Alimentary factors in the epidemiology of gastric cancer

One hundred and sixty male patients and 68 female patients with gastric cancer at Roswell Park Memorial Institute were matched with control patients with no neoplastic or gastrointestinal disease on age, country of birth, and nationality background of parents and grandparents. The matching was accomplished to control socioeconomic factors previously found related to gastric cancer. Item‐by‐item analysis of diet for both sexes showed that patients more frequently ingested potatoes, ate lettuce less often, ate more irregularly, and used cathartics frequently. Because diets typically do not consist of single items of food, using scaling techniques, we attempted to distinguish differences between cases and controls in frequency of use of groups of foods previously hypothesized to be related to the disease. We found no relationships on: 1. duration of use of the several fats used in frying; 2. frequency of ingestion of the various foods consumed fried, considered separately or together, weighting for the number used and the frequency of each; 3. frequency of eating all meats and fish considered separately or together, and 4. frequency of use of the various alcoholic beverages. We found a rather substantial difference in eating vegetables raw. Control patients ate larger numbers of vegetables raw than cancer patients. Low risk of gastric cancer was associated with ingesting in the uncooked state lettuce, tomatoes, carrots, coleslaw, and red cabbage, and risk declined with increases in the number of these vegetables eaten raw.