Wearables Reveal Early Cognitive Impairment Risk
This systematic review (Jan 2020–Dec 1, 2025) evaluates wearable-derived digital biomarkers for early detection and prevention of cognitive impairment and dementia. It synthesizes 49 studies (median N=145, total >200,000), finding actigraphy-dominant evidence linking disrupted sleep, circadian fragmentation, and activity irregularity to worse cognition; ML/deep-learning models reported AUCs ≈0.70–0.95. The review notes heterogeneity and limited external validation, urging longitudinal prediction studies.
Key Points
- 1Identify 49 studies linking wearable-derived sleep and activity markers to cognitive outcomes across >200,000 participants
- 2Highlight that ML/deep learning models achieved AUCs ≈0.70–0.95, indicating predictive potential but variable validation
- 3Recommend longitudinal, externally validated studies to translate passive monitoring into scalable early detection and prevention
Scoring Rationale
Comprehensive systematic review with peer-reviewed methods; limited by heterogeneity and scarce external validation across studies.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
