Prenatal phthalate exposures and adiposity outcomes trajectories: a multivariate Bayesian factor regression approach

2025-06-03

AI Summary

Prior studies examining prenatal phthalate exposure and childhood obesity have typically analyzed single anthropometric outcomes (e.g., BMI z-score, body fat percentage, waist circumference) in isolation and struggled with collinearity in chemical mixture exposures. Method: We developed a novel multivariate Bayesian factor regression model that jointly models multiple correlated obesity-related outcomes while accommodating time-varying and sex-specific exposure effects—enabling cross-outcome information sharing. Using data from the Mount Sinai Children’s Environmental Health Cohort, we implemented Markov Chain Monte Carlo (MCMC) inference and simulation-based validation. Results: Among boys, prenatal phthalate exposure exhibited an inverted U-shaped association with adiposity—negative in early childhood (ages 4–5) and positive later (ages 6–7)—suggesting critical windows of metabolic programming during development. This work establishes a new methodological paradigm for causal inference on environmental mixtures and multidimensional health outcomes.

Technology Category

• Machine Learning: Bayesian Learning
• Reasoning under Uncertainty: Relational Probabilistic Models
• Cognitive Modeling & Cognitive Systems: Adaptive Behavior

Application Category

• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Web Mining and Content Analysis: Models for Web evolution

Abstract

We aim to assess the longitudinal effects of prenatal exposure to phthalates on the risk of childhood obesity in children aged 4 to 7, with potential time-varying and sex-specific effects. Multiple body-composition-related outcomes, such as BMI z-score, fat mass percentage, and waist circumference, are available in the data. Existing chemical mixture analyses often look at these outcomes individually due to the limited availability of multivariate models for mixture exposures. We propose a multivariate Bayesian factor regression that handles multicollinearity in chemical exposures and borrows information across highly correlated outcomes to improve estimation efficiency. We demonstrate the proposed method's utility through simulation studies and an analysis of data from the Mount Sinai Children's Environmental Health Study. We find the associations between prenatal phthalate exposures and adiposity outcomes in male children to be negative at early ages but to become positive as the children get older.

Problem

• Assess prenatal phthalate effects on childhood obesity risk
• Analyze multiple adiposity outcomes jointly with multivariate model
• Evaluate time-varying and sex-specific phthalate exposure impacts

Innovation

• Multivariate Bayesian factor regression model
• Handles multicollinearity in chemical exposures
• Borrows information across correlated outcomes