it before release.

Source measurements and linked outcomes

Analyte	Period	Effect	Converted values	Original means (SD)	Basis	Locator
Lead (Pb)	As reported	About 69% lower (experimental)	68.67 → 21.33 µg/kg	Absolute values unextracted	Cassava flour, dry weight	Table 6 F11/Control A
Cadmium (Cd)	As reported	About 57% lower (experimental)	63 → 27.33 µg/kg	Absolute values unextracted	Cassava flour, dry weight	Table 6 F11/Control A
Total chromium (Cr)	As reported	About 19% lower (experimental)	68 → 55.33 µg/kg	Absolute values unextracted	Cassava flour, dry weight	Table 6 F11/Control A

Additives untested for metals; one lot per arm; arsenic results held.

Percentages are calculated from the source means shown in outcomes.

Basis as stated in outcomes; no mass balance reported.

Trade-offs and implementation

Cyanide is outside this table.

Not assessed.

Not quantified.

Not assessed.

Applicability and verification

Laboratory experiment only; not an established or approved processing method.

Lot testing.

See source.

Extraction limits and next actions

Unresolved field	Next action
controlled field or commercial-scale intervention effect	No controlled field or commercial cassava intervention study in the library; keep effects unquantified unless measured.

Provenance and assessment

As disclosed by the source.

Source facts and HMI calculations are distinguished above. These selected source findings do not constitute a systematic review; outcome-level appraisal remains incomplete.

Ask for origin and lot lead results on imported rootssourcing and procurement · As reported

Cassava

Pb · Increased

1

WHO GEMS/Food

HMI source record · Original source

One dataset family. Linked analytes, repeated seasons and reports are not independent studies.

Effects and comparators

LeadIncreasedImported roots higher (observational)Median root leadImported against domestic labelCassava root, as soldHigh · HMI calculates the relative change from the retained source means: (140 − 64) / 64 × 100 = 118.8%. This is a descriptive contrast of means; no confidence interval or significance test is inferred.

Imported against domestic labels within European cassava root submissions, 2008–2019.

How magnitude is classified

HMI editorial magnitude bands: Low is a change greater than 0% and below 25%; Moderate is 25% to below 50%; High is 50% or more, relative to that control’s own comparator. Direction is shown separately. These are descriptive display bands, not validated efficacy cutoffs, occurrence-baseline positions, statistical significance, evidence certainty, safety or net benefit. A range spanning bands retains both labels; a qualitative label has no implied percentage.

Control in context

In European submissions, roots labelled imported had a median of 140 µg/kg lead against 64 µg/kg for those labelled domestic, and 12 of 21 imported roots exceeded 100 µg/kg.

Context	Selected control
Material	Cassava root submitted to GEMS/Food from Europe
Stage	sourcing and procurement
Audience	Ingredient suppliers · Brands
Period	As reported
Conditions	European market submissions; country of growth not identified.
Tested scale	Monitoring dataset
Sample units	Source sampling units; no independent intervention units unless stated.
Trade-offs	Other metals, yield, taste and cost not quantified unless stated.
Verification	Confirm with lot-specific lead and cadmium results on the form sold, using an adequate quantification limit.

Design, applicability and operating conditions

Observational survey. European market submissions; country of growth not identified.

One study family.

As reported by the source

Reduces metal reaching or remaining in cassava, or detects it before release.

Source measurements and linked outcomes

Analyte	Period	Effect	Converted values	Original means (SD)	Basis	Locator
Lead (Pb)	As reported	Imported roots higher (observational)	64 → 140 µg/kg	Absolute values unextracted	Cassava root, as sold	primary/gemsfood-cassava-records.csv EURO root lead

Observational; labels describe origin relative to the reporting country only.

Percentages are calculated from the source means shown in outcomes.

Basis as stated in outcomes; no mass balance reported.

Trade-offs and implementation

Cyanide is outside this table.

Not assessed.

Not quantified.

Not assessed.

Applicability and verification

European market submissions; country of growth not identified.

Lot testing.

See source.

Extraction limits and next actions

Unresolved field	Next action
controlled field or commercial-scale intervention effect	No controlled field or commercial cassava intervention study in the library; keep effects unquantified unless measured.

Provenance and assessment

As disclosed by the source.

Source facts and HMI calculations are distinguished above. These selected source findings do not constitute a systematic review; outcome-level appraisal remains incomplete.

Test lots with a method able to see well below 100 µg/kglot acceptance and release · As reported

Cassava

Pb · Context

Context

WHO GEMS/Food

HMI source record · Original source

Contextual literature cited by this report; not a new intervention dataset.

Effects and comparators

LeadDetection, not removalMethod reporting limitReported limits versus required sensitivityAs soldContext · No supported numerical treatment contrast is retained for this finding. A direction, market association, source-selection strategy or guidance statement alone does not establish a percentage change.

Reporting limits in submissions against the 100 µg/kg lead limit.

How magnitude is classified

HMI editorial magnitude bands: Low is a change greater than 0% and below 25%; Moderate is 25% to below 50%; High is 50% or more, relative to that control’s own comparator. Direction is shown separately. These are descriptive display bands, not validated efficacy cutoffs, occurrence-baseline positions, statistical significance, evidence certainty, safety or net benefit. A range spanning bands retains both labels; a qualitative label has no implied percentage.

Control in context

Laboratories reporting cassava lead from the Americas used limits of 1–5 µg/kg, while many European arsenic non-detects carried limits of 80–300 µg/kg, too high to show whether levels of interest are present; testing detects metals and does not remove them.

Context	Selected control
Material	Cassava root, flour and starch
Stage	lot acceptance and release
Audience	Ingredient suppliers · Brands
Period	As reported
Conditions	Any lot-release testing for cassava.
Tested scale	Laboratory method
Sample units	Source sampling units; no independent intervention units unless stated.
Trade-offs	Other metals, yield, taste and cost not quantified unless stated.
Verification	Confirm with lot-specific lead and cadmium results on the form sold, using an adequate quantification limit.

Design, applicability and operating conditions

Regulatory or author guidance. Any lot-release testing for cassava.

One study family.

As reported by the source

Reduces metal reaching or remaining in cassava, or detects it before release.

Source measurements and linked outcomes

Analyte	Period	Effect	Converted values	Original means (SD)	Basis	Locator
Lead (Pb)	As reported	Detection, not removal	Numerical baseline not extracted	Absolute values unextracted	As sold	primary/gemsfood-cassava-records.csv LOD/LOQ fields

Method fitness only.

Percentages are calculated from the source means shown in outcomes.

Basis as stated in outcomes; no mass balance reported.

Trade-offs and implementation

Cyanide is outside this table.

Not assessed.

Not quantified.

Not assessed.

Applicability and verification

Any lot-release testing for cassava.

Lot testing.

See source.

Extraction limits and next actions

Unresolved field	Next action
controlled field or commercial-scale intervention effect	No controlled field or commercial cassava intervention study in the library; keep effects unquantified unless measured.

Provenance and assessment

As disclosed by the source.

Source facts and HMI calculations are distinguished above. These selected source findings do not constitute a systematic review; outcome-level appraisal remains incomplete.

Boiling peeled cassava before eatinghome preparation · As reported

Cassava

Pb · Decreased

tHg · Decreased

1

Adjei-Mensah et al. 2021

HMI source record · Original source

One dataset family. Linked analytes, repeated seasons and reports are not independent studies.

