Harvard Reports on Cancer Prevention
Volume I: Human Causes of Cancer
Cancer Causes & Control:
An International Journal of Studies of Cancer in Human Populations
Official Journal of the International Association of Cancer Registries
Volume 7 Supplement November 1996 ISSN 0957-5243


Dietary Factors

Introduction
Dietary differences have long been suspected as a cause of the large international variations in cancer rates. However, until the last two decades, there has been little scientific evidence with which to evaluate this suspicion. This is a field of active research, and many associations between particular foods and nutrients and specific cancer s are controversial. Nevertheless, a consistent picture has emerged of a prudent dietary pattern that, if maintained for many years, is likely to reduce risk of cancer.

Breast cancer
Most attention has focused on whether dietary fat increases breast cancer risk. Although breast cancer rates are higher in countries with higher average fat consumption, these tend to be more economically developed countries in which the prevalence of established risk factors for breast cancer (such as earlier onset of menstrual periods, later age at first birth, and lower number of children) is also higher.¹ High dietary fat intake may be simply a marker of relative affluence and other breast cancer risk factors.

Several studies that measured the individual dietary fat intake of large groups of women and then compared the intake of participants who developed breast cancer in the next five to 10 years with those who did not develop breast cancer have all reached the same conclusion: The average fat intake of the women who developed breast cancer was similar to that for those who did not.¹ Recent analyses of these studies have examined the group of women whose fat intake was less than 25 percent of calories from fat, which is very low by Western standards, and observed no reduction in risk. While these studies do not exclude the possibility that fat intake may relate to breast cancer many years later or that fat intake during childhood or adolescence may be important, they do indicate that women in middle life who reduce dietary fat intake cannot expect to lower their breast cancer risk substantially in the 10 years after making this change.

A growing body of evidence suggests that excess caloric intake in relation to physical activity may be associated with breast cancer risk. Restricting total food intake in animal experiments is one of the best ways to reduce the incidence of many types of cancer, including mammary cancer. Breast cancer rates around the world are correlat ed positively with average adult height, which is a surrogate for caloric balance during growth, and even in studies conducted in the United States and Europe, a weak positive association between height and breast cancer is apparent. The possibilities that an excess of calories during childhood and adolescence increases the risk for breast cancer or that a deficit of calories is protective could explain much of the international variation in breast cancer risk but do not lead to obvious interventions - few parents would wish to stunt their daughters' growth. If lesser degrees of caloric restriction or increasing energy expenditure through exercise could be linked to protection against breast cancer, then these may be practical interventions.

In contrast to the lack of association with dietary fat, many studies have indicated that higher consumption of vegetables may be associated with a modest reduction in breast cancer risk of the order of about 20 percent.^2,3 Which types of vegetables or nutrients are responsible for this association is unclear - carrots, broccoli, and cabbage all have their proponents. At this point, sensible advice would be to follow the US National Cancer Institute guidelines that suggest eating at least three servings of vegetables per day and varying these vegetables from day to day.

A large body of evidence suggests that even moderate alcohol consumption can increase breast cancer risk: An average of two drinks per day increases risk by about 25 percent.⁴ Recent studies show that alcohol increases the levels of estrogens in the bloodstream,⁵ which would be expected to increase breast cell division and increase cancer risk. Recommendations are complicated by the fact that modest alcohol consumption almost certainly decreases the risk of coronary heart disease (perhaps related to the same mechanism of increasing circulating estrogens or alternatively due to the effect of alcohol on blood clotting). The balance of risks and benefits for individual women is complex. The fewer heart-disease risk factors a woman has, the more likely it is that avoiding regular alcohol consumption is beneficial overall.

An intriguing possibility is that while total dietary fat bears little relation to breast cancer risk, monounsaturated fats actually may be protective. Recent studies in Europe demonstrate that women who consume or cook with olive oil, a rich source of monounsaturated fats, may be at lower risk than women in the same countries who use little olive oil.³ More research is needed before olive oil can be recommended for breast cancer prevention; however, it is worth noting that substituting monounsaturated fat for saturated fat has a favorable effect on blood lipid levels and is likely to reduce coronary heart disease risk.

Colon cancer
Colon cancer is infrequent in countries with low meat intake and less frequent in Western populations, such as Seventh-day Adventists, with a vegetarian diet.⁶ In follow-up studies,⁷ higher intake of red meat has been linked with colon cancer risk with reasonable consistency. The specific factors in red meat responsible for this are unclear —candidates include an effect due to animal fat, which may increase the rate of division of the cells of the colonic epithelium, and/or carcinogens, which may be created during the high-temperature cooking of animal protein.

A protective effect of dietary fiber has been less consistently observed. Vegetable and fruit fiber do appear to be associated with decreased risk, whereas grain fiber is not consistently protective, raising the possibility that other nutrients in fruit and vegetables are responsible for the protective association. Folic acid, which is found in green vegetables and in many multivitamin supplements, has been associated with decreased risk of both colon polyps (a precancerous lesion) and colon cancer itself. High alcohol intake, particularly when combined with low folic acid intake⁸ also may increase risk of colon cancer. Many studies also have linked alcohol intake to increased risk of rectal cancer.

The summary of the evidence suggests that reducing red meat intake to less than one serving per week and increasing the intake of vegetables, particularly green vegetables and perhaps fruit, is likely to reduce colon cancer risk.

Lung cancer
Numerous studies have shown consistently that among smokers, those who eat more vegetables, particularly green and yellow vegetables, are at lower risk of lung cancer. However, cigarette smoking is the overwhelming cause of lung cancer, and a smoker consuming the largest tolerable amounts of vegetables is still at much higher risk of lung cancer than a nonsmoker, underscoring the fact that diet can only modestly ameliorate, not abolish, the harmful effect of smoking.

There is little consensus about which specific nutrients in green and yellow vegetables are responsible for the anticancer action. A prominent candidate was ß-carotene, but in several studies, risk of lung cancer was not lower, and was possibly even higher, among those given ß-carotene supplements. These studies reinforce the belief among nutritionists that it is premature to recommend consumption of specific nutrient supplements for cancer prevention. It may be decades before it is established which of the many thousands of chemical compounds in vegetables are responsible for the protection they confer against some types of cancer.

Prostate cancer
Relatively little is known about risk factors for prostate cancer in general, and the nutritional epidemiology of the disease is equally enigmatic. Two recent follow-up studies have observed an increased risk among men with higher meat consumption;^9,10 however, too few studies are available to reach firm conclusions. Evidence on vitamin A has been contradictory, showing associations with both decreased risk and increased risk, and few data are available for fiber or alcohol. Given that prostate cancer is the most common cancer among men in the US, it is surprising that so few large studies of diet and prostate cancer have been done. Ongoing studies should help clarify whether diet is related to prostate cancer.

Other types of cancer
Alcohol, particularly in conjunction with smoking, increases risk of cancer of the esophagus, while vegetables and fruits appear to be protective. Similarly, vegetables and fruits have been shown consistently to reduce risk of stomach cancer. In Asia, consumption of salted foods may account for higher rates of stomach cancer.

Conclusion
The role of diet in cancer causation is complex and difficult to unravel, but ample evidence does exist to make some recommendations. Reducing meat intake and perhaps animal fat intake, increasing consumption of vegetables and fruits, and reducing alcohol intake, almost certainly will reduce risk of cancer overall, although the amount of risk reduction is uncertain.

Summary Points

Suggestions

Suggested Further Reading

References