Background/Purpose: Bone mineral density (BMD) at the 3 central sites usually correlates with each other but the effect of disease states on this correlation is not known. Ankylosing spondylitis (AS) with spinal involvement may increase BMD at spine but not at hip. Alternatively, RA or osteoporosis may not have any effect the on it. These aspects have not been well quantified and this study aims to assess the correlation amongst patients with the three disorders.

Methods: Retrospective chart review of patients who had been seen in rheumatology clinics at VA medical center with diagnosis of AS, RA or OP and had a DEXA scan done was performed. Data including age, race, gender, BMI, BMD with T-scores was tabulated. Patients with BMD at all 3 sites- lumbar spine, hip and arm were included in analysis. Correlation amongst the amongst the T-scores at three sites was calculated using pearson coefficient.

Results: A total of 350, patient charts were reviewed. After excluding patients with missing data, 241 were included in analysis. There were 16 patients with AS (Mean age = 63 years, Average BMI= 28.81, all males), 117 patients with RA (Mean age = 63.92 years, Average BMI= 28.38, 104 males and 13 females), and 98 patients with osteoporosis (Mean age = 70.42 years, Average BMI= 25.17, 93 males and 5 females). The correlation between the total hip and femoral neck was strong in all three groups. We found moderate correlation among the t-scores between spine – total hip and forearm – total hip t-scores in both AS and RA. The correlation was poor at these 2 sites in patients with osteoporosis. The spine BMD was found to be lower than other sites in 1/16 (6.25%) of AS and 5/117 (4.3%) of RA patients as compared to 16/98 (16.4%) of patients with OP, where it was also likely to influence treatment decision. Also additional forearm testing did not yield a better result in all 3 groups. These results are summarized in the table 1.

Conclusion: The results show that there was correlation among the three sites of BMD in the whole cohort and no significant difference was found among the three disorders.  Having AS where spine is frequently involved did not seem to impact the correlation coefficient when compared to RA and OP patients. This suggests that spinal involvement in AS may not have any significant impact on the BMD at LS spine in majority of patients. The correlation was lowest among the patients with osteoporosis who were older than the other 2 groups and osteophyte formation may have influenced the BMD at spine. Interestingly the BMD at spine was most useful in this group impacting the treatment decisions more than other 2. The results suggest that evaluation of BMD at both spine and hip is important regardless of the underlying disorder including AS and should be done collectively. On the other hand, forearm BMD evaluation was not useful in majority.

Table 1: Correlation among BMD at 3 sites (Hip, Lumbar spine and forearm) in patients with AS, RA and OP.