Cooking does not reduce methylmercury — and why this matters for consumer messaging
A widespread consumer misconception is that thorough cooking — frying, grilling, baking, broiling, smoking, canning, pickling — reduces the methylmercury (MeHg) content of fish. It does not. Methylmercury binds covalently to cysteine residues in muscle protein and survives every conventional cooking method intact. This means species selection at the point of sourcing is the only lever that meaningfully reduces dietary MeHg exposure; no kitchen practice substitutes for it. The certification implication is direct: HMTc consumer-facing messaging on seafood products should explicitly head off the cooking-as-mitigation misconception, because the consumer-protection failure mode is a household that assumes their tuna-heavy diet is safe because they cook it well.
This synthesis page consolidates the mechanism evidence and the comparative measurements across cooked-and-processed forms that establish the finding.
The mechanism
Methylmercury (CH₃Hg⁺) in fish muscle is not free or loosely bound. Approximately 80-95% of total fish-muscle mercury is in the methylated form, and essentially all of that fraction is covalently bound to the thiol (-SH) group of cysteine residues in muscle protein. The Hg-S bond is among the strongest covalent bonds in biochemistry (binding energy ~70 kcal/mol). Heat-denaturing the protein matrix during cooking changes the protein’s tertiary structure but does not cleave the Hg-S bond. The MeHg remains in the cooked muscle, available for intestinal absorption (>90% bioavailability in mammals per Farina et al. 2011) upon consumption.
This mechanism explains why canning, smoking, pickling, and other industrial processes also fail to reduce MeHg. Even autoclave-grade thermal treatment in commercial canning processes (typically 116-121°C for 30-90 minutes) leaves MeHg-protein binding intact. The matrix change is structural, not chemical.
The comparative evidence
BfR 2024 (German Federal Institute for Risk Assessment MEAL study) measured tHg and MeHg in pool samples of ready-to-eat fish across 30+ species and preparation forms. The MeHg/tHg ratio across the entire dataset cluster between 76% and 113% (the >100% values reflect measurement uncertainty at LOQ, not real >100% fractions). Comparing across preparation forms for the same species:
- Cod, baked: 0.02 mg/kg tHg, 0.02 mg/kg MeHg (ratio 95-96%)
- Cod, fish fingers: 0.01 mg/kg tHg, MeHg <LOQ (ratio 76%, below-LOQ noise)
- Eel, fresh: 0.10 mg/kg tHg, 0.10 mg/kg MeHg (ratio 94-96%)
- Eel, smoked: 0.08 mg/kg tHg, 0.08 mg/kg MeHg (ratio 97%)
- Herring, fresh: ~0.05 mg/kg tHg, MeHg in same range
- Herring, smoked: 0.08 mg/kg tHg, 0.07 mg/kg MeHg (ratio 91%)
- Herring, fried/pickled: 0.05 mg/kg tHg, 0.05 mg/kg MeHg (ratio 95%)
- Herring in sauce: 0.04 mg/kg tHg, 0.03 mg/kg MeHg (ratio 77%)
- Herring, pickled (Matjes/Bismarck): 0.03 mg/kg tHg, 0.03 mg/kg MeHg
- Halibut, fresh: 0.08 mg/kg tHg, 0.08 mg/kg MeHg (ratio 103%)
- Halibut, smoked: 0.11 mg/kg tHg, 0.09 mg/kg MeHg (ratio 108%)
The pattern is consistent: smoked vs fresh, pickled vs raw, fried vs baked, canned vs fresh — the MeHg content does not measurably shift with preparation. Where small variations appear (smoked halibut tHg slightly higher than fresh; herring-in-sauce slightly lower than fresh), they reflect lot-level sampling variability across different processing batches, not a process-driven reduction in MeHg per se.
Aguilar-Miranda et al. 2024 documents canned-tuna tHg across 60 samples in three Ecuadorian brands (means 0.14, 0.25, 0.41 mg/kg). The literature range cited by Aguilar-Miranda spans 0.005-1.4725 mg/kg across Latin American canned-tuna studies. Per the same source, ~89% of canned-tuna tHg is MeHg (citing Burger and Gochfeld 2004, Médieu et al. 2022). Canning — a rigorous thermal-and-pressure process — does not change this fraction.
Across the wiki’s broader fish-and-seafood evidence base, the MeHg/tHg ratio in muscle of finfish consistently sits at 80-95% regardless of preparation form. Where preparation does shift the apparent MeHg content (e.g., a sauce that dilutes the per-gram concentration by adding non-MeHg matrix mass), the per-serving MeHg intake is similarly diluted because the consumer is eating a sauce-and-fish mixture, not pure fish.
What does reduce MeHg exposure
Three levers are documented; none of them are cooking practices:
- Species substitution — replacing apex-predator fish (tuna, swordfish, shark, marlin, king mackerel, tilefish-Gulf of Mexico, bigeye tuna) with low-trophic species (salmon, sardine, anchovy, herring, mackerel-Atlantic, tilapia, US-farmed catfish) shifts the per-serving MeHg load by 5-10× to as much as 100× depending on the pair. This is the dominant lever for any HMTc certification standard targeting a meaningful MeHg ceiling.
