Prenatal Air Pollution Exposure and Autism Spectrum Disorder in the ECHO Consortium
Recommended Citation
Ghassabian A, Dickerson AS, Wang Y, Braun JM, Bennett DH, Croen LA, LeWinn KZ, Burris HH, Habre R, Lyall K, Frazier JA, Glass HC, Hooper SR, Joseph RM, Karr CJ, Schmidt RJ, Friedman C, Karagas MR, Stroustrup A, Straughen JK, Dunlop AL, Ganiban JM, Leve LD, Wright RJ, McEvoy CT, Hipwell AE, Giardino AP, Santos HP, Jr., Krause H, Oken E, Camargo CA, Jr., Oh J, Loftus C, O'Shea TM, O'Connor TG, Szpiro A, and Volk HE. Prenatal Air Pollution Exposure and Autism Spectrum Disorder in the ECHO Consortium. Environ Health Perspect 2025.
Document Type
Article
Publication Date
6-11-2025
Publication Title
Environmental health perspectives
Abstract
BACKGROUND: The relationship between prenatal exposure to low-level air pollution and child autism spectrum disorder (ASD) is unclear.
OBJECTIVE: To examine associations of prenatal air pollution exposure with autism.
METHODS: We analyzed data from 8,035 mother-child pairs from 44 United States cohorts in the Environmental influences on Child Health Outcomes (ECHO) Cohort. Fine particulate matter (PM(2.5)), nitrogen dioxide (NO(2)), and 8-hour-max ozone (O(3)) levels were estimated at residential addresses during pregnancy. Parents rated children's autism-related traits using the Social Responsiveness Scale (SRS) (mean age 9.4 years, SD=3.6) and reported physician-diagnosed ASD. We examined associations of the three air pollutants with SRS scores (10(th), 50(th), and 90(th) quantiles) using quantile regression and with ASD diagnosis using logistic regression. Models were run within census divisions, and coefficients were pooled in a meta-analysis.
RESULTS: Average (SD) pregnancy exposures were 9.3 µg/m3 (2.7) for PM(2.5), 21.8 ppb (8.8) for NO(2), and 40.3 ppb (5.5) for O(3), with variations across census divisions. The median SRS T-score was 46 (IQR=41 to 52), and 444 children (5.5%) had an ASD diagnosis. Higher PM(2.5) was associated with higher SRS scores at the 10(th) quantile (β=0.74, 95% CI: 0.09, 1.40) but not at the median or highest quantile. The association between PM(2.5) and ASD diagnosis was highly heterogeneous, with associations present in the South Central, Mountain, and Pacific census divisions. Heterogeneity was also high in the association between NO(2) and SRS at the median and only in the mid-Atlantic, West North Central, and South Atlantic census divisions. Higher O(3) was associated with higher SRS scores at the median (β per IQR increment=0.83, 95% CI: 0.05, 1.61) and highest quantile (β=2.19, 95% CI: 0.06, 4.32) in the meta-analysis. Higher O(3) also was associated with ASD.
DISCUSSION: Associations with ASD outcomes were present even at low levels of air pollutants. https://doi.org/10.1289/EHP16675.
PubMed ID
40498638
ePublication
ePub ahead of print
