Evaluating the Mediating Role of Gestational Phthalate and Replacement Biomarkers in the Associations Between Race/ethnicity and Birthweight.

Background:Gestational exposure to phthalates is higher among non-Hispanic black (NHB) women than non-Hispanic white (NHW) women and may contribute to racial/ethnic differences in birthweight z-scores. Objective:We aimed to investigate whether gestational phthalate and replacement biomarkers mediate the association between race/ethnicity and birthweight z-scores. Methods:We analyzed data from 271 pregnancies in the Human Placenta and Phthalates study, a prospective cohort recruited from 2017 to 2020. Urinary phthalate metabolite concentrations were measured at up to eight time points during pregnancy and averaged to represent gestational exposure. Metabolites were grouped by parent compound. Linear regression models estimated associations in phthalate biomarkers by race and ethnicity, between race and ethnicity and birthweight z-scores, and between phthalate biomarkers and birthweight z-scores. Mediation analyses assessed whether phthalate biomarkers mediate the association between race/ethnicity and birthweight z-scores. Results:NHB participants exhibited higher concentrations of most phthalate biomarkers than NHW participants. For example, monoethyl phthalate concentrations were 250% higher (95% confidence interval [CI] = 173, 349). NHB participants had 0.43 lower birthweight z-scores (95% CI = -0.67, -0.19) than NHW participants. Monoethyl phthalate and summed metabolites of di-iso-butyl phthalate (ΣDiBP) and di-n-butyl phthalate were inversely associated with birthweight z-scores. ΣDiBP was associated with 0.16 (95% CI = -0.32, -0.002) lower birthweight z-scores. Mediation by phthalate biomarkers was modest and not statistically significant, with a ≤5% mediated effect. Conclusions:Racial/ethnic differences in phthalate exposure and birthweight z-scores were evident, and some phthalate biomarkers were associated with lower birthweight z-scores. However, phthalate exposures explained only a small, nonsignificant portion of the association between race and ethnicity and birthweight z-scores, underscoring the need to investigate additional contributors.