Effects and comparators

LeadDecreasedLower lead per kilogram, ObuasiCassava leadBoiling versus unprocessed, ObuasiCassava as prepared; basis not statedHigh · HMI calculates the relative change from the retained source means: (610 − 2110) / 2110 × 100 = -71.1%. This is a descriptive contrast of means; no confidence interval or significance test is inferred.

LeadDecreasedLower lead per kilogram, AkwatiaCassava leadBoiling versus unprocessed, AkwatiaCassava as prepared; basis not statedHigh · HMI calculates the relative change from the retained source means: (600 − 2110) / 2110 × 100 = -71.6%. This is a descriptive contrast of means; no confidence interval or significance test is inferred.

Total mercuryDecreasedLower mercury per kilogram, ObuasiCassava total mercuryBoiling versus unprocessed, ObuasiCassava as prepared; basis not statedHigh · HMI calculates the relative change from the retained source means: (60 − 120) / 120 × 100 = -50.0%. This is a descriptive contrast of means; no confidence interval or significance test is inferred.

Total mercuryDecreasedLower mercury per kilogram, AkwatiaCassava total mercuryBoiling versus unprocessed, AkwatiaCassava as prepared; basis not statedHigh · HMI calculates the relative change from the retained source means: (50 − 160) / 160 × 100 = -68.8%. This is a descriptive contrast of means; no confidence interval or significance test is inferred.

Boiling (30 minutes in deionised water) against unprocessed portions of the same peeled roots, Obuasi and Akwatia.

How magnitude is classified

HMI editorial magnitude bands: Low is a change greater than 0% and below 25%; Moderate is 25% to below 50%; High is 50% or more, relative to that control’s own comparator. Direction is shown separately. These are descriptive display bands, not validated efficacy cutoffs, occurrence-baseline positions, statistical significance, evidence certainty, safety or net benefit. A range spanning bands retains both labels; a qualitative label has no implied percentage.

Control in context

Boiling peeled roots in the laboratory lowered lead and mercury per kilogram in both mining towns; results were not corrected for water lost or gained or oil taken up.

Context	Selected control
Material	Peeled cassava from Obuasi and Akwatia mining towns, Ghana, 2019
Stage	home preparation
Audience	Ingredient suppliers · Brands · Consumers
Period	As reported
Conditions	Peeled roots from gold-mining towns; laboratory preparation.
Tested scale	Laboratory
Sample units	Source sampling units; no independent intervention units unless stated.
Trade-offs	Other metals, yield, taste and cost not quantified unless stated.
Verification	Confirm with lot-specific lead and cadmium results on the form sold, using an adequate quantification limit.

Design, applicability and operating conditions

Laboratory processing experiment. Peeled roots from gold-mining towns; laboratory preparation.

One study family.

As reported by the source

Reduces metal reaching or remaining in cassava, or detects it before release.

Source measurements and linked outcomes

Analyte	Period	Effect	Converted values	Original means (SD)	Basis	Locator
Lead (Pb)	As reported	Lower lead per kilogram, Obuasi	2110 → 610 µg/kg	Absolute values unextracted	Cassava as prepared; basis not stated	Tables 2–3, 6
Lead (Pb)	As reported	Lower lead per kilogram, Akwatia	2110 → 600 µg/kg	Absolute values unextracted	Cassava as prepared; basis not stated	Tables 2–3, 6
Total mercury (tHg)	As reported	Lower mercury per kilogram, Obuasi	120 → 60 µg/kg	Absolute values unextracted	Cassava as prepared; basis not stated	Tables 2–3, 6
Total mercury (tHg)	As reported	Lower mercury per kilogram, Akwatia	160 → 50 µg/kg	Absolute values unextracted	Cassava as prepared; basis not stated	Tables 2–3, 6

One study; n per cell not reported; no moisture correction.

Percentages are calculated from the source means shown in outcomes.

Basis as stated in outcomes; no mass balance reported.

Trade-offs and implementation

Cyanide is outside this table.

Not assessed.

Not quantified.

Not assessed.

Applicability and verification

Peeled roots from gold-mining towns; laboratory preparation.

Lot testing.

See source.

Extraction limits and next actions

Unresolved field	Next action
controlled field or commercial-scale intervention effect	No controlled field or commercial cassava intervention study in the library; keep effects unquantified unless measured.

Provenance and assessment

As disclosed by the source.

Source facts and HMI calculations are distinguished above. These selected source findings do not constitute a systematic review; outcome-level appraisal remains incomplete.

Frying peeled cassava before eatinghome preparation · As reported

Cassava

Pb · Decreased

tHg · Decreased

1

Adjei-Mensah et al. 2021

HMI source record · Original source

One dataset family. Linked analytes, repeated seasons and reports are not independent studies.

Effects and comparators

LeadDecreasedLower lead per kilogram, ObuasiCassava leadFrying versus unprocessed, ObuasiCassava as prepared; basis not statedHigh · HMI calculates the relative change from the retained source means: (70 − 2110) / 2110 × 100 = -96.7%. This is a descriptive contrast of means; no confidence interval or significance test is inferred.

LeadDecreasedLower lead per kilogram, AkwatiaCassava leadFrying versus unprocessed, AkwatiaCassava as prepared; basis not statedHigh · HMI calculates the relative change from the retained source means: (80 − 2110) / 2110 × 100 = -96.2%. This is a descriptive contrast of means; no confidence interval or significance test is inferred.

Total mercuryDecreasedLower mercury per kilogram, ObuasiCassava total mercuryFrying versus unprocessed, ObuasiCassava as prepared; basis not statedHigh · HMI calculates the relative change from the retained source means: (20 − 120) / 120 × 100 = -83.3%. This is a descriptive contrast of means; no confidence interval or significance test is inferred.

Total mercuryDecreasedLower mercury per kilogram, AkwatiaCassava total mercuryFrying versus unprocessed, AkwatiaCassava as prepared; basis not statedHigh · HMI calculates the relative change from the retained source means: (30 − 160) / 160 × 100 = -81.3%. This is a descriptive contrast of means; no confidence interval or significance test is inferred.

Frying (25 minutes in oil) against unprocessed portions of the same peeled roots, Obuasi and Akwatia.

How magnitude is classified

HMI editorial magnitude bands: Low is a change greater than 0% and below 25%; Moderate is 25% to below 50%; High is 50% or more, relative to that control’s own comparator. Direction is shown separately. These are descriptive display bands, not validated efficacy cutoffs, occurrence-baseline positions, statistical significance, evidence certainty, safety or net benefit. A range spanning bands retains both labels; a qualitative label has no implied percentage.

Control in context

Frying peeled roots in the laboratory lowered lead and mercury per kilogram in both mining towns; results were not corrected for water lost or gained or oil taken up.

Context	Selected control
Material	Peeled cassava from Obuasi and Akwatia mining towns, Ghana, 2019
Stage	home preparation
Audience	Ingredient suppliers · Brands · Consumers
Period	As reported
Conditions	Peeled roots from gold-mining towns; laboratory preparation.
Tested scale	Laboratory
Sample units	Source sampling units; no independent intervention units unless stated.
Trade-offs	Other metals, yield, taste and cost not quantified unless stated.
Verification	Confirm with lot-specific lead and cadmium results on the form sold, using an adequate quantification limit.

Design, applicability and operating conditions

Laboratory processing experiment. Peeled roots from gold-mining towns; laboratory preparation.

One study family.

