Prenatal bisphenols and ambient air pollutants are differentially associated with early childhood length and weight from birth to 36 months: a multi-domain mixture analysis in a Korean birth cohort.
Prenatal exposure to environmental chemicals has been implicated in impaired fetal and early childhood growth. However, most previous studies have focused on single exposure domains, limiting understanding of the combined effects of co-occurring chemical mixtures. We investigated the joint effects of a prenatal multi-domain chemical mixture comprising endocrine-disrupting chemicals (EDCs), heavy metals, and ambient air pollutants on infant growth from birth to 36 months using data from the Korean Children's Environmental Health Study (Ko-CHENS), a nationwide prospective birth cohort of 5,056 mother-child pairs. We applied weighted quantile sum (WQS) regression and Bayesian kernel machine regression (BKMR) to estimate mixture effects and identify key contributing exposures, WQS and BKMR were fitted separately at each of five assessment ages, and a latent class mixed model (LCMM) was used to characterize prenatal exposure trajectory subgroups, whose association with repeated growth measures was then evaluated in a linear mixed model. Ten mixture-outcome tests (two outcomes x five ages) were pre-specified as primary and were corrected for multiplicity using the Benjamini-Hochberg false discovery rate (FDR) procedure. WQS regression identified a significant inverse association between the prenatal chemical mixture and infant length-for-age z-scores at birth and 6 months, with late-pregnancy BPF and BPS consistently emerging as the top-weighted contributors; both associations remained below a 5% FDR threshold. Overall mixture effects estimated by BKMR were null at every assessment age, although BKMR suggested nonlinear and exposure-specific dose-response relationships involving bisphenol analogues and ambient air pollutants across multiple assessment time points. No significant associations were observed between LCMM-derived exposure trajectory class and longitudinal growth outcomes through 36 months. Domain-specific patterns were observed across both analyses: bisphenol analogues showed more consistent associations with length outcomes, whereas ambient air pollutants were more frequently identified as key contributors in weight-related models. These exploratory findings suggest that prenatal exposure to bisphenol analogues and ambient air pollutants may be associated with early childhood growth and warrants further investigation. They support the use of mixture-based approaches in environmental health research but should not be interpreted as evidence of causal effects.