Fetal exposure to persistent organic pollutants and childhood autism risk.

Abstract

Exposure to persistent organic pollutants (POPs) during fetal development may increase the risk of autism spectrum disorder, but previous studies have relied on maternal pregnancy samples as proxies instead of directly measuring fetal exposures. Here, we used naturally shed deciduous teeth from autism cases and typically developing controls born between 1998 and 2014 from the California, USA, population-based case-control Childhood Autism Risk from Genetics and the Environment study. Using gas chromatography-tandem mass spectrometry, we quantified fetal second and third trimester and postnatal POPs in teeth dentine. In total, seven polychlorinated biphenyls, two organochlorine pesticides, and four brominated diphenyl ethers were quantified in deciduous teeth. We compared children clinically confirmed to have autism with typically developing children from the general population using logistic regression for exploratory single chemical analysis and Weighted Quantile Regression Analysis for mixture analysis. In the third trimester, the POPs mixture was positively associated with autism (median Odds Ratio (OR) = 1.54 (95% CI=[0.99, 2.83]). In sex-stratified interaction analysis, the OR for males (median=1.75) was significant (95% CI: [1.13, 3.2]), while the OR for females was small (1.16) and not significant (95% CI = [0.33, 3.5]). The POPs mixtures during the second trimester or postnatally were not significantly associated with autism risk. Therefore, direct fetal and early-life measures of POPs indicate the third trimester as a potential high-risk biological window for POPs exposures that increase autism risk, particularly in males.