Shim et al. 2004 — Mercury and fatty acids in canned tuna, salmon, and mackerel

This 2004 US study analyzed total mercury (tHg) and polyunsaturated fatty acids in canned tuna (n=240), salmon (n=16), and mackerel (n=16) purchased from stores in Lafayette, Indiana. Mercury was measured as total Hg by thermal decomposition amalgamation atomic absorption spectrophotometry (TDA/AAS); the paper does not speciate MeHg from inorganic Hg, but notes that up to 90% of Hg in fish is typically found as MeHg. Overall mean tHg was 188 ppb in tuna, 45 ppb in salmon, and 55 ppb in mackerel — all below the FDA action level of 1000 ppb. The paper’s central trade-off is that white/albacore tuna rows had higher tHg and higher EPA+DHA than light tuna in water, while salmon and mackerel had lower tHg and higher EPA+DHA than the tuna category average.

Key numbers

Total Hg and EPA+DHA in canned fish, aggregated by product form

Each tuna row in the source is the grand mean of duplicate analyses from 3 lots with 4 composites. Salmon and mackerel rows are duplicate analyses of 4 composites. Limit of detection for total Hg: 0.01 ng total Hg.

Summary by product category:

• Overall tuna mean (n=240): 188 ppb tHg
• Salmon mean (n=16): 45 ppb tHg
• Mackerel mean (n=16): 55 ppb tHg
• Light tuna in water was the lowest-tHg tuna group (mean 54 ppb).
• White/albacore tuna in water or pouch had higher tHg (227–330 ppb) and higher EPA+DHA (mean 902 mg/100 g) than light tuna in water.

FDA action level for commercial fish under CPG 540.600: 1000 ppb methylmercury expressed as mercury in the edible portion. No sample exceeded this level.

EPA + DHA content by product type:

• Light tuna in water: mean ~289 mg/100 g wet tissue (DHA 181–300; EPA 32–39)
• White/albacore tuna in water: mean ~902 mg/100 g wet tissue (DHA 555–741; EPA 190–333) — highest EPA+DHA among tuna types
• Salmon: mean ~1,623 mg/100 g wet tissue (DHA 564–874; EPA 884–925)
• Mackerel: mean ~851 mg/100 g wet tissue (DHA 282–649; EPA 218–553)

Exposure estimate at current tuna consumption for a 60-kg woman:

• At 95th percentile consumption (20 g/d): all tuna products below USEPA RfD of 0.1 µg MeHg/kg bw/day
• At 99th percentile consumption (35 g/d): light tuna in oil and white/albacore tuna deliver 129–153% of RfD
• Author-calculated fish-tissue concentration needed to stay below the RfD at 8 oz/wk (32.43 g/d): below 185 ppb
• The authors concluded that the FDA action level should be reduced to 185 ppb or that products above this level should carry an appropriate warning. This is the source authors’ risk-management recommendation, not an HMT&C threshold.

Methods (brief)

Total mercury measured using Thermal Decomposition Gold Amalgamation Atomic Absorption Spectrophotometry (TDA/AAS, DMA-80 analyzer), calibrated against NRCC certified reference materials Tort-2 (0.270 ppm Hg) and Dorm-2 (4.64 ppm Hg). Two calibration ranges used: 0–35 ng Hg (R²=0.9929) and 30–470 ng Hg (R²=0.9958). LOD: 0.01 ng total Hg. Composite samples were prepared by pooling and grinding the entire contents of 2 cans from each lot. Fatty acids quantified by GC/FID using AOAC method 991.39 on lipids extracted by modified Folch method. SRM recovery for total fat: 95%. Standard deviation <±10% of mean for all mercury values (SD not shown). Mercury is reported as total Hg; no speciation of MeHg vs inorganic Hg performed, but authors note ~90% of Hg in fish is typically MeHg.

Implications

Certification: The substantial difference between light tuna in water (mean 54 ppb tHg) and white/albacore tuna in water/pouch rows (227–330 ppb tHg) provides a concrete example of how species/form descriptors within the canned-fish category determine mercury exposure. The source’s 185 ppb calculation is tied to a 60-kg woman consuming 8 oz/week and should be treated as source-side risk-context, not as a wiki or HMT&C threshold.

Courses: The mercury/omega-3 trade-off quantified here is a concrete teaching example: white/albacore tuna provides the highest EPA+DHA among tuna products but at ~4× the tHg of light tuna in water. Salmon and mackerel offer a better ratio (higher EPA+DHA at lower tHg).

App: Product-form-level tHg values (light tuna in water ~54 ppb, white/albacore tuna water/pouch rows 227–330 ppb, salmon ~45 ppb, mackerel ~55 ppb) are useful source inputs for canned-fish risk estimation. Use product-form and species descriptors rather than brand identity.

Wiki pages this source may touch

• canned-tuna
• tinned-fish
• fish
• seafood
• canned-fish
• canned-seafood
• seafood
• mercury-total
• mercury-methyl
• fda-cpg-540-600-methylmercury-fish
• epa-iris-methylmercury-rfd

Verification notes

• 2026-05-18 Codex merge-enhance: matched DOI/raw path to P0151, added DOI access URL and PDF SHA-256, corrected invalid ingredient slugs to existing broader fish/seafood slugs, replaced [[metals/mercury]] with total-mercury and methylmercury routing, and created the FDA CPG 540.600 regulation page for the source-cited 1000 ppb action level.
• Strict Part 12 brand-firewall cleanup: source Table 1 names sampled brands, but this wiki page aggregates rows by product form/species and omits all sampled brand names.
• Numerical correction: prior page gave light-tuna-in-water EPA+DHA as ~268 mg/100 g; Table 1 values sum to 213, 313, and 339 mg/100 g, mean ~289 mg/100 g.
• The authors’ 185 ppb recommendation and label proposal are reported only as source-side risk-management statements; they are not adopted here as HMT&C standards or consumer advice.
• 2026-06-08 merge-enhance: removed the products/fresh-fish slug from frontmatter (paper studied only canned tuna, salmon, and mackerel — no fresh fish samples were analyzed; the slug was over-specifying frontmatter and fanning broad-context rows to fish-freshwater, fish-marine-predatory, and fish-marine-non-predatory rows for which the paper provides no direct evidence). Added the products/canned-seafood slug to reflect that the three product forms studied are all canned-seafood Cat 6 row members (direct evidence rather than seafood-broad-context only).

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.