Approaches to understanding the relationship between diet and colon cancer

Epidemiological research into the etiology of colon cancer has failed to provide consistent evidence on the role of diet. This may have arisen, in part, from two methodological limitations of the studies: the difficulty of obtaining accurate measurements of diet and the lack of variability in diet among the subjects studied. In case-control studies, diet measurement might be improved if current rather than past diet could be studied and this might be appropriate if studies were based on patients with precursor lesions rather than cancer. Precursor lesions might also serve as endpoints for randomized trials of dietary interventions. In both types of studies, laboratory measurements of specific dietary variables should be used to supplement diet questionnaires. Further, controls should not be selected from patients with diet-related diseases and studies should be conducted in communities with sufficient dietary variability to ensure the presence of an appropriate gradient in risk.