Why should the relationship between diet and cancer be investigated in prospective studies?

Prospective studies on diet and cancer are needed for five reasons: (i) unlike xenobiotic agents which, if found to be carcinogenic, can at least in principle be dispensed with, food of some kind is indispensable; (ii) our present knowledge on the role of diet in cancer is limited and permits only tentative recommendations; (iii) plausible biological hypotheses for this role are available, as well as biological markers allowing their exploration in epidemiological studies; (iv) these studies demand a prospective approach--a number of prospective studies, each involving several tens of thousands of adults, are now being started in Europe: and (v) diet affects not only cancers but also a spectrum of other diseases, which need to be investigated in parallel, to acquire knowledge on which to base firm recommendations on a diet capable of maximizing benefits and minimizing the risk to health.