Lehel et al. 2023 - Potentially toxic elements in yellowfin tuna from the Indian Ocean

Lehel and colleagues measured arsenic, cadmium, mercury, and lead by inductively coupled plasma optical emission spectrometry (ICP-OES) in the flesh of 40 yellowfin tuna (Thunnus albacares) purchased at a fishery market in Hungary but originating from Sri Lanka in FAO Fishing Area 57 (Indian Ocean, Eastern), subarea 57.1 (Bay of Bengal). Mean total arsenic was 0.98 ± 0.47 mg/kg wet weight, but total As is not regulated; estimating inorganic As as 5% of total As gave a mean of 0.05 ± 0.02 mg/kg, with estimated daily intake (EDI) values (0.03-0.09 µg/kg/day) well below the 0.3 µg/kg/day dietary reference. Cadmium was detected above the limit of detection (0.05 mg/kg) in only 20% of samples and all detected values stayed below the EU 2023/915 maximum level (0.10 mg/kg for tuna species); EDI values (0.05-0.07 µg/kg/day) were below the 1 µg/kg/day reference. Mercury was below the limit of detection (0.50 mg/kg) in 90% of samples and below the EU 2023/915 maximum (1.0 mg/kg for tuna species) in all samples; EDI was below the 0.3 µg/kg/day reference in 90% of samples but exceeded it in 2 samples (0.69 and 0.82 µg/kg/day). Lead exceeded the EU 2023/915 maximum level (0.30 mg/kg, muscle meat of fish) in 40% of samples, with detected values from 0.30 to 1.59 mg/kg; EDI values (0.28-1.49 µg/kg/day) were above the dietary reference values for both adults (0.16 µg/kg/day) and children (0.26 µg/kg/day) in 100% of samples, which the authors flagged as a potential harm and a call to attention for environmental pollution and consumer protection.

Key numbers

Wet-weight concentrations (Table 3, mg/kg w.w.; n=40 yellowfin tuna flesh samples):

• Total arsenic (tAs):
  • Mean ± SD: 0.98 ± 0.47 mg/kg.
  • Minimum (LOD): 0.50; minimum (measured): 0.51; maximum (measured): 2.05 mg/kg.
  • Regulated maximum: not regulated for total As in fish.
  • Detected above LOD in 100% of samples.
• Inorganic arsenic (iAs, estimated as 5% of total As per the authors’ convention):
  • Mean ± SD: 0.05 ± 0.02 mg/kg.
  • Minimum (LOD-derived): 0.03; minimum (measured-derived): 0.03; maximum (measured-derived): 0.10 mg/kg.
• Cadmium (Cd):
  • Mean ± SD: 0.03 ± 0.01 mg/kg.
  • Minimum (LOD): 0.05; minimum (measured): 0.05; maximum (measured): 0.07 mg/kg.
  • Regulated maximum: 0.10 mg/kg (EU 2023/915, tuna species).
  • Detected above LOD in 20% of samples; 100% of all samples below the regulated limit.
• Mercury (tHg):
  • Mean ± SD: NA (90% of samples below the LOD of 0.50 mg/kg, so statistical analysis was not performed).
  • Minimum (LOD): 0.50; minimum (measured): 0.73; maximum (measured): 0.88 mg/kg.
  • Regulated maximum: 1.0 mg/kg (EU 2023/915, tuna species).
  • All samples below the regulated limit.
• Lead (Pb):
  • Mean ± SD: 0.39 ± 0.37 mg/kg.
  • Minimum (LOD): 0.20; minimum (measured): 0.30; maximum (measured): 1.59 mg/kg.
  • Regulated maximum: 0.30 mg/kg (EU 2023/915, muscle meat of fish).
  • 47.5% of samples below the LOD; among detected samples, 40% of all samples exceeded the regulated limit.

