The potential effects of hypothetical PM2.5 interventions on childhood autism in different neighborhood socioeconomic contexts.

Abstract

Particulate air pollution is associated with autism spectrum disorder (ASD), with disadvantaged neighborhoods potentially increasing vulnerability due to stress or other social determinants of health. Understanding the impact of air pollution interventions on ASD incidence across neighborhood disadvantage levels can guide policies to protect vulnerable populations. We examined two sets of hypothetical PM2.5 interventions: percentage reduction and regulatory standards as thresholds, to assess their potential effects on ASD cumulative incidence. Using G-computation under a counterfactual framework, we estimated changes in the cumulative incidence of ASD by age 5 under hypothetical interventions compared to observed exposures. Our study involved a birth cohort of 318,298 children born between 2001-2014 in Southern California, with 4,548 diagnosed with ASD by age 5. Pregnancy average PM2.5 and neighborhood disadvantage were assigned to residential addresses. Adjusted Cox regression models were applied to estimate ASD cumulative incidence. Reducing pregnancy average PM2.5 by 30% or below 9 μg/m3 would have prevented 10.6 (95% CI, 3.6-19.2) and 12.5 (2.7-23.6) ASD cases per 10,000 children, respectively. The decreases in ASD cumulative incidence under hypothetical interventions were similar across neighborhood disadvantage levels. These findings suggest that reducing ambient PM2.5 levels to meet or surpass current standards could help prevent ASD.