Objectives: The aim of this study is to evaluate the effects of multicomponent rehabilitation on physical activity, sedentary behavior, and mobility in older people recently discharged from hospital. Design: Randomized controlled trial. Setting: Home and community. Participants: Community-dwelling people aged ⩾60 years recovering from a lower limb or back musculoskeletal injury, surgery, or disorder were recruited from local health center hospitals and randomly assigned into an intervention ( n = 59) or a control (standard care, n = 58) group. Intervention: The six-month intervention consisted of a motivational interview, goal attainment process, guidance for safe walking, a progressive home exercise program, physical activity counseling, and standard care. Measurements: Physical activity and sedentary time were assessed using an accelerometer and a single question. Mobility was evaluated with the Short Physical Performance Battery, self-reported use of a walking aid, and ability to negotiate stairs and walk outdoors. Intervention effects were analyzed with generalized estimating equations. Results: Daily physical activity was 127 ± 78 minutes/day and 121 ± 70 at baseline and 167 ± 81 and 164 ± 72 at six months in the intervention and control group, respectively; mean difference of 3.4 minutes (95% confidence interval (CI) = −20.3 to 27.1). In addition, no significant between-group differences were shown in physical performance. Conclusion: The rehabilitation program was not superior to standard care for increasing physical activity or improving physical performance. Mobility-limited older people who had recently returned home from hospital would have needed a longer and more frequently monitored comprehensive geriatric intervention.

Objective: To examine whether pre-admission community mobility explains the effects of a rehabilitation program on physical performance and activity in older adults recently discharged from hospital. Design: A secondary analysis of a randomized controlled trial. Setting: Home and community. Participants: Community-dwelling adults aged ⩾60 years recovering from a lower limb or back injury, surgery or other disorder who were randomized to a rehabilitation ( n = 59) or standard care control ( n = 58) group. They were further classified into subgroups that were not planned a priori: (1) mild, (2) moderate, or (3) severe pre-admission restrictions in community mobility. Interventions: The 6-month intervention consisted of a motivational interview, goal attainment process, guidance for safe walking, a progressive home exercise program, physical activity counselling, and standard care. Measurements: Physical performance was measured with the Short Physical Performance Battery and physical activity with accelerometers and self-reports. Data were analysed by generalized estimating equation models with the interactions of intervention, time, and subgroup. Results: Rehabilitation improved physical performance more in the intervention ( n = 30) than in the control group ( n = 28) among participants with moderate mobility restriction: score of the Short Physical Performance Battery was 4.4 ± 2.3 and 4.2 ± 2.2 at baseline, and 7.3 ± 2.6 and 5.8 ± 2.9 at 6 months in the intervention and control group, respectively (mean difference 1.6 points, 95% Confidence Interval 0.2 to 3.1). Rehabilitation did not increase accelerometer-based physical activity in the aforementioned subgroup and did not benefit those with either mild or severe mobility restrictions. Conclusions: Pre-admission mobility may determine the response to the largely counselling-based rehabilitation program.