The European Cancer Prevention Organization (ECP), like many organizations, offers dietary advice on cancer prevention to the general population (summarized by Hill, 2001). The advice is predicated on the belief that eating good food is one of the great joys of life, and should remain so. In summary the advice is:
- 1. Enjoy your food! There is no individual food that is so healthy that it is worth eating even if you do not enjoy it. There are always alternative sources of the vital healthy components.
- 2. Maintain a healthy body profile with a body mass index (BMI) between 20 and 25, by balancing energy intake against energy expenditure in the form of physical activity. That gives you an ‘energy budget’, and within that one should:
- 3. Eat plenty of the fruits and vegetables that you enjoy (5 to 8 portions per day);
- 4. Eat plenty of whole grain cereal (to give 10–15 g cereal fibre per day). And finally, just in case you have forgotten already, always remember!
- 5. Enjoy your food!
Unfortunately this dietary advice, which should benefit health in general as well as cancer in particular and which is solidly supported by epidemiological data, is still not based on the more solid evidence which comes from well-designed intervention trials. In consequence, this is a rapidly evolving field and so we will need to re-examine the advice regularly as new data emerge. Sometimes it is going to be necessary to update it.
Last April there was an issue of New England Journal of Medicine which contained two important papers describing intervention trials. Schatzkin et al. (2000) showed that an intervention diet rich in fruit and vegetables failed to decrease the risk of new colorectal adenoma formation. Alberts et al. (2000) showed that dietary supplements of wheat bran also failed to decrease new adenoma formation. In a recent paper, Bonithon-Kopp et al. (2000) have shown again that fibre supplements failed to decrease the risk of new adenoma formation. These papers received a great deal of publicity in the media, who delighted in the contradiction of well-known dietary advice promoting fibre consumption. They concluded that the advice to eat fibre should be ignored. Therefore, the question that we health scientists need to address is whether we should follow the lead of those hyper-intelligent beings who write for the media? Should we, in fact, abandon the inclusion of plant foods in our advice on cancer prevention? Our answer must be no, and for a number of very good reasons.
The first reason is that our advice in support of the five-a-day campaign on fruit and vegetables and in support of whole grain cereals was based on the totality of the scientific evidence. This was so strong that it is not weakened by just one paper on fruit and vegetables and two on fibre, however good these papers are individually. We cannot keep changing our advice every time a new report is published and hyped by the media. Maybe that would be good for newspaper sales but it would be poor science.
The second reason is that although colorectal cancer is the second commonest cancer in Europe, there is much more to cancer prevention than simply preventing colorectal cancer. The evidence from epidemiology overwhelmingly indicates that fruit and vegetables prevent cancer at a wide range of sites in the body (Block et al., 1992;Negri et al., 1994;Hill, 1995La Vecchia and Tavani, 1998;Table 1
). The evidence is similarly strong that cereal fibre (Hill, 1997;La Vecchia and Chatenoud, 1998;Table 3
) prevents cancer at a wide range of sites in the body, not just the large bowel. Even if it had been proved in these three papers that these plant foods have no protective effect against colorectal cancer, the advice from public health bodies would still be valid for cancer in general.
However, the third and most important reason is that these three papers do not address the question of colorectal cancer prevention, they are only concerned with the precursor lesion – the adenoma. Colorectal cancers arise through a multi-stage process called the adenoma-carcinoma sequence (Hill et al., 1978) or more recently the dysplasia-carcinoma sequence (Morson et al., 1983). The first step in this sequence (Fig. 1
) is adenoma formation, followed by adenoma growth, and increasing severity of epithelial dysplasia. The dysplastic lesion finally crosses the muscularis mucosae (Fig. 2
) to be capable of metastasis and is then a cancer (Morson et al., 1983).
