Comparison of Femoral Neck and Lumbar Spine BMD as Correlates of Atlantoaxial Subluxation in Rheumatoid Arthritis

Objective: Lumbar spine and femoral neck dual-energy X-ray absorptiometry (DXA) measurements are known to be influenced by degenerative changes to different extents. This study evaluated which site shows a stronger association with atlantoaxial subluxation (AAS) in rheumatoid arthritis (RA). Methods: We retrospectively identified 952 RA patients who had both cervical radiographs and dual-site DXA. AAS was defined as an anterior atlantodental interval exceeding 3 mm. We constructed a logistic regression model for each DXA site and compared their discriminative accuracy using the DeLong test, the Akaike information criterion (AIC), and subgroup analyses stratified by body mass index (BMI) tertile. Results: Of 952 patients, 381 (40.0%) had AAS. Each one-standard deviation (SD) increase in femoral neck T-score reduced AAS odds by 32% (adjusted odds ratio [OR] 0.68; 95% confidence interval [CI] 0.58-0.80; p < 0.001); the lumbar spine yielded 15% (adjusted OR 0.85; 95% CI 0.73-0.99; p = 0.035). The femoral neck model had a higher area under the curve (AUC) (0.77 vs. 0.71; DeLong p = 0.008) and lower AIC (1125 vs. 1152). Disparity was widest in the lowest BMI tertile (adjusted OR 0.55 vs. 0.91). Isolated femoral neck osteoporosis raised AAS risk 3.1-fold; lumbar spine osteoporosis alone did not. Conclusion: Femoral neck bone mineral density (BMD) was more strongly associated with AAS than lumbar spine BMD in RA patients. When screening for cervical instability, the femoral neck T-score may be more informative than the lumbar spine value, particularly in patients with low BMI.