Dynamic CT-based body composition analysis predicts surgical risk in Crohn’s disease with small bowel stenosis: a retrospective cohort study

Background Reliable predictors of surgical risk in Crohn’s disease (CD) with small bowel stenosis are lacking. Longitudinal CT enterography (CTE) derived body composition parameters may improve risk stratification. Aims To evaluate whether longitudinal CTE-derived body composition changes predict surgical risk in CD patients with small bowel stenosis. Methods This retrospective cohort study analyzed 385 CD patients between January 2018 and June 2022 with paired CTE scans. High-risk patients ( n = 96) required surgery for complications; low-risk patients ( n = 289) achieved medical remission. Skeletal muscle (SM), subcutaneous adipose tissue (SAT), visceral adipose tissue (VAT), and intermuscular adipose tissue (IMAT) metrics at L3–L5 levels were measured and normalized by vertebral height. Gender-stratified analyzes and Cox regression identified predictors. Results There were 289 cases in the low-risk group and 96 cases in the high-risk group. Interaction terms (time and gender) were tested, males showed significant reductions in L3–L4 skeletal muscle index (SMI) ( p < 0.001), L3-L4 IMAT index ( p < 0.001, p = 0.04), and L4-L5 VAT density ( p = 0.008, p = 0.005). Independent predictors of surgical risk included SAT density at L5 level in baseline ( p = 0.005), SMI at L3 level in follow up ( p < 0.001), VAT/total adipose tissue index (VTR) ( p = 0.004), delta SMI at L4 level ( p < 0.001), age ( p < 0.001), platelet count ( p = 0.010), erythrocyte sedimentation rate ( p < 0.001), and stenosis length ( p = 0.001). Conclusion Dynamic body composition parameters, particularly delta SMI and adipose tissue parameters, serve as valuable imaging biomarkers for predicting surgical necessity in CD patients with small bowel stenosis.