Summary
The design of any experiment, especially with animals, to relate nutritional factors to cancer incidence must take into account several critical theoretical and practical considerations. At the theoretical level, susceptibility to cancer is a combination of genetic, physiological and metabolic factors, and any one or all of these may be influenced by nutritional variables. Similarly the development of cancer in any one individual is a progressive event and it must be recognized that diet may influence any one or more of these stages. At the practical level, problems can arise because any experiment with animals is also a nutritional experiment, and because different answers to the same nutritional question can be obtained simply by choosing different experimental tumor models.
The neglect of these factors may mean that much of current nutrition/cancer research can progress no further than a “feed them and count the dead”; stage, and can only give answers that are at best equivocal. It is also possible that significant progress in elucidating the functional reja‐tionship between cancer and any nutritional variable will depend on the consideration of the biology of the intermediate stages between diet and disease.