Potential of whole-body dual-energy X-ray absorptiometry to predict muscle size of psoas major, gluteus maximus and back muscles.

BACKGROUND: Measurement of trunk muscle cross-sectional area (CSA) using axial magnetic resonance imaging (MRI) is considered clinically meaningful for understanding several spinal pathologies, such as low back pain and spinal sagittal imbalance. However, it remains unclear whether trunk muscle mass (TMM) measured using dual-energy X-ray absorptiometry (DXA) can predict the trunk muscle CSA. The aim of this study is to determine if DXA-derived TMM is associated and predicts with CSA of paraspinal muscles and gluteus maximus measured using MRI in healthy volunteers. METHODS: A total of 48 healthy volunteers underwent whole-body DXA and MRI of the spinopelvic region. The CSA of the psoas major, back muscles, and gluteus maximus were measured on axial MRI. Correlations and linear regressions between the TMM measured using DXA and the CSA of each musculature were investigated. RESULTS: There was a weak correlation between TMM and CSA of the psoas major in men (r = 0.39, P = 0.0678), and the linear regression was y = 301.74x - 401.24 (R2 = 0.2976, P = 0.0070). A moderate correlation was found in women (r = 0.58, P = 0.0021), and the linear regression was y = 230.21x - 695.29 (R2 = 0.4445, P = 0.0003). Moderate correlations were observed between TMM and CSA of the back muscles in both men (r = 0.63, P = 0.0012) and women (r = 0.63, P = 0.0007), the linear regression was y = 468.52x + 3688.5 (R2 = 0.5505, P < 0.0001) in men and y = 477.39x + 2364.1 (R2 = 0.564, P < 0.0001) in women. There was a strong correlation between TMM and CSA of the gluteus maximus in men (r = 0.72, P < 0.0001), and the linear regression was y = 252.69x - 880.5 (R2 = 0.6906, P < 0.0001). A moderate correlation was found in women (r = 0.69, P < 0.0001), and the linear regression was y = 230.74x - 231.32 (R2 = 0.6542, P < 0.0001). CONCLUSIONS: The DXA-derived TMM was able to predict the CSA of the psoas major, back muscles, and gluteus maximus, and significantly correlated with the CSA of the back muscles and gluteus maximus. It might be a safer and cheaper alternative for evaluating the size of the back muscles and gluteus maximus.