RESUMO
The present case-control study was designed mainly to estimate the risk of gastric cancer associated with consumption of alcoholic beverages. Seventy-four cases were selected from patients undergoing endoscopy with a histologically confirmed diagnosis of gastric cancer and matched by age and sex to a pool of controls (193) from three different sources, at the same hospital ward. After adjustment for several confounders, consumption of more than one bottle of red wine per day 20 years prior to the interview showed an odds ratio (OR) = 2.61 (P = 0.049), with a dose-response relationship increasing from 1.36 for those consuming less than one glass of wine per meal, up to 3.67 if the daily consumption exceeded one bottle of red wine. Other statistically significant determinants of the disease occurrence were low education level (OR = 4.33; P = 0.0008) and low social class (OR = 2.77; P = 0.016), as well as the regular practice of preserving food by smoke at home (OR = 1.91; P = 0.032).