Baseline measurement of bone mass predicts fracture in white women. Ann Int Med 111

Study Objective: To determine if a single bone mass mea-surement of the radius is predictive of future fractures at any site. Design: Observational study of a cohort of free-living sub-jects and a cohort of retirement-home residents with an av-erage follow-up of 6.7 years and 5.5 years, respectively (range, 1 to 15 years for both cohorts). Setting: General community and a retirement home. Subjects: Volunteer sample of white women (386 free-living and 135 living in a retirement home) who were free of disease and were not receiving medication known to affect bone metabolism. In terms of physical condition subjects ranged from the totally independent to the wheelchair-bound. Measurements and Main Results: A radial bone mass measurement was done at the initial visit. Subsequent non-spine fractures were reported by the subjects at follow-up visits, which were less than a year apart in most cases, and verified with medical records. Cox regression was used to model time to first fracture as a function of age and bone mass. These analyses showed that for every 0.1 g/cm decre-ment in bone mass, the relative risk of fracture was 2.2 (CI, 1.7 to 2.8) for the free-living and 1.5 (CI, 1.2 to 1.9) for the retirement-home residents. Baseline age did not predict the risk of fracture in either cohort, and controlling for baseline age did not reduce the relative-risk estimates of bone mass. Similar analyses also showed that bone mass was a statisti-cally significant predictor for first hip fractures (n = 30) among the nursing-home residents (relative risk, 1.9; CI, 1.4 to 2.7) and first forearm fractures (n = 10) among the fre