Associations Between Smartwatch-Derived Measures and Cognitive Function: Findings From the Electronic Framingham Heart Study.
BACKGROUND Daily heart rate (HR) and physical activity are biomarkers of cardiovascular disease. A smartwatch-derived metric, daily HR per step (DHRPS), has also been associated with cardiovascular disease. Integrating smartwatch data with neuropsychological assessments in older adults makes it possible to capture subclinical cardiovascular factors relevant to brain health. METHODS This cross-sectional study combined the electronic Framingham Heart Study smartwatch data (2021-2024) with neuropsychological assessments (2018-2023). The primary exposure was DHRPS defined as the average daily HR divided by steps per day. The dependent measure was assessed by composite neuropsychological scores including memory, executive function, language, and visuospatial cognitive domains. Linear mixed-effects models were used to examine the association between each exposure and repeated cognitive measures, adjusting for covariates. RESULTS A total of 248 older adults (mean age=73 years; median Mini-Mental State Examination score=29; 56% women; 14% multiethnic [including participants of Black, Hispanic, Asian, Indian, Pacific Islander, and Native American origins]) were assessed for a median of 662 smartwatch wear days, during which daily average HR (median 74 bpm [25th-75th percentiles: 69-78]) and daily steps (median 4810 [25th-75th percentiles: 3270-6831]) were recorded. Each 1-SD higher DHRPS was associated with 0.06-SD lower composite neuropsychological score (adjusted 95% CI, -0.11 to -0.01; P=0.01). Every 1000 more steps per day were associated with 0.03-SD higher composite neuropsychological score (adjusted 95% CI, 0.004-0.05; P=0.02). CONCLUSIONS Higher DHRPS was associated with lower cognitive function in older adults beyond HR or steps alone. Longitudinal evaluation of DHRPS exposures is needed to determine its utility for predicting cognitive decline trajectories.