Boonen S et al. (2005) Effect of osteoporosis treatments on risk of non-vertebral fractures: review and meta-analysis of intention-to-treat studies. Osteoporos Int 16: 1291–1298

To date, evidence that treatments for osteoporosis reduce the risk of nonvertebral fractures has mainly come from post-hoc subgroup analyses. International guidelines state, however, that subgroup analyses should not be the sole basis for conclusions on treatment efficacy. A new systematic review and meta-analysis by Boonen et al. aimed to overcome the limitations of earlier such analyses by stringent selection of trials that evaluated nonvertebral fracture risk in the intention-to-treat analysis, and that met the regulatory registration criteria for osteoporosis treatments.

The authors identified 11 phase III, randomized, placebo-controlled efficacy trials of ≥3 years' duration. Statistically significant reductions in the risk of nonvertebral fracture were seen in three trials: two with risedronate (P ≤0.03) and one with strontium ranelate (P = 0.04). Data from the six eligible bisphosphonate trials were then pooled; a meta-analysis showed that both alendronate and risedronate significantly reduced the relative risk of nonvertebral fracture (alendronate, P = 0.012; risedronate, P = 0.001).

The results of meta-analyses should not be used to compare the efficacies of different treatments, say the authors, because disease severity, patient demographics and definitions of nonvertebral fracture differ between trials. Not all trials included fractures caused by trauma, for example. There is a need for trials that compare different osteoporosis treatments directly, and that have prevention of nonvertebral fracture as a primary endpoint.