To main content
Norsk
Publications

Dietary mercury exposure in a population with a wide range of fish consumption - self-capture of fish and regional differences are important determinants of mercury in blood

Academic article
Year of publication
2012
Journal
Science of the Total Environment
External websites
Cristin
Doi
Contributors
Marthe Torunn Solhaug Jenssen, Anne Lise Brantsæter, Margaretha Haugen, Helle Margrete Meltzer, Thorjørn Larssen, Helen Engelstad Kvalem, Bryndis Eva Birgisdottir, Yngvar Thomassen, Dag Ellingsen, Jan Alexander, Helle Katrine Knutsen

Summary

Human, low level, chronic exposure to mercury (Hg) from fish is of concern because of potential neurodevelopmental and cardiovascular toxicity. The purpose of the study was to 1) measure total mercury (THg) in blood and estimate dietary exposure in a population group with a wide range of seafood consumption, 2) assess the intake and blood concentration in relation to tolerable intake values, 3) characterise dietary sources, and 4) to investigate the relationship between dietary THg with THg in blood (BTHg), including factors that can explain the variance in BTHg concentrations. The participants (n = 184) filled in an extensive food frequency questionnaire which was combined with a database on THg concentrations in Norwegian food, and donated blood and urine. Median consumption of seafood was 65 g/day (range 4 to 341 g/day). The calculated mean dietary THg exposure was 0.35 (median 0.30) μg/kg body weight/week. Seafood contributed on average 95% to the exposure. The JECFA Provisional Tolerable Weekly Intake (PTWI) of 1.6 μg MeHg/kg bw/week was not exceeded by any of the participants. BTHg ranged from 0.6 to 30 μg/L, with a mean of 5.3 (median 4.0 μg/L). There was a strong relationship between total seafood consumption and BTHg concentrations (r = 0.58 95%CI: 0.48, 0.67) and between estimated THg dietary exposure and BTHg (r = 0.46 95%CI: 0.35, 0.57). Fish consumption, sex, catching > 50% of their seafood themselves, and living in coastal municipalities were significant factors in linear regression models with lnBTHg. Including urinary Hg in the regression model increased the explained variance from 54% to 65%. In a toxicokinetic model, the calculated dietary intake appeared to moderately underestimate the measured BTHg among the participants with the highest BTHg. Only two of the participants had BTHg slightly above a value equivalent to the JECFA PTWI, but none of them were women in fertile age.