Establishment and validation of reference values for abdominal skeletal muscle compartments in children

Abstract

Sarcopenia, resulting from skeletal muscle loss, is a serious concern in children, for whom nutrition and physical activity are essential for growth. In sarcopenia research, the skeletal muscle mass is often obtained from computed tomography (CT) cross-sectional images in both adults and children; however, the method of evaluating skeletal muscle using CT, especially the area range to be measured, has not yet been validated in children. The aim of this study is to establish reference values for three compartments of the abdominal skeletal muscle area (aSMA) obtained from cross-sectional CT images, and to validate the differences among them by assessing correlations with physical development. We conducted a single-centre, retrospective, cross-sectional study and included patients aged 1-17 years who underwent abdominal CT at Chiba University Hospital from 2007 to 2020. Patients with chronic diseases such as malignant tumours and inflammatory bowel disease were excluded from the analysis. aSMA was obtained from CT cross-sections at the lumbar L3-L4 intervertebral disc level by dividing it into three compartments: the psoas muscle area (PMA), paraspinal muscle area (PSMA), and total skeletal muscle area (TSMA). Quartile regression curves by sex and muscle compartment were generated using non-crossover and nonparametric regression quantile estimation. The correlation between each compartment of aSMA with height and weight was assessed using Spearman's rank correlation coefficient. We analysed the abdominal CT images of 593 children (male: n=335, female: n=258) and generated curves predictive for the 5th, 25th, 50th, 75th, and 95th percentiles for each compartment of aSMA by sex. In patients aged 13 years and older, boys had significantly larger aSMA areas than girls in all three compartments. Among the three compartments, PSMA had the strongest correlation coefficient with height and weight for both boys and girls. We generated quantile regression curves for three aSMA compartments obtained from cross-sectional CT images and established reference values in children. PSMA values were most strongly correlated with height and weight among the three aSMA compartments.