GLP-1 Analog Use is Associated With Improved Disease Course in Inflammatory Bowel Disease: A Report from the Epi-IIRN
Gorelik Y, Ghersin I, Lujan R, Shlon D, Loewenberg Weisband Y, Ben-Tov A, Matz E, Ledderman N, +4 more
Journal of Crohn's and Colitis (2025) · n=3,737
In 3,737 patients with IBD and type 2 diabetes drawn from the Israeli Epi-IIRN nationwide cohort, GLP-1-analog use was associated with reduced composite poor IBD outcomes (adjusted HR 0.74) — and the protective signal was carried specifically by the obese subgroup (aHR 0.61), with no detectable effect in non-obese patients.
This is the Israeli Epi-IIRN (Epidemiology Group of the Israeli IBD Research Nucleus) retrospective cohort analysis that constitutes the largest single nationwide-population study of GLP-1-analog use in inflammatory bowel disease as of 2026. The Epi-IIRN dataset captures essentially all IBD-diagnosed patients across the four Israeli health funds (covering more than 98% of the national population), and the analysis exploits the dataset's combination of pharmacy-claims linkage, hospitalization records, surgical procedures, and corticosteroid-prescription history to assess whether GLP-1-receptor-agonist exposure in IBD patients with type 2 diabetes was associated with differential IBD-outcome trajectories.
The cohort comprised 3,737 patients carrying both IBD and type 2 diabetes diagnoses, of whom 633 (16.9%) were treated with GLP-1 analogs during the observation period; 50.4% had ulcerative colitis and the balance had Crohn's disease. The composite primary outcome captured IBD-related hospitalization, IBD-related surgery, corticosteroid courses, and treatment-escalation events across 24,338 patient-years of follow-up. Adjusted Cox proportional-hazards models controlled for age, sex, socioeconomic status, smoking, BMI category, IBD subtype, disease duration, prior advanced-therapy exposure, and diabetes-treatment context.
GLP-1-analog treatment was associated with reduced composite poor outcomes in the full cohort, adjusted hazard ratio 0.74 (95% CI 0.62–0.89), with directionally consistent effects in each IBD subtype: ulcerative colitis aHR 0.71 (95% CI 0.52–0.96) and Crohn's disease aHR 0.78 (95% CI 0.62–0.99). The most consequential finding was the obesity-stratified interaction: the protective signal was concentrated in obese patients (aHR 0.61, 95% CI 0.50–0.77) and not detected in non-obese patients (aHR 0.94, no significant difference from null). The authors framed the result as evidence that GLP-1-analog use in IBD patients with comorbid type 2 diabetes is associated with improved IBD-disease-course endpoints, with the magnitude of the association consistent with what would be expected if the favorable effect were operating through obesity-related disease-modifier pathways rather than a direct GLP-1-receptor anti-inflammatory effect on intestinal mucosa.
Cross-link to the semaglutide peptide page and to the IBD-and-peptides dossier, where this paper sits alongside Levine et al. 2025 Inflamm Bowel Dis as the two load-bearing 2024–2025 cohort signals that have driven the IBD-and-GLP-1 conversation into the editorial-mainstream gastroenterology literature.
The fundamental limitation is retrospective observational design. The cohort is selected — every patient has both IBD and type 2 diabetes, and GLP-1-analog prescription is a clinician-and-patient decision conditioned on metabolic-disease severity, weight, prior diabetes-medication tolerance, and access. Confounding by indication is the principal threat to interpretation: patients prescribed GLP-1 analogs in 2018–2024 differed from non-GLP-1-treated patients in ways the propensity-adjustment framework cannot fully address. The obesity-stratified subgroup analysis is informative about the mechanism but also raises the alternative interpretation that the effect operates through weight-loss-mediated reduction in IBD-relevant inflammatory load rather than a GLP-1-class direct effect on intestinal immune biology. The Israeli population is genetically and demographically distinct from US and Western-European populations, with implications for generalizability. The cohort does not include IBD patients without type 2 diabetes — the population in which a hypothesis-generating prospective IBD-specific RCT would most plausibly enroll — and the analysis therefore does not address the relevant counterfactual of GLP-1-analog use in non-diabetic IBD patients. The composite primary endpoint combines events of varying clinical significance; component-level analyses are reported and directionally consistent but reduce statistical precision. No mucosal-healing, endoscopic, or biomarker (calprotectin, CRP) data are available at population-level granularity. The signal is hypothesis-generating; the prospective IBD-specific RCT that would establish causality has not yet read out. The 2025 Bayoumy et al. systematic review and meta-analysis pooled this and ten other observational studies covering 16,242 IBD patients and reported directionally consistent results. The Levine et al. 2025 independent US cohort is the closest available replication.
07·Member discussion
No member discussion yet.
Member-only conversation lives here — cycle notes, practitioner commentary, pattern-matching. Be the first paying member to start the thread.