J Crohns Colitis. 2021 Oct 7;15(10):1658-1678. doi: 10.1093/ecco-jcc/jjab047.

Daniele Piovani ^1 2, Claudia Pansieri ^1 2, Soumya R R Kotha ¹, Amanda C Piazza ¹, Celia-Louise Comberg ¹, Laurent Peyrin-Biroulet ³, Silvio Danese ^1 2, Stefanos Bonovas ^1 2

Author information

¹Department of Biomedical Sciences, Humanitas University, Milan, Italy.

²IRCCS Humanitas Research Hospital, Milan, Italy.

³Department of Gastroenterology and Inserm NGERE U1256, University Hospital of Nancy, University of Lorraine, Vandoeuvre-lès-Nancy, France.

Abstract

Background and aims: The association between smoking and inflammatory bowel disease [IBD] relies on old meta-analyses including exclusively non-Jewish White populations. Uncertainty persists regarding the role of smoking in other ethnicities.

Methods: We systematically searched Medline/PubMed, Embase, and Scopus for studies examining tobacco smoking and the risk of developing IBD, ie, Crohn's disease [CD] or ulcerative colitis [UC]. Two authors independently extracted study data and assessed each study's risk of bias. We examined heterogeneity and small-study effect, and calculated summary estimates using random-effects models. Stratified analyses and meta-regression were employed to study the association between study-level characteristics and effect estimates. The strength of epidemiological evidence was assessed through prespecified criteria.

Results: We synthesised 57 studies examining the smoking-related risk of developing CD and UC. Non-Jewish White smokers were at increased risk of CD (29 studies; relative risk [RR]: 1.95, 95% confidence interval [CI]: 1.69?2.24; moderate evidence). No association was observed in Asian, Jewish. and Latin-American populations [11 studies; RR: 0.97; 95% CI: 0.83-1.13], with no evidence of heterogeneity across these ethnicities. Smokers were at reduced risk of UC [51 studies; RR: 0.55, 95% CI: 0.48-0.64; weak evidence] irrespectively of ethnicity; however, cohort studies, large studies, and those recently published showed attenuated associations.

Conclusions: This meta-analysis did not identify any increased risk of CD in smokers in ethnicities other than non-Jewish Whites, and confirmed the protective effect of smoking on UC occurrence. Future research should characterise the genetic background of CD patients across different ethnicities to improve our understanding of the role of smoking in CD pathogenesis.