Prenatal, but not Postnatal, Exposure to Polychlorinated Biphenyls (PCBs) is associated with ADHD-Related Behaviors in 8-Year-Old Children

Abstract

Background: The neurotoxicity of prenatal PCB exposure has been established in both animal and human studies. Conversely, few human studies have replicated the behavioral effects of postnatal exposures that have been observed in monkeys, possibly because of challenges to postnatal exposure estimation. To refine exposure assessment, we previously developed and validated a pharmacokinetic model to estimate children’s PCB-153 levels across development. Aims: To assess whether postnatal exposure to PCBs, estimated using a pharmacokinetic model, is associated with ADHD-related behaviors in 8-year-olds enrolled in a birth cohort study of prenatal contaminant exposure and subsequent child development. Methods: We used a published pharmacokinetic model to estimate monthly PCB-153 levels in children through 12 months of age based on parameters such as breastfeeding, cord serum PCB-153 levels and growth. Behavior was evaluated using the Conner’s Rating Scales for teachers (CRS-T). The associations between PCB-153 levels and 4 CRS-T indices were assessed using multivariable quantile regression at the 50th and 75th percentiles, where higher percentiles reflect worse behavior. Results: Cord serum PCB-153 levels (median: 39 ng/g lipids, 5th-95th percentile: 7-317) were associated with ADHD-related behaviors, although statistical significance was only observed with quantile regression models at the 75th percentile. This association decreased in strength with later postnatal levels. For example, the 1.5 point (95 % CI: 0.3, 2.1) increase in Conner’s ADHD index score associated with an inter-quartile range change in cord serum PCB-153 levels weakened to 0.9 (95 % CI: 0.1, 3.0) and 0.1 (95 % CI:-0.3, 2) for estimated children’s levels at 2 and 12 months, respectively. Conclusions: Results suggest that the association between PCBs and behavior in 8-year-olds is mainly attributable to prenatal exposure and is more evident in children with worse behavioral scores.