Letters Balance training programme

Authors’ reply to Bonten

BMJ 2015; 351 doi: https://doi.org/10.1136/bmj.h5116 (Published 30 September 2015) Cite this as: BMJ 2015;351:h5116
  1. Fabienne El-Khoury, doctoral researcher1,
  2. Bernard Cassou, professor of public health and geriatrician2,
  3. Patricia Dargent-Molina, senior researcher in epidemiology3
  1. 1Inserm, Centre de Recherche Epidémiologie et Statistique Sorbonne Paris Cité (CRESS), U1153, F-94807, Villejuif, France
  2. 2Inserm, VIMA: Vieillissement et Maladies Chroniques, U1168, F-94807, Villejuif
  3. 3Inserm CRESS-Equipe ORCHARD, Hôpital Paul Brousse, 94807 Villejuif
  1. patricia.dargent{at}inserm.fr

Bonten’s comments and questions on how our results can be translated into clinical practice and integrated into clinical guidelines give us the opportunity to talk about the place that exercise programmes such as Ossébo may have in a comprehensive strategy of falls prevention in older people.1 2

Two approaches to falls prevention in community dwelling older people have been shown to be effective in randomised controlled trials. The first is the multifactorial and multidisciplinary approach, which involves individual clinical assessments of risk factors for falling, followed by multiple interventions targeting all remediable risk factors (normally within a specialist falls service). The second is single community based interventions targeting common risk factors (such as impaired strength and balance).

Multifactorial intervention community programmes are resource and staff intensive, and there is no evidence that they are more effective than single intervention programmes, such as balance and strength training.3 Hence, the multifactorial approach, especially if delivered by a multidisciplinary team, is generally reserved for those at high risk, such as frail patients who have had recurrent falls or those presenting to the emergency department with a fall or fracture.4 5 Single interventions, particularly community based exercise programmes designed to improve balance and reduce falls and injuries, will help gain population coverage at lower individual cost.6 Referral to this type of programme must be based on simple criteria or tests that can be easily used by GPs and other experienced and appropriately trained professionals in the health or physical activity domains. For this intervention to have an impact at the population level (number of falls and injuries prevented), the eligibility criteria must allow a large proportion of “at risk” people, who are likely to benefit from the proposed intervention, to be selected.

In the Ossébo experiment, the recruitment strategy was based on voting registration lists (rather than GPs’ patient lists, for instance) because we wanted the population source to be as broad and unselected as possible. We used two functional tests (gait speed and the tandem walk test) rather than a history of fall as the primary selection tool because this enabled us to select a larger number of at risk older women who were likely to benefit from our group based exercise programme. About 42% of the women examined had the required functional criteria (“not too fit, not too frail”) and 27% were selected for the trial (after excluding those with medical contraindications or other exclusion criteria). About 40% of the women selected for the trial reported having fallen in the past year. But the functional tests allowed us to identify a large number of other women who, although they had not fallen yet, were at increased risk of falls and injuries because of impaired balance and gait (and were likely to benefit from a balance training programme such as Ossébo).

We believe that exercise programmes that have been shown to be safe and effective, like the Ossébo programme, should be mostly recommended on the basis of simple criteria that are directly linked to the particular intervention. However, we acknowledge that some functional tests (such as the measure of gait speed or the timed up and go test, which is often recommended in guidelines as a second line test for older people who have fallen in the previous year) may not be easy to implement in routine clinical care. We agree that future research should assess the value of simpler clinical tests (such as single leg stance and tandem stance) that would be more acceptable to GPs and it should aim to develop simple test criteria that could be used by patients for self referral, thereby increasing uptake of fall prevention exercise programmes in older people.7

As Bonten suggests, the participants’ primary care physicians might have provided advice or care during follow-up that may have helped lower the risk of falling. However, there is no reason to believe that this happened more in the intervention group than in the control group, and that it resulted in an overestimation of the intervention’s effect. As mentioned in the article, at the end of the baseline examination, all women, whether included in the trial or not, were offered brochures about fall and fracture prevention, which covered the importance of physical activity, a balanced diet, and vitamin D supplementation and offered suggestions for assessing home hazards and managing drugs. Participants in both groups also received newsletters twice a year reminding them about major risk factors for falls and prevention measures. We monitored changes in risk factors for falling during follow-up by including questions about serious health problems that had begun or become worse since the last visit (such as newly diagnosed diseases, hospital admissions, vision problems, move to an institution) and by recording drug use at the one and two year examinations. We have also checked that the “contrast” between the two comparison groups regarding physical activity was maintained over time (same physical activity questionnaire at baseline, one, and two years, which included a question about participation in fall prevention exercise classes).


Cite this as: BMJ 2015;351:h5116


  • Competing interests: None declared.


View Abstract