Prediction of Alzheimer’s Disease Progression from Mild Cognitive Impairment Using Polygenic Risk Scores

Mild cognitive impairment (MCI) represents the prodromal stage of Alzheimer’s disease (AD), and not all MCI patients progress to AD. Accurate stratification of MCI clinical outcomes is therefore crucial. Although polygenic risk scores (PRS) can distinguish AD patients from cognitively normal individuals, their utility in predicting heterogeneous MCI outcomes remains unclear. This study evaluated the predictive ability of PRS for both MCI progression and reversion. PRS were constructed using four algorithms and showed strong inter-method correlations. When divided into quartiles, MCI patients in the highest PRS quartile had a significantly greater risk of progression to AD, while lower PRS were asso-ciated with increased likelihood of reversion to normal cognition. We developed stepwise prediction models incorporating PRS, demographic variables, clinical assessments, and cerebrospinal fluid (CSF) biomarkers. Prediction performance did not differ significantly across PRS algorithms. The best-performing model combined PRS, demographic variables, and clinical assessments, while the addition of CSF biomarkers provided no further im-provement. These findings highlight the potential of PRS, integrated with routine clinical information, to enhance individualized risk prediction for MCI outcomes.