Epidemiology and mechanisms relating diet to risk of colorectal cancer

Cross-sectional comparisons, case-control studies and trends in food intakes show high rates of colorectal cancer in populations consuming diets high in meat and fat, and low in starch, NSP and vegetables. These studies suggest that the potential for prevention of colorectal cancer by diet is very great. Attributable population risk estimates from case-control studies suggest that 25-35% of colorectal cancers might be prevented by high intake of vegetables and fibre and that 15-25% of colorectal cancers could be attributed to a high fat intake (Tomatis er al. 1990). There are plausible physiological reasons for a protective effect of starch and NSP in large bowel cancer. Cross-sectional comparisons suggest that the recommendation (Department of Health, 1991) to increase starch and NSP consumption in the UK by 50% from 12 to 18 g/d will increase stool weight by 25% and reduce large bowel cancer incidence on a population level by 15%. Existing results from cohort investigations show only weak associations and more data are required from large studies using accurate dietary assessment techniques in populations where there is extensive dietary variation, and which have collected biological samples in order that interaction between diet, biomarkers of diet, and different genotypes that may determine risk can be examined. A major problem in epidemiological studies of large bowel cancer is the absence of an easily accessible intermediate risk marker, known to alter in response to diet in metabolic studies, that can be used to link dietary intake and the presence of the disease in either intervention or prospective studies. To meet the Department of Health (1991) recommendations for NSP intake, vegetable and fruit consumption needs to double in the UK. Vegetables may have added benefits in reducing colorectal cancer rates because they contain antioxidant nutrients and flavonoids, sulphur-containing compounds, and folate, all of which can be shown to favourably affect factors thought to be important in reducing risk. Intervention studies are generally held to be the most robust way of testing hypotheses but those already conducted with β-carotene and vitamin E have not reduced risk of either large bowel cancer or recurrence of precursor lesions such as adenomatous polyps. Results from other interventions with calcium, NSP and resistant starch are awaited. A protective effect of starch and NSP probably arises from their marked effect on bacterial metabolism in the large bowel, which leads to an increase in stool weight, and in butyrate production, and reduced pH, and levels of secondary bile acids, diacylglycerol, and free ammonia. In rodents given known carcinogens, 'insoluble' sources of NSP are generally protective, and high fat (or energy) diets increase tumorigenesis. Tumorigenesis may be reduced by n-3 fatty acids, but the effects of fat in rodent models of colon cancer are less consistent than those found in mammary tumorigenesis. In humans, a role for fat in increasing risk via increased bile acid excretion has been proposed. A risk from red and processed meat seems to be emerging from prospective studies, and high levels of meat increase faecal ammonia and NOC concentrations, and intakes of HAA. Some NOC and HAA are known carcinogens and some of the chromosomal mutations found in colorectal cancer are consistent with effects of NOC and HAA. Meat consumption should not increase. There has also been some interest in the past in the effect of bacterial enzymes in modifying carcinogen metabolism but the relevance of this to human diets and carcinogenesis is uncertain. The effect of diet on fecapentaene excretion is unknown. There are numerous different compounds in vegetables actively under investigation for their effect on carcinogenesis at different sites, including the large bowel, but intervention studies already conducted with β-carotene and vitamin E suggest that the active ingredients in vegetables involved in colorectal cancer protection are not antioxidant nutrients. As these attempts to alter risk with supplements have so far not been successful, they are not recommended for the general population.