A systematic literature review of epidemiologic studies of developmental manganese exposure and neurodevelopmental outcomes.

BACKGROUND: Neurotoxic effects of high-level occupational exposure to manganese (Mn) are well established; however, whether lower-level environmental exposure to Mn in early life causes neurodevelopmental toxicity in children is unclear.
METHODS: A systematic literature review was conducted to identify and evaluate epidemiologic studies of specific Mn biomarkers assessed during gestation, childhood, or adolescence in association with neurodevelopmental outcomes, focusing on quantitative exposure-response estimates with specific endpoints that were assessed in multiple independent study populations. Study quality was evaluated using the revised RTI item bank and the Cochrane Risk of Bias tool, and the overall weight of epidemiologic evidence for causality was evaluated according to the Bradford Hill considerations.
RESULTS: Twenty-two epidemiologic studies were identified that estimated associations between early-life Mn biomarkers and neurodevelopmental outcomes. Seven of these studies provided adjusted estimates for the association with child intelligence assessed using versions of the Wechsler Intelligence Scales for Children; no other specific neurodevelopmental endpoints were assessed in more than three independent study populations each. Among the studies of child intelligence, five studies in four independent populations measured blood Mn, three studies measured hair Mn, and one measured dentin Mn. Overall, cross-sectional associations between Mn biomarkers and measures of child intelligence were mostly statistically nonsignificant but in a negative direction; however, the lone prospective cohort study found mostly null results, with some positive (favorable) associations between dentin Mn and child intelligence. Studies were methodologically limited by their cross-sectional design and potential for confounding and selection bias, as well as unaddressed questions on exposure assessment validity and biological plausibility.
CONCLUSIONS: The statistical associations reported in the few studies of specific Mn biomarkers and specific neurodevelopmental endpoints do not establish causal effects based on the Bradford Hill considerations. Additional prospective cohort studies of Mn biomarkers and validated neurodevelopmental outcomes, and a better understanding of the etiologic relevance of Mn biomarkers, are needed to shed light on whether environmental exposure to Mn causes adverse neurodevelopmental effects in children.