Epidemiologic Studies of Diet and Cancer 1

Abstract

The hypothesis I want to discuss have to do with fats, fiber, and vitamin A. First, a word about the various methods of study, particularly in epidemiology. Experiment is the scientific instrument par excellence and as utilized in intervention trials in epidemiology has been called the most powerful epidemiologic tool. An example in dietary epidemiology is the current trial going on among American doctors whereby they are taking carotene or a placebo on a random basis.

The criteria for intervention trials have to be observed very carefully: there has to be evidence in favor of a salutary effect of the substance being tried. There has to be evidence that there is no long-term toxicity, and there has to be the likelihood that the human participants will comply with the stated regimen for years, sometimes, and that the individuals on the placebo will stay on the placebo. For various reasons, past large-scale human intervention trials, although costing many millions of dollars and absorbing huge amounts of time of investigators, have yielded convincing results in only a few instances.

The least powerful tool of epidemiology is the international correlational study. I don't have to dwell on the drawbacks of these. A typical study relates the mortality from colon cancer to the ingestion of fats in populations in different countries (1). Even though there is a correlation, this does not prove that fats cause colon cancer. The proportion of people involved in manufacturing can be related to colon cancer in a similar fashion. There is growing evidence that exercise reduces risk of colon cancer. We might hypothesize that those nations requiring a great deal of physical labor; that is, those nations in which large proportions are not involved in manufacturing subject their populations to much more exercise, and this could account for their low mortality from colon cancer.