Abstract
Background. Nutritional status impairments accompanying inflammatory bowel disease (IBD) adversely affect both the achievement and maintenance of remission and patients’ quality of life. Despite the potential utility of computed tomography (CT) as an adjunctive method for body composition assessment, its application in this context remains insufficiently explored. Aim. To evaluate the applicability of CT-derived body composition parameters for the objective characterization of nutritional status in patients with IBD. Materials and methods. A retrospective analysis was conducted using data from patients who underwent evaluation and treatment in a gastroenterology department. Outcomes included laboratory parameters (hemoglobin, total protein, albumin, serum iron, and total cholesterol concentrations) and body composition metrics assessed on CT using a computer vision algorithm: Visceral Fat Area (VFA), Subcutaneous Fat Area (SFA), Intramuscular Fat Area (IFA), VFA/SFA ratio, intramuscular fat fraction, Skeletal Muscle Area (SMA), Skeletal Muscle Index (SMI), SMA/VFA ratio, and CT-derived Skeletal Muscle Radiodensity (SMRD). Results. The study included 142 patients with IBD: 45 (31.7%) with ulcerative colitis and 97 (68.3%) with Crohn’s disease. IBD exacerbation was associated with lower serum iron levels (odds ratio – 1.08, 95% confidence interval – 1.02–1.15; p=0.005). No associations were identified between clinical disease activity and body composition parameters. In analyses of laboratory markers and body composition, higher albumin and hemoglobin levels were associated with higher SMI and SMRD. Total protein was positively associated with SMA (β=0.024; p=0.003), SMI (β=0.028; p=0.003), and SMRD (β=0.024; p=0.020), and inversely associated with intramuscular fat fraction (β=-0.027; p=0.043). Conclusion. Opportunistic CT analysis may enable objective assessment of protein nutritional status in patients with IBD without increasing radiation exposure. The robustness of muscle-related parameters as morphological markers is supported by their independence from disease activity phase and their correlation with biochemical indicators of protein metabolism.

This publication has 30 references indexed in Scilit: