Study design: A retrospective cross-sectional study.
Objective: The aim of this study was to identify the prevalence of and risk factors for lumbar scoliosis in patients with rheumatoid arthritis (RA) using lumbar images obtained from dual-energy x-ray absorptiometry (DXA).
Summary of background data: The prevalence of lumbar scoliosis in the normal adult population has been reported, but that in patients with RA remains unclear.
Methods: Subjects comprised 241 patients with RA who underwent annual DXA. Cobb angles of the lumbar spine were measured by lumbar anteroposterior DXA and the prevalence of lumbar scoliosis (curvature ≥10°) was calculated. Correlations between lumbar scoliosis and potential risk factors (age, sex, duration of RA, T score of lumbar spine and hip, medications for RA [daily predonisolone dose, use of biological agents] and osteoporosis, disease activity score-C-reactive protein, progression stage and functional classification of RA, and severity of hand deformity were analyzed.
Results: The prevalence of lumbar scoliosis in patients with RA was 32.0%. Mean Cobb angle was 7.1º ± 5.5º among all subjects, compared with 13.6º ± 4.4º (range, 10º-32º) among subjects with scoliosis. Subjects with scoliosis were significantly older (67.8 yr) than those without (61.6 yr, P < 0.0001). The daily prednisolone dose was significantly higher in subjects with scoliosis (4.14 mg) than in those without (3.46 mg, P = 0.0389). The T score of the hip was significantly smaller in subjects with scoliosis (-1.79) than in those without (-1.26, P = 0.0005). A multivariate logistic regression analysis revealed age as the sole risk factor for lumbar scoliotic changes in patients with RA (odds ratio, 1.068; 95% confidence interval, 1.031-1.107; P = 0.0003).
Conclusion: The prevalence of lumbar scoliosis in patients with RA was 32.0%, about 3- or 4-times higher than its prevalence as obtained from previous reports of DXA cohorts irrespective of RA. Increased age represented an independent risk factor for lumbar scoliosis in patients with RA.
Level of evidence: 4.