Air Pollution, Developmental Delays and Autism Spectrum Disorder in the Early Markers of Autism (EMA) Study

Background:  Many epidemiologic studies have suggested that air pollutants may play a role in impaired neurodevelopment and more specifically in the etiology of ASD However, it is unclear whether the air pollution effects are ASD-specific or extend to individuals with developmental delays (DD).  Studies examining exposures during the gestational period are also needed to better understand the role of air pollution in autism risk.

Objectives: To examine the relationship between prenatal and first year of life air pollution exposure and ASD or DD risk in the Early Markers for Autism (EMA) study.  

Methods: This study includes 420 children with ASD, 169 with DD, and 428 general population (GP) controls from a population-based case-control study of pregnant women in San Diego, Orange, and Imperial counties created through the linkage of California birth records, Department of Developmental Services (DDS) records, and the California Prenatal and Newborn Screening Program records.  All diagnoses were derived from DDS records and verified by expert medical record review.  The maternal residence on the birth certificate and addresses reported during prenatal screening visits were used to estimate exposure for each trimester of pregnancy and first year of life. Regional air pollutant measures (NO₂, PM₁₀, PM_2.5, Ozone) were based on the Environmental Protection Agency’s Air Quality System data and near roadway air pollution (NRAP) estimates from the CALINE-4 traffic dispersion model. Logistic regression models were used to determine the association between estimated air pollutant levels and ASD or DD risk, compared to GP controls.  We further examined if this relationship differed among children with ASD and comorbid intellectual disability (ID) (n=184), as compared to GP controls.  Models were adjusted for maternal education, method of prenatal payment, and county of birth.

Results: Comparison of the highest and lowest quartiles of the air pollution distribution showed non-statistically significant increased ASD risk for prenatal and first year of life exposure to PM_2.5, PM₁₀, and NO₂.  Analyses with continuous measures of air pollutants during pregnancy or the first year of life showed no significant associations with ASD risk after adjusting for confounders, in particular county of residence at birth.  Examination of the subset of children with ASD+ID indicated an increased risk per 2 standard deviation (2 SD) increase in ozone exposure in the first year of life [aOR=1.52 (1.04-2.23) per 6.7 ppb] relative to GP controls.  Increasing PM₁₀ exposure during the first trimester was significantly associated with DD risk [aOR=2.09 (1.10-4.04 per 16.8 ug/m³ (per 2SD increase))], after adjusting for all confounders.

Conclusions: Air pollutant exposure may have broad neurodevelopmental effects, which manifest differently across ASD, DD, and ID phenotypes. In contrast to previous studies we did not observe a statistically significant association of air pollutants with ASD risk overall. Differences in case ascertainment, ASD severity, and pollutant exposure range may have affected our results and should be considered in further examinations of geographically-defined exposures for ASD.