Association between Alcohol Consumption and Diverticulosis and Diverticular Bleeding: A Systematic Review and Meta-analysis

Abstract

There have been conflicting reports on the association of alcohol use and diverticular disease. We aimed to determine the odds of developing diverticular disease and diverticular bleeding in patients who consumed alcohol on a regular basis compared with those who did not. MEDLINE and PUBMED were searched up until February 2017 on observational trials, which investigated the effect of alcohol use on two outcomes of diverticular disease: diverticulosis and diverticular bleeding. Quantitative estimates (odds ratios [OR] and confidence intervals [CI]) from included studies were pooled by using a random-effects model. Heterogeneity across studies was assessed by the I² statistic. In 6 studies including 53,644 subjects and 6 studies including 3,404 subjects, alcohol consumption on a regular basis was not associated with either diverticulosis (OR=1.99; 95% CI 0.99-4.03, I²=99%) or diverticular bleeding (OR=1.39; 95% CI 0.84-2.32, I²=45%) compared to subjects who did not consume alcohol on a regular basis, respectively. Increased odds of diverticulosis or diverticular bleeding among individuals who consume alcohol on a regular basis were not observed in these meta-analyses.

Keywords: Alcohol; Diverticular bleeding; Diverticulosis; Meta-analysis.

Figure 1

Forest plot of the included studies with adjusted analysis assessing odds of diverticulosis in individuals who consumed alcohol, square data markers represent ORs; horizontal lines, the 95% CIs with marker size reflecting the statistical weight of the study using random-effects meta-analysis. A diamond data marker represents the overall OR and 95% CI for the outcome of interest.

Figure 2

Forest plot of the included studies with adjusted analysis assessing odds of diverticular bleeding in individuals who consumed alcohol, square data markers represent ORs; horizontal lines, the 95% CIs with marker size reflecting the statistical weight of the study using random-effects meta-analysis. A diamond data marker represents the overall OR and 95% CI for the outcome of interest.