Nutritional aspects of human carcinogenesis.

Evidence that nutritional factors are involved in human carcinogenesis has come from studies of migrants, correlational studies of various dietary factors with incidence and mortality, internationally and nationally, associations with host factors which may be nutritionally mediated, and case-control studies in various populations. For breast cancer, migrant studies suggest that dietary factors may operate in early life; correlational studies show strong associations with total fat intake; associated host factors include age at menarche and height and weight; and two case-control studies have suggested an association with total fat intake. For colorectal cancer, migrant studies suggest that dietary factors can operate later in life; correlational studies have shown associations with total fat, total protein, beef and certain fibre fractions; associated host factors include serum cholesterol and intestinal excretion of bile acids and salts; and case-control studies have shown the importance of western type meals and beef intake in Hawaiian Japanese (although this was not replicated in Japan), a protective effect of fibre in Israel, a protective effect of cruciferous vegetables in the United States and an association with saturated fat intake but not with crude fibre intake in Canada.