Associations of Autism Spectrum Disorder with Residential Air Pollution Exposure in a Large Southern California Pregnancy Cohort

Background:  Autism spectrum disorder (ASD) prevalence has increased over several decades to its currently estimated 1 in 68 children in the United States. Recent research suggests that adverse child neurodevelopmental outcomes, including ASD, may occur in association with regional and near-roadway air pollution exposures. There have been limited longitudinal analyses of population-based cohort studies that have addressed the relationship between air pollution and ASD, and results from previous case-control studies have not been consistent.

Objectives:  This retrospective cohort study evaluated the associations of ASD risk with prenatal exposure to regional air pollutants, including particulate matter (PM) ≤2.5 μm in aerodynamic diameter (PM_2.5), ≤10 μm (PM₁₀), nitrogen dioxide (NO₂), and ozone (O₃).

Methods:  We retrospectively identified a cohort of children born in Kaiser Permanente Southern California (KPSC) hospitals between 1995 and 2009 (n=294,637), reviewing electronic medical records from birth until the first date of clinical diagnosis of ASD, last date of continuous KPSC health plan membership, death from any cause, or age 5. Birth certificate addresses were geocoded and used as the residential locations where birth-year annual average exposure to regional particulate and gaseous air pollutants (PM_2.5, PM₁₀, NO₂, and O3) were estimated by inverse distance weighting to the network of Southern California monitoring stations, and distances to major roadways and freeways were calculated for 294,637 mother-child pairs. ASD diagnosis was defined based on ICD-9 diagnostic codes from at least 2 separate KPSC hospital visits. Relative risks of ASD were estimated by hazard ratios (HRs) using Cox regression models adjusted for potential confounders.

Results:  There were 2414 children with a diagnosis of ASD by age 5. After adjusting birth year, for maternal age, parity, education, household income, race/ethnicity, history of comorbidities (≥1 diagnosis of heart, lung, kidney, or liver disease; cancer), sex of the child, and KPSC medical center service areas, HRs scaled to interquartile range of the distribution of each pollutant were 1.18 (95% confidence interval [CI], 1.05-1.34) per 8.74 μg/m³ PM_2.5 1.10 (95% CI, 1.00-1.21) per 11.3 μg/m³ PM₁₀; 0.95 (95% CI, 0.85-1.07) per 11.9 ppb NO₂; and 0.79 (95% CI, 0.71-0.88) per 10.8 ppb O₃. No associations of residential distance to major roadways and freeways were observed. HRs for air pollutants did not change substantially after additionally adjusting for preeclampsia, maternal diabetes, and preterm birth.

Conclusions:  Prenatal exposure to PM_2.5 and PM₁₀, were associated with risk of ASD in children from a large, prospectively analyzed birth cohort in Southern California. Inverse associations with O₃ merit further investigation.