Prenatal Phthalate Exposure and Social Responsiveness Scores in Early Childhood Using Quantile Regression: The HOME and EARLI Studies

Abstract

Background/Aim: Prenatal exposure to phthalates, non-persistent chemicals used in some consumer products, may adversely affect neurodevelopment. However, linear regression estimates effects at the mean of a neurodevelopmental trait and the effect of phthalates may be underestimated if sub-populations at the ‘tails’ of distributions are more susceptible. To address this gap, we assessed the relation between prenatal phthalate exposure and children’s Autism Spectrum Disorder (ASD)-related behaviors using quantile regression. Methods: We used harmonized data from the Health Outcomes and Measures of the Environment (HOME) Study, a general population cohort (n= 276) and Early Autism Risk Longitudinal Investigation (EARLI) (n=145) Study, an enriched risk cohort of moms who had a child with ASD. We measured maternal concentrations of 9 phthalate metabolites in urine samples collected twice during pregnancy. Caregivers reported children’s behaviors associated with ASD on the Social Responsiveness Scale (SRS) at age 3-8 years; higher scores indicate more ASD-related behaviors. We estimated covariate-adjusted differences in continuous SRS T-scores with increasing log10-transformed phthalate metabolite concentrations using quantile regression. Results: In HOME, monobenzyl phthalate, monoethyl phthalate, and di-2-ethylhexyl phthalate (ΣDEHP) metabolite concentrations were modestly associated with increased SRS T-scores at the 50th or 75th percentile compared to the 25th. For example, each 10-fold increase in ΣDEHP concentrations was associated with higher SRS T-scores at the 50th (β:3; 95% CI:2, 7) and 75th (β:3; 95% CI:-1, 7) percentiles, but not the 25th (β:1; 95% CI: 0, 5). In EARLI, there were predominately inverse and less precise associations between phthalate metabolite concentrations and SRS T-scores across percentiles.Conclusion: In the HOME Study, concentrations of some phthalate were associated with more autistic behaviors among children with higher SRS T-scores; inconsistencies with the EARLI Study may be due to differences in cohort specific participant characteristics.