Exercise in girls during growth seems to confer a high peak bone mineral density (BMD). Exercise in adulthood, in the peri- and postmenopausal period, and in old age prevents bone loss or increases BMD with a magnitude of minor biological significance. However, these changes must be regarded as beneficial compared to the age-related bone loss, which inevitably will occur if no interventions are implemented. Prospective intervention studies also suggest that exercise improves muscle strength, coordination and balance, even in elderly women, all of which are improvements with a potential of reducing the number of falls. A randomised, controlled, prospective, blinded study (the only study design that tests a hypothesis) of exercise with fracture as end point is extremely difficult to conduct, due to the large sample sizes needed. At present, no such studies exist. Retrospective and prospective observational and case–control studies suggest that physical activity in women is associated with reduced fracture risk. This may be correct, but we must never forget that a consistently replicated sampling bias may produce the same outcome.
The Achilles heel of exercise is the reduction or the cessation of physical activity, which commonly occurs among middle-aged women when family and work demands reduce the time available for exercise. A higher BMD or improvement in muscle size and muscle strength achieved by exercise during adolescence seems to be eroded on retirement, leaving virtually no remaining benefits in old age, the period when fragility fractures begin to be a problem of increasing magnitude. However, recreational activities seem to maintain some of the musculo-skeletal benefits, but to date we do not know the level of activity needed to retain these benefits. Dose–response relationships need to be quantified, as also the effects on bone size, shape and architecture. Another essential question that we must address is how many fewer fractures will be the result of a community-based exercise campaign. Will efforts by the community to encourage a higher level of physical exercise, with the aim of reducing bone fractures, be cost-effective? The higher level of proof, suggesting that exercise does reduce fragility fractures and thus reduces the total cost for the society, must come from well-designed and well-executed, prospective, randomised, controlled trials. The responsibility of executing these studies lies in the hands of both researchers and the community.