Abstract
The results of a retrospective survey of possible etiological factors in a series of 93 patients with carcinoma of the stomach, 340 patients with carcinoma of the colon and rectum, and 1,020 controls are discussed. Gastric cancer patients showed a dietary pattern indicating an increased use of animal fats, cooked fats, fried foods, bacon, and a decreased use of dairy produce. The differences between patients and controls, however, were not statistically significant. No significant differences regarding marital status, religion, occupation, bowel habits, tobacco and beverage habits, and usage of individual foodstuffs were observed between cases of gastric cancer and controls. Apart from a slight negative correlation between cigarette smoking and cancer of the colon and rectum, and a more frequent history of constipation and use of laxatives, which were of doubtful significance, no obvious differences in dietary patterns and personal habits were observed between patients with intestinal cancer and controls. It is concluded that although these cancers are almost certainly environmental in origin, the retrospective method may be unsatisfactory for demonstrating etiological factors in a relatively homogeneous population, since many of the suspected carcinogenic stimuli are widely distributed and are difficult to quantitate.