- Trophic-level differentiation within species — within a single species (e.g., tuna), MeHg burden varies by individual size and age. Specifying skipjack (smaller, shorter-lived) over albacore or bluefin (larger, longer-lived) within the canned-tuna sub-category reduces per-serving MeHg by 50-70%.
- Consumption-frequency reduction — reducing the number of fish meals per week reduces cumulative weekly MeHg intake proportionally. FDA-EPA “Best Choices / Good Choices / Avoid” framework operationalizes this as species-stratified meal-frequency guidance. EFSA PTWI (1.3 µg MeHg/kg body weight/week) and EPA RfD (0.1 µg MeHg/kg body weight/day) calibrate the safe upper bound; consumption-frequency caps below these levels are the operative consumer-protection lever in advisory frameworks.
What this synthesis does not yet rest on
The mechanism is well-established and the comparative MeHg-vs-preparation-form measurements are consistent across multiple jurisdictions (German MEAL, Latin American canned tuna, US/Canadian freshwater fish surveys). What is less well-characterized:
- Selenium-MeHg interaction during cooking. Whether cooking releases Se from the matrix in a way that shifts the Se:Hg molar ratio at the point of intestinal absorption is an open question. See se-hg-molar-ratio-certification-debate for the broader Se:Hg framework discussion.
- Industrial fermentation of fish products (fish sauce, garum, anchovy paste, surstromming, hákarl). MeHg in fermented fish products has limited measurement in the wiki’s current corpus; fermentation may produce small shifts via microbial activity (the gut-microbiome MeHg-demethylation work in Coe et al. 2023 is suggestive at low orders of magnitude), but no systematic body of evidence yet establishes whether fermented fish products differ meaningfully from fresh.
- High-temperature super-extended cooking (>4 hours at >100°C, e.g., long-pressure-canning of bone broths) is not well-represented in the measured-cooked-fish literature. The default assumption is that even these conditions do not cleave Hg-S bonds, but direct measurement would strengthen the claim.
Consumer-protection implication
HMTc consumer-facing messaging on certified seafood products should treat the cooking-doesn’t-reduce-MeHg finding as one of the three load-bearing claims (alongside “species selection is the dominant lever” and “sensitive populations need stricter consumption-frequency caps”). The misconception that “cooking destroys the mercury” is widespread enough across general-population dietary advice that any HMTc certification messaging that does not explicitly head it off leaves a consumer-protection gap.
The narrow exception worth naming: certain Pb and Cd content in finfish, primarily concentrated in skin and subcutaneous fat, can be modestly reduced (10-20%) by skinning and trimming before cooking. This is a small effect that primarily matters for whole-fish preparations where skin would be consumed; it does not generalize to MeHg.
Downstream pages updated
- seafood — Levers to reduce contamination section includes the “cooking does not reduce MeHg” finding under Processing levers.
- canned-fish — same, with explicit note that canning is a thermal-and-pressure process that does not reduce MeHg.
- fresh-fish — same, with note that freezing/thaw cycles and all cooking methods leave MeHg intact.
- fish-containing-baby-foods — same finding applies; infant-feeding messaging especially benefits from explicit treatment.
- mercury-methyl — Toxicology and Analytical methods sections cover the mechanism.
Anchor sources
- bfr2024-methylmercury-fish-seafood-germany — BfR MEAL study; pooled-sample MeHg-vs-tHg ratios across 30+ species and preparation forms; load-bearing comparative evidence.
- aguilar-miranda2024-thg-canned-tuna-ecuador — Canned-tuna tHg/MeHg fraction evidence; literature-range citation across Latin American studies.
- farina-rocha-aschner-2011-mehg-mechanism — Mechanism reference for MeHg-cysteine binding, intestinal absorption rate.
- coe2023-gut-microbiome-methylmercury-demethylation — Gut-microbiome MeHg demethylation evidence; relevant for fermented-fish open question.
How this page was promoted
Established 2026-05-18 from the seafood-axis ingest and synthesis work. The mechanism evidence (cysteine-binding, intestinal absorption) is decades-established in the toxicology literature; the comparative MeHg-vs-preparation-form measurements were consolidated in the wiki’s evidence base during the 2026-05-18 manual-fetch corpus ingest. The page exists primarily for consumer-protection certification messaging and for HMTc Levers sections on seafood-related product pages.
Peer review state
This synthesis claim has not yet been evaluated by external reviewers. Verdicts will be added here as named domain experts (listed at curators) complete their review. The verdict log is data/peer-review/<reviewer-slug>.jsonl and is part of the public corpus.
| Reviewer | Verdict | Review date | Notes |
|---|---|---|---|
| no reviews yet |
The Heavy Metal Index publishes synthesis claims before external review completes, with the review state visibly tracked. This is the same model Cochrane uses for its protocols: the claim is published, the review accumulates over time, and the credibility of the claim is partly the cumulative result of visible review.
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.