Association between beta-blockers and fracture risk: a Bayesian meta-analysis.

Abstract

The association between beta-blockers (BB) and fracture risk is controversial, due largely to conflicting findings from previous studies. The present study sought to evaluate the effect of BB on fracture risk by using a Bayesian meta-analysis approach. We systematically retrieved 13 observational studies on the association between BB use and fracture risk. This meta-analysis involved more than 907,000 men and women with mean/median age of individual studies ranging from 43 to 81 years. We used a hierarchical Bayesian random effects model to synthesize the results. BB use was associated with an average 17% reduction in the risk of any fracture (risk ratio [RR] 0.83; 95% credible interval [CrI]: 0.71-0.93), hip fracture (RR 0.83; 95% CrI: 0.70-0.92) and vertebral fracture (RR 0.81; 95% CrI: 0.61-0.99). The probability that BB use reduces fracture risk by at least 10% was 0.91. Beta-blockers are associated with reduced risk of fracture in older adults, but the effect size is likely to be modest.