Remote detection of Alzheimer's disease pathology using a digital sleep and circadian biomarker: an InSleep46 study

Deploying a remote sleep and circadian monitoring device in a countrywide population-based cohort in their late 70s is feasible. A model based on iterative feature selection was able to identify individuals with significant Tau (Braak3+) pathology with AUROC similar to plasma pTau217. This provides proof-of-concept that digital sleep biomarkers may be useful in identifying individuals at high risk of developing clinical AD. Work is underway to refine the model further, replicate these results in other cohorts, and identify the shortest duration of recording required for robust prediction.