Effect of pravastatin on frequency of fracture in the LIPID study: secondary analysis of a randomised controlled trial

Background Statins inhibit the same biochemical pathway as aminobisphosphon ates, therefore these cholesterol-lowering agents may have a beneficial eff ect on osteoporosis. This possibility has been supported by the finding tha t some statins also stimulate bone formation, and by observational studies suggesting that patients using statins have higher bone densities and lower fracture rates than controls. To assess whether statins have clinically si gnificant effects on bone, we studied the frequency of fractures in a large randomised controlled trial of these agents. Methods 9014 patients (17% women, median age 62 years) with ischaemic heart disease were randomly assigned pravastatin 40 mg daily or placebo and foll owed up for a mean of 6.0 years. Fractures were ascertained from adverse-ev ent reports. Findings 101 patients in the placebo group were admitted to hospital for fr acture compared with 107 in the pravastatin group (hazard ratio 1.05 [95% C I 0.80-1.37]). When patients with fractures not necessitating hospital admi ssion were added, the total number of patients having a fracture was 183 in the placebo group and 175 in the pravastatin group (0.94 [0.77-1.16]). Sep arate analyses for women alone and for individuals aged 65 years and over g ave similar results. Interpretation These findings offer no support for the hypothesis that stat ins have a significant effect on fracture risk. However, this study was not of an osteoporotic population, and fracture rate, although clinically impo rtant, is an insensitive index of effects on bone. Statins should not be us ed to prevent osteoporosis until there is evidence for their efficacy based on randomised controlled trials.