Preventing osteoporosis, falls and fractures among older people. Few exercise programmes studied have prevented falls

Kannus promotes physical activity for the prevention of injurious falls among elderly people.1 He acknowledges the uncertainty that surrounds the effectiveness of specific exercise programmes tested as interventions in randomised controlledtrials. This is in contrast to the epidemiological evidence from longitudinal cohort or case-control studies. We agree with him that regular physical activity outside formal exercise programmes islikely to be beneficial to both younger and older people. In developing evidence based guidelines for the prevention of falls in older people we found good evidence that exercise programmes for unselected older people living in the community do not prevent falls,2 with the possible exception of balance training (tai chi).3 Two trials found that selected older people (those aged over 804 or with mild deficits in strength and balance5) benefit from individually tailored exercise programmes administered by qualified professionals. Our guidelines recommend that the implementation of exercise programmes for unselected older people should not be a priority. By contrast, multifaceted intervention programmes, including the identification and treatment of postural hypotension, review of drug treatment, modification of the home environment, and possibly exercise training, do reduce the incidence of falls. Multiple risk assessment and modification have also proved successful in the context of an accident and emergency department. The main research challenge is to develop these interventions further and test them in pragmatic implementation trials rather than to search for the optimal exercise activity