Diagnostic Value of Sarcopenia Computed Tomography Metrics for Older Patients with or without Cancers with Gastrointestinal Disorders

The diagnostic utility of poor body composition measures in sarcopenia remains unclear. We hypothesize that the skeletal muscle gauge [combination of skeletal muscle index (SMI) and skeletal muscle density (SMD); SMG = SMI × SMD] would have significant diagnostic and predictive value in certain muscle regions and populations.Prospective cross-sectional study.We examined inpatients age ≥60 years with or without cancer and with gastrointestinal disorders.We used computed tomography (CT) image metrics in the 12th thoracic (T12), third lumbar (L3), erector spinae muscle (ESM), and psoas muscle (PM) regions to establish correlations with the 2019 Asian Working Group for Sarcopenia Consensus and used receiver operating characteristic area under the curve (AUC) to compare differences between metrics. Associations between CT metrics and mortality were reported as relative risk after adjustments.We evaluated 385 patients (median age, 69.0 years; 60.8% men) and found consistent trends in cancer (49.6%) and noncancer (50.4%) cohorts. SMG had a stronger correlation with muscle mass than SMD [mean rho: 0.68 (range, 0.59‒0.73) vs 0.39 (range, 0.28‒0.48); all P < .01] in T12, L3, and PM regions and a stronger correlation with muscle function than SMI [mean rho: 0.60 (range, 0.50‒0.77) vs 0.36 (range, 0.22‒0.58); all P < .05] in T12, ESM, and L3 regions. SMG outperformed SMI in diagnostic accuracy in all regions, particularly for L3 (AUC: 0.87‒0.88 vs 0.80‒0.82; both P < .05). PMG (PM gauge) and L3SMG did not differ, whereas EMG (ESM gauge) or T12SMG and L3SMG did (AUC: 0.80‒0.82 vs 0.87‒0.88; all P < .05). L3SMI, L3SMD, T12SMG, EMG, and PMG showed no association with 1-year cancer-related mortality after adjusting for confounders; however, L3SMG [relative risk = 0.92 (0.85‒0.99); P = .023) was.L3SMG covers all features of sarcopenia with more diagnostic value than other metrics, allowing a complete sarcopenia assessment with CT alone and not just in populations with cancer.