Long-term increases in urinary metals are associated with lipid peroxidative but not oxidative DNA damage in adolescents and young adults: A 10-year prospective cohort study.

Metal levels have been linked to oxidative stress, yet longitudinal evidence on how baseline levels and long-term changes influence oxidative pathways remains limited. In this 10-year prospective analysis of the Young Taiwanese Cohort (YOTA), 397 participants aged 12-30 years completed paired assessments in 2006-2008 and 2017-2019. Urinary concentrations of eight metals (Pb, Cd, Ni, Cr, Fe, Zn, Cu, and Mn) and oxidative stress biomarkers, including 8-hydroxy-2'-deoxyguanosine (8-OHdG) and 8-iso-prostaglandin F2α (8-iso-PGF2α), were measured at both time points. An overall metal mixture index was calculated as the mean of standardized ln-transformed metal concentrations. Joint linear regression models simultaneously incorporated baseline exposure levels and decade-long changes (Δ) to estimate their independent associations with follow-up oxidative stress biomarkers. In joint models, neither baseline nor Δ metals were significantly associated with follow-up 8-OHdG after false discovery rate correction. In contrast, follow-up 8-iso-PGF2α was positively associated with both baseline and long-term increases in the metal mixture index. A 1-SD higher baseline metal mixture index was associated with 62.74% higher follow-up 8-iso-PGF2α (95% CI: 25.48% to 111.28%; q = 0.001), while a 1-SD higher Δ metal mixture index was associated with 100.77% higher follow-up 8-iso-PGF2α (95% CI: 53.42% to 162.74%; q < 0.001). Several individual metals-particularly Pb, Cd, Fe, and Cu at baseline and multiple metals (Pb, Cd, Cr, Fe, Zn, Cu, and Mn) for Δ-also showed positive associations with follow-up 8-iso-PGF2α after FDR correction. Restricted cubic spline (RCS) analyses supported largely linear dose-response relationships. Findings were consistent across subgroup and sensitivity analyses, including alternative mixture specifications using principal component analysis (PCA) and quantile g-computation (qgcomp). In conclusion, long-term increases in metal mixtures were associated with higher lipid peroxidation but not oxidative DNA damage, indicating pathway-specific oxidative responses from adolescence to early adulthood. These findings highlight the importance of considering exposure trajectories beyond baseline levels.