ABSTRACT
Abstract Background: To compare the incidence of osteoporotic fractures in patients with rheumatoid arthritis (RA) with matched controls from a university hospital. Methods: Consecutive RA patients (n = 100) were matched (age and sex) with controls (1:2). The follow-up period began at the index date, defined as the date of diagnosis for RA patients and the date of the first medical claim at the Health Management Organization (HMO) for non-RA patients. Fracture incidence rates per 1000 persons-years (PY) for distinct types of fractures were calculated. Multivariate cox regression analysis was performed to identify factors associated with fractures. Results: One hundred RA patients were followed for a total of 975.1 patients-years and 200 controls for 1485.7 patients-years. No difference was found in the overall fracture incidence rate per 1000 PY between RA and controls (19.5, 95% CI 12.7-28.6 vs 12.1, 95% CI 7.7-18.7, p =0.07). In the Cox regression analysis, only age (HR 1.06, 95% CI 1.02-1.11, p = 0.006) and history of a prior fracture (HR 9.85, 95% CI 2.97-32.64, p < 0.001) were associated with fractures after the index date. The stratified analysis of the fractures by location showed that only the vertebral fractures were more frequent in RA patients compared with controls (12.9 per 1000 PY, 95% CI 8.9-25.8, vs. 3.4, 95% CI 1.4-8.1, respectively, p =0.01). Conclusion: Patients with RA didn't show an overall increased risk of osteoporotic fractures compared with matched controls, but vertebral fractures were more frequently observed in patients with RA.