Eight trials (n=843) were included in the review: 6 randomised controlled trials (RCTs) and 2 quasi-experimental studies in which the treatment centre, rather than the patient, was randomly assigned.
Five trials assessed compliance with the exercise programme. The proportion of patients who adhered to the entire programme ranged from 42 to 93% (4 trials), while the proportion of sessions that the participants attended ranged from 73 to 95% (4 trials).
Exercise had a marginally significant effect on fall rate, reducing it by 6.2% (95% CI: 0, 12.7, P=0.06) with no evidence of significant heterogeneity (Q=6.11, P=0.53). The pooled OR was not statistically significant (0.79, 95% CI: 0.59, 1.05, P=0.10).
There was a statistically significant improvement in balance after exercise therapy (7 studies: effect size 0.32, 95% CI: 0.17, 0.47, P<0.0001). However, there was evidence of statistically significant heterogeneity between the studies (P<0.0001).
There were statistically significant improvements in ankle muscle strength (effect size 0.42, 95% CI: 0.25, 0.59, P<0.0001) and knee muscle strength (effect size 1.11, 95% CI: 0.90, 1.32, P<0.0001) after exercise therapy. However, there was evidence of statistically significant heterogeneity between the studies (P<0.001 for both).
There was no significant association between post-intervention fall risk and compliance rate, or between post-intervention fall rate and baseline fall rate.