Article Text


SAT0153 Vertebral deformities in rheumatoid arthritis: the oslo, truro, and amsterdam collaborative study
  1. MC Lodder1,
  2. G Haugeberg2,
  3. WF Lems1,
  4. RE Orstavik2,
  5. BA Dijkmans1,
  6. TK Kvien2,
  7. AD Woolf3
  1. 1Rheumatology, Vrije Universiteit Medical Centre, Amsterdam, The Netherlands
  2. 2Rheumatology, Diakonhjemmet Hospital, Oslo, Norway
  3. 3Rheumatology, Royal Cornwall Hospital, Truro, UK


Objectives To determine the occurrence of vertebral deformities (VD) in a cross-sectional study of postmenopausal women with rheumatoid arthritis (RA). Secondly, associates for VD will be identified.

Methods A total of 150 female patients (50 per centre), aged between 50 and 70 years and a disease duration of 5 or more years, attending a general rheumatology clinic were consecutively included. Demographic and clinical data were collected according to a standardised procedure; bone fracture history, disease activity measures, hand X-rays (Larsen) and bone mineral density by means of dual energy X-ray absorptiometry of hip and lumbar spine. Lateral X-rays of thoracal and lumbar spine were read by one observer. Assessment was done as proposed by Genant.1Associations between the presence of VD, and demographic, clinical and BMD variables were examined bivariately. The multivariate analysis was carried out using logistic regression.

Results Spine X-rays of 137 women were included in the analysis. Of these women 20 out of 137 (14.6%) had at least one VD grade one or higher; 18% in Oslo, 10% in Amsterdam, and 16% in Truro. The table shows results of the bivariate analysis. Ever use of corticosteroids was not associated with VD (p = 0.33).

Previous non-vertebral fracture after age 25 (as recalled by the patient) was the independent associate for VD in a logistic regression model containing age, weight, total Larsen score, non-vertebral fracture after age 25 and either femoral neck or spine BMD. Femoral neck BMD was independent variable in a model with age, weight, total Larsen score and femoral neck BMD. Age was the only independent associate of VD if femoral neck BMD was replaced by spine BMD in the latter model. The inclusion of corticosteroid use in either one of the models did not change the results.

Abstract SAT0153 Table 1

Conclusion VD in a clinic population of RA patients was associated with either previous clinical fracture, femoral neck BMD or age as single independent variable dependent on the regression model applied. Studies on more patients allowing for larger regression models should further clarify the true associates of VD.


  1. Genant HK, et al. J Bone Miner Res. 1993;9:1137–48

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.