Determination of toxic elements (mercury, cadmium, lead, tin and arsenic) in fish and shellfish samples. Risk assessment for the consumers

Overview

Olmedo and colleagues measured mercury, cadmium, lead, arsenic, and tin in fresh, canned, and frozen fish and shellfish products commonly consumed in Andalusia, Spain. The study analyzed 485 samples, reported medians with 5th and 95th percentiles in mg/kg wet weight, and performed methylmercury speciation in the high-total-mercury predatory fish subset. Arsenic was measured as total arsenic, not inorganic arsenic.

Key numbers

The study collected samples between 2009 and 2010. Table 2 reports fresh fish and shellfish medians as mg/kg wet weight:

For the highest fresh-fish mercury rows, Table 2 reports cat shark mercury (Hg) 0.698 (0.153-1.406), swordfish 0.540 (0.177-1.227), tuna 0.470 (0.298-0.779), and blue shark 0.350 (0.238-0.963). For arsenic, the highest fresh medians are shrimp 0.739 (0.388-0.884), sardine 0.561 (0.169-0.959), red mullet 0.427 (0.230-0.729), cat shark 0.340 (0.183-0.450), and clam 0.307 (0.189-0.415).

Table 3 reports canned products as mg/kg wet weight:

Table 3 reports frozen products as mg/kg wet weight: codfish Hg/cadmium (Cd)/lead (Pb)/arsenic (As) 0.000/0.002/0.004/0.036; common sole 0.000/0.003/0.544/0.059; hake 0.000/0.001/0.086/0.028; prawn 0.000/0.001/0.033/0.045; shrimp 0.070/0.010/0.004/0.509; and squid 0.000/0.073/0.004/0.044.

The methylmercury speciation section states that no Hg2+ peak was detected in analyzed fish samples and, with Hg2+ LOD set at 0.5 µg/L, more than 99.9% of total mercury in fresh blue shark, cat shark, swordfish, tuna, and canned tuna was inferred to be methylmercury (MeHg).

Table 4 category summaries report total Hg median 0.017 across all 485 samples, with 6 samples above the applicable maximum level (1.24%). The category listed as predatory/specified fish had Hg median 0.237 and 95th percentile 0.974. Total Cd median was 0.004, with 3 samples above maximum level (0.62%). Total Pb was reported with lower/upper bounds because 346 samples were below LOD (71.34%), with 25 samples above maximum level (5.15%). Canned-food tin (Sn) had lower-bound median 0.000, upper-bound median 0.001, and no samples above the 200 mg/kg maximum level.

The risk-assessment table reports estimated weekly intakes for an average 60 kg person: Hg 63.63 µg/week, Cd 17.57 µg/week, Pb 75.95 µg/week, and Sn 15.40 µg/week. The text estimates inorganic arsenic intake at 0.42 µg/day (2.94 µg/week) by assuming 2% of fish arsenic and 3.5% of shellfish arsenic as inorganic arsenic, but the measured arsenic results themselves are total arsenic.

Methods (brief)

Fresh and frozen seafood samples were collected from the central market of Granada, and canned species were purchased from major supermarkets in Granada. About 0.7 g of edible sample was microwave digested for Hg, Cd, Pb, As, and Sn analysis. Arsenic, total mercury, cadmium, lead, and tin were determined by atomic absorption methods; MeHg speciation used HPLC-ICP-MS after tetramethylammonium hydroxide extraction. Medians and P5-95th percentile were used because concentrations were non-normal; lower and upper bounds were used when more than 60% of samples were below LOD.

Implications

This source contributes primary wet-weight occurrence evidence for fresh fish, canned fish, frozen fish, and shellfish rows. It is especially important for seafood mercury because it directly checked MeHg speciation in high-Hg predatory fish and found the total-Hg signal to be overwhelmingly MeHg in that subset. Total arsenic and the paper’s modeled inorganic-arsenic intake assumption must remain separate from measured inorganic arsenic data.

Update history

The five most recent substantive edits to this page, classified major (evidence or structure moved), correction (a published value or statement was wrong and has been fixed), or minor (narrative rewritten without changing the underlying evidence). Each description is derived from what the edit did to this page; the linked commit is the authoritative record, routine regeneration passes are excluded, and the full version history lives in git.