Physical Activity Program Participation and the Risk of Falls for Older Group Health Members

Abstract

Introduction: Falls are one of the biggest health concerns for aging adults. Despite evidence suggesting the importance of regular physical activity (PA) for reducing fall risk, few older adults engage in fall-prevention-oriented exercise. Regular PA through exercise programs offered as a Medicare or health-plan-covered benefit may be one method to increase PA and reduce fall risk. Here we investigate the effectiveness of participating in EnhanceFitness (EF) and Silver Sneakers (SS), two nationally-disseminated senior exercise programs, in reducing risk of falls resulting in medical care. Methods: A population-based, retrospective cohort study was conducted using data from Group Health Cooperative (GHC) members over age 65. Participants were classified as consistent users (having used EF/SS 2 or more times each year they were enrolled in GHC during the study period [2005-2011]); intermittent users (having used EF/SS two or more times in one or more years they were enrolled in GHC during the study period but not all years), or non-users of the EF/SS. A time-to-first fall requiring medical treatment (identified via ICD-9 code and E-codes in the medical record) analysis using Cox proportional hazards models was used for both programs to generate hazard ratios (HR) comparing consistent and intermittent users with non-users of either program. Hierarchical adjustment was used to address confounding by demographic characteristics and comorbidities (measured by ICD-9 codes in electronic health records). Results: In fully adjusted models, there was evidence of a dose-response relationship between EF participation and decreased fall risk compared to non-users (consistent EF user HR= 0.75, 95% CI = 0.64-0.89 and intermittent EF user HR = 0.87, 95% CI = 0.80-0.94). Participation in SS was not significantly associated with a decrease in risk for consistent users (HR= 0.97, 95% CI = 0.90-1.04), but a small significant reduction in risk was seen for intermittent users (HR= 0.93 95% CI= 0.90-0.97). Analyses evaluating effect modification showed that SS use was related to significantly lower fall risk among individuals over age 75 or with a BMI of 28 or below. Conclusion: Participation in EF provides a protective effect against falls resulting in medical care, with an indication of a dose-response relationship wherein this effect is strongest for consistent users. Results are less clear for SS participation, suggesting a small protective effect against medical falls for consistent and intermittent users that is potentially stronger for older and lower-BMI users.