Prenatal-to-childhood urinary phenol and metal biomarkers in relation to BMI trajectories in a 10-year birth cohort.

Early-life exposure to phenols and metals is widespread, but evidence linking repeated exposure across development to long-term growth patterns remains limited. We analyzed 509 mother-child pairs from the Sheyang Mini Birth Cohort Study, in which urinary concentrations of three phenols (BPA, triclosan, and BP-3) and four metals (Pb, Cd, As, and Cr) were measured during pregnancy and at ages 1, 3, 8, and 10 years. BMI z-score trajectories were identified using latent class trajectory modeling, and associations were evaluated using Firth's logistic regression. Bayesian kernel machine regression and weighted quantile sum regression were used as complementary mixture analyses, and continuous-time Markov models were used secondarily to examine transitions between normal weight and overweight. Two trajectories were identified: stable lower (90.4%) and late-emerging higher (9.6%) BMI z-score trajectories. In the primary analysis, age-8 Cr showed a positive association with the late-emerging higher BMI z-score trajectory in the total population (OR = 3.35, 95% CI: 1.38-8.12), with an FDR-supported association among boys (OR = 6.41, 95% CI: 1.99-20.58; q < 0.05). Prenatal As in boys (OR = 1.57, 95% CI: 1.04-2.38) and age-1 BP-3 in girls (OR = 1.65, 95% CI: 1.01-2.71) showed suggestive associations (P < 0.05 but FDR-adjusted q ≥ 0.05) and were therefore interpreted as exploratory signals. Secondary Markov analyses provided borderline evidence linking age-1 BP-3 to lower recovery from overweight. Overall, age-8 Cr showed a positive trajectory-associated signal in the total population, and this association was FDR-supported among boys, whereas late-pregnancy As and age-1 BP-3 showed suggestive signals that merit further evaluation. These findings suggest that specific phenol and metal biomarkers may be related to adverse childhood BMI z-score patterning, but timing-specific interpretations remain exploratory and require confirmation in larger longitudinal studies.