As reported by the source

Reduces metal reaching or remaining in cassava, or detects it before release.

Source measurements and linked outcomes

Analyte	Period	Effect	Converted values	Original means (SD)	Basis	Locator
Lead (Pb)	As reported	Lower lead per kilogram, Obuasi	2110 → 70 µg/kg	Absolute values unextracted	Cassava as prepared; basis not stated	Tables 2–3, 6
Lead (Pb)	As reported	Lower lead per kilogram, Akwatia	2110 → 80 µg/kg	Absolute values unextracted	Cassava as prepared; basis not stated	Tables 2–3, 6
Total mercury (tHg)	As reported	Lower mercury per kilogram, Obuasi	120 → 20 µg/kg	Absolute values unextracted	Cassava as prepared; basis not stated	Tables 2–3, 6
Total mercury (tHg)	As reported	Lower mercury per kilogram, Akwatia	160 → 30 µg/kg	Absolute values unextracted	Cassava as prepared; basis not stated	Tables 2–3, 6

One study; n per cell not reported; no moisture correction.

Percentages are calculated from the source means shown in outcomes.

Basis as stated in outcomes; no mass balance reported.

Trade-offs and implementation

Cyanide is outside this table.

Not assessed.

Not quantified.

Not assessed.

Applicability and verification

Peeled roots from gold-mining towns; laboratory preparation.

Lot testing.

See source.

Extraction limits and next actions

Unresolved field	Next action
controlled field or commercial-scale intervention effect	No controlled field or commercial cassava intervention study in the library; keep effects unquantified unless measured.

Provenance and assessment

As disclosed by the source.

Source facts and HMI calculations are distinguished above. These selected source findings do not constitute a systematic review; outcome-level appraisal remains incomplete.

Do not rely on peeling to remove mercuryhome preparation · As reported

Cassava

tHg · Decreased

tHg · Increased

1

Adjorlolo-Gasokpoh et al. 2012

HMI source record · Original source

One dataset family. Linked analytes, repeated seasons and reports are not independent studies.

Effects and comparators

Total mercuryDecreasedOuter skin lower than flesh (Bogoso entrance)Cassava mercury by tissueOuter skin versus fleshDried tissue; basis not statedModerate · HMI calculates the relative change from the retained source means: (66.6 − 127.45) / 127.45 × 100 = -47.7%. This is a descriptive contrast of means; no confidence interval or significance test is inferred.

Total mercuryIncreasedInner cortex higher than flesh (Prestea)Cassava mercury by tissueInner cortex versus fleshDried tissue; basis not statedModerate · HMI calculates the relative change from the retained source means: (195.47 − 152.11) / 152.11 × 100 = 28.5%. This is a descriptive contrast of means; no confidence interval or significance test is inferred.

Outer peel against flesh of the same roots at four sites 1.5–14 km from mines.

How magnitude is classified

HMI editorial magnitude bands: Low is a change greater than 0% and below 25%; Moderate is 25% to below 50%; High is 50% or more, relative to that control’s own comparator. Direction is shown separately. These are descriptive display bands, not validated efficacy cutoffs, occurrence-baseline positions, statistical significance, evidence certainty, safety or net benefit. A range spanning bands retains both labels; a qualitative label has no implied percentage.

Control in context

Peeled flesh held 122–168 µg/kg mercury; the outer skin held less than the flesh at all four sites, so peeling did not remove mercury from the part that is eaten.

Context	Selected control
Material	Cassava near goldmines at Bogoso and Prestea, Ghana
Stage	home preparation
Audience	Ingredient suppliers · Brands · Consumers
Period	As reported
Conditions	Gold-mining areas using mercury amalgamation.
Tested scale	Field survey
Sample units	Source sampling units; no independent intervention units unless stated.
Trade-offs	Other metals, yield, taste and cost not quantified unless stated.
Verification	Confirm with lot-specific lead and cadmium results on the form sold, using an adequate quantification limit.

Design, applicability and operating conditions

Laboratory processing experiment. Gold-mining areas using mercury amalgamation.

One study family.

As reported by the source

Reduces metal reaching or remaining in cassava, or detects it before release.

Source measurements and linked outcomes

Analyte	Period	Effect	Converted values	Original means (SD)	Basis	Locator
Total mercury (tHg)	As reported	Outer skin lower than flesh (Bogoso entrance)	127.45 → 66.6 µg/kg	Absolute values unextracted	Dried tissue; basis not stated	Table 1
Total mercury (tHg)	As reported	Inner cortex higher than flesh (Prestea)	152.11 → 195.47 µg/kg	Absolute values unextracted	Dried tissue; basis not stated	Table 1

Four sites; number of independent plants unclear.

Percentages are calculated from the source means shown in outcomes.

Basis as stated in outcomes; no mass balance reported.

Trade-offs and implementation

Cyanide is outside this table.

Not assessed.

Not quantified.

Not assessed.

Applicability and verification

Gold-mining areas using mercury amalgamation.

Lot testing.

See source.

Extraction limits and next actions

Unresolved field	Next action
controlled field or commercial-scale intervention effect	No controlled field or commercial cassava intervention study in the library; keep effects unquantified unless measured.

Provenance and assessment

As disclosed by the source.

Source facts and HMI calculations are distinguished above. These selected source findings do not constitute a systematic review; outcome-level appraisal remains incomplete.

Boil cassava from mining areas (basis caveat)home preparation · As reported

Cassava

tHg · Context

Cd · Context

1

Fobi et al. 2025

HMI source record · Original source

One dataset family. Linked analytes, repeated seasons and reports are not independent studies.

Effects and comparators

Total mercuryEffect size not established (bases differ)Cassava mercuryBoiled versus rawRaw: dry weight; boiled: not statedContext · No supported numerical treatment contrast is retained for this finding. A direction, market association, source-selection strategy or guidance statement alone does not establish a percentage change.

CadmiumEffect size not established (bases differ)Cassava cadmiumBoiled versus rawRaw: dry weight; boiled: not statedContext · No supported numerical treatment contrast is retained for this finding. A direction, market association, source-selection strategy or guidance statement alone does not establish a percentage change.

Boiled against raw cassava from the same sampling points; bases differ.

How magnitude is classified

HMI editorial magnitude bands: Low is a change greater than 0% and below 25%; Moderate is 25% to below 50%; High is 50% or more, relative to that control’s own comparator. Direction is shown separately. These are descriptive display bands, not validated efficacy cutoffs, occurrence-baseline positions, statistical significance, evidence certainty, safety or net benefit. A range spanning bands retains both labels; a qualitative label has no implied percentage.

Control in context

The authors report large drops in mercury and cadmium after boiling, but raw cassava was measured on a dry-weight basis and the boiled basis is not stated, so the size of the effect cannot be established.

Context	Selected control
Material	Cassava from six points in the Kade gold-mining area, Ghana, May 2024
Stage	home preparation
Audience	Ingredient suppliers · Brands
Period	As reported
Conditions	Gold-mining area; one composite per sampling point.
Tested scale	Field and laboratory
Sample units	Source sampling units; no independent intervention units unless stated.
Trade-offs	Other metals, yield, taste and cost not quantified unless stated.
Verification	Confirm with lot-specific lead and cadmium results on the form sold, using an adequate quantification limit.

Design, applicability and operating conditions

Laboratory processing experiment. Gold-mining area; one composite per sampling point.

One study family.

As reported by the source

Reduces metal reaching or remaining in cassava, or detects it before release.

