Ab0341 comparison of candidate key factors for incident fragility fracture in postmenopausal female patients with rheumatoid arthritis

Background Bone fragility fracture (BFF) in patients with rheumatoid arthritis (RA) is still a significant comorbidity that deteriorates the quality of life after that. Some reports clarified risk factors and the keys to preventing them. Objectives The aim of this study is to evaluate the factors statistically and advocate a novel key to preventing the incidence of BFF. Methods A retrospective cohort study data was used. Postmenopausal RA patients were picked up. They have been treating RA and measured bone mineral density (BMD) with dual-energy X-ray absorptiometry. The initial target of therapy is the attainment of remission with SDAI within 6 months of initiation. The primary outcome was incident BFF (inc-BFF). Follow-up started at BMD measurement (baseline) and continued until the development of the first fracture or censoring at death, loss to follow-up or end of the study. In these subjects, T-scores and pr-BFF were combined, and T-score < -2.5 in the lumbar spine (LS) or femoral neck (FN), or presence of pr-BFF was evaluated as primary osteoporosis criteria matched (pr-OP). Potential key factors for incident BFF, such as age, disease duration of RA, anti-citrullinated polypeptide antibodies (ACPA), rheumatoid factor (RF), remission rate using clinical disease activity index (CDAIRR), C-reactive protein, Health Assessment Questionnaire Disability Index, pain score using a visual analog scale (PS-VAS), vector value of coordinate on X-axis and Y-axis (Vxy) using Joint Index Vector (JIV), body mass index (BMI), presence of comorbidities such as hyper-fallability (Fall), lifestyle-related diseases (LSD), cognitive impairment (CI), estimated glomerular filtration rate calculated with cystatin C (eGFR_CysC), anti-osteoporotic drug administration and glucocorticoid administration, and serum albumin level (ALB) were chosen as variants. Receptor operation characteristics (ROC) curve was examined for each variant in regard to inc-BFF and the Cut-off index (COI) for each variant was determined. A Cox regression analysis with a multivariate mode in the variants that had statistical significance with the ROC. Kaplan-Meier survival curve (K-M) was examined for each variant in regard to COI. Finally, the K-M study examined the chi-square test by dividing it into positive pr-OP, positive all single significant variants, and positive combined conditions, and simultaneously calculated inc-BFF rates for each subgroup. Differences in BFF rates were compared for each matching pair. Statistical significance was set at less than 5% for all statistical methods. Results A total of 239 patients were recruited. The mean age was 73.6 years and the mean follow-up period was 52.4 months. In the ROC study, pr-OP, ACPA, CDAIRR, PS-VAS, Vxy, Fall, LSD, CI, eGFR_CysC, and ALB demonstrated statistical significance with the COI of presence, 0.9 (U/mL), 0.52, 25.0 (mm), 0.012, presence, presence, presence, 50.7 (ml/min/1.41m2), and 4.0 (g/dL), respectively. In these, PS-VAS, Vxy, and ALB had significant risk ratios with values of 1.04, 0.07, and 0.20, respectively. Hazard ratios of each variant in the K-M study were 3.51, 4.56, and 1.81, however, p-values were <0.001, <0.01, and 0.07 for PS-VAS, Vxy, and ALB, respectively. The BFF rate in the pr-OP and PS-VAS ≥ 25.0 subgroups were 4.5%, 12.9%, 14.7%, and 37.9% in the negative/negative, negative/positive, positive/negative, and positive/positive groups, respectively, whereas the BFF rate in the pr-OP and Vxy ≥ 0.012 subgroups were 4.3%, 8.3%, 6.3%, and 32.4% in the negative/negative, negative/positive, positive/negative, and positive/positive group, respectively, When PS-VAS and Vxy were combined to matching either of these, the BFF rate in the subgroups were 5.6%, 12.5%, 12.8%, and 49.0% in the negative/negative, negative/positive, positive/negative, and positive/positive group, respectively. Conclusion These results suggest that Vxy > 0.012 and PS-VAS > 25 mm are available risk indicators for inc-BFF. The composite indicator should be more predictable. REFERENCES: NIL. Acknowledgements: NIL. Disclosure of Interests None Declared.