Effect of a Simple Exercise Program on Hospitalization-Associated Disability in Older Patients: A Randomized Controlled Trial

Abstract

ObjectiveHospitalization-associated disability [HAD, ie, the loss of ability to perform ≥1 basic activities of daily living (ADLs) independently at discharge] is a frequent condition among older patients. The present study assessed whether a simple inpatient exercise program decreases HAD incidence in acutely hospitalized very old patients. DesignIn this randomized controlled trial (Activity in Geriatric Acute Care) participants were assigned to a control or intervention group and were assessed at baseline, admission, discharge, and 3 months thereafter. Setting and ParticipantsIn total, 268 patients (mean age 88 years, range 75–102) admitted to an acute care for older patients unit of a public hospital were randomized to a control (n = 125) or intervention (exercise) group (n = 143). MethodsBoth groups received usual care, and patients in the intervention group also performed simple supervised exercises (walking and rising from a chair, for a total duration of ∼20 minutes/day). We measured ADL function (Katz index) and incident HAD at discharge and after 3 months (primary outcome) and Short Physical Performance Battery, ambulatory capacity, number of falls, rehospitalization, and death during a 3-month follow-up (secondary outcomes). ResultsMedian duration of hospitalization was 7 days (interquartile range 4 days). The intervention group had a lower risk of HAD with reference to both baseline [odds ratio (OR) 0.36; 95% confidence interval (CI) 0.17–0.76, P = .007] and admission (OR 0.29; 95% CI 0.10–0.89, P = .030). A trend toward an improved ADL function at discharge vs admission was found in the intervention group compared with controls (OR 0.32; 95% CI ‒0.04 to 0.68; P = .083). No between-group differences were noted for the other endpoints (all P > .05). Conclusion and ImplicationsA simple inpatient exercise program decreases risk of HAD in acutely hospitalized, very old patients.