Adenoma formation is an extremely common event in Western countries. The overall prevalence in males more than 50% by the age of 70 years. The vast majority of these adenomas will be small and asymptomatic; they will remain so, and will only be detected, if at all, at post mortem. Not only are they very common, but removal is only a short-term treatment; after polypectomy, 30% of patients will develop new adenomas within 3 years (Table 5
). Because they are silent lesions, we know very little about their aetiology. However, we know that the epidemiology of colorectal adenomas differs from that of colorectal carcinomas in a number of crucial ways summarized by Hill (1991). In brief (Table 6
), the subsite distribution is different (Table 7
); the sex-ratio male/female cases is 1.5–2.0 for adenomas but <1 for colon cancers; there are populations with very different adenoma prevalence but the same colon cancer risk, and populations with very similar adenoma prevalence but very different colon cancer risk (Table 8
); tobacco and alcohol are strong risk factors for colon adenomas but not for carcinomas; in a recent study vitamin E intake was a risk factor for colon adenomas but protective against carcinomas (Rennert, 2001).
In retrospect, therefore, it is not surprising that the factors controlling adenoma formation differ from those controlling progression. Adenomas originate at the base of the crypts and well away from the colonic lumen (Fig. 2). Large adenomas are much more prevalent in the distal colon, as are severely dysplastic adenomas. In fact, their distribution in the colon is very different from that of small adenomas and is more similar to that of carcinomas. The even subsite distribution and the site of origin suggest that the causal agents for adenoma formation are not metabolites from the lumen but are delivered via the vascular system. In contrast, the concentration of large/dysplastic adenomas and carcinomas in the distal colon suggests that the factors causing adenoma growth and progression to cancer are not delivered by the vascular system. It is more consistent with the factors being formed in the proximal colon and concentrated during transit through the colon (Hill et al., 1978;Hill, 1991).
The concept that adenoma growth is supported by luminal factors is supported by the observation that adenomas regress in size when the faecal stream is diverted (Cole and Holden, 1959). A macroscopic adenoma protrudes into the lumen and is bathed in luminal contents. It is known that luminal butyrate is a major energy source for the normal colonocyte, so it would not be surprising if other luminal factors also affect colon tumour growth.
Thus there is no reason at all to consider that the (vascular) factors responsible for adenoma formation will be in any way related to those (luminal) factors causing progression from adenoma to cancer. The conclusion that they are not related was arrived at independently, on the basis of epidemiological data by Yoon et al. (2000). Because small adenomas are usually asymptomatic, we know very little about their causation. In contrast, we have a wealth of information on possible risk factors for the overall process of carcinoma formation, and it is this that tells us of the protective effect of vegetables and cereal fibres.
So which is more important to overall cancer risk? Is it adenoma formation or adenoma growth? It is well established that in a multi-stage process, the overall rate is determined by the rate of the slowest step in that process. This will be true of colorectal carcinogenesis. The diet factors associated with colorectal cancer will necessarily be associated with adenoma formation only if adenoma formation is the slowest step. The slowest step is easily recognized; the substrate accumulates (because it is formed faster than it is used up) and the product is at very low concentration (because it is rapidly taken on to the next stage). In Western countries, adenoma formation is very common indeed. It is clear that only a fraction of small adenomas progress to malignancy. The formation of adenomas cannot be the rate-limiting step and so there is no reason, in retrospect, to suspect that cereal fibre or vegetables would prevent adenoma formation.
It would be very useful to be able to prevent adenoma formation. However the protection would need to be overwhelming, because of their very high prevalence, and because of their high rate of recurrence (Table 1), but how to do that? Although we suspect that diet is important in their aetiology, we have little reliable information on which aspects of the diet are important. Perhaps for this reason, the diet components that affect the overall process of carcinogenesis have been tested in adenoma formation. It would have been exciting and valuable if they had worked. Apart from calcium they have all proved negative.
These three papers were well-designed intervention studies of colorectal adenoma formation. They did not claim to, and were never designed to, address the problem of prevention of cancer in general or colorectal cancer in particular. The fact that there was no protection in these studies, therefore, does not affect the case that they protect against cancer in general and colorectal cancer in particular by one iota.
1. Alberts DS , Martinez ME , Roe DJ et al. ( 2000 ). and the Phoenix Colon Cancer Prevention Physicians Network. Lack of effect of a high-fiber cereal supplement on the recurrence of colorectal adenomas . N Engl J Med 342 : 1156 – 1162 .
