Background/Purpose: Rheumatoid arthritis (RA) and osteoporosis are both chronic, potentially debilitating conditions that frequently co-occur, affecting millions of people annually. Among those with osteoporosis, RA is a known risk factor for fracture and is associated with greater morbidity and mortality post fracture. However, variation in the association of RA with specific types of osteoporotic fracture has been less examined, and there is little data among patients receiving osteoporosis therapy. In this study, we examined associations between RA and fractures of the hip, proximal humerus, and wrist among older women who initiated bisphosphonate (BP) therapy. 

Methods: The study cohort included women aged 50-89y who initiated bisphosphonate (BP), therapy during 1998-2019 within a large integrated healthcare system in Northern California, excluding those with select bone disorders, myeloma, metastatic cancer, and end stage renal disease. RA was defined by ≥2 diagnoses in the 5 years prior to BP initiation. Incident hip, proximal humerus, and wrist fractures diagnoses were ascertained using electronic health record data (fracture events with prior fracture diagnosis in the past year not included) up to 3 years post-BP initiation. Cox proportional hazard models were used to examine the association of RA and incident fracture, reporting hazard ratios (HR) with 95% confidence intervals [CI], adjusted for age and race and ethnicity. 

Results: Among 158,404 older women (mean age 70.1 ± 9.2 years; 66.4% White, 3.3% Black, 11.1% Hispanic, 17.5% Asian/Pacific Islander, 1.9% other/unknown) who initiated BP therapy, 3.0% had diagnosed RA. During up to 3 years follow-up, 1.8%, 1.2%, and 1.8% had a hip, proximal humerus, and wrist (distal radius/ulna) fracture. Compared to women without RA, those with RA had a higher incidence of hip (2.9% vs 1.8%, p< 0.001) and proximal humerus (1.8% vs 1.2%, p< 0.001) but not wrist (1.5% vs 1.8%, p=0.08) fractures. Adjusting for age and race and ethnicity, RA was associated with greater risk of hip fracture (HR 1.9 [1.6-2.3]) and humerus fracture (HR 1.6 [1.3-2.0]) but not wrist fracture (HR 0.8 [0.7-1.1]). 

Conclusion: Among women initiating osteoporosis therapy, our study identified site-specific differences in the association of RA and risk of non-vertebral major osteoporotic fracture, with nearly two-fold higher risk of hip fracture associated with RA diagnosis. The risk of proximal humerus fracture was also increased but that of wrist fracture was not. Prior studies also document a strong association between RA and hip fracture but comparably weaker associations between RA and wrist fracture. The lack of association of RA with wrist fracture may be attributable to baseline differences in functional status and level of physical activity. While pharmacologic treatment factors (including BP duration) and RA severity were not examined, these findings contribute to the growing recognition that RA is associated with differential fracture risk that varies by skeletal site. Future research should investigate the relationship between RA symptom severity and skeletal health outcomes.

« Back to ACR Convergence 2024

ACR Meeting Abstracts - https://acrabstracts.org/abstract/site-specific-differences-in-the-association-of-non-vertebral-major-osteoporotic-fractures-and-rheumatoid-arthritis/