Hand dexterity and mobility independently predict cognition in older adults: a multi-domain regression analysis.

Background: Motor function is a sensitive indicator of cognitive aging but the unique contributions of different motor domains are unclear when assessed together.

Methods: We evaluated 98 community-dwelling older adults (median age: 74). From a neuropsychological battery, a primary Global Cognitive Composite score (GCCS) and three secondary domain scores were derived using Principal Component Analysis (PCA). Motor predictors included the Nine-Hole Peg Test (NHPT), grip strength, Apraxia Screen of TULIA (AST), SPPB sub-tests (5-chair-rises time (5CRT), 4 m-walk time (4MWT), balance), and inertial measurement unit (IMU)-based gait parameters. Stepwise regression controlling for age and sex identified robust predictors of the GCCS.

Results: The final model identified several significant, independent motor predictors of the GCCS. Poorer hand dexterity (NHPT; β = -0.29, p < 0.01), slower 5CRT (β = -0.28, p < 0.01), and slower 4MWT (β = -0.17, p = 0.03) were associated with worse cognitive performance, while greater minimum toe clearance was associated with better performance (β = 0.19, p = 0.01). In contrast, grip strength, balance, usual gait speed, and measures of gait variability were not retained. The model explained 50.3% of the variance (Adjusted R²) in global cognitive performance.

Conclusion: Hand dexterity (NHPT) and specific functional mobility tasks (5CRT, 4MWT) are robust, independent predictors of cognition in older adults. Grip strength, balance, usual gait speed, and gait variability offer limited additional value when assessed together. The NHPT and timed SPPB components are accessible, pragmatic tools for motor-cognitive research and screening.