2. Block G , Patterson B , Subar A ( 1992 ). Fruit, vegetables and cancer prevention. A review of the epidemiological evidence . Nutr Cancer 18 : 1 – 29 .
3. Bonithon-Kopp C , Kronborg O , Giacosa A , Rath U , Faivre J ( 2000 ). for the European Cancer Prevention Organisation Study Group. Calcium and fibre supplementation in prevention of colorectal adenoma recurrence: a randomised intervention trial . Lancet 356 : 1300 – 1306 .
4. Cole JW , Holden WD ( 1959 ). Post-colectomy regression of adenomatous polyps in the human colon . Arch Surg 79 : 385 – 392 .
5. Hill MJ ( 1986 ). Microbes and human carcinogenesis Edward Arnold , London .
6. Hill MJ ( 1989 ). Aetiology of colorectal cancer: current concepts . Bailliere's Clinical Gastroenterology 3 : 567 – 592 .
7. Hill MJ ( 1991 ). Bile acids and colorectal cancer . Eur J Cancer Prev 1 ( suppl 2 ): 69 – 75 .
8. Hill MJ ( 1995 ). Diet and cancer: A review of the scientific evidence . Eur J Cancer Prev 4 ( suppl 2 ): 3 – 42 .
9. Hill MJ ( 1997 ). Cereals, cereal fibre and colorectal cancer risk; a review of the epidemiological literature . Eur J Cancer Prev 6 : 219 – 225 .
10. Hill MJ (2001). ECP Dietary Advice for cancer prevention. Eur J Cancer Prev
(accepted for publication).
11. Hill MJ , Morson BC , Bussey HJR ( 1978 ). Etiology of the adenoma-carcinoma sequence . Lancet i : 245 – 247 .
12. Kronborg O , Hage E , Ademsen S , Deichgraeber E ( 1983 ). Follow-up after colorectal polypectomy . Scand J Gastroenterol 18 : 1089 – 1093 .
13. La Vecchia C , Chatenoud L ( 1998 ). Fibres, whole grain foods and breast and other cancers . Eur J Cancer Prev 7 ( suppl 2 ): S25 – S28 .
14. La Vecchia C , Tavani A ( 1998 ). Fruit and vegetables and human cancer . Eur J Cancer Prev 7 : 3 – 8 .
15. Morson BC , Bussey HJR , Day DW , Hill MJ ( 1983 ). Adenomas of the large bowel . Cancer Surveys 2 : 451 – 478 .
16. Negri E , D'Avanco B , Tavani A ( 1994 ). The role of vegetables and fruit in cancer risk . In Epidemiology of Diet and Cancer (eds Hill MJ , Giacosa A , Caygill C ); Ellis Horwood , Chichester , pp. 327 – 334 .
17. Neugat AI , Jacobson JS , Ahsan H et al. ( 1995 ). Incidence and recurrence rates of colorectal adenomas: A prospective study . Gastroenterology 108 : 402 – 408 .
18. Rennert G (2001). Dietary intervention studies in cancer prevention. Eur J Cancer Prev
(accepted for publication).
19. Schatzkin A , Lanza E , Corle D et al. ( 2000 ). The Polyp Prevention Trial Study Group. Lack of effect of a low-fat, high-fiber diet on the recurrence of colorectal adenomas . N Engl J Med 342 : 1149 – 1155 .
20. Wegener M , Borsch G , Schmidt G ( 1986 ). Colorectal adenomas: Distribution, incidence of malignant transformation, and rate of recurrence . Dis Colon Rectum 29 : 383 – 387 .
21. Williams CB , Macrae FA ( 1986 ). The St Mark's neoplastic polyp follow-up study . Front Gastro Res 10 : 226 – 242 .
22. Yoon H , Benamouzig R , Little J , Francois-Collange M , Tome D ( 2000 ). Systematic review of epidemiological studies on meat, dairy products and egg consumption and risk of colorectal adenomas . Eur J Cancer Prev 9 : 151 – 164 .