This trial is the first exercise study that focuses on fracture incidence as a primary study endpoint. Although we marginally failed to determine significant effects on “overall” fracture risk (p = .074) or rate ratio (p = .095), our findings further increased the evidence that exercise relevantly prevents fractures in the elderly.

The purpose of this study is to determine the effect of strictly supervised long-term exercise training on “overall” fracture incidence and bone mineral density (BMD) in postmenopausal osteopenic women.

Eighty-five early postmenopausal (1–8 years), osteopenic women living in the area of Erlangen–Nuremberg, Germany without any medication or diseases affecting bone metabolism were assessed after 12 years of supervised exercise (EG) or unvarying lifestyle (control, CG).

“Overall” fracture risk ratio in the EG was 0.32 (95% confidence interval (CI), 0.08 to 1.05; p = .074), and the rate ratio for “overall” fractures was 0.38 (95% CI, 0.11 to 1.15; p = .095). BMD changes at lumbar spine (EG, −0.8%; 95% CI, 0.8% to −2.7% vs. CG, −4.0%; 95% CI, −2.4% to −5.7%; p = .011) and femoral neck (EG, −3.7%; 95% CI, −2.4% to −5.0% vs. CG, −6.7%; 95% CI, −5.3% to −8.2%; p = .003) significantly differed between both groups.

Although we marginally failed to determine significant effects on overall fracture risk or rate ratio, our study increased the body of evidence for the fracture prevention efficiency of exercise programs, with special regard on bone strength (as assessed by bone mineral density measurement). Future studies should focus on subjects more prone to fractures to generate enough statistical power to clearly determine this issue.