• fractures;
  • falls;
  • osteoporosis;
  • aged;
  • quantitative ultrasound

OBJECTIVES: To investigate which factors best predict whether a fall will result in a fracture.

DESIGN: Prospective cohort.

SETTING: Residential care facilities.

PARTICIPANTS: One thousand three hundred forty-two older men and women (mean age 86.0) who had a fall.

MEASUREMENTS: Clinical risk factors and bone fragility by calcaneus broadband ultrasound attenuation (BUA) were assessed at baseline, and falls and fall-related fractures were recorded for up to 2 years. All fractures were validated using X-ray reports.

RESULTS: During a median follow-up of 1.97 years, 6,646 falls resulted in 308 fractures. Fracture rates were 6.7, 4.8, and 3.1 per 100 falls for BUA in the lowest (10.5–39.7 dB/MHz), middle (39.8–58.9 dB/MHz), and highest (≥58.9dB/MHz) tertiles, respectively. In multivariate analysis, the odds ratio for any fracture was 1.17 (95% confidence interval=1.08–1.27; P<.001) for every 10-dB/MHz reduction in BUA. Other significant risk factors in the model were lower weight, longer leg length, better balance, no severe cognitive impairment, intermediate care residence, lack of a history of falls in the previous year, and a history of fracture since age 50. For hip fracture, the only significant risk factors were lower weight, longer leg length, intermediate care residence, and lack of falls history.

CONCLUSION: Important risk factors for whether a fracture occurs immediately after a fall include direct measures of bone fragility and a number of anthropometric and clinical factors that may increase or attenuate the mechanical impact of a fall. By identifying those at highest risk of fracture after a fall, these findings should have important implications for fracture prevention strategies.