Letters

Preventing osteoporosis, falls, and fractures among elderly people

Few exercise programmes studied have prevented falls

Getting younger and older people moving may seem sensible, but evidence is lacking

Author's reply

Few exercise programmes studied have prevented falls

EDITORKannus 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 controlled trials. 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 is likely 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.

Gene Feder, Senior lecturer. 
g.s.feder{at}mds.qmw.ac.uk

Yvonne Carter, Professor. 
Sheila Donovan, Facilitator. 
Department of General Practice and Primary Care, St Bartholomew's and the Royal London School of Medicine and Dentistry, Queen Mary and Westfield College, London E1 4NS

Colin Cryer, Statistician. 
South East Institute of Public Health, King's College London, Tunbridge Wells, Kent TN3 0XT

Getting younger and older people moving may seem sensible, but evidence is lacking

EDITORKannus's editorial promoting lifelong physical activity as essential in preventing osteoporosis, falls, and fractures among elderly people is not evidence based.¹ Several findings need to be considered.

Firstly, no randomised trials provide evidence that regular exercise reduces fractures.

Secondly, vigorous exercise during growth increases bone mineral density by 10-20%² and moderate exercise during growth also increases bone mineral density.³ But these benefits may be lost after activity is stopped.⁴ Although residual benefits in bone mineral density have been reported in retired athletes aged under 50, falls and fractures are common in elderly people (those over 65). Whether elderly people who were athletes or exercised moderately before retiring from sport have high bone mineral density is unknown.

Thirdly, most exercise intervention studies in adults report no change in bone mineral density or increases of 1-3% (small changes of questionable biological significance given that the baseline bone mineral density is already low). One study suggests that exercise actually decreases bone mineral density.⁵

Fourthly, the lower prevalence of past or current physical activity in patients with hip fracture than in controls is hypothesis generating, not hypothesis testing.

It may seem sensible to get "younger and older people moving," but evidence is lacking. How many fewer fractures would occur if lifelong exercise was widely taken up by younger and older people? Is the hypothesis one sided? Could lifelong exercise increase the risk of falls and fractures?

E Seeman, Associate professor of medicine. 
Austin and Repatriation Medical Centre, University of Melbourne, Melbourne, Australia ego{at}austin.unimelb.edu.au

Author's reply

EDITORFeder et al emphasise that exercise in unselected groups of older people does not prevent falls while that in selected groups does. I do not know of any comparative study indicating which specific groups of older adults benefit from an exercise programme, but the randomised trials that have shown a benefit have included clearly different age groups, from a mean of 65¹ to one of 84.² Thus age does not seem to be a strong predictor of success or failure of the programme. I personally believe that the programme's quality and the motivation of its leaders are crucial.

Feder et al point out that multifaceted intervention programmes can reduce falls in older adults. I agree. Tinetti et al and Close et al have provided strong evidence for this, ^{3 4} Tinetti et al's study including balance and strengthening exercises.³ Unfortunately, neither study could separate the independent role of each of the modified factors, although such analyses are important. Thus these multifactorial interventions do not exclude exercise as one way of reducing the risk of falling.

Seeman is worried about the level of evidence concerning exercise in the prevention of osteoporosis and related fractures. As in any medical condition, few actions have been verified by randomised trials. This does not mean that we should stop trying to obtain better evidence, but at a certain point we have to summarise current knowledge and give recommendations. I agree that no randomised trial has proved that regular exercise reduces the risk of fracture; probably such a large and long trial will never be conducted successfully. This does not mean, however, that exercise is ineffective in preventing fractures related to age.

Many well controlled studies of physical activity and areal bone mineral density indicate that exercise during growing years is more beneficial than exercise in adulthood. What is now needed is long term follow up of those who once obtained clear bone gain by exercise in early life. The studies showing small increases in areal bone mineral density obtained by exercise in adulthood (1-3%) may have underestimated the true effect of mechanical loading on bone strength. A recent experimental intervention showed that mechanical loading can improve bone strength by reshaping the bone structure without increasing its areal bone mineral density.⁵ If this finding is repeated the current pessimistic attitudes towards the possibilities of improving adults' bone strength with exercise must be re-evaluated.

The final answer to whether lifelong physical activity decreases the risk of falls and fractures will probably never be available, but the evidence summarised in my editorial speaks strongly for it.

Pekka Kannus, Chief physician. 
Accident and Trauma Research Centre, UKK Institute, PO Box 30, FIN-33501 Tampere, Finland Klpeka{at}uta.fi

A longer version of this letter is available on the BMJ's website www.bmj.com