Neurobehavioral Endpoints in Children Affected by Environmental Polychlorinated Biphenyls

Abstract

MAA2-PD-13 Introduction: The aim of this study was to contribute to the knowledge of neurobehavioral changes in children 8 to 9 years of age (n = 435). Methods: Neurobehavioral examinations included sensomotor integration, eye–hand coordination, attention, and memory and cognitive processes. A general health/social questionnaire and behavioral scale were filled in by parents and the behavior at school was evaluated by teacher. Statistical analysis involved 3 types of data evaluation: The PCBs serum concentrations and selected heavy metals (Pb, Mn, Hg, in whole blood) also were determined. The PCB level was positively skewed and normalized by a natural logarithmic transformation (Ln) and calculated in quartiles for descriptive purposes. Correlation coefficients were calculated between dependent neurobehavioral variables and both the exposure level of PCBs and confounding variables (<0.05). To test for the validity of these correlations, we conducted multiple linear regression analyses to adjust for confounding. Results: The σPCB-concentrations in serum ranged from 1692.8 to 257.5 ng/g lipids; higher levels were found close to the PCBs source, a former PCB-production plant. Significantly poorer sensomotor performances were found in preliminary evaluations in children with higher PCB levels (fourth quartile of PCB) but were no longer apparent after adjustment for schools or residential areas. Older age and greater maternal intelligence were associated with better performances in cognitive and some sensomotor tests. Previous findings of an inverse influence of preperinatal PCBs on neurobehavioral performances were confirmed only if study area and schools were not included in the full regression model. The outcome of neuropsychological tests and scales of behavior were included into this analysis as dependent variables, and a selected set of covariates, including PCBs in serum concentrations, were used as independent predictors. Because the 2 areas and the schools within areas differed for both exposure and several of the outcome variables exposure-outcome associations were additionally adjusted for “area” “schools” and “area plus schools” in the multifactor regression analysis. The final regression models included the set of variables (σPCBs, neurobehavioral and social). Discussion and Conclusions: The results suggest that postnatal or concurrent exposure to PCBs does not exhibit inverse association with neurobehavior if study area is taken into consideration in the full regression model. This outcome must be compared with well-established findings of neurobehavioral adversity of environmental PCB exposure in relation to preperinatal exposure and emphasizes the higher vulnerability of the developing relative to the mature brain.