Estimated daily intake (Table 4, µg/kg/day; calculated as concentration × 65.7 g/day average EU fish consumption / 70 kg body weight, per EFSA convention):

• Inorganic arsenic (5% of total As): reference 0.3 µg/kg/day. Average EDI 0.05; range 0.03 (LOD) to 0.09 (measured maximum). 0% of samples above reference.
• Cadmium: reference 1 µg/kg/day. Average EDI 0.03; range 0.05 (LOD) to 0.07 (measured maximum). 0% of samples above reference.
• Lead: reference 0.16 µg/kg/day (adults) and 0.26 µg/kg/day (children). Average EDI 0.37; minimum (LOD) 0.19; minimum (measured) 0.28; maximum (measured) 1.49. 100% of samples above both adult and child reference values.
• Mercury (from Section 3.3 prose, not Table 4, which omits Hg): reference 0.3 µg/kg/day. The source parenthetically reports 0.23 µg/kg/day for the 90% of samples below reference (computed using half-LOD substitution for below-LOD samples per Section 2.4); 2 samples were above reference at EDI 0.69 and 0.82 µg/kg/day.

Comparator regulation cited by the authors: EU Commission Regulation 2023/915 (Pb 0.30 mg/kg for muscle meat of fish; Cd 0.10 mg/kg and Hg 1.0 mg/kg for tuna species Thunnus spp., Katsuwonus pelamis, and Euthynnus spp.). Dietary reference values from EFSA scientific opinions and from Wong et al. 2022 (lead).

Per capita context (cited from EU Commission and FAO data): EU average annual fish consumption 23.97 kg/person (65.7 g/day); Hungary 6.28 kg/person (well below EU average); Portugal 59.91 kg/person (2-3 times the EU average). The 65.7 g/day value was used for the EDI calculation throughout.

Methods (brief)

40 yellowfin tuna (Thunnus albacares) were collected from a fishery market in Hungary; the fish originated from Sri Lanka and were caught in FAO Fishing Area 57 (Indian Ocean, Eastern), subarea 57.1 (Bay of Bengal). The paper specifies the boundary of the fishing zone in detail. From each fish, 5 g of flesh was taken from the muscle of the back side using a metal-free tool, placed in labelled plastic bags after shredding and homogenization (Potter S, B. Braun Biotech International GmbH, Melsungen, Germany), and frozen at −70 °C (So-Low Ultra-Low Freezer, Model C85-9, Environmental Equipment Co. Inc., Cincinnati, OH, USA) until analysis.

Analytical work was performed at the Department of Animal Hygiene, University of Veterinary Medicine. 0.5 g of sample was weighed into a CEM MARS XPreSS teflon vessel and decomposed by microwave digestion (CEM MARS6, CEM Corporation, Matthews, NC, USA) after addition of 5 mL of 65% nitric acid (Aristar, VWR) and 5 mL of 30% hydrogen peroxide (Normapur, VWR), with the program ramp 35 min, hold 50 min at 200 °C, energy 1700 W. The digest was diluted to 25 mL with deionized water, then twice further with a 1 mg/L Y internal standard solution and a 0.25 mg/L Au solution (VWR) for mercury stabilization, and analyzed on a Perkin Elmer Optima 8300 DV ICP-OES (40 MHz RF generator, 1300 W RF power, BURGENER PEEK MIRA MIST nebulizer, plasma 12 dm³/min, auxiliary 0.2 dm³/min, nebulizer 0.7 dm³/min, observation height 15 mm).

Method validation parameters (Table 1) gave: total arsenic LOQ 1.67 mg/kg, LOD 0.50 mg/kg, precision 12.7%, trueness 13.6%; cadmium LOQ 0.17 mg/kg, LOD 0.05 mg/kg, precision 8.4%, trueness −10.9%; mercury LOQ 1.67 mg/kg, LOD 0.50 mg/kg, precision 12.3%, trueness 8.1%; lead LOQ 0.67 mg/kg, LOD 0.20 mg/kg, precision 3.5%, trueness −8.4%. Acceptance thresholds were precision below 20% and trueness within ±15%. Quality control used certified reference materials ERM-CE278k (mussel tissue, JRC Geel, Belgium) for As/Cd/Pb and ERM-CE464 (tuna fish) for Hg; recoveries were As 102.5%, Cd 106.5%, Pb 91.7%, Hg 98.1% (Table 2). For statistical evaluation, values below LOD were replaced by half the LOD; analysis was performed in Microsoft Excel and R (version 3.3.2). Mercury was not statistically analyzed because 90% of samples were below the LOD.

