Skip to main content
Research Highlights

Body Composition Predicts Biologic Treatment Failure in Crohn’s Disease

Edited by 

Key Clinical Summary

  • Higher visceral adipose tissue was associated with greater 52-week loss of response to infliximab or ustekinumab in patients with Crohn’s disease.

  • Lower skeletal muscle index was associated with both primary nonresponse and subsequent loss of response to biologic therapy.

  • A model combining body composition and clinical parameters predicted 52-week loss of response, with an externally validated area under the curve (AUC) of 0.902.

A multicenter retrospective cohort study published in Inflammatory Bowel Diseases found that elevated visceral adipose tissue (VAT) and reduced skeletal muscle (SM) were associated with biologic treatment failure in Crohn’s disease. The findings suggest that computed tomography–derived body composition measures may help identify patients at increased risk for loss of response to infliximab or ustekinumab.

Study Findings

Investigators enrolled 248 patients with Crohn’s disease who initiated infliximab or ustekinumab between January 2018 and December 2023. The multicenter study included investigators from institutions in Nanjing, Yangzhou, and Suzhou, China. Visceral fat index (VFI) and skeletal muscle index (SMI) were measured using computed tomography. The primary outcome was loss of response at 52 weeks, while primary nonresponse after induction was the secondary outcome. Patients were stratified into 3 groups according to VFI and SMI levels.

Primary nonresponse occurred in 15.7% of patients in the lowest SMI group, compared with 7.2% and 3.7% in the middle and highest groups, respectively (P=.021). Rates of loss of response followed a similar pattern, reaching 38.0% in the lowest SMI group versus 17.1% and 16.5% in the other groups (P<.001). 

Higher visceral fat also correlated with poorer outcomes. Loss of response occurred in 12.8%, 17.1%, and 41.7% of patients across increasing VFI groups (P<.001). Mucosal healing rates decreased as VFI increased, from 63.9% in the lowest group to 40.0% and 26.9% in the middle and highest groups, respectively (P<.001). 

Elevated VFI—defined as greater than 0.887 in men and 0.679 in women—and reduced SMI—less than 40.2 in men and 31.0 in women—were independent risk factors for 52-week loss of response. A prediction model incorporating body composition and clinical data achieved an externally validated AUC of 0.902 (95% CI, 0.828-0.975). 

Clinical Implications

The findings identify body composition as a potential marker of biologic treatment outcomes in Crohn’s disease. In particular, the associations between high visceral fat, low skeletal muscle, and treatment failure suggest that conventional measures may not fully characterize factors associated with response to biologic therapy.

CT-derived VFI and SMI could potentially contribute to risk stratification when imaging is already available. The strong externally validated discrimination of the combined predictive model also supports further evaluation of body composition alongside clinical characteristics. 

However, the study was retrospective and evaluated patients treated with infliximab or ustekinumab. The reported associations therefore should not be interpreted as evidence that modifying visceral fat or skeletal muscle will improve biologic response, nor do the findings establish the model as a prospective treatment-selection tool.

The authors concluded that elevated VAT and reduced SM were associated with loss of response to biologic therapy in Crohn’s disease and reported that their predictive model integrating body composition parameters showed good performance. 

Prospective validation will be important before CT-derived body composition measures or the predictive model can be incorporated into routine clinical decision-making.

Reference: Guo Q, Wang Q, Chen J, et al. Impact of visceral adipose tissue and skeletal muscle on early and long-term biologic treatment failure in Crohn’s disease: a multicenter retrospective cohort study. Inflamm Bowel Dis. 2026;32(7):1268-1278. doi:10.1093/ibd/izag023

© 2026 HMP Global. All Rights Reserved