The Microbiome, Crohn’s Disease Risk and Childhood Exposure to Siblings with Crohn’s Disease

October 27, 2026

Issue 20

Clinical Question

Does exposure to a sibling with Crohn’s disease (CD) influence CD risk, and can gut microbiome alterations explain this association?

Editor’s Bottom Line

These data provide additional evidence that early-life environmental exposures, perhaps mediated through changes in the microbiome, drive the development of Crohn’s disease in susceptible individuals.

Reference

Chen R, Kim HJ, Bushra M, et al. Childhood exposure to a sibling with Crohn’s disease alters gut microbiome and Crohn’s disease susceptibility. Gut. Epub ahead of print Aug 25, 2026. https://doi.org/10.1136/gutjnl-2026-339257

Synopsis

This prospective multicentre cohort study enrolled 3,126 healthy siblings of patients with CD from the Crohn’s and Colitis Canada Genetic Environmental Microbial (CCC-GEM) Project, of whom 2,265 had childhood exposure (sibling was under 18 years of age when the CD sibling, or proband, was diagnosed) and 861 had adult exposure (18 years or older). Baseline stool, blood and urine samples were collected for gut microbiome profiling (16S rRNA sequencing), fecal calprotectin (FCP) and lactulose-to-mannitol ratio (LMR). Statistical analyses assessed the association between childhood exposure and CD onset, with external validation in the nationwide South Korean National Health Insurance Service (NHIS) database (10,085 siblings of CD patients). Additional analyses examined whether microbiome perturbations and gut barrier dysfunction mediated the exposure-to-CD-onset relationship. To test mechanistic causality, stool from childhood- and adult-exposed donors was transplanted into germ-free mice in a T-cell transfer colitis model. Finally, a two-step risk stratification model integrating FCP and gut microbial enterotypes was developed in a discovery cohort and internally validated.

Results showed that childhood exposure to a proband was linked with a fourfold higher risk of CD compared with adult exposure in the GEM cohort (adjusted Hazard Ratio [aHR]: 4.0; 95% Confidence Interval [CI], 1.83–8.75; p<0.001), a finding validated in the South Korean NHIS database (aHR: 2.54; 95% CI, 1.70–3.81; p<0.001). The association remained significant after adjusting for the CD polygenic risk score and neither duration of exposure nor the age gap between siblings modified the risk.

Other results showed childhood-exposed siblings had significantly higher rates of gut barrier dysfunction (17.4% vs 10.3%) while microbiome analysis identified 32 genera significantly associated with childhood exposure, predominantly from the Lachnospiraceae family, with Lachnospira, Roseburia and Colidextribacter showing the strongest and most consistent associations. All were reduced in childhood-exposed siblings and also independently reduced in patients with established CD. Analysis found these three genera to be partial mediators of the childhood exposure–CD onset relationship, with proportions mediated of 8.92%, 7.34% and 11.56%, respectively.

In the germ-free mouse model, transplanting stool from childhood-exposed donors produced significantly greater body mass loss, higher fecal lipocalin-2 and elevated colonic interleukin (IL)-22 mRNA expression, compared with adult-exposed donor microbiota.
A two-step risk stratification strategy categorized siblings into those with or without subclinical gut inflammation (FCP>100 μg/g) and gut microbial enterotype (Bacteroides-dominant vs Blautia/Prevotella-dominant). Siblings with both subclinical gut inflammation and a Blautia/Prevotella enterotype had a 10-year cumulative CD incidence of 22.5% in the discovery cohort (aHR vs no gut inflammation: 14.30; 95% CI, 5.89–34.75), confirmed at 19.8% in the internal validation cohort (HR: 9.25; 95% CI, 2.98–28.73).

Details

Study Design: Prospective multicentre cohort study with external validation
Funding: The Crohn’s and Colitis Canada, the Canadian Institutes of Health Research, the Helmsley Charitable Trust, the Biocodex Microbiota Foundation and the Weston Family Foundation
Allocation: Non-randomized
Setting: Multicentre
Level of Evidence: 2b