A Multidomain Longitudinal Analysis of Frailty, Functional Limitation, Balance and Falls in Older Adults

Background/Objectives: Falls are a major threat to healthy ageing and the preservation of independence, yet the risk of falls in older adults commonly arises from multiple interacting domains rather than a single factor. This longitudinal analysis assessed whether frailty, functional limitation, and balance performance were associated with future falls among older adults. Methods: This longitudinal analysis assessed older adults aged 65 years and older over a three-year follow-up interval, using baseline data collected in 2015 and follow-up falls data collected in 2018. Baseline health, functional, and physical performance measures were evaluated in relation to self-reported falls at follow-up. Candidate predictors comprised sociodemographic, health, functional, and physical performance variables. Univariable logistic regression, adjusted screening models, and a final complete-case multivariable logistic regression model were used. Results: The eligible analytic sample comprised 1932 participants, of whom 480 (24.8%) reported falls at follow-up. In adjusted analyses, greater frailty index, greater ADL limitation, female sex, shorter full-tandem balance time, depressive symptoms, and greater self-rated mobility severity were associated with higher odds of future falls. However, only frailty index, ADL limitation, full-tandem balance time, and female sex were retained in the final multivariable model. In the final multivariable model of 1451 participants, frailty index per 0.1-unit increase (OR 1.26, 95% CI 1.07-1.48), ADL limitation score per 1-point increase (OR 1.30, 95% CI 1.16-1.45), female sex (OR 1.62, 95% CI 1.17-2.25) and longer full-tandem balance time per 10 s increase (OR 0.89, 95% CI 0.82-0.96) remained associated with falls. The final model showed modest discrimination and acceptable apparent calibration. Conclusions: Higher frailty, greater ADL limitation, poorer tandem balance and female sex were associated with subsequent falls. These results reinforce the value of a multidomain approach to fall-risk assessment.