Associations of pre- and postnatal per- and polyfluoroalkyl substance exposure with adolescents' eating behaviors

Per- and polyfluoroalkyl substances (PFAS), persistent environmental chemicals, may act as obesogens by interacting with neuroendocrine pathways regulating energy homeostasis and satiety signals influencing adolescent eating behaviors.

In 211 HOME Study adolescents (Cincinnati, OH; recruited 2003-2006), we measured PFAS concentrations in serum collected during pregnancy, at delivery, and at ages 3, 8, and 12 years. Caregivers completed the Child Eating Behavior Questionnaire (CEBQ) at age 12, and we calculated food approach and food avoidance scores. Using quantile-based g-computation, we estimated covariate-adjusted associations between a mixture of four gestational PFAS and CEBQ scores. We identified high (n = 76, 36%) and low (n = 135, 64%) longitudinal PFAS mixture exposure profiles between delivery and age 12 years using latent profile analysis and related these to CEBQ scores. We examined whether child sex or physical activity modified these associations.

We observed no association of gestational PFAS mixture with food approach or food avoidance scores. Children in the higher longitudinal PFAS mixture profile had slightly higher food approach scores (β: 0.47, 95% CI: -0.27, 1.23) and similar food avoidance scores (β: -0.15, 95% CI: -0.75, 0.46) compared with children in the lower profile. We found some evidence that higher physical activity favorably modified the association between longitudinal PFAS mixture profiles and emotional overeating (interaction P value = 0.13). Child sex did not consistently modify any associations.

Serum PFAS concentrations were not consistently linked to adolescent eating behaviors in this study, suggesting alternative pathways, such as metabolic rate, may underlie previously observed associations between PFAS exposure and childhood obesity.