Discordant sibling analysis of autism risk associated with prenatal exposure to tailpipe and non-tailpipe particulate matter pollution.

Background: We previously assessed associations of prenatal exposure to fine particulate matter (PM2.5) tracers reflecting tailpipe (elemental carbon [EC] and organic carbon [OC]) and non-tailpipe emissions (copper [Cu], iron [Fe] and manganese [Mn]) with risk of autism spectrum disorder (ASD) in a large pregnancy cohort. To address genetic and family environment confounding, we conducted an ASD-discordant sibling study.

Methods: Data included 4024 children (1837 with and 2187 siblings without ASD) born to 1801 unique mothers who had at least one child diagnosed with ASD by age 5, and one child without ASD. Prenatal exposures to total PM2.5, trace elements Cu, Fe, Mn, EC, and OC and dispersion-modeled near-roadway-air-pollution (NRAP) from freeway and non-freeway source were obtained using maternal addresses during pregnancy. Conditional logistic regression was used to assess ASD risk associated with exposures adjusting for covariates. Results were reported as odds ratio (OR, 95 % CI) per inter-quartile increment of each exposure.

Results: In single-pollutant models, child ASD risk (OR; 95 % CI) was associated with gestational exposures to non-tailpipe source Cu (1.17; 1.03-1.33), Fe (1.26; 1.07-1.48), Mn (1.29; 1.11-1.50); but not likewise associated with tailpipe source EC (1.10; 0.92-1.32) and OC (1.10; 0.91-1.32). Total PM2.5 and non-freeway NRAP were both associated with ASD risk. Adjusting for total PM2.5 or NRAP attenuated the ASD associations with Cu, Fe, and Mn but they remained largely statistically significant. By trimester analysis showed the associations with Cu, Fe, and Mn were significant in the first two trimesters.

Conclusions: This ASD-discordant sibling study confirmed previously reported ASD risk associated with prenatal exposure to PM2.5, NRAP and non-tailpipe particulate trace-element Cu, Fe, and Mn, particularly in the first two trimesters, thus, increasing evidence of causality.