Background/Purpose: Conventional DXA imaging of spine and hip to measure bone mineral density (BMD) has limitations in patients with ankylosying spondylitis (AS) as their spinal DXA measurements may be falsely elevated due to syndesmophyte formation and ligament ossification.  Therefore, we investigated the correlation of hip and spine BMD measurements in patients with AS to determine if hip DXA will prove clinically useful while avoiding the confounding effect of spinal disease. We also studied risk factors for osteoporosis (OP) in AS.

Methods: We identified patients from our AS registry ≥ 18 years of age who met the ASAS (Assessment of Spondyloarthritis International Society) classification criteria for AS. Patients with thyroid and parathyroid disease, chronic liver or kidney disease, on anti-convulsant medications, surgical spinal fusions and hip arthroplasties were excluded. Demographic and clinical data were recorded and included HLA- B27 status, disease duration >5 years, presence of syndesmophytes on lumbar X-ray. Disease activity was measured using Bath AS Disease Activity Index (BASDI). BMD was measured using a Hologic machine and interpreted using ISCD 2005 guidelines and WHO criteria for definition of OP. Patients were divided into 3 groups- osteoporosis ( T-score< -2.5), osteopenia (T-score of - 1.0 and -2.5) and normal ( T-score of  >-1.0). In addition, ESR and blood levels of CRP, and 25-hydroxy vitamin D were measured.

Correlation between the lowest T-scores of hip (total hip or femoral neck) and lumbar spine was measured using Spearman’s correlation coefficient (rho). Chi-square and odds ratio using logistic regression were used to assess the association of the purported risk factors for OP in these patients.

Results: We identified 101 patients with AS; 26.2 % females, and 25.2 % African-Americans (AA). The mean age was 43.0 years (±13.7) in patients with normal BMD versus 47.8 years (±14.4) with OP and osteopenia (p= 0.0867), and 40.5 % of patients had syndesmophytes on lumbar Xray. Prevalence of OP = 16.8%, osteopenia = 36.6% and 46.5% had normal BMD. There was moderate correlation between the lowest T-values of hip and lumbar spine (AP view), rho= 0.59 (figure 1). The AA with AS had higher odds of having osteoporosis compared to Whites; Odds ratio (OR) = 5.3 (1.03-26.84) (95% CI), p = 0.045 and also AS patients with high CRP levels had higher odds of having OP, OR= 4.1(1.22-13.97) (95% CI), p=0.0226. There was no association between OP and age, sex, BASDI, vitamin D levels, HLAB27 positivity, and ESR.

Conclusion: Our results demonstrated a correlation between T-scores of hip and lumbar spine in patients with AS with disease duration of > five years, suggesting that DXA of the hip and the lumbar spine are useful in diagnosing OP in patients with AS. Elevated CRP level increases the risk of OP in patients with AS. African-Americans with AS are at higher risk of developing OP compared to whites.