Body composition analysis using abdominal scans from routine clinical care in patients with Crohn's Disease

Crohn's Disease is associated with body composition changes, which have important treatment and prognostic implications. Measurement of body composition usually requires dedicated scanning or measurement, with retrospective analysis of existing datasets impossible. We sought to determine whether single slice analysis of abdominal scans, obtained during routine clinical care, in patients with Crohn's Disease accurately predicts body composition compartments.

Abdominal CT images of patients with Crohn's disease were analyzed and comparison was made with total body fat-free mass, total body fat mass, femoral neck t-score, and other parameters reported from DXA, the reference method.

Thirty-seven subjects were identified, 15 male and 22 female, with a mean age of 43.8 years. There was significant correlation (Pearson r = 0.923, p < 0.001) between skeletal muscle area from CT and total fat-free mass measured by DXA. Similarly, total body fat mass correlated strongly (r = 0.928, p < 0.0001) with subcutaneous fat area. In this cohort of ambulatory Crohn's Disease patients, low muscle mass/sarcopenia was prevalent and predictive of lower bone mineral density.

Fat mass, fat-free mass, and appendicular skeletal muscle index can be predicted by analysis of a single CT slice in patients with Crohn's Disease. Similar to published data from healthy subjects, the L3 vertebral body level provided the most robust correlation with most parameters. This study represents the first published use of routinely obtained abdominal imaging to demonstrate this relationship - and to predict body composition components - in patients with inflammatory bowel disease.