OBJECTIVES

To assess the effects of bisphosphonates for the prevention and treatment of corticosteroid-induced osteoporosis.

SEARCH STRATEGY

We searched the Cochrane Musculoskeletal Group trials register, Medline up to 1997 and Embase1988-1997), and selected hand searching of reference lists was conducted. Hand searching of scientific abstracts from relevant meetings for the last five years was also done. An electronic search in Current Contents was done for the last six months. The Cochrane Controlled Trials Register (CCTR) will be searched for future updates. All languages were included in the search. For practical reasons only those in English were included, but all languages will be retrieved and translated for future updates.

SELECTION CRITERIA

All controlled clinical trials (CCTs) dealing with prevention or treatment of corticosteroid-induced osteoporosis with bisphosphonates of any type and reporting the outcomes of interest were assessed. Trials had to involve adults only, and subjects had to be taking a mean steroid dose of 7.5 mg/day or more.

DATA COLLECTION AND ANALYSIS

All data extraction was performed by two independent reviewers. Outcomes of interest included change in bone mineral density (BMD) at the lumbar spine and femoral neck at six and 12 months. If present, data on number of new fractures and withdrawals due to adverse effects were also extracted. All data extraction was performed by two independent reviewers. Both continuous and dichotomous data were analyzed using fixed effects models. When significant heterogeneity was present, a random effects model was used.

MAIN RESULTS

A total of 13 trials, including 842 patients are included in this meta-analysis. Results are reported as a weighted mean difference of the percent change in BMD between the treatment and placebo groups, with trials being weighted by the inverse of their variance. The 95% confidence intervals (95% CI) are presented. At the lumbar spine, the weighted mean difference of BMD between the treatment and placebo groups was 4.3% (95% CI 2.7, 5.9). At the femoral neck, the weighted mean difference was 2.1% (95%CI 0. 01, 3.8). Although there was a 24% reduction in odds of spinal fractures [OR 0.76 (95%CI 0.37, 1.53)], this result was not statistically significant.

REVIEWER'S CONCLUSIONS: Bisphosphonates are effective at preventing and treating corticosteroid-induced bone loss at the lumbar spine and femoral neck. Efficacy regarding fracture prevention cannot be concluded from this analysis, although bone density changes are correlated with fracture risk.