Hsi et al. 2016 - Methylmercury Concentration in Fish and Risk-Benefit Assessment of Fish Intake among Pregnant versus Infertile Women in Taiwan

This study measured methylmercury in fish and used those values in a risk-benefit assessment for pregnant and infertile women in Taiwan.

Key numbers

Fish MeHg values are reported as mean ± SD across triplicate samples per species in mg/kg wet weight (Table 3, body text page 6). Hair MeHg values are geometric mean ± standard error in mg/kg (Table 1 footnote e, body text page 6).

Fish MeHg by species (mg/kg wet wt., n = 3 per species):

• Swordfish (Xiphias gladius): 0.28 ± 0.23.
• Tuna (Thunnus alalunga): 0.14 ± 0.13.
• Cod: 0.130.
• Greater amberjack: 0.110.
• Salmon: 0.097.
• Hairtail: 0.065.
• Tilapia: 0.050.
• Mackerel: 0.017.
• Milkfish: 0.006.
• Anchovy: 0.002.

Diet-group aggregates (mg/kg wet wt.):

• Carnivorous fish: 0.11 ± 0.13.
• Omnivorous fish (milkfish only): 0.006 ± 0.01.
• Filter-feeding fish (anchovy only): 0.002 ± 0.001.

All measured fish MeHg concentrations were below the Codex guideline level of 1 mg/kg (CAC/GL 7, FAO/WHO 1991).

Hair MeHg (mg/kg, geometric mean ± SE):

• Infertile women (n = 162): 1.82 ± 0.14.
• Pregnant women (n = 62): 1.24 ± 0.18.
• P = 0.01 (Wilcoxon rank-sum).
• 80% of infertile women and 68% of pregnant women had hair MeHg exceeding the USEPA reference-dose hair concentration of 1 mg/kg.

Daily MeHg exposure dose (µg/kg body weight/day, arithmetic mean ± SD):

• Pregnant women: 0.34 ± 0.34.
• Infertile women: 0.51 ± 0.49.
• 76.7% of pregnant women and 90.9% of infertile women exceeded the USEPA RfD of 0.1 µg/kg/day.

Methods

Hair samples (~2 cm, occipital scalp, stainless-steel surgical scissors) and 1-g fish muscle samples were prepared for MeHg analysis following the Brooks Rand BRL Method BR-0011 and USEPA Method 1630 with minor modifications. Hair was sonicated 30 min in neutral detergent, rinsed three times with deionized water, and oven-dried at 37 °C for 24 h. Fish was alkaline-digested in 25% KOH/methanol at 75 °C for 5.5 h, then liquid-liquid extracted with CH2Cl2 and HCl, filtered, and purged. Sodium acetate buffer and 1.0% NaBEt4 were added for ethylation prior to analysis on a MERX integrated automated MeHg analyzer (Brooks Rand). Quality control used IAEA-085 (human hair) and BCR-465 (tuna fish) certified reference materials; average recovery 94.7%, coefficient of variation 4.27%, accuracy 89.2%. Each sample was analyzed in triplicate.

The daily MeHg exposure dose was estimated from a species-weighted intake calculation referencing USEPA Eq. 1, with hazard quotient and allowable-daily-intake calculations against the USEPA MeHg RfD of 0.1 µg/kg/day. Statistical analysis used SPSS version 17.0; Wilcoxon rank-sum, chi-square, Fisher exact, Kruskal-Wallis, and Spearman rank correlation tests as appropriate.

Implications

Per-species MeHg values for all 10 species are routeable to seafood occurrence context, with the predatory species (swordfish, tuna, cod, greater amberjack, salmon, hairtail) most relevant to fish-marine-predatory occurrence routing. Human hair MeHg and daily-exposure-dose values are biomonitoring/exposure context for Taiwanese women of childbearing age, not product occurrence.

Wiki pages this source may touch

• Swordfish
• Fish
• Seafood
• Fresh Fish
• Fish — marine, predatory (tuna, swordfish, shark, king mackerel)
• Methylmercury

Verification notes

• Source identity checked against DOI 10.1371/journal.pone.0155704 and the downloaded PDF.
• Speciation: paper analyzed MeHg directly via USEPA Method 1630 ethylation/GC-CVAFS workflow; metals: ["MeHg"] is correct, do not collapse to tHg.
• Statistic-type discipline: fish values are mean ± SD across triplicate samples in mg/kg wet weight (Table 3); hair values are geometric mean ± standard error in mg/kg (Table 1 footnote e). The two statistic types must not be conflated in downstream pooling.
• Method/vendor names retained in Methods (Brooks Rand BR-0011, MERX analyzer, IAEA-085, BCR-465, SPSS) are scientific-reproducibility identifiers under the locked 2026-05-17 Part 12 reading; none are attached to a contamination value.
• 2026-06-09: GPT-style 5-check audit applied. Findings applied: fish-basis qualifier and statistic-type labels added; Methods expanded with Brooks Rand BR-0011 / USEPA Method 1630 / MERX / IAEA-085 / BCR-465 detail; per-species values for all 10 species added; fish-marine-predatory added to products. Findings rejected: Check 2 ❌ severity (every slug verified present in the taxonomy snapshot; the auditor’s concern was under-routing, not vocabulary, so the level was reduced to ⚠️ and addressed by adding fish-marine-predatory). Independent verification caught: sampling_year_range was null; updated to 2007-2014 to span both fish (2014) and women (2007-2010) sampling windows; daily MeHg exposure dose and RfD-exceedance percentages were absent from Key numbers; added.

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.