Prenatal exposure to mixtures of endocrine-disrupting chemicals and autism spectrum disorder

Abstract

BACKGROUND AND AIM: Mixtures methods have the potential to address the unclear relationship between prenatal exposure to endocrine-disrupting chemicals (EDCs) and risk of autism spectrum disorder (ASD). However, mixtures studies in this area are currently sparse. This study examines the relationship of prenatal exposure to a mixture of 25 EDCs, representing 4 classes of persistent pollutants, and ASD. METHODS: We analyzed data from the Early Markers for Autism (EMA) Study, a population-based, case-control study of children born during 2000-2003 in Southern California. Children diagnosed with ASD (n=491) were identified from clinical records of the Department of Developmental Services and were confirmed by expert review. General population (GP) controls (n=373) were randomly sampled from birth certificate files and frequency matched to children with ASD by sex, birth month, and birth year. Eleven polychlorinated biphenyls (PCBs), 6 brominated flame retardants (BFRs), 2 organochlorine pesticides (OCPs), and 6 polyfluoroalkyl substances (PFAS) were detectable in maternal serum samples collected during the second trimester. To examine the relationship between ASD and EDC mixtures, both within each chemical subclass and across all EDCs, we applied two METHODS: weighted quantile sums regression (WQSR) with repeated-holdout validation and Bayesian kernel machine regression (BKMR). RESULTS:In WQSR adjusted for covariates, PCB, PFAS, and BFR mixtures were not associated with ASD when constrained to the positive direction, but BFRs were associated with ASD in the negative direction (aOR: 0.82; 95%CI: 0.71, 0.95 per quartile increase). Results from BKMR were generally consistent with WQSR, with PBDE 153 contributing the most to both BFR mixture models. WQSR and BKMR analyses showed null associations between the aggregate EDC mixture and ASD. BKMR analyses did not indicate interaction between any of the EDC biomarkers. CONCLUSIONS:Results suggest that prenatal exposure to the mixture of these EDCs is not adversely associated with overall ASD risk. KEYWORDS: autism, mixtures, endocrine-disrupting compounds, prenatal