Abstract
INTRODUCTION: Alzheimer's disease (AD) is characterized by the progressive accumulation of amyloid-beta (Aβ) plaques, tau tangles, and neurodegeneration (ATN framework). While neuroimaging detects these changes, it is costly and not widely accessible. Blood-based biomarkers offer scalable alternatives for early detection. METHODS: We integrated plasma biomarkers, neuroimaging, and cognitive data from two large cohorts. We examined multivariate patterns of plasma Aβ42/Aβ40, phosphorylated tau (p-tau) 217, p-tau181, neurofilament light (NfL), and glial fibrillary acidic protein (GFAP) that predict amyloid, tau, neurodegeneration, and memory decline. RESULTS: Specific plasma biomarker combinations predicted amyloid burden in the precuneus, tau in the entorhinal cortex, hippocampal atrophy, and memory decline. Models trained in one cohort accurately predicted outcomes in the other, confirming cross-cohort generalizability. DISCUSSION: Distinct plasma signatures reflect underlying AD pathways, supporting the use of blood-based biomarkers as cost-effective, noninvasive tools for early detection, disease monitoring, and individualized risk profiling.