Objectives: Alcohol is a carcinogen that may increase the risk of Barrett's esophagus (BE) through direct contact with esophageal mucosa. However, few studies have investigated this association and findings have been inconsistent. We sought to examine the association between measures of total and beverage-specific alcohol consumption and BE risk.
Methods: We conducted a large population-based case-control study that collected information on lifetime alcohol consumption and other exposures from 285 patients with nondysplastic BE, 108 patients with dysplastic BE, and two separate control groups: 313 endoscopy patients with acute inflammatory changes ("inflammation controls") and 644 population controls. We calculated odds ratios (ORs) and 95% confidence intervals (CIs) for categories of average alcohol consumption using unconditional multivariate logistic regression.
Results: Relative to life-long nondrinkers and consumption of <1 drink/week, consumption of 7-20 drinks/week (OR=0.53, 95% CI: 0.31-0.91) and 21-41 drinks/week (OR=0.37, 95% CI: 0.19-0.73) of total alcohol throughout the life was inversely associated with nondysplastic BE, for comparisons with population controls. Lifetime total alcohol consumption was also inversely associated with dysplastic BE (7-20 drinks/week OR=0.52, 95% CI: 0.19-1.43; 21-41 drinks/week OR=0.22, 95% CI: 0.07-0.73). Similarly, reduced risk estimates were found for comparisons with inflammation controls. The inverse associations were observed separately for beer and wine consumption, with a significant linear trend observed with beer consumption. The risks associated with liquor consumption were up to twofold higher; however, they were not statistically significant. We found no evidence for effect modification by factors known (or suspected) to cause BE.
Conclusions: Overall, alcohol consumption does not increase the risk of BE. Significant inverse associations were observed for beer consumption, the underlying reasons for which remain unclear.