Neurotoxic risk caused by stable and variable exposure to methylmercury from seafood.

OBJECTIVES To examine whether the dose-effect relationship for developmental mercury neurotoxicity is affected by variable mercury exposure during pregnancy. METHODS The study was based on a birth cohort of 1022 children born in the Faroe Islands between March 1986 and December 1987. Neurobehavioral performance of 917 children (90%) was assessed at age 7. Intrauterine methylmercury exposure was determined from mercury concentrations in cord blood and 2 sets of maternal hair. Complete exposure information was available for 614 children (67%). RESULTS In children with complete exposure data, 8 of 16 neuropsychological tests showed deficits significantly associated with the cord-blood mercury concentration after confounder adjustment. Variable intrauterine exposure was suggested by disagreement between mercury concentrations in the 2 maternal hair samples. Removal of the 61 children (10%) with the greatest degree of variable exposure had a minimal effect on most exposure-effect relationships. However, the effect of the cord-blood concentration on verbal learning and memory was greater after this exclusion. CONCLUSION The study supports previous findings from this cohort study that maternal mercury exposure during pregnancy is associated with neuropsychological deficits detectable at age 7 years and that this association is evident in women with stable exposures throughout pregnancy. Thus the association is not the result of variable exposures.