Longitudinal modeling of per- and polyfluoroalkyl substance exposure in relation to changes in oxidative DNA damage in Taiwanese youth: A 10-year prospective cohort study

Per- and polyfluoroalkyl substances (PFAS) are persistent environmental contaminants implicated in oxidative stress–related health outcomes. However, most epidemiological studies rely on single baseline measurements, limiting insight into how longitudinal changes in PFAS exposure relate to oxidative DNA damage over time. We conducted a longitudinal analysis among 521 participants from the Young Taiwanese Cohort (YOTA), who were aged 12–30 years at baseline (2006–2008) and re-examined at follow-up (2017–2019), with paired measurements of 11 serum PFAS and urinary 8-hydroxy-2′-deoxyguanosine (8-OHdG). PFAS exposures were evaluated using both baseline concentrations and annualized changes (Δln-PFAS/Δt), including an overall PFAS exposure index (the mean of these standardized ln-PFAS concentrations). Baseline PFAS concentrations were generally not associated with changes in oxidative DNA damage. In contrast, annualized increases in PFAS were consistently associated with greater increases in 8-OHdG. Specifically, each 1-SD increase in annualized changes in ln-PFAS concentrations was associated with 2.33%, 3.77%, and 3.05% increases in ln-8-OHdG for linear perfluorooctane sulfonic acid (PFOS), branched PFOS, and perfluoroheptanoic acid (PFHpA), respectively, whereas a 1.98% decrease was observed for perfluorododecanoic acid (PFDoA); the PFAS exposure index was also positively associated with ln-8-OHdG (3.36% increase). Dose–response analyses revealed compound-specific non-linear patterns, including quadratic and cubic relationships for PFOS isomers. Results were robust across alternative model specifications, creatinine adjustment strategies, and exclusion of participants with cardiometabolic conditions, with directionally consistent associations across subgroups. In conclusion, longitudinal increases in PFAS exposure are associated with progressive oxidative DNA damage in young adults. These findings highlight the importance of dynamic exposure assessment in evaluating the health effects of persistent environmental chemicals. • PFAS increases over time were linked to higher urinary 8-OHdG in youth. • Longitudinal PFAS changes were more predictive than baseline PFAS levels. • Linear and branched PFOS and PFHpA showed the strongest associations. • Results were robust across sensitivity analyses and model specifications.