Mediating effect of social determinants of health between race and phthalate biomarkers in pregnancy.

INTRO: Prenatal phthalate exposure is linked to adverse reproductive and developmental outcomes and is consistently higher among non-Hispanic Black compared to non-Hispanic White pregnant individuals, yet the factors responsible for these disparities are poorly characterized. We sought to quantify Black-White differences in repeated urinary concentrations of phthalate biomarkers during pregnancy and to estimate the extent to which neighborhood-level social determinants of health (SDoH) mediate these disparities.

METHODS

This secondary analysis used data from the Human Placenta Project Study, a prospective observational cohort study conducted at two academic institutions between 2016 and 2018. Eligible participants had singleton pregnancies, were recruited before 14 gestational weeks, and self-identified as Black or White; individuals identifying as Hispanic were excluded. Urine specimens were collected at up to 8 visits (12-38 weeks' gestation). Statistical analyses were performed from January to April 2025. Our exposures included self-identified race (Black vs White) and neighborhood SDoH, principally the 2018 Social Vulnerability Index (SVI) and the National Walkability Index, linked by geocoded residential address. Specific-gravity-adjusted concentrations of 14 phthalate metabolites and 4 replacement metabolites were quantified by high-performance liquid chromatography-tandem mass spectrometry. We used Bayesian multivariate mediation models to estimate adjusted geometric mean ratios (GMRs) with 95% credible intervals (95% CrI) for the total effect of race and the joint indirect effects via SVI and Walkability, and we reported the proportion mediated.

RESULTS

Among 249 participants (mean age, 26.5 years; 131 [52.6%] Black), 1562 urine samples were analyzed. Black participants had higher pregnancy-average geometric means of mono-ethyl phthalate (MEP; total GMR, 3.87; 95% CrI, 3.03-4.93), Σdi-n-butyl phthalate (ΣDnBP; 1.86; 1.52-2.27), and mono-benzyl phthalate (MBzP; total GMR, 1.41; 95% CrI, 1.04-1.90). The proportion mediated for SVI and Walkability jointly was 10% for MEP, 18% for ΣDnBP, and 44% for MBzP. No significant racial differences were observed for other phthalate biomarkers.

CONCLUSIONS

Black pregnant individuals had higher phthalate biomarker concentrations for several phthalates, and up to nearly half of the disparity for MBzP was explained by neighborhood-level SDoH. Interventions aimed at improving community social and built environments may help reduce racially patterned chemical exposures during pregnancy.