Fig. 2: Covariate-adjusted associations between self-reported racial and ethnic identity and urinary chemical concentrations (µg/L) among pregnant ECHO participants.

Effect estimates are ratios of geometric means and 95% confidence intervals from individual linear mixed effect censored-response regression models of specific gravity/creatinine-corrected urinary phenol concentrations as outcomes and maternal racial and ethnic identity categories as predictors (non-Hispanic White = reference category), a random intercept on pregnancy to account for multiple urinary measurements and adjusted for maternal age (years), pre-pregnancy body mass index (kg/m2), educational level (completed vs. did not complete bachelor’s degree), gestational age at biospecimen collection (weeks), season of biospecimen collection (fall vs. winter vs. spring vs. summer), and ECHO study cohort (11 cohorts). Abbreviations: BPA bisphenol A, BPF bisphenol F, BPS bisphenol S, ECHO Environmental Influences on Child Health Outcomes, Mult/Oth/Asian non-Hispanic multiple races, “Other,” and Asian.