EDI was calculated as the metal concentration × 65.7 g/day fish consumption (EU average from EFSA) / 70 kg body weight.

Speciation and interpretation notes

• The measured arsenic analyte is total As (tAs); inorganic As (iAs) is not directly speciated but estimated from the convention that “the remaining amount (5%) is incorporated as inorganic derivatives” (Section 3.1). The frontmatter metals array therefore lists only tAs, not arsenic-inorganic, per CLAUDE.md Part 14: the iAs values in the Key numbers are derived from the 5% convention and must be read as estimates, not measurements.
• The measured mercury analyte is total mercury (tHg); methylmercury was not speciated. The page reports metals: [tHg] rather than MeHg, per CLAUDE.md Part 14. The authors compare the tHg result against the EU 2023/915 limit for tuna species (1.0 mg/kg total Hg). MeHg-specific reference values (e.g., FDA 0.46 µg/kg/day RfD or 1.0 mg/kg JECFA PTWI equivalent) are not directly addressed.
• Cadmium and lead were quantified as total Cd and total Pb (no speciation reported); the page uses the umbrella Cd and Pb slugs.
• The mercury EDI computation in the discussion treats below-LOD samples as half the LOD (0.25 mg/kg), which is why the Hg EDI mean of 0.23 µg/kg/day is reported even though the concentration mean is “NA” in Table 3. Two samples exceeded the 0.3 µg/kg/day reference at EDI 0.69 and 0.82 µg/kg/day; these correspond to the 0.73-0.88 mg/kg measured maximum range in Table 3 multiplied by the standard 65.7 g/day / 70 kg factor.
• The lead detection rate (47.5% of samples below LOD 0.20 mg/kg) combined with the 40% exceedance of the 0.30 mg/kg regulated limit implies a sharply bimodal distribution: roughly half the samples were undetectable, while a substantial fraction of the detected samples cluster above the regulatory limit. The 1.59 mg/kg maximum is more than five times the EU 2023/915 limit. The authors interpret this as evidence of environmental Pb pollution and a consumer-protection concern, while noting that mean EDI (0.37 µg/kg/day) exceeds both the adult (0.16 µg/kg/day) and child (0.26 µg/kg/day) dietary reference values in 100% of samples - the latter calculation uses the average concentration, so even non-detect samples (substituted at half-LOD) contribute to the EDI ratio.
• The fish originated in the Bay of Bengal (FAO Fishing Area 57.1, off Sri Lanka) but were purchased at retail in Hungary; the exposure inferences are framed for European/Hungarian consumers, while the contamination source is the Indian Ocean fishery. The jurisdictions field carries HU (market/consumer), LK (catch origin), and EU (regulatory comparator).
• The paper’s Sections 3.1-3.4 include the authors’ own comparisons of their tuna results to a range of prior tuna heavy-metal studies (Núñez et al. 2018, Koesmawati and Arifin 2015, da Silva et al. 2020 / Aracaju, Bae et al. 2023, Olmedo et al. 2013, and others). Cross-source comparisons belong to the Part 9 synthesis pass and not to this source page; the comparator references are listed here as pointers, not consolidated into a comparative statement.

Implications for wiki use

This page contributes a regulatory-grade ICP-OES dataset for fresh yellowfin tuna flesh sampled from a Hungarian retail-market chain whose product originated in the Bay of Bengal. The dataset is direct occurrence evidence for fresh marine predatory fish (specifically yellowfin tuna). The lead findings (40% of samples above the EU 2023/915 muscle-meat-of-fish maximum; 100% of EDI values above adult and child dietary reference values) are the most consequential signal from this study. The arsenic dataset (total As 0.98 mg/kg w.w. mean, with iAs estimated as 5% of total rather than speciated) is less directly comparable to speciated-iAs studies. The mercury dataset is left-censored at the LOD (0.50 mg/kg): 90% of yellowfin tuna samples were below 0.50 mg/kg total Hg. The cadmium dataset is similarly left-censored (80% of samples below the LOD of 0.05 mg/kg).