Source measurements and linked outcomes

Analyte	Period	Effect	Converted values	Original means (SD)	Basis	Locator
Total mercury (tHg)	As reported	Effect size not established (bases differ)	301 → null µg/kg	Absolute values unextracted	Raw: dry weight; boiled: not stated	Tables 4–5
Cadmium (Cd)	As reported	Effect size not established (bases differ)	28 → null µg/kg	Absolute values unextracted	Raw: dry weight; boiled: not stated	Tables 4–5

Basis mismatch between raw and boiled results; no detection limits.

Percentages are calculated from the source means shown in outcomes.

Basis as stated in outcomes; no mass balance reported.

Trade-offs and implementation

Cyanide is outside this table.

Not assessed.

Not quantified.

Not assessed.

Applicability and verification

Gold-mining area; one composite per sampling point.

Lot testing.

See source.

Extraction limits and next actions

Unresolved field	Next action
controlled field or commercial-scale intervention effect	No controlled field or commercial cassava intervention study in the library; keep effects unquantified unless measured.

Provenance and assessment

As disclosed by the source.

Source facts and HMI calculations are distinguished above. These selected source findings do not constitute a systematic review; outcome-level appraisal remains incomplete.

Trade-offs and limits of controls

Soaking that leaches metals also moves them into the soaking water, which then needs to be treated as waste rather than reused.^{9Albishi TS, Adaigho D, Alqahtani NK, Sami R, Uguru H, Helal M, Alaoufi SH, Alkaltham MS, Osiga-Aibangbee D, Okolotu GI, Emumejaye EP, Qari SH · 2026In vitro evaluation of bio-fortification effects on the nutritional quality, toxicological safety, and antioxidant of cassava (Manihot esculenta) flour, and environmental safety of processing water, using natural additivesFrontiers in NutritionRead source →} The plant-extract treatments added clove, turmeric and moringa material that was not tested for lead, so a lower measured value in the flour could hide a contribution from the additive.^{9Albishi TS, Adaigho D, Alqahtani NK, Sami R, Uguru H, Helal M, Alaoufi SH, Alkaltham MS, Osiga-Aibangbee D, Okolotu GI, Emumejaye EP, Qari SH · 2026In vitro evaluation of bio-fortification effects on the nutritional quality, toxicological safety, and antioxidant of cassava (Manihot esculenta) flour, and environmental safety of processing water, using natural additivesFrontiers in NutritionRead source →}

Cooking changes the food as well as its metal content. Frying gave the lowest lead and mercury per kilogram in the Ghanaian comparison, but it adds oil and removes water in ways the study did not measure, and boiling moves metals into water that is then discarded.^{4Adjei-Mensah R, Ofori H, Tortoe C, Johnson PT, Aryee D, Frimpong SK · 2021Effect of home processing methods on the levels of heavy metal contaminants in four food crops grown in and around two mining towns in GhanaToxicology Reports 8 (2021) 1830–1838Read source →}

Choosing a crop for contaminated land involves trade-offs between metals. In the Ecuadorian mining area cassava was one of three crops with arsenic below detection but carried more nickel than most of the nine crops and cadmium above the EU tropical-root limit.^{2European Commission · 2023Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →}^{5Romero-Crespo P, Jiménez-Oyola S, Salgado-Almeida B, Zambrano-Anchundia J, Goyburo-Chávez C, González-Valoys A, Higueras P · 2023Trace elements in farmland soils and crops, and probabilistic health risk assessment in areas influenced by mining activity in EcuadorEnvironmental Geochemistry and HealthRead source →} Drying cassava into flour or gari concentrates whatever metal the root carried, so a fresh-weight root result cannot be read directly against a dry product.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}^{9Albishi TS, Adaigho D, Alqahtani NK, Sami R, Uguru H, Helal M, Alaoufi SH, Alkaltham MS, Osiga-Aibangbee D, Okolotu GI, Emumejaye EP, Qari SH · 2026In vitro evaluation of bio-fortification effects on the nutritional quality, toxicological safety, and antioxidant of cassava (Manihot esculenta) flour, and environmental safety of processing water, using natural additivesFrontiers in NutritionRead source →}

Testing cassava for metals

A useful cassava lead test must be able to see well below 100 µg/kg.^{2European Commission · 2023Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →}^{7Codex Alimentarius Commission · 1995General Standard for Contaminants and Toxins in Food and Feed (CXS 193-1995)Codex Alimentarius (Joint FAO/WHO Food Standards Programme)Read source →} Laboratories reporting to GEMS/Food from the Americas used limits of 1–5 µg/kg for root lead, but many European arsenic non-detects carried limits of 80–300 µg/kg, too high to show whether arsenic is present at levels of interest.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →} Several Nigerian and Ghanaian studies reported no detection limit, reference material or recovery, one reported non-detects as 0.00 without a limit, and a garri study used X-ray fluorescence that gave implausible chromium and nickel values.^{6Cecilia Owusu-Agyemang Fobi, Guy Eshun, Twumasi Ankrah Kwarteng, Afia Sakyiwaa Amponsah, Prince Owusu Adoma · 2025Effect of Boiling on Heavy Metal and Cyanide Concentrations and Associated Health Risks in Cassava and Cocoyam From a Gold Mining Area in Kade, GhanaInternational Journal of Food ScienceRead source →}^{8AuthorS: Kingsley O. Omeje in common food tubers from Nigeria Benjamin O. Ezema, Sabinus O.O. Eze AFFILIAtIoNS: Pesticide residues · 2023Evaluation of pesticide residues and heavy metalsPeer-reviewed articleRead source →}^{9Albishi TS, Adaigho D, Alqahtani NK, Sami R, Uguru H, Helal M, Alaoufi SH, Alkaltham MS, Osiga-Aibangbee D, Okolotu GI, Emumejaye EP, Qari SH · 2026In vitro evaluation of bio-fortification effects on the nutritional quality, toxicological safety, and antioxidant of cassava (Manihot esculenta) flour, and environmental safety of processing water, using natural additivesFrontiers in NutritionRead source →}^{10EDITED BY Arun K. Bhunia, Purdue University, United States REVIEWED BY Marco E. Mng’ong’o, sold in selected markets—taken Mbeya University of Science, Technology · 2023OPEN ACCESS potential toxic levels of cyanidePeer-reviewed articleRead source →}^{13Christian Ebere Enyoh, Qingyue Wang, Mominul Haque Rabin, Rasheed Oluwafemi Bakare, Joseph Longji Dadiel, Wu Shangrong, Senlin Lu, Ifenna Ilechukwu · 2023Preliminary characterization and probabilistic risk assessment of microplastics and potentially toxic elements (PTEs) in garri (cassava flake), a common staple food consumed in West AfricaEnvironmental Analysis Health and ToxicologyRead source →}

State the form and basis on every result: raw root on a fresh-weight basis, flour or gari as sold, and laboratory flour on a dry-weight basis are not interchangeable.^{2European Commission · 2023Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →}^{9Albishi TS, Adaigho D, Alqahtani NK, Sami R, Uguru H, Helal M, Alaoufi SH, Alkaltham MS, Osiga-Aibangbee D, Okolotu GI, Emumejaye EP, Qari SH · 2026In vitro evaluation of bio-fortification effects on the nutritional quality, toxicological safety, and antioxidant of cassava (Manihot esculenta) flour, and environmental safety of processing water, using natural additivesFrontiers in NutritionRead source →} Arsenic in cassava should be read as total arsenic unless speciated; only six inorganic arsenic results exist, all for starch, and all but one were below detection.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}

