Article Text

OP0039 A Five Year Prospective Study of Bone Mineral Density in Ankylosing Spondylitis
  1. A. Deminger1,
  2. E. Klingberg1,
  3. M. Lorentzon2,
  4. M. Hedberg3,
  5. E. Rehnberg4,
  6. H. Carlsten1,
  7. L.T. Jacobsson1,
  8. H. Forsblad-d'Elia1
  1. 1Department of Rheumatology and Inflammation Research
  2. 2Center for Bone Research, Sahlgrenska Academy at University of Gothenburg, Göteborg
  3. 3Section of Rheumatology, Borås Hospital, Borås
  4. 4Section of Rheumatology, Alingsås Hospital, Alingsås, Sweden


Background In ankylosing spondylitis (AS) two different bone remodeling processes are taking place; new bone formation with the development of syndesmophytes and loss of bone leading to increased risk of osteoporosis and vertebral fractures. Most studies of bone mineral density (BMD) in AS have been cross sectional and only a few studies have investigated the changes in BMD over time.

Objectives Our aim was to investigate how BMD changes over five years in a group of patients with AS and in addition search for predictors for the change in BMD.

Methods 204 patients from rheumatology departments in western Sweden were included 2009 in a longitudinal study. The patients met the modified New York criteria for AS. At baseline and after five years BMD was measured with dual-energy x-ray absorptiometry (DXA) in hip, radius and lumbar spine in anteroposterior (AP) and lateral projections. Other methods of assessment were questionnaires, back mobility tests, blood samples and lateral spinal radiographs for syndesmophyte grading (mSASSS).

Results At baseline 204 patients were examined with DXA, 117 (57%) men and 87 (43%) women. The median age was 49 (17-78) years and median duration of symptoms was 24 (2-55) years. 21% were treated with TNF inhibitors (TNFi), 30% used conventional DMARDs and 77% NSAIDs. At the 5-year follow-up 165 patients (80%) were reexamined with DXA, 90 (55%) men and 75 (45%) women.

In the table the results of the changes in BMD (ΔBMD) at different measuring sites in total and divided by gender are shown. ΔBMD in different locations were small but significant in the whole group. Divided by gender all sites changed significantly in men and in women significantly in radius and femoral neck. BMD increased in the spine and decreased in femoral neck and radius.

As secondary aims we investigated how ΔBMD at various locations were correlated to baseline characteristics as age, duration of symptoms, duration of diagnosis, BASMI, BASFI, BASDAI, ASDASCRP, BAS-G1, BAS-G2, mSASSS, CRP, ESR and mean ESR 5 years before baseline and during the follow up period. Analyses showed that higher mean ESR 5 year prior to the follow-up correlated with decreased BMD in hip total, radius and lumbar spine lateral and volumetric. Higher BASMI at baseline correlated with increase in BMD in femoral neck and lumbar spine AP. We also studied if there were any differences in ΔBMD in relation to treatment with TNFi and to presence of syndesmophytes at baseline. Treatment with TNFi showed significant difference in lumbar spine AP and hip total with higher ΔBMD in the treatment group. Presence of syndesmophytes at baseline correlated with increase in BMD in hip total and femoral neck. High age correlated with decrease in BMD in radius.

Conclusions Over all there were small but significant changes in BMD during five years. BMD increased significantly in the spine and decreased significantly in radius and hip neck. Treatment with TNFi seems to increase BMD in both the spine and hip. Prolonged high inflammation measured by mean ESR was a risk factor for bone loss.

Disclosure of Interest None declared

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.