Skip to content

EFSA 2024 - Small organoarsenic species in food

EFSA’s 2024 CONTAM opinion is the current EU risk assessment for small organoarsenic species, especially dimethylarsinic acid DMA(V) and monomethylarsonic acid MMA(V), in food. It matters for the wiki because it separates total arsenic, inorganic arsenic, and methylated organic arsenic instead of treating all arsenic in rice, fish, shellfish, and seaweed as equivalent. EFSA found that DMA(V), not MMA(V), is the risk-driving small organoarsenic species under the available data: DMA(V) exposures produce margins of exposure below 10,000 in many dietary surveys, especially for high-percentile infants, toddlers, other children, fish consumers, rice consumers, and rice-based infant-formula scenarios. MMA(V) values can be high in oysters, mussels, smoked herring, canned herring, and some demersal fish scenarios, but EFSA’s calculated MOEs remained well above its low-concern threshold.

Key numbers

Hazard reference points and risk thresholds:

SpeciesCritical endpointReference pointEFSA low-concern MOE screenEFSA conclusion
MMA(V)Decreased body weight linked to diarrhoea in ratsBMDL10 = 18.2 mg MMA(V)/kg bw/day, equivalent to 9.7 mg As/kg bw/day>= 500All modelled MOEs were above 500; no health concern identified
DMA(V)Increased urinary bladder tumours in male ratsBMDL10 = 1.1 mg DMA(V)/kg bw/day, equivalent to 0.6 mg As/kg bw/day>= 10,000Many MOEs below 10,000; EFSA says this raises a health concern
Other small organoarsenic speciesNot resolvedNo RP derivedNot assessedToxicology data insufficient

Occurrence data scope:

Data pointValue
DMA analytical results after cleaning1,260
MMA analytical results after cleaning988
DMA left-censored results44%
MMA left-censored results94%
Countries represented8 EU countries for DMA; 4 for MMA
Dominant sampling countryItaly, 67% of DMA samples and 85% of MMA samples
Sampling years used in occurrence dataMostly 2012-2021; figure includes 2012-2022 distribution
Results with uncertain unit reporting654 DMA and 653 MMA results; EFSA assumed ug As/kg
Thio-analogue occurrence data submittedNone for thio-DMA(V), dithio-DMA(V), or thio-MMA(V)

Selected DMA occurrence values used for exposure assessment, expressed as ug As/kg whole weight:

Food / commodityNLeft-censoredMean LB-UBHigh/reliable percentile noted
Grains and grain-based products, FoodEx2 level 153315%30.4-33.1P95 75.7
Vegetables and vegetable products, FoodEx2 level 117438%127.2-133.2P95 387.3
Fish/seafood/amphibians/reptiles/invertebrates, FoodEx2 level 122741%58.4-62.9P95 not calculated; pooled category
Rice grain, polished221not stated in summary40.0LB=UB
Rice grain, brown110not stated in summary27.8LB=UB
Rice and similar4394%33.5-34.179.2
Rice flour4414%39.3-40.089.4
Puffed/extruded rice textured bread3135%32.6-43.550.2
Rice-based children biscuits/rusks/cookies identified as rice cakes729%22.1-25.223.5
Algae/prokaryotes organisms1169%143-145449
Algae/prokaryotes organisms, dried11210%605-6151,706
Algae-based formulations1984%104-1180-20
Seaweed-containing salads30%1,356Not reported
Marine fish3471%36.5-45.012.0-20.0
Diadromous fish1942%4.9-10.310.0-10.9
Freshwater fish7not stated in table rowNo DMA foundNot reported
Molluscs10224%63.0-66.9170
Canned/marinated/pickled seafood12336%53.0-57.1170
Marinated/pickled fish50%83.7Not robust
Canned/jarred fish1861%10.5-14.7Not robust

Selected MMA occurrence values, expressed as ug As/kg:

Food / commodityNLeft-censoredMean LB-UBNotes
Fish/seafood/amphibians/reptiles/invertebrates, FoodEx2 level 115253%46.3-53.4P95 251
Grains and grain-based products, FoodEx2 level 1463100%0-6.0P95 0-20
Rice samples submitted to EFSA411100%No quantified MMAEFSA says this matches literature
Fish samples submitted to EFSA35100%No quantified MMALiterature data used for fish scenarios
Oysters270%227EFSA occurrence data
Mussels4846%54.0-61.5EFSA occurrence data
Unspecified molluscs9944%88.6-95.6EFSA occurrence data
Miscellaneous demersal marine fish35not stated87.7Literature, CALIPSO/Leblanc 2006
Mackerel20not stated78.0Literature, CALIPSO/Leblanc 2006
Canned herring20 pooled samplesnot stated152Literature, Hackethal et al. 2021
Smoked herring20 pooled samplesnot stated397Literature, Hackethal et al. 2021
Canned mackerel1 compositenot stated82.0Literature, CALIPSO/Leblanc 2006
Fish fingers, breadedderivednot stated2.7From cod/hake/haddock data

Dietary exposure and MOE results:

ScenarioExposure / MOE result
Highest chronic DMA exposure across European surveysToddlers: mean 0.130-0.157 ug As/kg bw/day; P95 0.397-0.477
Adult chronic DMA exposure maximaMean 0.038-0.044 ug As/kg bw/day; P95 0.133-0.158
Rice consumers only, highest DMA mean exposureOther children: 0.076-0.078 ug As/kg bw/day
Rice consumers only, highest DMA P95 exposureOther children: 0.186-0.190 ug As/kg bw/day
Fish-meat consumers only, highest DMA mean exposureToddlers: 0.165-0.205 ug As/kg bw/day
Fish-meat consumers only, highest DMA P95 exposureToddlers: 0.415-0.511 ug As/kg bw/day
Rice-based infant formula, average DMA level and high consumption0.25 ug As/kg bw/day
Rice-based infant formula, maximum DMA level and high consumption0.62 ug As/kg bw/day
DMA MOE range, mean chronic exposure150,000 to 3,800
DMA MOE range, P95 chronic exposure35,300 to 1,300
Rice-based infant formula DMA MOEs3,200, 2,400, 1,250, and 970 across the four scenarios
MMA fish-meat consumers MOEs3,233,000 down to 28,000
MMA mollusc consumers MOEs9,700,000 down to 52,000
MMA processed/preserved fish consumers MOEs9,700,000 down to 28,000
Probability DMA MOE < 10,000, chronic P95 exposureInfants 0.65, toddlers 0.72, other children 0.59
Probability DMA MOE < 10,000, rice-based infant formula0.98 for mean scenario and 1.0 for high scenario
Probability DMA MOE < 10,000, fish consumers P95Infants 0.59, toddlers 0.61, other children 0.53

Main contributors to DMA dietary exposure:

Population groupNotable contributors
InfantsFish meat 3-78% (median 13%); reconstituted infant cereals 0-72% (median 33%)
ToddlersFish meat 14-77% (median 29%); rice 11-65% (median 29%)
Other childrenRice 17-72% (median 43%); fish meat 16-61% (median 30%)
Adults / elderly / very elderlyFish meat and rice dominate; fish-meat contribution in very elderly up to 92% (median 47%)
Adult subgroupsDried/smoked/salted fish up to 27% in elderly; molluscs up to 36% in adults

Methods (brief)

EFSA combined a public-literature review, cleaned occurrence submissions, EFSA Comprehensive European Food Consumption Database surveys, and specific consumers-only scenarios. All occurrence values used for exposure assessment were converted to ug As/kg whole weight. EFSA assumed reported DMA and MMA were the pentavalent species DMA(V) and MMA(V) when providers did not confirm valency, and assumed ug As/kg where unit reporting did not specify whether values were elemental arsenic or species mass. Chronic DMA exposure used 49 dietary surveys from 22 European countries; MMA exposure focused on consumers-only scenarios because nearly all submitted MMA data outside molluscs were left-censored. Occurrence values for some processed commodities were derived from raw primary commodities with EFSA Raw Primary Commodity processing factors and recipes, assuming no DMA loss during processing.

Implications

This source supports arsenic-speciation pages and downstream rows where total arsenic can be misleading: rice and rice products, rice-based infant foods, fish, shellfish, canned/preserved fish, seaweed/kelp foods, algae-based supplements, and rice-based formula-like products. The DMA(V) results are especially important because organic arsenic is often treated as lower concern, yet EFSA’s DMA(V) MOEs fall below 10,000 in many high-exposure scenarios. For rice and seafood matrices, this source should be used alongside inorganic-arsenic sources rather than merged into them: DMA/MMA are not iAs, but they are also not toxicologically irrelevant.

The infant-formula relevance is narrow and should remain visibly qualified. EFSA’s rice-based infant-formula scenario used hydrolysed-rice formula powder data and default infant intake assumptions; it is not evidence for cow-milk formula or soy formula unless the destination page explicitly treats the rice-based product as specialty/dairy-free context.

Wiki pages this source may touch

Verification notes

  • Read the PDF text across the full opinion, with close checks on the summary, occurrence sections 3.6.1-3.6.2, exposure sections 3.7.1-3.7.2, risk-characterisation sections 3.8.1-3.8.2, uncertainty section 3.9.2, conclusions 4.11-4.15, and recommendations.
  • DOI verified from the PDF title page as 10.2903/j.efsa.2024.8844; no matching source page or DOI record was found before ingest.
  • EFSA reports DMA/MMA occurrence as elemental arsenic concentration (ug As/kg) for exposure assessment. Do not reinterpret these as ug DMA/kg or ug MMA/kg.
  • The source uses assumptions that should travel with extracted values: 654 DMA and 653 MMA results had uncertain species-mass versus elemental-arsenic unit reporting; EFSA assumed ug As/kg. EFSA also assumed DMA/MMA were pentavalent species when valency was not confirmed.
  • Thio-DMA(V), dithio-DMA(V), thio-MMA(V), DMA(III), and MMA(III) are important data gaps here, not routeable occurrence findings from the submitted EFSA dataset.
  • Brand names do not appear in this source page. Instrument/software names are omitted because the page is not about validating a specific analytical method.

Page history

The five most recent substantive edits to this page. The full version history lives in git; when DOI minting comes online (see schema docs), each entry below will also link to a version-pinned DataCite DOI.

CommitDateDescription
140e84e2026-06-03refresh manual fetch generated outputs
10b548d2026-06-03repair June 2 tracker: zlotko2021-black-soldier-fly-chitin-nickel-sorption