Abstract / Summary
Abstract Purpose To determine whether dual-energy X-ray absorptiometry (DXA)-derived trunk lean mass corresponds to computed tomography (CT)-based local muscle size and attenuation, and to describe 2-year recovery after surgery for sagittal adult spinal deformity (ASD). Methods This single-center cohort study included 121 women aged 60 years or older with sagittal ASD and no major coronal curve. Ninety-two had analyzable preoperative L3 CT and paired DXA data. CT cross-sectional area (CSA) represented local muscle quantity and CT attenuation represented local radiodensity. Associations between DXA trunk lean mass and five prespecified CT metrics were tested with false-discovery-rate correction. Radiographic, physical-performance, and disability outcomes were assessed preoperatively and at 2 years. Results DXA trunk lean mass correlated with psoas CSA (r = 0.49; adjusted p = 0.005) but not with multifidus CSA or attenuation of the multifidus, erector spinae, or psoas. Sagittal alignment improved substantially. Stride length, one-leg standing, stand-up performance, and Locomo-25 improved, whereas gait speed, grip strength, and two-step outcomes did not. Oswestry Disability Index and Roland-Morris Disability Questionnaire scores improved in 91 matched patients. Conclusion DXA trunk lean mass and CT attenuation provide complementary rather than interchangeable information. Two-year recovery after sagittal ASD correction was clinically meaningful but domain specific. Multimodal muscle assessment may help characterize preoperative muscle phenotype, although prognostic use requires prospective validation.