**4. Discussion**

The clinical importance of TMM–BIA was first reported by Hori et al. [12]. They conducted a multicenter, cross-sectional study of 1738 patients (mean age, 70.2 ± 11.0 years; 781 males and 957 females) and found that TMM–BIA was significantly associated with various spinal pathologies, including low back pain, quality of life related to low back pain, and spinal sagittal imbalance, indicating that TMM–BIA is a useful indicator for

understanding the pathology of the spine in clinical settings. However, only one study [19] has investigated the association between TMM–BIA and other pre-existing assessment methods for trunk muscles; therefore, the accuracy of TMM–BIA has ye<sup>t</sup> to be validated.

The present study is the first validation study of TMM–BIA, and the results of this study indicate a strong correlation between TMM–BIA and the CSA of PVM. Furthermore, we clarified that TMM–BIA is more strongly correlated with the CSA of PVM, excluding fat infiltration, than the total PVM. Our results sugges<sup>t</sup> that TMM–BIA is a valid index of trunk muscle mass.

The CSA of PVM and fat infiltration of the PVM measured via MRI or CT has been widely used for quantitative evaluation of the trunk muscles. Many studies have sought to investigate the association between the CSA, or fat infiltration, and spinal pathologies [2,23,24]. Takahashi et al. [24] reported that a decrease in PVM in patients with osteoporotic vertebral fractures was significantly related to low back pain and delayed union after fracture onset. Kjaer et al. determined that fat infiltration of the MF was associated with low back pain in adults [25]. Sasaki et al. [26] found that the fatty infiltration ratio of the ES in the upper lumbar spine was significantly associated with low back pain. However, the widespread use of MRI or CT for the evaluation of trunk muscle mass is impractical, as it is time-consuming and expensive, and CT exposes patients to radiation. In contrast, TMM–BIA is a straightforward, non-invasive, and reliable method for large-scale measurements.

Functional assessment of the trunk muscles was performed via BMS. Several studies have reported that BMS may be a useful index for spinal pathology and function, such as spinal sagittal alignment [27], thoracic kyphosis [28], and range of motion of the spine [29]. Despite its clinical importance, the measurement of BMS is difficult in patients with low back pain, and has the potential risk of vertebral fracture in patients with severe osteoporosis [30]. We found a strong correlation between TMM–BIA and BMS using a relatively large sample size, which indicates that TMM–BIA is an accurate tool for the functional assessment of trunk muscles without any risk of adverse effects.

Our study had several limitations. First, TMM–BIA includes the total volume of all trunk muscles (not only PVM); however, we could only measure the CSA of PVM. Therefore, other trunk muscles were not evaluated using TMM–BIA in this study. Second, in this study, we measured trunk muscle mass using only one type of BIA device. It has been reported that the ASM varies depending on the type and manufacturer of the BIA device [31]. Therefore, trunk muscle masses may differ when another BIA device is used. A conversion formula that shows the same ASM across BIA devices has been reported [32]. Future studies to develop a similar conversion formula for trunk muscle mass are needed. Third, we did not analyze the influence of sex or age in this study. There were significant differences in CSA, BMS, and TMM–BIA between male and female study participants (Table 1). As the purpose of this study was first to verify whether TMM–BIA correlates with the quantitative and functional assessments traditionally used for trunk muscles, an examination of the influence of sex or age on the relationship of TMM–BIA and the CSA of PVM will be the subject of our next research work. Last, this study was a cross-sectional analysis of the relationship between TMM–BIA and the CSA of PVM using data collected on the same day. Therefore, the relationship between changes in TMM–BIA and those in the CSA of PVM was not studied. Future studies should focus on analyzing the changes in these parameters via a longitudinal study design.
