Cigarette Smoking and the Risk of Barrett's Esophagus

A Systematic Review and Meta-analysis

Juliana Andrici; Michael R Cox; Guy D Eslick


J Gastroenterol Hepatol. 2013;28(8):1258-1273. 

In This Article

Abstract and Introduction


Background and Aim Barrett's esophagus (BE) is a premalignant condition to esophageal adenocarcinoma. It is currently not clear whether cigarette smoking increases the risk of developing BE, and no meta-analysis has been performed on the topic. We conducted a systematic review and meta-analysis, providing a quantitative estimate of the increased risk of BE associated with cigarette smoking, to help clarify whether a relationship exists between smoking and BE.

Methods Four electronic databases (Medline, PubMed, Embase, and Current Contents Connect) were searched to May 17, 2013, for observational studies of BE patients. We calculated pooled odds ratios (ORs) and 95% confidence intervals (CIs) using a random effects model for the association of smoking with BE. BE patients were compared with non-gastroesophageal reflux disease (GERD) controls as well as with population-based and GERD controls.

Results Thirty-nine studies comprising 7069 BE patients were included in the meta-analysis. Having ever-smoked was associated with an increased risk of BE compared with non-GERD controls (OR 1.44; 95% CI 1.20–1.74), population-based controls (OR 1.42; 95% CI 1.15–1.76), but not GERD controls (OR 1.18; 95% CI 0.75–1.86). The meta-analyses of the studies reporting the lowest and highest number of pack-years smoked showed an increased risk of BE (OR 1.41; 95% CI 1.22–1.63) and (OR 1.53; 95% CI 1.27–1.84), respectively.

Conclusion Cigarette smoking was associated with an increased risk of BE. Being an ever-smoker was associated with an increased risk of BE in all control groups. A greater number of pack-years smoked was associated with a greater risk of BE.


Barrett's esophagus (BE) involves the replacement of the normal squamous esophageal lining by specialized or intestinal columnar epithelium.[1,2] The main clinical significance of BE is its association with an increased risk of developing esophageal adenocarcinoma,[3] which, although historically an uncommon disease, has been experiencing a dramatic increase in incidence in the United States and other Western countries over recent decades.[4–7] The prevalence of BE in the general population is uncertain mainly because BE subjects are often asymptomatic and therefore do not present for diagnostic endoscopy. However, a study from Sweden found a prevalence of 1.6% in a random sample of 3000 individuals from the general population.[8] Among patients with gastroesophageal reflux disease (GERD), which is a common complaint, the prevalence of BE has been reported to be between 3 and 15%.[9,10]

Risk factors for BE include white race, male sex, older age, obesity,[11] and GERD.[12] While cigarette smoking is a well-recognized risk factor in the development of esophageal squamous cell carcinoma,[13] and has been associated with esophageal adenocarcinoma in some studies,[14] it has not been definitely linked with an increased risk of developing BE. The literature results are currently mixed, with some studies showing a positive association,[8,11,15–17] while others report no association.[18–22] A recent analysis of five case–control studies from the International Barrett's and Esophageal Adenocarcinoma Consortium (BEACON) consortium[23] found a positive relationship between cigarette smoking and BE, thus providing a strong indicator that an association is present.

With a rapidly increasing incidence of adenocarcinoma, which carries a poor prognosis,[24] the importance of identifying modifiable risk factors for its precursor lesion, BE, is obvious in terms of patient education of preventative measures. To date, no meta-analysis of the relationship between cigarette smoking and BE has been performed. To confirm the relationship between smoking and BE found in the recent BEACON consortium analysis,[23] we conducted a meta-analysis combining the results of studies reporting the prevalence of cigarette smoking in BE subjects, and thus provided a quantitative estimate of the increased risk of BE associated with smoking.