Postgrad Med J 85:38-39 doi:10.1136/pgmj.2007.065748
  • Original article

Search for secondary osteoporosis: are Z scores useful predictors?

  1. K Swaminathan1,
  2. R Flynn2,
  3. M Garton3,
  4. C Paterson4,
  5. G Leese1
  1. 1
    Department of Endocrinology, Ninewells Hospital, Dundee, UK
  2. 2
    MEMO, Division of Medicine & Therapeutics, Ninewells Hospital, Dundee, UK
  3. 3
    Department of Rheumatology, Perth Royal Infirmary, Perth, UK
  4. 4
    Department of Medicine, Ninewells Hospital, Dundee, UK
  1. Dr K Swaminathan, Wards 1&2, Ninewells Hospital, Dundee DD1 9SY, UK; krishnan.swaminathan{at}
  • Received 30 October 2007
  • Accepted 26 September 2008


Aim: To determine whether Z scores can be used to predict the likelihood of patients having a secondary cause of low bone mineral density.

Methods: A retrospective cross sectional study was conducted among 136 consecutive patients with osteoporosis at Ninewells Hospital, Dundee, UK, between 1998–2002.

Results: 20.5% of female patients in this study were identified with previously unrecognised contributors to the low bone mineral density. In women, at a Z score cut-off of −1, the sensitivity of detecting a secondary cause for osteoporosis is 87.5% with a positive predictive value of 29.2%.

Conclusion: In women, a Z score of −1 would identify a majority of patients with a secondary cause for low bone mineral density and identifies patients who would especially benefit from a thorough history and clinical examination.


  • Competing interests: None.

  • Funding: None

Responses to this article