Assessment of the role of diet in cancer prevention

Population studies have demonstrated a clear relationship between the incidence of various cancers and consumption of different classes of foods, indicating that diet can have a protective effect. Diets, however, contain many hundreds of nutrient and non-nutrient compounds. Moving beyond the generalities provided by population studies requires short-term studies in which diets can be adjusted to test the effects of specific ingredients or compounds. The cost of this approach is the loss of direct outcome measures. Biomarkers related to early events in the cancer process which can show significant change within the duration of a trial, have to substitute. The choice of marker and the ability to validate its use are crucial both to the refining of dietary advice and to the search for the active principles of diet.