Buying and verifying cassava

Ask for lot-specific lead results on the form being bought, with the laboratory’s quantification limit, compared with the destination-market limit.^{2European Commission · 2023Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →}^{7Codex Alimentarius Commission · 1995General Standard for Contaminants and Toxins in Food and Feed (CXS 193-1995)Codex Alimentarius (Joint FAO/WHO Food Standards Programme)Read source →} The EU sets 0.10 mg/kg lead, 0.050 mg/kg cadmium and 0.90 mg/kg nickel for fresh root and tuber vegetables including tropical roots, on a wet-weight basis after washing and separating the edible part.^{2European Commission · 2023Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →} Codex sets 0.1 mg/kg for both lead and cadmium in root and tuber vegetables.^{7Codex Alimentarius Commission · 1995General Standard for Contaminants and Toxins in Food and Feed (CXS 193-1995)Codex Alimentarius (Joint FAO/WHO Food Standards Programme)Read source →} Neither sets a lead or cadmium limit specific to cassava flour, gari or tapioca starch.^{2European Commission · 2023Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →}^{7Codex Alimentarius Commission · 1995General Standard for Contaminants and Toxins in Food and Feed (CXS 193-1995)Codex Alimentarius (Joint FAO/WHO Food Standards Programme)Read source →}

Record where the cassava was grown as well as where it was packed or bought, because monitoring datasets identify only the reporting market and an imported or domestic label.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →} Avoid roots grown on mining waste or in gold-mining districts unless lot testing shows otherwise; Ghanaian and Ecuadorian mining-area cassava carried the highest mercury, lead and cadmium in the evidence.^{3A. Adjorlolo-Gasokpoh, A. A. Golow, J. Kambo-Dorsa · 2012Mercury in the Surface Soil and Cassava, Manihot esculenta (Flesh, Leaves and Peel) Near Goldmines at Bogoso and Prestea, GhanaBull Environ Contam ToxicolRead source →}^{4Adjei-Mensah R, Ofori H, Tortoe C, Johnson PT, Aryee D, Frimpong SK · 2021Effect of home processing methods on the levels of heavy metal contaminants in four food crops grown in and around two mining towns in GhanaToxicology Reports 8 (2021) 1830–1838Read source →}^{5Romero-Crespo P, Jiménez-Oyola S, Salgado-Almeida B, Zambrano-Anchundia J, Goyburo-Chávez C, González-Valoys A, Higueras P · 2023Trace elements in farmland soils and crops, and probabilistic health risk assessment in areas influenced by mining activity in EcuadorEnvironmental Geochemistry and HealthRead source →}^{6Cecilia Owusu-Agyemang Fobi, Guy Eshun, Twumasi Ankrah Kwarteng, Afia Sakyiwaa Amponsah, Prince Owusu Adoma · 2025Effect of Boiling on Heavy Metal and Cyanide Concentrations and Associated Health Risks in Cassava and Cocoyam From a Gold Mining Area in Kade, GhanaInternational Journal of Food ScienceRead source →}

Exposure from cassava

Cassava is a staple in much of Africa, Latin America and Asia, so intake can be large where it is eaten daily, but no source in the ledger estimated lead or cadmium exposure from cassava with verified concentrations. A Colombian study reported a cadmium hazard quotient of 1.32 for cassava in one municipality, but the quotient cannot be reproduced from the reference dose the paper states and its concentrations are only in a supplement that is not in the library.^{16Unknown authors · 2023Health Risk Assessment for Human Exposure to Heavy MetalsPeer-reviewed articleRead source →} A Ghanaian mining-area study reports daily intake estimates for boiled cassava, but they cannot be reproduced from its own tables.^{6Cecilia Owusu-Agyemang Fobi, Guy Eshun, Twumasi Ankrah Kwarteng, Afia Sakyiwaa Amponsah, Prince Owusu Adoma · 2025Effect of Boiling on Heavy Metal and Cyanide Concentrations and Associated Health Risks in Cassava and Cocoyam From a Gold Mining Area in Kade, GhanaInternational Journal of Food ScienceRead source →} In the United States, one children’s dry food whose main ingredient was cassava flour had 44 µg/kg total mercury, which the authors treated as an outlier needing follow-up rather than a property of cassava.^{15Charles A. Barber, Marc Boyer, Jake A. Carter, Patrick J. Gray, Jennifer Fong Sam, Mesay M. Wolle · 2025Toxic elements in baby and young children’s foods in the US and correlation to ingredientsFood Additives & Contaminants: Part BRead source →}

Evidence review and remaining gaps

This assessment covers every source record in the HMI library that names cassava, plus every cassava, manioc and tapioca row in the WHO GEMS/Food exports; each has a recorded decision based on its content.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →} Twelve sources contribute results for cassava itself, including four open-access papers added in a targeted gap search, and six are held in part because their values could not be reconciled with their own methods.^{3A. Adjorlolo-Gasokpoh, A. A. Golow, J. Kambo-Dorsa · 2012Mercury in the Surface Soil and Cassava, Manihot esculenta (Flesh, Leaves and Peel) Near Goldmines at Bogoso and Prestea, GhanaBull Environ Contam ToxicolRead source →}^{4Adjei-Mensah R, Ofori H, Tortoe C, Johnson PT, Aryee D, Frimpong SK · 2021Effect of home processing methods on the levels of heavy metal contaminants in four food crops grown in and around two mining towns in GhanaToxicology Reports 8 (2021) 1830–1838Read source →}^{6Cecilia Owusu-Agyemang Fobi, Guy Eshun, Twumasi Ankrah Kwarteng, Afia Sakyiwaa Amponsah, Prince Owusu Adoma · 2025Effect of Boiling on Heavy Metal and Cyanide Concentrations and Associated Health Risks in Cassava and Cocoyam From a Gold Mining Area in Kade, GhanaInternational Journal of Food ScienceRead source →}^{9Albishi TS, Adaigho D, Alqahtani NK, Sami R, Uguru H, Helal M, Alaoufi SH, Alkaltham MS, Osiga-Aibangbee D, Okolotu GI, Emumejaye EP, Qari SH · 2026In vitro evaluation of bio-fortification effects on the nutritional quality, toxicological safety, and antioxidant of cassava (Manihot esculenta) flour, and environmental safety of processing water, using natural additivesFrontiers in NutritionRead source →}^{10EDITED BY Arun K. Bhunia, Purdue University, United States REVIEWED BY Marco E. Mng’ong’o, sold in selected markets—taken Mbeya University of Science, Technology · 2023OPEN ACCESS potential toxic levels of cyanidePeer-reviewed articleRead source →}^{11Iyabo OO, Ogundiran MB, Adebayo OL · 2015Toxic and Essential Metals in Staple Foods Commonly Consumed by Students in Ekiti State, South West, NigeriaInternational Journal of ChemistryRead source →}^{12Bortey-Sam N, Nakayama SMM, Akoto O, Ikenaka Y, Fobil JN, Baidoo E, Mizukawa H, Ishizuka M · 2015Accumulation of Heavy Metals and Metalloid in Foodstuffs from Agricultural Soils around Tarkwa Area in Ghana, and Associated Human Health RisksInternational Journal of Environmental Research and Public Health (2015)Read source →}^{13Christian Ebere Enyoh, Qingyue Wang, Mominul Haque Rabin, Rasheed Oluwafemi Bakare, Joseph Longji Dadiel, Wu Shangrong, Senlin Lu, Ifenna Ilechukwu · 2023Preliminary characterization and probabilistic risk assessment of microplastics and potentially toxic elements (PTEs) in garri (cassava flake), a common staple food consumed in West AfricaEnvironmental Analysis Health and ToxicologyRead source →}^{14Orisakwe OE, Nduka JK, Amadi CN, Dike DO, Bede O · 2012Heavy metals health risk assessment for population via consumption of food crops and fruits in Owerri, South Eastern, NigeriaChemistry Central Journal 2012, 6:77Read source →} Remaining gaps are the concentrations behind a Colombian exposure study, held only in its supplement, and several Nigerian cassava surveys cited in reviews that are not yet in the library.^{16Unknown authors · 2023Health Risk Assessment for Human Exposure to Heavy MetalsPeer-reviewed articleRead source →}

