Alcohol drinking and gastric cancer risk: a meta-analysis of observational studies

Abstract

Background: Many studies investigated the association between alcohol drinking and gastric cancer risk, but the results were controversial. We performed a meta-analysis of observational studies to explore the association.

Materials and methods: We searched PubMed to identify the relevant studies that reported the association between alcohol drinking and gastric cancer risk up to December 31, 2016. We pooled relative risks (RRs) in random effects model and performed dose-response analysis to quantify the association. Cochran Q test and I² analyses were used to evaluate the heterogeneity. Meta-regression, subgroup, sensitivity and publication bias analyses were also performed.

Results: 75 studies were included in our study. The pooled RR of high vs low total alcohol drinking was 1.25 (95% CI, 1.15-1.37, P < 0.001), and a nonlinear association was further observed. Subgroup analysis showed that alcohol drinking significantly associated with the risk of gastric noncardia cancer (RR, 1.19; 95% CI, 1.01-1.40, P = 0.033), but not with the risk of gastric cardia cancer (RR, 1.16; 95% CI, 0.98-1.39, P = 0.087). Notably, the pooled RRs of high vs low analyses were 1.13 (95% CI, 1.03-1.24, P = 0.012) for beer drinking, 1.22 (95% CI, 1.06-1.40, P = 0.005) for liquor drinking, and 0.99 (95% CI, 0.84-1.16, P = 0.857) for wine drinking.

Conclusions: Our meta-analysis found a nonlinear association between alcohol drinking and gastric cancer risk, and heavy drinking level was strongly related to gastric cancer risk. Beer and liquor had significant positive associations with gastric cancer risk, while wine drinking would not increase gastric cancer risk. These results need to be verified in future research.

Keywords: cancer prevention; dietary habit; epidemiology; gastric cancer.

Figure 1. Flow chart of the systematic search of literature on alcohol consumption and the risk of gastric cancer

Figure 2. Relatives risk of gastric cancer for the high vs low category of alcohol consumption

Studies are grouped according to study design. The pooled RRs were calculated using the random-effects models. Open diamond denote the pooled RR. The size of gray box is positively proportional to the weight assigned to each study (inverse of variance), and horizontal lines represent the 95% confidence intervals. RR, relative risk; CI, confidence interval.

Figure 3. Nonlinear dose-response analysis of the association between total alcohol drinking and gastric cancer risk in studies assessed by restricted cubic spline model with three knots

The solid line represented the estimated relative risk and the dashed lines represented the 95% confidence intervals.

Figure 4

Relatives risk of gastric cancer for the high vs low categories of (A) beer, (B) liquor, and (C) wine drinking. The pooled RRs were calculated using the random-effects models. Open diamond denote the pooled RR. The size of gray box is positively proportional to the weight assigned to each study (inverse of variance), and horizontal lines represent the 95% confidence intervals. RR, relative risk; CI, confidence interval.

Figure 5

Dose-response analysis of the association between (A) beer, (B) liquor, and (C) wine drinking and gastric cancer risk in studies assessed by restricted cubic spline model with three knots. The solid line represented the estimated relative risk and the dashed lines represented the 95% confidence intervals.

Figure 6. Subgroup analysis of the association between total alcohol drinking and gastric cancer risk based on cancer sites

The pooled RRs were calculated using the random-effects models. Open diamond denote the pooled RR. The size of gray box is positively proportional to the weight assigned to each study (inverse of variance), and horizontal lines represent the 95% confidence intervals. RR, relative risk; CI, confidence interval.