Abstract
BACKGROUND: A series of mobility and concerns about falling measurement instruments are recommended by guidelines for fall prevention in older adults for fall risk assessment. However, evidence regarding their value in evaluating and predicting fall risk is inconsistent. This study aims to examine the associations and predictive ability between these assessments and falls in community-dwelling older adults in China. METHODS: Mobility of 2544 participants aged [Formula: see text]60 years was assessed at baseline using the Timed Up and Go (TUG), Eyes-Closed Unipedal Stance Test (ECUST), 30-second Chair Stand (30s-CS), and concerns about falling were assessed by the Falls Efficacy Scale International (FES-I). Fall events were followed for nearly 12 months. Associations of mobility assessments and concerns about falling with falls were examined using Cox proportional hazard and restricted cubic spline regression. The C-statistic, net reclassification improvement (NRI) and integrated discrimination improvement (IDI) were employed to evaluate the predictive value of mobility assessments, concerns about falling, and additional risk factors. RESULTS: During the follow-up period, 236 (9.28%) participants reported at least one fall. Each standard deviation increase in TUG time was associated with a 1.154-fold greater risk of falls (95% CI: 1.035-1.286). Compared with participants with FES-I score ≤ 45, those with FES-I score > 45 had a 1.346-fold greater fall risk, with marginal statistical significance(95% CI: 0.995-1.822, P = 0.054). The ECUST and the 30s-CS were not associated with fall risk in the overall sample. Nonlinear relationship was not observed between mobility assessments and fall risk. And the non-linear correlation was marginally significant between FES-I and fall risk (P = 0.056). The predictive accuracy of the TUG and FES-I used in isolation or in combination was poor, with C statistics ranging from 0.545 to 0.568. The predictive performance was significantly enhanced but still insufficient after baseline characteristics were incorporated into the model, with a C-statistic of 0.631 (95% CI: 0.594-0.667), an IDI of 0.013 (95% CI: 0.010-0.040) and a NRI of 0.178 (95% CI: 0.111-0.260). CONCLUSION: The TUG test is an useful tool for evaluating fall risk in community-dwelling older adults. The FES-I has potential to identify older adults at an elevated risk of fall with a cut-off value of 45. However, their ability to predict falls is limited. Incorporating additional fall-related factors improves the predictive performance but is still inadequate.