References identify the sources used in these answers. The Sources index preserves additional discovery links; inclusion there does not mean a source has been reviewed for this assessment. Corrections, completed searches, new direct evidence and changes in requirements should reopen the relevant answer.

Cassava FAQs

What heavy metals are found in cassava?

Lead is the metal most often found at levels that matter; cadmium, arsenic and mercury are usually low or below detection in monitored cassava.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →} Nickel and chromium have been measured only in a few small studies, with one mining-area root reaching 2,350 µg/kg nickel.^{5Romero-Crespo P, Jiménez-Oyola S, Salgado-Almeida B, Zambrano-Anchundia J, Goyburo-Chávez C, González-Valoys A, Higueras P · 2023Trace elements in farmland soils and crops, and probabilistic health risk assessment in areas influenced by mining activity in EcuadorEnvironmental Geochemistry and HealthRead source →}^{8AuthorS: Kingsley O. Omeje in common food tubers from Nigeria Benjamin O. Ezema, Sabinus O.O. Eze AFFILIAtIoNS: Pesticide residues · 2023Evaluation of pesticide residues and heavy metalsPeer-reviewed articleRead source →}^{9Albishi TS, Adaigho D, Alqahtani NK, Sami R, Uguru H, Helal M, Alaoufi SH, Alkaltham MS, Osiga-Aibangbee D, Okolotu GI, Emumejaye EP, Qari SH · 2026In vitro evaluation of bio-fortification effects on the nutritional quality, toxicological safety, and antioxidant of cassava (Manihot esculenta) flour, and environmental safety of processing water, using natural additivesFrontiers in NutritionRead source →}

How do heavy metals get into cassava?

Cassava takes up metals from soil, and lead can also come from soil clinging to the root or from handling after harvest.^{2European Commission · 2023Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →}^{5Romero-Crespo P, Jiménez-Oyola S, Salgado-Almeida B, Zambrano-Anchundia J, Goyburo-Chávez C, González-Valoys A, Higueras P · 2023Trace elements in farmland soils and crops, and probabilistic health risk assessment in areas influenced by mining activity in EcuadorEnvironmental Geochemistry and HealthRead source →} On mining-waste soil in Ecuador cassava took up cadmium and nickel but stayed below detection for arsenic and lead.^{5Romero-Crespo P, Jiménez-Oyola S, Salgado-Almeida B, Zambrano-Anchundia J, Goyburo-Chávez C, González-Valoys A, Higueras P · 2023Trace elements in farmland soils and crops, and probabilistic health risk assessment in areas influenced by mining activity in EcuadorEnvironmental Geochemistry and HealthRead source →}

Which forms of cassava have the lowest levels?

For lead, starch and flour submitted from the Americas were lowest at a median of 10 µg/kg, while European tapioca of unstated form was highest at 115 µg/kg.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →} For cadmium, monitored root was lowest, below 4 µg/kg median, and Nigerian flours measured 34–65 µg/kg.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}^{10EDITED BY Arun K. Bhunia, Purdue University, United States REVIEWED BY Marco E. Mng’ong’o, sold in selected markets—taken Mbeya University of Science, Technology · 2023OPEN ACCESS potential toxic levels of cyanidePeer-reviewed articleRead source →}

Which countries or regions produce cassava with the lowest lead?

Cassava root submitted to GEMS/Food from the Americas had the lowest lead, a median of 9 µg/kg, against 46–92 µg/kg in European submissions.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →} These are reporting markets, mostly of imported roots, so the evidence cannot name the producing countries with the lowest lead.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}

Do foods made with cassava also contain heavy metals?

Yes, at levels that depend on the food: Nigerian market lafun carried 28–53 µg/kg lead, and vendor garri, fufu and amala 140–300 µg/kg on an unstated basis.^{10EDITED BY Arun K. Bhunia, Purdue University, United States REVIEWED BY Marco E. Mng’ong’o, sold in selected markets—taken Mbeya University of Science, Technology · 2023OPEN ACCESS potential toxic levels of cyanidePeer-reviewed articleRead source →}^{11Iyabo OO, Ogundiran MB, Adebayo OL · 2015Toxic and Essential Metals in Staple Foods Commonly Consumed by Students in Ekiti State, South West, NigeriaInternational Journal of ChemistryRead source →} One US children’s dry food made mainly from cassava flour had 44 µg/kg mercury, which its authors treated as an outlier.^{15Charles A. Barber, Marc Boyer, Jake A. Carter, Patrick J. Gray, Jennifer Fong Sam, Mesay M. Wolle · 2025Toxic elements in baby and young children’s foods in the US and correlation to ingredientsFood Additives & Contaminants: Part BRead source →}

What do the levels in cassava mean for health?

No source in the ledger estimated exposure from cassava with verified concentrations, so the evidence cannot say how much cassava contributes to lead or cadmium intake. A Colombian study reported a cadmium hazard quotient above 1 for cassava in one municipality, but the figure cannot be reproduced from its own stated reference dose.^{16Unknown authors · 2023Health Risk Assessment for Human Exposure to Heavy MetalsPeer-reviewed articleRead source →}

Are there legal limits for heavy metals in cassava?

Yes, for fresh root: the EU allows 0.10 mg/kg lead, 0.050 mg/kg cadmium and 0.90 mg/kg nickel in tropical roots and tubers, and Codex sets 0.1 mg/kg for lead and cadmium in root and tuber vegetables.^{2European Commission · 2023Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →}^{7Codex Alimentarius Commission · 1995General Standard for Contaminants and Toxins in Food and Feed (CXS 193-1995)Codex Alimentarius (Joint FAO/WHO Food Standards Programme)Read source →} Neither sets a metal limit specific to cassava flour, gari or tapioca, although both regulate cyanide in those products.^{2European Commission · 2023Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →}^{7Codex Alimentarius Commission · 1995General Standard for Contaminants and Toxins in Food and Feed (CXS 193-1995)Codex Alimentarius (Joint FAO/WHO Food Standards Programme)Read source →}

Can processing reduce heavy metals in cassava?

