Critical windows of susceptibility in the association between manganese and neurocognition in Italian adolescents living near ferro-manganese industry

Introduction: Understanding the neurodevelopmental effects of manganese (Mn) is complicated due to its essentiality for growth and development. While evidence exists for the harmful effects of excess Mn, pediatric epidemiologic studies have observed inconsistent associations between Mn and child cognition. Objective: We sought to estimate prospective associations between Mn measured in three different early-life time windows with adolescent cognition using deciduous teeth biomarkers. Methods: Deciduous teeth were collected from 195 participants (ages 10–14 years) of the Public Health Impact of Manganese Exposure (PHIME) study in Brescia, Italy. Measurements of tooth Mn represented prenatal (∼14 weeks gestation – birth), early postnatal (birth – 1.5 years) and childhood (∼1.5 – 6 years) time windows. Neuropsychologists administered the Wechsler Intelligence Scale for Children, 3rd edition (WISC-III), to obtain composite IQ and subtest scores. Associations between tooth Mn at each time window and adolescent WISC-III scores were estimated using multivariable linear regression. We tested differences in associations between Mn and outcomes across time windows using multiple informant models. Sex-specific associations were explored in stratified models. Results: Adjusted associations between tooth Mn and composite IQ scores were positive in the prenatal period and negative in the childhood period. Associations were strongest for subtest scores that reflect working memory, problem solving, visuospatial ability and attention: prenatal Mn was positively associated with Digits backward [SD change in score per interquartile range increase in Mn: β = 0.20 (95 % CI: 0.02, 0.38)] and Block design [β = 0.21 (0.01, 0.41)] and early postnatal Mn was positively associated with Digits forward [β = 0.24 (0.09, 0.40)], while childhood Mn was negatively associated with Coding [β = -0.14 (-0.28, -0.001)]. Sex-stratified analyses suggested different Mn-cognition associations for boys and girls and was also dependent on the time window of exposure. Conclusion: Our results suggest that exposure timing is critical when evaluating Mn associations between Mn and cognition. Higher prenatal Mn was beneficial for adolescent cognition; however, these beneficial associations shifted towards harmful effects in later time windows. Cognitive domains most sensitive to Mn across time windows included visuospatial ability, working memory, attention and problem-solving.