ABSTRACT
PubMed,Embase,Cochrane Library,and ISI web were searched without any limitations with regard to publication date or language,as well as the references of qualifying articles.All kinds of cohort studies comprising the risk of gastric cancer between diabetic patients and control subjects were included.We excluded studies that investigated only mortality but not incidence.25 studies met our criteria,and the qualities of these studies were evaluated using the Newcastle-Ottawa Quality Assessment Scale.Statistical analyses were carried out with STATA software version 12.0.A random-effects model meta-analysis showed an increased risk of gastric cancer in diabetic patients (RR =1.20,95 ％ CI 1.12-1.28).Subgroup analyses indicated that this result persisted in studies of both male(RR =1.15,95％ CI 1.02-1.29) and female (RR =1.29,95％ CI 1.09-1.53) subjects,in studies of European countries(RR =1.25,95％ CI 1.07-1.46) and Asia countries (RR =1.18,95％ CI 1.09-1.28).Compared with nondiabetics,the incidence of gastric cancer may be increased by approximately 20％ in diabetics.Thus diabetes may be an independent risk factor for gastric cancer.This effect tends to be more significant in type 1 and female patients.