Wiki pages this source may touch

• tuna
• fish
• seafood
• fresh-fish
• fish-marine-predatory
• seafood
• arsenic
• arsenic-total
• arsenic-inorganic
• cadmium
• mercury
• mercury-total
• lead
• eu-2023-915-cadmium
• eu-reg-2022-617-mercury-fish

Sources

• Lehel J, Papp Z, Bartha A, Palotás P, Szabó R, Budai P, Süth M. 2023. “Metal Load of Potentially Toxic Elements in Tuna (Thunnus albacares) - Food Safety Aspects.” Foods 12(16):3038. DOI: 10.3390/foods12163038.
• Manual-fetch PDF: raw/manual-fetch/Kimi_Agent_Download Corruption Issue/seafood_papers/03_Marine_Predatory/Metal Load of Potentially Toxic Elements in Tuna (Thunnus albacares) -- Food Safety Aspects.pdf (SHA-256 91f7d64f36f373c095f0049531679579e48682b17c26e81886320962e1425a24).

Verification notes

• 2026-06-03 Claude fresh ingest from manual-fetch/Kimi_Agent_Download Corruption Issue/seafood_papers/03_Marine_Predatory. Numbers cross-checked against Abstract, Section 2.1 (Sampling), Section 2.2 (Analytical Method), Section 2.3 (Exposure Calculation), Table 1 (validation), Table 2 (QC), Table 3 (concentrations), Table 4 (EDI), Sections 3.1-3.4 (per-metal results and literature comparison), and Conclusions.
• Brand firewall (Part 12): the source describes a generic “fishery market in Hungary” and identifies the original catch origin (FAO 57.1) but does not name brands or specific market operators; no brand attribution to redact.
• Analyte speciation: Hg method is ICP-OES on total digest (no MeHg speciation), so metals: [tHg] rather than [MeHg] or [tHg, MeHg]. iAs is estimated at 5% of total As per the EFSA convention used by the authors, not directly speciated; the page lists both tAs and arsenic-inorganic in the metals array but flags the iAs estimation in the speciation notes.
• The regulation cross-link list includes the available eu-2023-915-cadmium page and the eu-reg-2022-617-mercury-fish page. The paper itself cites only EU 2023/915 (reference 29); these wiki-link references are jurisdictional context for the regulatory framework, not a claim that the paper analyzes these specific sub-rules. There is no Pb-in-fish-muscle sub-page in the current taxonomy, so the EU 2023/915 lead context for muscle meat of fish (0.30 mg/kg) is described in prose only.
• 2026-06-03 Claude audit subagent (fresh context) verdict: REVISE. Findings applied: (1) removed arsenic-inorganic from metals array (paper estimates iAs at 5% per EFSA convention rather than speciating, so per CLAUDE.md Part 14 only tAs belongs in the array; the iAs estimation discussion remains in this section); (2) removed the eu-2023-915-lead-infant-young-child-foods cross-link (paper’s Pb comparator is the EU 2023/915 muscle-meat-of-fish limit at 0.30 mg/kg, not the infant-foods sub-rule); (3) trimmed the cross-source comparator paragraph in the speciation notes to a pointer-only list (the synthesis sentence “Pb in the Lehel dataset is at the upper end… while Cd and Hg are at the lower end” was a my-voice synthesis claim, removed); (4) trimmed routing-layer instruction in Implications (“Routing should treat this source as direct evidence…”) per CLAUDE.md Part 5b (routing is the system’s job, not the model’s); (5) clarified the Hg EDI 0.23 µg/kg/day figure as the source’s parenthetical for the 90% below-reference group computed via half-LOD substitution, since Table 4 omits Hg. Findings rejected as false positives: (a) audit flagged [[ingredients/tuna]] as missing from the taxonomy snapshot, but the routing audit’s routing_unresolved.csv does NOT flag it and 5+ existing source pages (bae2023, bfr2024, blanco2023, chamorro2024, fsai2016) already use this slug, so it is accepted convention; (b) audit flagged the Hg EDI 0.23 µg/kg/day as a Claude-derived number not in the source, but the value appears verbatim in Section 3.3 of the PDF as the parenthetical “(0.23 µg/kg/day)” immediately following “(90%)“.

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.