Objectives: To characterize paraspinal muscle asymmetry using quantitative CT parameters in adolescent idiopathic scoliosis (AIS) and to examine the associations among muscle asymmetry, vertebral rotation, and curve severity. Methods: A retrospective analysis of 68 AIS patients was conducted. Quantitative CT measured the fatty infiltration rate (FIR) of paraspinal muscles at apical and stable vertebral levels. A muscle asymmetry index was calculated based on the FIR difference between concave and convex sides. Pearson and Spearman correlations and linear regression were used as the main analyses. AVR across upper, pedicle, and lower levels was evaluated using repeated-measures analysis, and SEM was performed as an additional analysis. Inter-observer repeatability of the apical muscle measurements was additionally assessed using an independently repeated segmentation dataset. Results: Paraspinal muscles at the apical region exhibited significant asymmetry, with concave FIR (33.4 ± 14.0%) being significantly higher than convex FIR (15.8 ± 6.8%; p < 0.001). In contrast, the stable vertebra showed no significant asymmetry. Muscle asymmetry was significantly associated with Cobb angle in both Pearson and Spearman analyses (r = 0.456, p = 0.0001; rho = 0.430, p = 0.0003). Its association with AVR was weaker (Pearson r = 0.232, p = 0.058; Spearman rho = 0.302, p = 0.013). In multivariable linear regression, both AVR (β = 1.222, 95% CI 0.827 to 1.617, p < 0.001) and muscle asymmetry (β = 0.375, 95% CI 0.167 to 0.583, p = 0.0006) remained independently associated with Cobb angle. Inter-observer repeatability for the apical muscle asymmetry index remained excellent (ICC(2,1) = 0.958, 95% CI 0.933 to 0.974). Conclusions: Significant asymmetric CT-derived low-attenuation change was observed in the apical paraspinal muscles of patients with AIS, predominating on the concave side. In this severe AIS cohort, paraspinal muscle asymmetry was consistently associated with Cobb angle and showed a weaker, more method-dependent association with AVR. These findings suggest a relationship between paraspinal muscle asymmetry and the severity of three-dimensional deformity.
Zhao et al. (Fri,) studied this question.