Distinct longitudinal trajectories were observed across the biomarkers. Five of the six biomarkers demonstrated significant increases over time (all P<0.001), whereas the annual change for p-tau181 was not statistically significant (P=0.122). Specifically, p-tau217 (β=0.09), Aβ40 (β=3.46), and Aβ42 (β=0.35) increased over time. Similarly, GFAP (β=5.58) and NfL (β=2.43) showed pronounced upward trends. All estimates were obtained from models adjusting for baseline age, sex, education, APOE e4 status, and MMSE score. Baseline age was positively associated with five of six biomarkers (all P<0.05, except Aβ42). APOE e4 status was associated with Aβ42 levels (non-carriers: β = 3.19 P=0.001; carriers: β=2.76 P=0.002). Lower MMSE scores were associated with higher GFAP (β=-2.43 P=0.041) and NfL(β=-2.13 P=0.009). Male sex was associated with lower GFAP (β= -21.12, P=0.003), and higher education was associated with higher GFAP (β=1.14, P<0.001).

Conclusion

Our findings demonstrate broadly similar longitudinal patterns among plasma AD biomarkers, with most markers showing progressive increases over time. These results underscore the importance of considering demographic, genetic, and cognitive factors when interpreting long-term biomarker trajectories in aging populations.