Longitudinal PFAS exposure and thyroid function trajectories in Taiwanese youth: a 10-year prospective cohort study

Evidence linking per- and polyfluoroalkyl substances (PFAS) to thyroid dysfunction is growing, yet longitudinal data spanning adolescence to early adulthood remained limited.

We analyzed 495 participants aged 12-30 years in the YOung TAiwanese Cohort (YOTA) followed for 9-12 years. 11 Plasma PFAS and serum thyroid-stimulating hormone (TSH) and free thyroxine (free T4) were measured at baseline and follow-up. PFAS exposure index was derived as the mean of these standardized ln-PFAS concentrations. Using multivariable ANCOVA-style linear regression models, we examined associations of baseline PFAS levels and changes in PFAS with follow-up ln-TSH and ln-free T4, and incident thyroid disease was examined using Cox models.

In multivariable models, both higher baseline PFAS exposure index and a more positive change in PFAS exposure index were associated with lower follow-up TSH (-17.22%, p = 0.003; and - 16.81%, p = 0.011, respectively). Several baseline PFAS compounds showed inverse associations with follow-up TSH, including linear PFOA, linear PFOS, N-MeFOSAA, and PFHpA; inverse associations for PFAS changes were most evident for linear PFOA and linear PFOS. For free T4, baseline PFAS measures were not significant after FDR correction, whereas a more positive change in PFAS exposure index was associated with higher follow-up free T4 (+5.55%, p = 0.001), with nominal positive associations observed for linear PFOA and PFDoA. These patterns were robust in sensitivity analyses (including baseline-only/change-only specifications, change-score models, and exclusions of baseline/follow-up thyroid dysfunction). In Cox regression analyses, greater annualized increases in the PFAS exposure index were associated with lower risks of hypothyroidism and thyroid nodules.

PFAS exposure was associated with lower TSH and higher free T4 over 10 years of follow-up, suggesting sustained thyroid-axis perturbation from adolescence into young adulthood. These findings underscore a susceptible developmental window and support long-term biomonitoring and stronger regulatory action where PFAS contamination persists.