At an individual level Helicobacter pylori was associated with the occurrence of gastric cancer but in some African and Asian countries its prevalence runs with low gastric cancer rates, the so-called African and Asian enigmas. We assessed whether the association between gastric cancer and H. pylori prevalence at an area level is modified by the level of exposure to fruits and vegetables, alcohol or tobacco. Regression models were fitted to data from 58 countries using as dependent variable log transformed gastric cancer rates and as independent covariables the H. pylori prevalence, fruits and vegetables consumption, cigarette smoking, alcohol intake and interaction terms. The levels of alcohol consumption or cigarette smoking modified the association between gastric cancer and H. pylori infection. Models including H. pylori prevalence, alcohol consumption, cigarette smoking and the interaction terms H. pylori x alcohol or H. pylori x tobacco were used to compute gastric cancer incidence multiplying regression coefficients by a H. pylori prevalence of 85% (the approximate median in African countries) and the median figures observed in each continent for alcohol and tobacco availability. The expected gastric cancer incidence per 100,000 would be 5.7 assuming the alcohol and tobacco availability in African countries, 7.0 in Asia and Oceania, 16.0 in America and 26.0 in Europe. The interaction between H. pylori and cigarette or alcohol consumption may contribute to further explain the international variation in gastric cancer and the so-called African and Asian enigmas.
Copyright 2003 Wiley-Liss, Inc.