Effect of infection on the risk of inflammatory bowel disease: an umbrella review and second-order meta-analysis

Background: Multiple meta-analyses have evaluated the effect of Helicobacter pylori ( H. pylori ) infection on the risk of inflammatory bowel disease (IBD), but their findings remain inconsistent. Objectives: To critically assess the existing meta-analyses regarding the effect of H. pylori infection on the risk of IBD. Data sources and methods: We systematically searched four electronic databases and manually screened the reference lists of relevant articles. A Measurement Tool to Assess systematic Reviews 2 (AMSTAR-2), Grade of Recommendation, Assessment, Development, and Evaluation (GRADE), and Graphical Representation of Overlap for OVErviews (GROOVE) tools were used to assess the methodological quality, evidence quality, and the overall overlap of primary studies among the included meta-analyses, respectively. A second-order meta-analysis was performed on primary studies derived from the included meta-analyses with very high overlap to explore the association of H. pylori infection with IBD risk and identify the potential influencing factors. Risk ratios (RRs) were pooled. Results: Twelve meta-analyses encompassing 55 primary studies were included. The methodological quality was very low to low. The evidence quality was very low to high. Moderate- to high-quality evidence suggested that H. pylori virulence factors and antibiotic exposure influenced the association of H. pylori infection with IBD. The overall overlap of primary studies among the included meta-analyses was very high. A second-order meta-analysis demonstrated a negative association of H. pylori infection with any type of IBD (RR: 0.60), Crohn’s disease (RR: 0.53), ulcerative colitis (RR: 0.67), and IBD-unclassified (RR: 0.73). Subgroup analyses demonstrated that both IBD subtypes ( P interaction < 0.01), regions ( P interaction < 0.01), and study designs ( P interaction = 0.05) significantly influenced this association. Sensitivity analyses indicated that the results were robust across IBD subtypes and regions, but not across the study designs. Discussion: H. pylori infection seems to exhibit a protective effect against the risk of IBD. This association may be influenced by IBD subtypes, regions, H. pylori virulence factors, and antibiotic exposure. Design: This is an umbrella review. Registration: The protocol was prospectively registered with the PROSPERO database (CRD42024559958).