Osteoporosis International

, Volume 23, Issue 1, pp 75–85

Fracture risk assessment without bone density measurement in routine clinical practice

  • W. D. Leslie
  • S. Morin
  • L. M. Lix
  • H. Johansson
  • A. Oden
  • E. McCloskey
  • J. A. Kanis
  • for the Manitoba Bone Density Program
Original Article

DOI: 10.1007/s00198-011-1747-2

Cite this article as:
Leslie, W.D., Morin, S., Lix, L.M. et al. Osteoporos Int (2012) 23: 75. doi:10.1007/s00198-011-1747-2



Fracture probability assessed without bone mineral density (BMD) could potentially be sufficient for clinical decision making in many individuals categorized as low or high fracture risk. For individuals falling in a moderate risk range, there is incremental value in using BMD in the probability calculation as this appropriately reclassifies risk in over one third of the individuals.


A new fracture risk assessment tool from the World Health Organization (FRAX®) estimates 10-year major osteoporotic and hip fracture probabilities from multiple clinical risk factors with or without hip BMD. The objective of this study is to determine whether fracture probability derived without BMD can be used to identify individuals who would be designated for treatment.


A historical cohort of 36,730 women and 2,873 men aged 50 years and older drawn from the Manitoba Bone Density Program database, which contains clinical BMD results for the Province of Manitoba, Canada, was included in the study.


When 10-year probability for major osteoporotic fracture estimated without knowledge of BMD was high (≥20%), the vast majority (92.8%) qualified for intervention under the National Osteoporosis Foundation (NOF) guidelines, whereas among those at low risk (<10%), the vast majority (80.5%) did not satisfy any NOF intervention criteria. The benefit of including BMD in the risk assessment was greatest among those initially at moderate risk (10–19%) when fracture probability was derived without BMD, but this represented only 29.4% of the cohort (9.3% of those aged <65 years and 48.7% of those ≥65 years).


Fracture probability derived without BMD is able to risk stratify women in terms of future fracture risk and could potentially be sufficient for clinical decision making in many of those designated at low or high fracture risk.


Administrative data Bone mineral density Dual-energy X-ray absorptiometry Fracture prediction FRAX Osteoporosis 

Supplementary material

198_2011_1747_MOESM1_ESM.doc (16 kb)
Supplementary Table 1Area under the curve (AUC) for fracture prediction. (DOC 15.5 kb)
198_2011_1747_MOESM2_ESM.doc (40 kb)
Supplementary Table 2Proportion of individuals satisfying intervention criteria according to risk categorization from hip fracture probability without BMD. (DOC 39.5 kb)
198_2011_1747_MOESM3_ESM.doc (202 kb)
Supplementary Fig. 1Receiver operating characteristic curves for identification of individuals meeting intervention criteria using major osteoporotic fracture probability (dark) and hip fracture probability (light) calculated without BMD. (DOC 202 kb)

Copyright information

© International Osteoporosis Foundation and National Osteoporosis Foundation 2011

Authors and Affiliations

  • W. D. Leslie
    • 1
  • S. Morin
    • 2
  • L. M. Lix
    • 3
  • H. Johansson
    • 4
  • A. Oden
    • 4
  • E. McCloskey
    • 5
  • J. A. Kanis
    • 6
  • for the Manitoba Bone Density Program
  1. 1.Department of Medicine (C5121)University of ManitobaWinnipegCanada
  2. 2.McGill UniversityMontrealCanada
  3. 3.University of SaskatchewanSaskatoonCanada
  4. 4.GothenburgSweden
  5. 5.Osteoporosis CentreNorthern General HospitalSheffieldUK
  6. 6.WHO Collaborating Centre for Metabolic Bone DiseasesUniversity of SheffieldSheffieldUK

Personalised recommendations