Prenatal mercury exposure, maternal seafood consumption and associations with child language at five years

Background: Methyl mercury (MeHg) is a well-known neurotoxin and evidence suggests that also low level exposure may affect prenatal neurodevelopment. Uncertainty exists as to whether the maternal MeHg burden in Norway might affect child neurodevelopment. Objective: To evaluate the association between prenatal mercury exposure, maternal seafood consumption and child language and communication skills at age five. Methods: The study sample comprised 38,581 mother-child pairs in the Norwegian Mother and Child Cohort Study. Maternal mercury blood concentration in gestational week 17 was analysed in a sub-sample of 2239 women. Prenatal mercury exposure from maternal diet was calculated from a validated FFQ answered in midpregnancy. Mothers reported children's language and communications skills at age five by a questionnaire including questions from the Ages and Stages Questionnaire (ASQ), the Speech and Language Assessment Scale (SLAS) and the Twenty Statements about Language-Related Difficulties (language 20). We performed linear regression analyses adjusting for maternal characteristics, nutritional status and socioeconomic factors. Results: Median maternal blood mercury concentration was 1.03 mu g/L, dietary mercury exposure was 0.15 mu g/kg bw/wk, and seafood intake was 217 g/wk. Blood mercury concentrations were not associated with any language and communication scales. Increased dietary mercury exposure was significantly associated with improved SLAS scores when mothers had a seafood intake below 400 g/wk in the adjusted analysis. Sibling matched analysis showed a small significant adverse association between those above the 90th percentile dietary mercury exposure and the SLAS scores. Maternal seafood intake during pregnancy was positively associated with the language and communication scales. Conclusion: Low levels of prenatal mercury exposure were positively associated with language and communication skills at five years. However, the matched sibling analyses suggested an adverse association between mercury and child language skills in the highest exposure group. This indicates that prenatal low level mercury exposure still needs our attention.