Abstract
Plasma glial fibrillary acidic protein (GFAP) and neurofilament light chain (NEFL) are promising dementia biomarkers, yet underexplored in community settings. We studied 36 153 UK Biobank participants (aged 39-70 years, median follow-up: 10.53 years) to assess the biomarkers’ predictive performance using receiver operating characteristic (ROC) analysis and their influencing factors using linear regression. GFAP and NEFL demonstrated specific, strong predictive value for Alzheimer’s disease (AD) and vascular dementia (VaD) (AUC = 0.76-0.83). Mechanistically, both biomarkers correlated with larger ventricular volumes and slower reaction times. Their levels were influenced by demographics and environmental factors such as smoking, comorbid chronic kidney disease, and diabetes. Genetic analysis showed GFAP levels were influenced by genetic polymorphisms, especially rs429358 mapped to APOE. Incorporating these factors significantly improved predictive accuracy for dementia (AUC = 0.86-0.89). These results could support the establishment of clinical reference ranges, enhancing screening and prediction utility of GFAP and NEFL in the community.