Prediction of cognitive decline and Alzheimer's disease conversion by a plasma biomarker panel in non-demented individuals.

Abstract

Alzheimer's disease (AD) represents a growing global health burden, underscoring the urgent need for reliable diagnostic and prognostic biomarkers. Although several disease-modifying treatments have recently become available, their effects remain limited, as they primarily delay rather than halt disease progression. Thus, the early and accurate identification of individuals at elevated risk for conversion to AD dementia is crucial to maximize the effectiveness of these therapies and to facilitate timely intervention strategies. Baseline plasma concentrations of amyloid beta 40 (Aβ40), amyloid beta 42 (Aβ42), glial fibrillary acidic protein (GFAP), neurofilament light chain (NFL), and tau phosphorylated at residue 181 and 217 (pTau181, pTau217) were quantified in 233 non-demented participants from the Alzheimer's Disease Neuroimaging Initiative (ADNI) and followed for up to 11 years. Covariates included demographic variables and neuropsychological measures. Longitudinal cognitive trajectories were modelled with linear mixed-effects models (LMM), and the risk of progression to AD dementia was evaluated with logistic regression and Cox proportional-hazards analyses. Longitudinally, higher baseline plasma levels of GFAP, NFL, pTau181, and pTau217 independently predicted steeper cognitive decline. Cross-sectional analyses demonstrated significant associations of pTau217, pTau181, and Aβ42 with baseline memory impairment. A logistic regression model incorporating five plasma biomarkers-Aβ42, GFAP, NFL, pTau181, and pTau217-demonstrated robust predictive accuracy for discrimination of future converters to AD dementia. The addition of demographic variables and a baseline memory score further improved model specificity and positive predictive value. These findings support the utility of a concise plasma biomarker panel comprising Aβ42, GFAP, NFL, pTau181, and pTau217 for predicting both longitudinal cognitive deterioration and conversion to AD dementia. This less-invasive blood-based panel could serve as a practical triage tool to enrich clinical trials and facilitate timely therapeutic interventions with emerging disease-modifying treatments.