Abstract Background Sibling pairs are demonstrated to share similar gut microbiome profiles, especially during childhood. We hypothesized that childhood exposure (versus adult exposure) to a sibling with Crohn’s disease (CD) represents a window of vulnerability during which microbial perturbations may shape long-term disease susceptibility to CD. Methods In the prospective GEM Project (n = 3,126 healthy siblings of CD patients), we assessed whether “childhood exposure” to CD (defined as having a sibling diagnosed before age 18) was associated with increased risk of future CD onset. This association was validated in an independent, nationwide South Korean cohort (n = 50 million). Associations between childhood exposure and gut microbiome composition (16S rRNA sequencing) were evaluated using MaAsLin3 in GEM cohort. We further leveraged a T-cell transfer model of colitis in germ-free mice to investigate the effects of stool from ‘childhood-exposed’ versus ‘adult-exposed’ siblings on colitis. Finally, an integrative risk model predicting CD risk, combining childhood exposure, fecal calprotectin (FCP), and microbial community features was trained (Canadian sub-cohort) and validated (sub-cohort from outside Canada) within GEM. Results In GEM, 2,265 siblings (72.5%) had childhood exposure to CD. Childhood exposure was associated with a four-fold higher risk of developing CD compared with adult exposure (adjusted HR aHR 95% CI: 4.00 1.83-8.75; P = 5.3 × 10-4), after adjusting for potential confounders (e.g, CD-polygenic risk score). This association was validated in the South Korean dataset (aHR 95%CI: 2.54 1.70-3.81; P = 6.0 × 10-6). In GEM, 33 bacterial genera were differentially abundant in childhood-exposed group (vs adult exposure). Mediation analysis revealed Lachnospira, Roseburia and Colidextribacter as partial mediators of the effect of childhood exposure on CD onset. In vivo, transplantation of stool from childhood-exposed siblings into a germ-free mouse model exacerbated colitis and increased mucosal expression of IL-22 and IL23p19 and stool lipocalin-2 compared to adult-exposed donors, suggesting a causal role of early-life microbial perturbation. The integrated model identified a high-risk subset characterized by childhood exposure, a Blautia- or Prevotella-dominant enterotype, and FCP100 ug/g, with 10-year cumulative CD incidence of 22.5% (training) and 19.8% (validation). Conclusion Childhood exposure to a sibling with CD is an independent risk factor associated with a four-fold increased risk of disease. Gut microbial community profiles can further refine risk stratification among these siblings and support a mechanistic role for early-life microbial perturbation in predisposing to CD. Conflict of interest: Chen, Rirong: No conflict of interest Kim, Hyun Jung: There is no conflict of interest. Dang, Cong Phi: No conflict of interest Bushra, Maham: No conflict of interest Li, Qilong: No conflict of interest Espin-Garcia, Osvaldo: No conflict of interest Dotan, Iris: No conflict of interest Steinhart, Hillary: No conflict of interest Huynh, Hien: No conflict of interest Jacobson, Kevan: No conflict of interest Griffiths, Anne: No conflict of interest Turpin, Williams: No conflict of interest Croitoru, Kenneth: No conflict of interest Lee, Sun-Ho: No conflict of interest
Chen et al. (Thu,) studied this question.