CANCER AND EPIDEMIOLOGICAL STUDIES
The Nurses’ Health Study started in the USA in 1976 to investigate risk factors for cancer and other diseases in women and has now involved over 90,000 nurses. Thus far among its major findings are that the risk of bowel cancer is increased by eating a lot of red meat and reduced by taking folic acid in multivitamin supplements. Breast cancer incidence is unaffected by fat and fibre intake but is increased by one third in response to moderate amounts of alcohol, as we noted.
Bowel cancer has been intensively studied in the context of diet and it is worth considering the results of some of the major surveys to illustrate why this is such an intractable problem. Apart from the fact that it is one of the three major cancer killers, colorectal carcinoma is much studied because it is a well-defined, multi-step process in which, without treatment, adenomas develop into full carcinomas (Fig. 1).
In the early stages, polyps form on the wall of the bowel that surgically removed (polypectomy). Individuals so treated have been particularly studied to determine the effect of diet on adenoma recurrence.
Three familiar dietary players have been to the fore in this context: folate, calcium and fibre. Folate (there are several naturally occurring forms) is a B vitamin that helps to shuttle carbon atoms around when the building blocks of DNA and proteins are being made. It is therefore important in DNA replication and in the repair of damaged DNA, so it’s easy to see that if you were short of folate you might be prone to cancer. A synthetic form of folate – folic acid – is used as a food supplement. A number of both retrospective and prospective studies suggest that folate has a protective effect, consistent with its importance in maintaining the integrity of DNA. Furthermore, other studies have measured the levels of folate in plasma and serum and found an inverse association with bowel cancer – that is, the more folate you have the better – consistent with the Nurses’ Health Study.
As usual in this field, there are conflicting reports, some of which conclude that dietary supplementation can actually increase the risk. For folate you need enough to maintain your DNA in a healthy state, thereby minimising your susceptibility to cancer, but if you have too much it may block DNA repair and thus help to drive carcinogenesis. So, supplementing the diet with folic acid might provide protection against colon carcinoma for those whose normal circulating levels of folate are low but might be a very unhelpful thing to do for individuals with higher levels. All of which indicates the dangers of giving diet supplementation to general populations.
As with folate, there is a general view that plenty of calcium is good, not only because it gives you strong bones but also because you are less likely to get bowel cancer. Again there are prospective studies that support this view but, as ever, there are others, including one of 36,000 cases, showing that calcium supplementation has no effect. These findings scarcely make the case for calcium supplementation and, as with folate, may be confounded by individual variation – that is, additional calcium intake will not benefit those on healthy diets who have normal levels of calcium.
Dietary fibre, sometimes called roughage, is the stuff we eat that can’t be digested but that does an important job in taking up water and generally helping our insides to work. The well-publicised advice is that eating plenty of fibre helps to prevent colon cancer and there are many supporting studies. Notable among these is the European Prospective Investigation into Cancer and Nutrition (EPIC) study showing that 35 grams per day reduced the risk by 40% compared with 15 grams per day. This is a particularly powerful contribution because it involved over half a million (520,000) people from ten European countries. However, there are, of course, other studies (e.g. the Polyp Prevention Trial and the Wheat Bran Fiber Trial: National Institutes of Health, 2000) that show no protective effect and at least one that concludes men are much better protected than women. In addition to the pitfalls mentioned earlier, a further cause of confusion may be the variation in study duration and the fact that follow-up periods are generally short relative to the many years over which cancers usually develop.
It is probably obvious that, when faced with a considerable number of independent studies about a specific factor with conclusions spanning all categories, it might be worth putting them together and seeing if a more focused message emerges. This is what epidemiologists call a ‘meta-analysis’ and it’s really a way of increasing your sample size and hence the statistical power of the data. It’s not absolutely straightforward because there’s no point in including a study you consider so badly designed as to be worthless. Thus, an element of judgement was required to combine data from 13 separate studies to show that there was no effect of a high fibre diet on the risk of developing colon cancer, a conclusion that was, of course, completely at odds with the EPIC finding.