Abstract
Objective To compare the accuracy of Bioelectrical impedance (BIA) and Computer tomography (CT) in evaluating body composition, specifically assessing visceral fat area (VFA) and skeletal muscle index (SMI), in Chinese gastric cancer patients.Method Spearman correlation analysis was employed to evaluate the relationship between the two methods, and the Bland-Altman method was used to assess their agreement. The performance of BIA in diagnosing low muscle mass and visceral obesity was evaluated based on the area under the receiver operating characteristic (ROC) curve.Result There was a significant positive correlation (r = 0.650) between BIA and CT methods for detecting SMI and VFA, p < 0.001;r = 0.772, p < 0.001). Bland Altman analysis revealed an average bias between the BIA and CT measurement systems of 0.920 ± 1.161 kg/m² (95% CI: -1.658-2.898 kg/m²). The average bias for the VFA measurement system was − 12.304 ± 24.973 cm² (95% CI: -61.252–36.644 cm²). The 95% limits of agreement (LOA) range for percentage deviation were − 19.837–33.702%, and for VFA it was − 90.138–57.034%. Significant differences were observed in the mean biases of SMI and VFA measurements between males and females (SMI: 1.008 ± 1.256 kg/m² vs 0.761 ± 0.954 kg/m², P = 0.036; VFA: 16.206 ± 23.807 cm² vs -5.225 ± 25.570 cm², P < 0.001). Using CT as a reference, the cutoff value for identifying low muscle mass using SMI measured by BIA was 9.754 kg/m² for males and 8.498 kg/m² for females. The critical value for identifying visceral obesity using VFA measured by BIA was 70.45 cm² (AUC = 0.895 (0.860–0.930)), AUCmen = 0.771 (0.708–0.833), AUCwomen = 0.770 (0.688–0.853).Conclusion Although a significant correlation was observed, caution should be exercised when using BIA and CT interchangeably for assessing SMI and VFA values. BIA tended to overestimate SMI and underestimate VFA, especially across different genders. This resulted in an underestimation of the prevalence of low muscle mass and an overestimation of visceral obesity.