Relation of prenatal low-level mercury exposure with early child neurobehavioral development and exploration of the effects of sex and DHA on it

Abstract Background The extent to which low-level, in utero mercury exposure affects child neurobehavioral development during early childhood has been inconclusive. In addition, the effects of sex and docosahexaenoic acid (DHA) on these relationships remain unclear and controversial. Objectives This study aimed to explore the associations between prenatal low-level mercury exposure and child neurobehavioral development during the neonatal and toddler periods and to examine the potential confounding or interactive effects of sex and DHA status on these relationships. Methods This longitudinal study included 286 mother-child pairs in Shanghai, China, whose cord blood samples were analyzed for total mercury, DHA, other nutrients and coexposure contaminants possibly due to maternal consumption of seafood. Children's neurobehavioral development was assessed with the Neonatal Behavioral Neurological Assessment (NBNA) three days after birth and the Bayley Scales of Infant and Toddler Development, Third Edition (BSID-III) at 18 months of age. Results Cord blood mercury concentration, with geometric mean of 2.00 μg/l, was related to poorer NBNA performance but unrelated to BSID-III scores with adjustment for DHA and other covariates. Cord serum DHA was positively associated with motor development assessed by the BSID-III. The interaction was found between mercury and DHA on the NBNA score, and the inverse relation of cord blood mercury with NBNA was significant only among the children with lower DHA levels ( Conclusions Our research provides initial evidence for the negative effects of prenatal low-level mercury exposure on neonates' neurobehavioral development. Prenatal DHA status may modify the relationship between cord blood mercury level and neonatal neurobehavioral development, but the confounding effects of DHA were not observed. Further studies are warranted before the causality of the observed associations can be determined.