Modestly, in the only measured case: soaking peeled roots in water for 48 hours lowered lead in the resulting flour by about 11% and cadmium by 8%.^{9Albishi TS, Adaigho D, Alqahtani NK, Sami R, Uguru H, Helal M, Alaoufi SH, Alkaltham MS, Osiga-Aibangbee D, Okolotu GI, Emumejaye EP, Qari SH · 2026In vitro evaluation of bio-fortification effects on the nutritional quality, toxicological safety, and antioxidant of cassava (Manihot esculenta) flour, and environmental safety of processing water, using natural additivesFrontiers in NutritionRead source →} Larger reductions with plant-extract soaks come from one laboratory lot and untested additives, so they are not an established method.^{9Albishi TS, Adaigho D, Alqahtani NK, Sami R, Uguru H, Helal M, Alaoufi SH, Alkaltham MS, Osiga-Aibangbee D, Okolotu GI, Emumejaye EP, Qari SH · 2026In vitro evaluation of bio-fortification effects on the nutritional quality, toxicological safety, and antioxidant of cassava (Manihot esculenta) flour, and environmental safety of processing water, using natural additivesFrontiers in NutritionRead source →}

Does cooking cassava get rid of heavy metals?

Heat does not destroy metals, but boiling and frying lowered lead and mercury per kilogram in one study of peeled roots from Ghanaian mining towns: lead fell from 2,110 µg/kg to about 600 µg/kg when boiled and 70–80 µg/kg when fried.^{4Adjei-Mensah R, Ofori H, Tortoe C, Johnson PT, Aryee D, Frimpong SK · 2021Effect of home processing methods on the levels of heavy metal contaminants in four food crops grown in and around two mining towns in GhanaToxicology Reports 8 (2021) 1830–1838Read source →} The study did not correct for water or oil, so some of the drop may be dilution or transfer to cooking water; roasting lowered mercury least.^{4Adjei-Mensah R, Ofori H, Tortoe C, Johnson PT, Aryee D, Frimpong SK · 2021Effect of home processing methods on the levels of heavy metal contaminants in four food crops grown in and around two mining towns in GhanaToxicology Reports 8 (2021) 1830–1838Read source →}

Does peeling cassava remove mercury?

Not reliably: near Ghanaian goldmines, the peeled flesh held 122–168 µg/kg mercury, and the outer skin held less than the flesh at all four sites.^{3A. Adjorlolo-Gasokpoh, A. A. Golow, J. Kambo-Dorsa · 2012Mercury in the Surface Soil and Cassava, Manihot esculenta (Flesh, Leaves and Peel) Near Goldmines at Bogoso and Prestea, GhanaBull Environ Contam ToxicolRead source →} Peeling remains standard practice and the EU lead and cadmium limits apply to the edible part after washing, but no study measured lead or cadmium in peel against flesh.^{2European Commission · 2023Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →}

Does soaking cassava overnight help?

Soaking for 48 hours in plain water lowered flour lead and cadmium by about a tenth in a laboratory study, and the metals moved into the water.^{9Albishi TS, Adaigho D, Alqahtani NK, Sami R, Uguru H, Helal M, Alaoufi SH, Alkaltham MS, Osiga-Aibangbee D, Okolotu GI, Emumejaye EP, Qari SH · 2026In vitro evaluation of bio-fortification effects on the nutritional quality, toxicological safety, and antioxidant of cassava (Manihot esculenta) flour, and environmental safety of processing water, using natural additivesFrontiers in NutritionRead source →} Overnight soaking and traditional fermentation have not been measured for metals.

Is the cyanide in cassava a heavy metal?

No; cyanide comes from natural compounds in the plant and is controlled separately, for example at 50 mg/kg in fresh peeled root and 10 mg/kg in cassava and tapioca flour in the EU.^{2European Commission · 2023Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →} Proper processing to remove cyanide does not by itself show anything about lead or cadmium.

Has lead in cassava been falling?

In European GEMS/Food submissions, 17 of 39 root results from 2008–2012 exceeded 100 µg/kg against 1 of 11 from 2013–2019.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →} These are national monitoring submissions with changing coverage, not a fixed panel of producers.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}

Is tapioca lower in metals than cassava root?

Starch labelled as cassava or tapioca starch from the Americas was low in lead (median 10 µg/kg), similar to root from the same region.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →} European results labelled only “Tapioca” were higher (median 115 µg/kg), but their form is not stated, so the evidence does not show that starch is reliably lower.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}

Is cassava grown near gold mines higher in metals?

Yes, in the evidence available: cassava from Ghanaian gold-mining districts carried 120–430 µg/kg mercury near Bogoso and Kade, peeled roots from two mining towns averaged 2,110 µg/kg lead, about 30% of roots around Tarkwa had 370–450 µg/kg lead, and an Ecuadorian mining-orchard root had 60 µg/kg cadmium, above the EU limit of 50 µg/kg for tropical roots.^{2European Commission · 2023Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →}^{3A. Adjorlolo-Gasokpoh, A. A. Golow, J. Kambo-Dorsa · 2012Mercury in the Surface Soil and Cassava, Manihot esculenta (Flesh, Leaves and Peel) Near Goldmines at Bogoso and Prestea, GhanaBull Environ Contam ToxicolRead source →}^{4Adjei-Mensah R, Ofori H, Tortoe C, Johnson PT, Aryee D, Frimpong SK · 2021Effect of home processing methods on the levels of heavy metal contaminants in four food crops grown in and around two mining towns in GhanaToxicology Reports 8 (2021) 1830–1838Read source →}^{5Romero-Crespo P, Jiménez-Oyola S, Salgado-Almeida B, Zambrano-Anchundia J, Goyburo-Chávez C, González-Valoys A, Higueras P · 2023Trace elements in farmland soils and crops, and probabilistic health risk assessment in areas influenced by mining activity in EcuadorEnvironmental Geochemistry and HealthRead source →}^{6Cecilia Owusu-Agyemang Fobi, Guy Eshun, Twumasi Ankrah Kwarteng, Afia Sakyiwaa Amponsah, Prince Owusu Adoma · 2025Effect of Boiling on Heavy Metal and Cyanide Concentrations and Associated Health Risks in Cassava and Cocoyam From a Gold Mining Area in Kade, GhanaInternational Journal of Food ScienceRead source →}^{12Bortey-Sam N, Nakayama SMM, Akoto O, Ikenaka Y, Fobil JN, Baidoo E, Mizukawa H, Ishizuka M · 2015Accumulation of Heavy Metals and Metalloid in Foodstuffs from Agricultural Soils around Tarkwa Area in Ghana, and Associated Human Health RisksInternational Journal of Environmental Research and Public Health (2015)Read source →} These studies used different bases and small samples, so roots from mining areas need lot testing rather than an assumed safe level.^{4Adjei-Mensah R, Ofori H, Tortoe C, Johnson PT, Aryee D, Frimpong SK · 2021Effect of home processing methods on the levels of heavy metal contaminants in four food crops grown in and around two mining towns in GhanaToxicology Reports 8 (2021) 1830–1838Read source →}^{6Cecilia Owusu-Agyemang Fobi, Guy Eshun, Twumasi Ankrah Kwarteng, Afia Sakyiwaa Amponsah, Prince Owusu Adoma · 2025Effect of Boiling on Heavy Metal and Cyanide Concentrations and Associated Health Risks in Cassava and Cocoyam From a Gold Mining Area in Kade, GhanaInternational Journal of Food ScienceRead source →}

Are gari and fufu higher in lead than cassava root?

