Randomised controlled trials (RCTs) of a physical activity or exercise intervention compared with control (activity or usual treatment), sham control or another intervention that did not include exercise were eligible for inclusion. Eligible studies were of community-dwelling (defined as ambulatory and not depending on someone else for basic activities of daily living or walking) older adults aged over 50 years.
The included studies had a female to male ratio of 3:1, 4:1 or 5:1 or were 100% female. Mean ages ranged from 72 to 84 years (where reported). Some studies were of healthy participants; others included participants with other conditions (such as visual impairment, stroke, osteoporosis, history of bed rest or hospitalisation). A wide variety of interventions were included (such as group, individual or combined programmes and land-based or water-based exercise) with the objective of improving strength and/or balance. Education and vitamin D supplementation were provided in some programmes. Some programmes provided follow-up with a health professional (physiotherapists, occupational therapists, nurses, physicians).
Outcomes of interest fell into six categories that described fall risk: muscle strength (isolated strength measurement; typically lower body strength), functional tasks that included timed up-and-go (rising from a chair or rising from a chair and walking), balance (static or dynamic), gait (velocity over a short distance), fear of falls or falls efficacy, self-reported functional status and a composite measure of several factors. Other falls outcomes were reported (such as fall rate, number of multiple fallers, number of injurious falls and home safety).
Two reviewers independently selected studies for inclusion. Disagreements were resolved by a third reviewer.