Sensor-derived multidimensional gait domains mediate the association between intrinsic capacity impairment and fall risk in community-dwelling older adults: a cross-sectional study

BACKGROUND: Falls are a leading cause of injury and functional decline in older adults. Intrinsic capacity (IC) impairment has emerged as an important predictor of fall risk, yet the biomechanical pathways linking IC decline to falls remain uncharacterized. Whether sensor-derived multidimensional gait domains serve as domain-specific indirect pathways in the IC-fall association has not been examined. METHODS: In this cross-sectional study of 813 community-dwelling adults (aged ≥60 years), IC impairment and a 12-month self-reported fall history were assessed. Gait was quantified using a wearable multi-sensor system, with nine parameters summarized into three domains: pace, gait phase, and kinetics. Logistic regression-based mediation analysis was used to test the indirect effects of gait domains on the association between IC and falls. RESULTS: Higher IC impairment was independently associated with greater odds of falling after full adjustment (OR = 1.36, 95% CI: 1.11-1.66). Among gait domains, pace (OR = 0.75, 95% CI: 0.56-0.99) and kinetics (OR = 0.76, 95% CI: 0.59-0.99) were independently associated with lower fall odds, whereas gait phase was not. Mediation analyses indicated that pace and kinetics partially mediated the IC-fall association, accounting for 18.76% and 10.64% of the total effect respectively, with gait phase showing no significant indirect effect. CONCLUSIONS: These exploratory findings suggest that multidimensional gait assessment may complement IC screening to improve fall risk stratification and identify domain-specific intervention targets in older adults.