TITLE: “Ankylosing Spondylitis is strongly related to clinical spine fractures independently of drugs use: a register-based case-control study.”

Background/Purpose:

Ankylosing Spondylitis (AS) is associated not only with systemic low bone mass, but also with locally increased fragility due to biomechanical changes in the spine. However, data on the impact of AS on clinical spine fracture risk is scarce. We used a large population-based registry from Denmark to investigate the association between AS and fractures, with a particular focus on the spine.

Methods:

We carried out a case control study. From the Danish National Health Service Registers, we identified 124,655 fracture cases and 373,962 age- and gender-matched controls. Prevalence of AS in cases and controls was estimated. Crude odds ratios (OR) and 95% confidence intervals (CI) according to AS status were calculated using conditional logistic regression. We further adjusted the analyses for: 1. use of oral corticosteroids; and 2.use of oral corticosteroids, NSAIDs and strong analgesics. Similar analyses were repeated separately for the spine, hip and forearm fracture cases and corresponding matched controls.

Results:

Among 124,655 cases, 139 (0.11%) had a diagnosis of AS, while 271 (0.07%) out of 373,962 controls had AS (crude OR 1.54 [95%CI 1.26-1.89]). Similarly, 18 (0.54%) out of 3,364 spine fracture cases compared to 10 (0.10%) matched controls had AS (crude OR 5.42 [2.50-11.70]). Among 10,530 hip fracture cases 7 (0.07%) AS patients were identified, and 27 (0.09%) of 31,356 controls suffered AS (crude OR 0.78 [0.34-1.78]). Finally, 20,035 forearm fracture cases were screened for AS, and a prevalence of 0.08% (n=16) was found, compared to 0.04% (23) in the controls: unadjusted OR 2.09 [1.10-3.95]. The observed associations remained significant after adjustment for use of oral corticosteroids [see Table]. Conversely, all of them were attenuated when adjusted for use of NSAIDs and strong analgesics but the association between AS and clinical spine fracture (adjusted OR 4.41 [1.90-10.20]).

Conclusion:

AS-affected patients are at increased risk for fractures independently of oral corticosteroid use. Clinical spine fractures are the most strongly related to AS, with a multivariate adjusted OR of almost 4.5, independent of use of drugs commonly used for the treatment of AS including oral corticosteroids and NSAIDs. Patients with AS should be fully assessed for fracture risk as part of their clinical management.

TABLE. Results for the association between fracture and AS status (conditional logistic regression).

OR = Odds Ratio; 95%CI = 95% Confidence Interval