Prenatal Serum Levels of Brominated Flame Retardants in Association with Autism Spectrum Disorder and Intellectual Disability: Potential Sex Differences

Background: Prior work suggests prenatal exposure to polybrominated diphenyl ethers (PBDEs) may have neurodevelopmental impacts. However, few human studies have prospectively examined levels of these chemicals during suspected critical windows of neurodevelopment in association with autism and other developmental disorders.

Objectives: To determine whether prenatal exposure to brominated flame retardants influences risk of autism spectrum disorder (ASD) or intellectual disability without autism (ID).

Methods:  We conducted a population-based case-control study. Participants were from the Early Markers of Autism (EMA) study. Children with ASD (n=545) and ID (n=181) were identified from the Department of Developmental Services (DDS), with diagnoses confirmed by expert clinician review, and ASD cases meeting Diagnostic and Statistical Manual of Mental Disorders (DSM)-IV-TR criteria. General population (GP) controls (n=418) were randomly selected from birth certificates of the same birth years as cases (2000-2003) after excluding those served by DDS. Concentrations of nine PBDE congeners and 2,2’,4,4’,5,5’-hexabromobiphenyl were measured in maternal 2^nd trimester serum samples. Logistic regression was used to calculate crude and adjusted odds ratios (AOR) for associations with ASD, and separately for ID, compared to GP controls, by quartiles of analyte concentrations in primary analyses. Stratified analyses examined potential effect modification by offspring sex as well as potential differences by ASD with and without comorbid ID. Sensitivity analyses were conducted to test the robustness of results to a number of alternate modeling strategies.

Results: Six congeners were frequently detected in this study population and evaluated in further analysis. Levels of most PBDE congeners were lower in children with ASD or ID relative to GP controls. In adjusted analyses, inverse associations were found for ASD relative to GP, comparing the highest quartile of concentration to the lowest (AOR=0.56, 95% CI 0.38, 0.84 for BDE-153, and AOR=0.54; 95% CI 0.36, 0.80 for the sum of detected congeners). Results differed for odds of ASD when stratified by child sex, with estimates above the null in girls, and confidence intervals closely overlapping 1.0 for BDE-28 and -47. Similar patterns were observed for ID. No differences in risk were noted for ASD with or without ID. Sensitivity analyses did not alter findings.

Conclusions: Contrary to expectation, results did not suggest increased risk of ASD or ID with higher levels of prenatal PBDEs. However, these findings suggest potential sexual dimorphism in neurodevelopmental effects of prenatal exposure to brominated flame retardants.