Smoking in irritable bowel syndrome: a systematic review

Objective: The aim of this review was to examine (1) the prevalence of smoking in subjects with irritable bowel syndrome (IBS), (2) whether smoking prevalence significantly differs between subjects with and without IBS, and (3) whether smoking significantly predicts the presence or the development of IBS. Methods: Articles were retrieved by systematically searching the Scopus, Web of Science, and PubMed electronic databases from inception to July 2016, using the keywords “smoking” and “tobacco” combined with “irritable bowel syndrome.” Reference lists of included articles were also searched. Articles were included if they (1) reported data on smoking prevalence in subjects with IBS and/or on the association (assessed by means of multivariate analyses) between smoking and IBS, (2) identified IBS according to Manning criteria or Rome I–III criteria, (3) were English-language articles, and (4) involved only adult subjects. Results: The electronic searches yielded a total of 1,637 records, and 42 articles met inclusion criteria. Another 13 articles were retrieved through manual search, leading to a total of 55 included articles. Smoking prevalence in subjects with IBS was assessed by 48 articles and ranged from 0% in university students to 47.1% in patients with microscopic colitis. Thirty-three articles compared smoking prevalence between subjects with and without IBS. In 25 articles no significant difference was found. In seven articles smoking was significantly more frequent in subjects with IBS compared to those without IBS, while one study found a significantly higher smoking prevalence in controls. Eighteen multivariate analyses assessing the association between smoking and IBS were presented in 16 articles. Only one study employed a prospective design. In 11 analyses, smoking was not significantly associated with IBS after adjusting for covariates. In seven studies smoking independently predicted the presence of IBS. Conclusions: According to the selected articles, a significant association between smoking and IBS cannot be confirmed. However, different shortcomings may hinder generalizability and comparability of many studies. A dimensional assessment of smoking, a prospective design, the differentiation between IBS subgroups, and the recruitment of patients in clinical settings, especially in primary care, are necessary to clarify the role of smoking in IBS.