Iron status as a covariate in methylmercury-associated neurotoxicity risk

Intrauterine methylmercury exposure and prenatal iron deficiency negatively affect offspring's brain development. Since fish is a major source of both methylmercury and iron, occurrence of negative confounding may affect the interpretation of studies concerning cognition. We assessed relationships between methylmercury exposure and iron-status in childbearing females from a population naturally exposed to methylmercury through fish intake (Amazon). We concluded a census (refuse <20%) collecting samples from 274 healthy females (12-49 years) for hair-mercury determination and assessed iron-status through red cell tests and determination of serum ferritin and iron. Reactive C protein and thyroid hormones was used for excluding inflammation and severe thyroid dysfunctions that could affect results. We assessed the association between iron-status and hair-mercury by bivariate correlation analysis and also by different multivariate models: linear regression (to check trends); hierarchical agglomerative clustering method (groups of variables correlated with each other); and factor analysis (to examine redundancy or duplication from a set of correlated variables). Hair-mercury correlated weakly with mean corpuscular volume (r = .141; P = .020) and corpuscular hemoglobin (r = .132; .029), but not with the best biomarker of iron-status, ferritin (r = .037; P = .545). In the linear regression analysis, methylmercury exposure showed weak association with age-adjusted ferritin; age had a significant coefficient (Beta = .015; 95% CI: .003-.027; P = .016) but ferritin did not (Beta = .034; 95% CI: −.147 to .216; P = .711). In the hierarchical agglomerative clustering method, hair-mercury and iron-status showed the smallest similarities. Regarding factor analysis, iron-status and hair-mercury loaded different uncorrelated components. We concluded that iron-status and methylmercury exposure probably occur in an independent way.