Alcohol drinking and tobacco smoking in gastric cancer. A case-control study.

This study compares 210 cases of cancer of the stomach (138 males and 72 females) with 630 controls (414 males and 216 females) afflicted with a wide variety of diseases. All patients (cases and controls) were admitted for treatment at the University Hospital of Montevideo, Uruguay in the time period July 1985-December 1988. They were submitted to the same detailed questionnaire by 3 social workers unaware of the objectives of the study. The analysis was performed at the Louisiana State University of New Orleans, using multiple logistic regression. The variables analyzed were cigarette smoking, alcohol ingestion, salted meat intake, total vegetable consumption, total fruit consumption and "mate" ingestion. Strong positive associations were found in both sexes for low fruit and vegetable consumption, high intake of salted meat and "mate" ingestion. Only males showed significantly elevated OR's for tobacco smoking and alcohol drinking. Of the main tobacco variables, time-dependent one's (age at start, duration) displayed significant gradients of increasing risks. Thus, a prolonged exposure seems more important than the amount smoked per day for the risk of developing gastric carcinoma. Both wine and hard liquor carried increased OR's, but heavy drinkers of wine displayed a six-fold increase in risk, replicating previous reports from France.