Meta-analysis on the relationship between nonsteroidal anti-inflammatory drug use and gastric cancer.

Although nonsteroidal anti-inflammatory drugs (NSAIDs), including aspirin, show chemopreventive effects on several types of cancer, their role in the development of gastric cancer remains controversial. The objective of this study is to assess the quantitative association between NSAIDs and gastric cancer by meta-analysis using both crude data and adjusted data. A search of Medline, Embase, Web of science, and citations from relevant articles was conducted for eligible articles published up to March 2009. A meta-analysis was developed. Pooled risk ratio and odds ratio with 95% confidence intervals were calculated, based on both crude and adjusted data from the selected studies and using random or fixed effects models based on heterogeneity analysis. Overall, 21 individual epidemiological studies (13 case-control studies with 29 560 participants and 8 cohort studies with 2 199 227 participants) were examined. The summary crude and adjusted risk ratios were 0.89 (0.83-0.97) and 0.81 (0.73-0.89) when the case-control and cohort studies were pooled, respectively. Significant protective effects of NSAIDs against gastric cancer appeared in all subgroups (study design, type of drug, site of cancer, and sample source), but only the site of cancer and sample source can substantially influence the estimate of effect. The chemopreventive effect of NSAIDs in gastric cancer was stronger based on the adjusted data than that based on crude data. Therefore, we conclude that NSAID use is associated with a decrease in the development of gastric cancer. The associations were more obvious after we adjusted for several risk factors that are known to contribute to the development of gastric cancer.