Power of Functional Mobility Tests in Predicting Fall Risk Among Older Adults

Background and Purpose No consensus exists on which commonly used functional test best predicts fall risk in community‐dwelling older adults, and a direct comparison of several widely used tests may support more reliable clinical screening decisions. This prospective study investigated the effectiveness of commonly used functional tests in predicting falls in older adults. Method This prospective observational cohort study included 41 participants (25%) aged ≥ 60 years classified as fallers (≥ 1 fall during 6‐month follow‐up; median age = 70 IQR = 10) and 122 non‐fallers (median age = 68, IQR = 13). The Timed Up and Go Test (TUG), Five Times Sit‐to‐Stand Test (FTSST), Four Square Step Test (FSST), 30‐Second Chair Stand Test (30s‐CST), and Functional Reach Test (FRT) were administered at baseline. Univariable and multivariable logistic regression and ROC analysis were performed with bootstrap internal validation. Results Fallers and non‐fallers did not differ significantly in age, sex, or BMI ( p > 0.05), but chronic disease was more prevalent among fallers (78.0% vs. 52.5%, p = 0.004). TUG (AUC = 0.777) and FSST (AUC = 0.771) showed the numerically highest, and nearly identical, discrimination, followed by 30s‐CST (AUC = 0.708) and FTSST (AUC = 0.693), whereas FRT (AUC = 0.588) did not discriminate significantly. All tests were significantly associated with falls in univariable models, but only TUG remained significant in the multivariable model (OR = 1.26 per second; 95% CI: 1.08–1.47). A history of falls alone yielded an AUC of 0.82; in post hoc exploratory models, adding each test to fall history improved model fit (bootstrap‐corrected AUC 0.89–0.97). Using dataset‐derived cut‐offs, FSST had the highest sensitivity (90.2%) but low specificity (35.2%); TUG showed the best sensitivity–specificity balance (80.5%/48.4%); FRT had the lowest specificity (24.6%). Discussion TUG and FSST showed the numerically highest discrimination, though only at an acceptable level, and no individual test demonstrated sufficiently strong stand‐alone performance to reliably predict future falls. Combining tests with fall history improved model fit in this sample, suggesting that functional tests may complement, rather than replace, fall‐history screening; these exploratory findings require external validation. These findings may guide clinicians in selecting fall‐risk screening tools in geriatric physical therapy.