Characteristics Associated with Improved Physical Performance among Community-Dwelling Older Adults in a Community-Based Falls Prevention Program

This was a retrospective analysis of quasi-longitudinal data from an ongoing, community-based falls prevention program. The purpose was to identify participant characteristics predicting improvement on physical performance measures associated with falls risk. Community-dwelling older adults >= 60 years old participated in a community-based implementation of the Otago Exercise Program (OEP). Participants with increased falls risk (n = 353) were provided with individualized exercises from OEP and were invited to return for monthly follow-up. One hundred twenty-eight participants returned for at least two follow-up visits within 6 months of their initial visit (mean time to second follow-up = 93 days with standard deviation = 43 days). Outcome measures assessed at initial and all follow-up visits included Four Stage Balance Test (4SBT), Timed Up and Go test (TUG), and Chair Rise Test (CRT). Distributions were examined, and results were categorized to depict improvement from initial visit (IVT) to second follow-up visit (F2). Key predictor variables were included in multivariable linear or logistic regression models. Improved 4SBT performance was predicted by greater balance confidence. Better TUG performance at F2 was predicted by no use of assistive device for walking, higher scores on cognitive screening, and better IVT TUG performance. Improvement on CRT was predicted by younger age and lower scores on cognitive screening. While improvements on each of the three measures were predicted by a unique combination of variables, these variables tended to be associated with less frailty.