Environ Int. 2026 May 12;212:110302. doi: 10.1016/j.envint.2026.110302. Online ahead of print.
ABSTRACT
BACKGROUND: Exposure to per- and polyfluoroalkyl substances (PFAS) and phthalates is widespread during adolescence, a critical developmental period for metabolic regulation.
AIM: To assess associations of serum PFAS and urinary phthalate/DINCH metabolites-individually and as mixtures-with a unified panel of metabolic biomarkers representing adipose-brain-liver cross-talk and oxidative stress in European teenagers.
METHODOLOGY: Serum PFAS and urinary phthalate/DINCH metabolites were measured in 1033 European teenagers (12-17 years) from the Human Biomonitoring Initiative for Europe (HBM4EU) Aligned Studies. Metabolic biomarkers representing adipose (HDL, LDL, cholesterol, and triglycerides)-brain (leptin, adiponectin, and kisspeptin)-liver (glucose, insulin) cross-talk and oxidative stress (8-hydroxy-2′-deoxyguanosine, 8OHdG) were measured. The Body Mass Index z-score (zBMI) was calculated. Single pollutant models, multivariate MANOVA, quantile g-computation, and BKMR models were fit, including interaction terms with sex.
RESULTS: Single pollutant models showed positive associations of PFAS and phthalate/DINCH metabolites with 8OHdG. PFAS were associated with higher leptin, HDL, LDL, and cholesterol, while some phthalate/DINCH metabolites were associated with lower kisspeptin, HDL, triglycerides, cholesterol, zBMI, and higher adiponectin. We observed weak but statistically significant associations between PFAS and phthalate/DINCH metabolites with the entire set of metabolic biomarkers in the MANOVA. The PFAS mixture was associated with higher kisspeptin, LDL, HDL, cholesterol, and 8OHdG. The phthalate/DINCH mixture was associated with lower HDL.
CONCLUSIONS: Exposure to these contaminants may be related to dyslipidemia in teenagers. PFAS and phthalate/DINCH metabolites may exert opposite associations on metabolism, with the exception of increasing oxidative stress. Given the cross‑sectional design and potential residual confounding, longitudinal studies are warranted.
PMID:42150235 | DOI:10.1016/j.envint.2026.110302
