The literature concerning international correlations between dietary items and cancer frequency is reviewed. An updated correlation of the most recent data on cancer mortality and food consumption is made. Strong and consistent correlations are reported between death rates of cancers of the colon and breast and the per capita consumption of total fat and of nutrients derived from animal sources, especially beef, pork, eggs, and milk. Similar but less consistent correlations have been reported with cancers of the prostate, ovary, and endometrium. In addition, correlations between precursor lesions of coronary heart disease and colon cancer in New Orleans autopsy populations are reported. Some studies suggest that milk intake correlates better with atherosclerotic disease while beef intake correlates better with colon cancer. Negative correlations of colon cancer rates and vegetable consumption are reported. Colon cancer rates also show negative correlations with stool weight, irrespective of the type of vegetable responsible for the increased bulk. Epidemiological data are consistent with the hypothesis that excessive beef and low vegetable consumption are causally related to colon cancer. These food items probably do not have a direct carcinogenic role but rather provide a microenvironment favorable to the actions of carcinogens.
Presented at the Workshop on Fat and Cancer, December 10 to 12, 1979, Bethesda, Md. Work supported by Contract N01-CP-53521 from the National Cancer Institute, USPHS.