Chest computed tomography is a valid measure of body composition in individuals with advanced lung disease.

Abstract

There is growing interest in evaluating body composition using routine clinical computed tomography (CT) scans; however, the validity of this technique in lung transplant patients has not been described. The study objectives were to determine the reliability of measuring fat compartments from thoracic CT and evaluate the validity of muscle and fat cross-sectional area (CSA) from thoracic CT by comparing to bioelectrical impedance analysis (BIA). Thoracic CT scans from lung transplant assessments were obtained for analysis. Total thoracic muscle CSA, pectoral muscle CSA, subcutaneous adipose tissue (SAT), and mediastinal adipose tissue (MAT) were manually segmented by two independent raters. Reliability was analysed using intra-class correlation coefficient (ICC). Correlations were determined between CT measures with fat-free mass index (FFMI), body fat mass index (BFMI) and per cent body fat (%BF) from BIA; and anthropometrics [body mass index (BMI) and waist circumference (WC)]. High inter- and intra-rater reliability were found for SAT and MAT (ICCs=0.99). Pectoral and total muscle CSA were correlated with FFMI (r=.41, p=.003 and r=.57, p<.001, respectively). SAT was associated with whole-body fat from BIA and with BMI and WC (r=.61 to .80, p<.001). MAT was associated with BMI (r=.58, p<.001) and WC (r=.61, p<.001). This study supports the reliability and validity of using thoracic CT to measure muscle and fat. Future studies are needed to investigate whether these CT-based measures are predictive of clinical and post-transplant outcomes in advanced lung disease.