Associations between Prenatal Perceived Stress and Child Autism-Related Traits in the ECHO Cohort.

Abstract

To examine associations of prenatal perceived stress with child autism-related traits and to evaluate effect modification by both child sex and co-occurring prenatal depressive symptoms. The sample was drawn from Environmental Influences on Child Health Outcomes cohort sites with data on prenatal perceived stress (Perceived Stress Scale [PSS]) and autism-related traits (Social Responsiveness Scale, second Edition [SRS-2]). We used linear and logistic regression models to determine associations between PSS and either SRS-2 T-scores or odds of moderate-to-severe autism-related traits (SRS-2 T-score ≥65) and evaluated effect modification by sex and prenatal depressive symptoms. Among 4115 mother-child pairs, higher PSS was associated with increased severity of autism-related traits (β = 0.15 [95% CI = 0.13-0.18]) and odds of moderate-to-severe traits (OR = 1.04 [1.04-1.05]). There was evidence for sex-by-exposure interaction for continuous (β_Males = 0.18 [0.14-0.22]; β_Females = 0.13 [0.10-0.16], P_interaction = 0.03) but not dichotomous trait severity (OR_Males = 1.05 [1.03-1.06]; OR_Females = 1.04 [1.03-1.05], P_interaction = 0.52). Associations did not differ by prenatal depressive symptoms (low: n = 2,538, b = 0.14 [0.11-0.18]; high: n = 889, b = 0.10 [0.04-0.15]; P_interaction = 0.66). We found small yet significant associations between prenatal stress and autism-related traits. Limited evidence for effect modification by sex or prenatal depressive symptoms suggests prenatal perceived stress influences autism-related traits after accounting for these factors.