A Serum Neurofilament Light-Based Risk Score Predicts Cognitive Decline in Early Parkinson’s Disease: A Longitudinal Parkinson’s Progression Markers Initiative Study
Background: The progression of Parkinson’s disease (PD), particularly cognitive decline, is substantially heterogeneous. Serum neurofilament light chain (NfL) is a useful biomarker of neurodegeneration, and PD patients with high serum NfL levels are more likely to develop dementia.Objective: We investigated whether a composite risk score (sNfL-Risk score) that combines serum NfL with age, cognition, motor severity, and RBD-related symptoms can predict cognitive decline in PD.Methods: We conducted a prospective cohort study using datasets from the Parkinson’s Progression Markers Initiative (PPMI). The sNfL-Risk score was constructed by integrating scores for serum NfL, age, MoCA, UPDRS part III, and RBDSQ. Associations between the sNfL-Risk score and cognitive decline were evaluated using logistic regression and net reclassification improvement (NRI) analyses.Results: A strong association was observed between higher sNfL-Risk scores and an increased risk of cognitive decline at 5 years. Each 1-point increase in the sNfL-Risk score was associated with 2.65-fold greater odds of cognitive decline at 5 years (95% Confidence interval (CI) 2.12–3.32; p < 0.0001). The higher the risk score was, the more significantly the probability of cognitive decline within 5 years increased, exceeding 48% in patients with scores ≥3. Compared with NfL alone, the sNfL-Risk score performed better in predicting cognitive decline (AUC 0.8 vs. 0.684; p < 0.0001) and demonstrated superiority in risk stratification (categorical NRI 0.255; p = 0.006).Conclusions: The sNfL-Risk score may be useful for stratifying patients at risk of PD-related cognitive decline and may contribute to the development of personalized management strategies.