Single vendor samples in Ado-Ekiti, Nigeria, reported 300 µg/kg lead in garri and 150 µg/kg in fufu, but the basis was not stated and the same 300 µg/kg figure is printed for six different tuber foods.^{11Iyabo OO, Ogundiran MB, Adebayo OL · 2015Toxic and Essential Metals in Staple Foods Commonly Consumed by Students in Ekiti State, South West, NigeriaInternational Journal of ChemistryRead source →} A garri study from Nigeria and Japan measured only chromium and nickel, with implausibly high values, so it adds no usable lead or cadmium evidence.^{13Christian Ebere Enyoh, Qingyue Wang, Mominul Haque Rabin, Rasheed Oluwafemi Bakare, Joseph Longji Dadiel, Wu Shangrong, Senlin Lu, Ifenna Ilechukwu · 2023Preliminary characterization and probabilistic risk assessment of microplastics and potentially toxic elements (PTEs) in garri (cassava flake), a common staple food consumed in West AfricaEnvironmental Analysis Health and ToxicologyRead source →}

How can I tell whether a cassava lead test is good enough?

Check that the quantification limit is far below 100 µg/kg; laboratories reporting cassava to GEMS/Food from the Americas used 1–5 µg/kg.^{1World Health Organization · 2026GEMS/Food Contaminants database heavy-metal exportsGEMS/Food Contamination Monitoring and Assessment ProgrammeRead source →}^{2European Commission · 2023Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →} Ask for the form and basis tested, because a dry flour result and a fresh root result are not directly comparable.^{2European Commission · 2023Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006Official Journal of the European UnionRead source →}^{9Albishi TS, Adaigho D, Alqahtani NK, Sami R, Uguru H, Helal M, Alaoufi SH, Alkaltham MS, Osiga-Aibangbee D, Okolotu GI, Emumejaye EP, Qari SH · 2026In vitro evaluation of bio-fortification effects on the nutritional quality, toxicological safety, and antioxidant of cassava (Manihot esculenta) flour, and environmental safety of processing water, using natural additivesFrontiers in NutritionRead 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
Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006European Commission · Official Journal of the European Union · 2023 · eur-lex.europa.euRegulation
Mercury in the Surface Soil and Cassava, Manihot esculenta (Flesh, Leaves and Peel) Near Goldmines at Bogoso and Prestea, GhanaA. Adjorlolo-Gasokpoh, A. A. Golow, and J. Kambo-Dorsa · Bull Environ Contam Toxicol · 2012 · doi.org/10.1007/s00128-012-0849-7Peer-reviewed
Effect of home processing methods on the levels of heavy metal contaminants in four food crops grown in and around two mining towns in GhanaAdjei-Mensah R, Ofori H, Tortoe C, Johnson PT, Aryee D, and Frimpong SK · Toxicology Reports 8 (2021) 1830–1838 · 2021 · doi.org/10.1016/j.toxrep.2021.11.001Peer-reviewed
Trace elements in farmland soils and crops, and probabilistic health risk assessment in areas influenced by mining activity in EcuadorRomero-Crespo P, Jiménez-Oyola S, Salgado-Almeida B, Zambrano-Anchundia J, Goyburo-Chávez C, González-Valoys A, et al. · Environmental Geochemistry and Health · 2023 · doi.org/10.1007/s10653-023-01514-xPeer-reviewed
Effect of Boiling on Heavy Metal and Cyanide Concentrations and Associated Health Risks in Cassava and Cocoyam From a Gold Mining Area in Kade, GhanaCecilia Owusu-Agyemang Fobi, Guy Eshun, Twumasi Ankrah Kwarteng, Afia Sakyiwaa Amponsah, and Prince Owusu Adoma · International Journal of Food Science · 2025 · doi.org/10.1155/ijfo/5137953Peer-reviewed
General Standard for Contaminants and Toxins in Food and Feed (CXS 193-1995)Codex Alimentarius Commission · Codex Alimentarius (Joint FAO/WHO Food Standards Programme) · 1995 · www.fao.orgGovernment
Evaluation of pesticide residues and heavy metalsAuthorS: Kingsley O. Omeje in common food tubers from Nigeria Benjamin O. Ezema and Sabinus O.O. Eze AFFILIAtIoNS: Pesticide residues · Peer-reviewed article · 2023 · doi.org/10.17159/sajs.2024/15969Peer-reviewed
In vitro evaluation of bio-fortification effects on the nutritional quality, toxicological safety, and antioxidant of cassava (Manihot esculenta) flour, and environmental safety of processing water, using natural additivesAlbishi TS, Adaigho D, Alqahtani NK, Sami R, Uguru H, Helal M, et al. · Frontiers in Nutrition · 2026 · doi.org/10.3389/fnut.2025.1731737Peer-reviewed
OPEN ACCESS potential toxic levels of cyanideEDITED BY Arun K. Bhunia, Purdue University, United States REVIEWED BY Marco E. Mng’ong’o, sold in selected markets—taken Mbeya University of Science, and Technology · Peer-reviewed article · 2023 · doi.org/10.3389/fsufs.2023.1165501Peer-reviewed
Toxic and Essential Metals in Staple Foods Commonly Consumed by Students in Ekiti State, South West, NigeriaIyabo OO, Ogundiran MB, and Adebayo OL · International Journal of Chemistry · 2015 · doi.org/10.5539/ijc.v7n1p155Peer-reviewed
Accumulation of Heavy Metals and Metalloid in Foodstuffs from Agricultural Soils around Tarkwa Area in Ghana, and Associated Human Health RisksBortey-Sam N, Nakayama SMM, Akoto O, Ikenaka Y, Fobil JN, Baidoo E, et al. · International Journal of Environmental Research and Public Health (2015) · 2015Peer-reviewed
Preliminary characterization and probabilistic risk assessment of microplastics and potentially toxic elements (PTEs) in garri (cassava flake), a common staple food consumed in West AfricaChristian Ebere Enyoh, Qingyue Wang, Mominul Haque Rabin, Rasheed Oluwafemi Bakare, Joseph Longji Dadiel, Wu Shangrong, et al. · Environmental Analysis Health and Toxicology · 2023 · doi.org/10.5620/eaht.2023005Peer-reviewed
Heavy metals health risk assessment for population via consumption of food crops and fruits in Owerri, South Eastern, NigeriaOrisakwe OE, Nduka JK, Amadi CN, Dike DO, and Bede O · Chemistry Central Journal 2012, 6:77 · 2012 · doi.org/10.1186/1752-153X-6-77Peer-reviewed
Toxic elements in baby and young children’s foods in the US and correlation to ingredientsCharles A. Barber, Marc Boyer, Jake A. Carter, Patrick J. Gray, Jennifer Fong Sam, and Mesay M. Wolle · Food Additives & Contaminants: Part B · 2025 · doi.org/10.1080/19393210.2025.2591749Peer-reviewed
Health Risk Assessment for Human Exposure to Heavy MetalsUnknown authors · Peer-reviewed article · 2023 · doi.org/10.3390/ijerph20010435Peer-reviewed

Sources

Browse Cassava 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 — First published from every cassava record in the library: WHO GEMS/Food monitoring results for root, cooked cassava, flour and starch, five field, market and processing studies, verified limits and stage-led controls.

Referenced by

Vegetables

Graph View

1

2

European Commission

3

4

5

6

7

8

9

10

11

12

13

14

15

Food Additives

Heavy metals in Cassava

Cassava overview

What counts as cassava

Heavy metal contamination profile of cassava

How metals reach cassava

Which forms and markets measured lower

Soaking, drying and cooking

Mitigation from field to flour

Trade-offs and limits of controls

Testing cassava for metals

Buying and verifying cassava

Exposure from cassava

Evidence review and remaining gaps

Cassava FAQs

References

Sources

Update history

Related ingredients

Referenced by

Graph View