PRISM-AD quantifies clinical and genetic contributions to cognitive risk prediction

PRISM-AD in 4,435 European-ancestry older adults links higher modeled genetic contributions to later AD dementia, with a genetic-factor ratio of 1.135 (BH q < 0.001).

• Why it matters: Overall risk scores do not reveal how clinical status and inherited genetic susceptibility jointly shape cognitive trajectories from normal cognition to impairment.
• What they did: The model combined nonlinear clinical functions with a polygenic risk score term, then internally validated CN-to-MCI in 908 and MCI-to-AD dementia in 714 using participant-level forecasts over three or five years.
• The result: Forecast accuracy for CN-to-MCI was average precision 0.2996 (three-year) and 0.4094 (five-year), while PRS increments did not reach BH q < 0.05; in a retrospective follow-up, genetic contributions were higher in later AD within shared risk deciles.

AI summary of the abstract · may contain errors · not medical advice · read the paper.