Overview
- Published in JAMA Network Open, the pooled analysis examined 7,105 U.S. adults who were dementia-free at baseline and followed for 3.5 to 17 years, during which about 1,000 developed dementia.
- For each 10-year increase in the sleep-derived brain age index over chronological age, future dementia risk rose by roughly 39%, even after adjusting for demographics, health factors, and APOE genetic risk.
- The machine-learning model integrated 13 microstructural EEG features, such as delta activity reflecting sleep depth and memory-linked sleep spindles, to estimate brain age.
- Because overnight EEG can be collected noninvasively at home, the authors highlight potential for risk screening with wearables, while cautioning that replication and causal studies are needed before clinical use.
- One EEG characteristic—higher kurtosis, reflecting occasional large signal spikes—was linked to lower dementia risk, and researchers note that treating sleep disorders may modify brain-wave patterns without proving causality.