Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Overview and Major Recommendations Many diseases now known to be associated with dietary imbalances or toxic contaminants in food were once thought to arise from other causes, ranging from 'bad air to disorders of the bodily humors. Their associations with diet have tended to be worked out by the same sequence of discoveries. First, a particular human disease is linked to some aspect of diet, e.g., a lack of fresh vegetables or the fungal contamination of grain. Alteration of the diet (by crude additions or subtractions) is then shown to prevent or alleviate the disease. Sub- sequently, someone succeeds in producing a similar disease in an animal model E.g., scurvy in guinea pigs or mycotoxicoses in rats and mice), which leads to the precise identification of the active components of the diet (e.g., a nutrient or a toxic ingredient). Finally, laboratory scientists discover the mechanism by which the nutrient or toxin exerts its effects, although by this stage such details have become mainly a matter of academic interest because the disease in question has already been eradicated. A similar sequence of discoveries has taken place in some branches of cancer research. Chronic exposure to coal tar in mineral oil was observed to cause skin cancer in humans more than a century before laboratory investigators succeeded in producing cancers in rabbits by painting their skin with coal tar. Once the cancer had been produced experimentally, the active carcinogens in the tar--polycyclic aromatic hydrocarbons--could be purified and identified, and later, their mecha- nisms of action explained. By this time, however, skin cancer in humans resulting from such exposure had long since been effectively abolished by general improvements in working conditions in factories; however, laboratory studies of these and other carcinogens have continued and have provided important insights into the mechanisms of carc~nogenesis. These historical examples may seem too simple for predicting the the course of research to unravel interactions between a multifactorial disease like cancer and a complex mixture like diet; but they may give us some idea of what to expect. Although interest in the study of diet and carcinogenesis can be traced to laboratory experiments performed more than half a century ago, it seemed in the 1960's that we were still in the first stages of the sequence described above. At that time, despite evidence from early experiments that modification of either total food intake or some dietary components could influence carcinogenesis, the possibility that diet per se was a significant factor in human cancer was still considered remote. Then epidemiolo- gists linked the incidence of several common cancers, e.g., breast cancer, with certain general dietary patterns. Laboratory scientists followed up these observations by developing animal models for cancers suspected of being affected by diet. Subsequently, epidemiologists observed that the high incidence of breast cancer and certain other 1
2 DIET, NUTRITION, AND CANCER: DIRECTIONS FOR RESEARCH cancers is associated with a diet high in fat or its components, and laboratory investigators found that mammary cancers in certain species were similarly modified by changes in the amounts and types of dietary fat. Further studies may eventually permit isolation of the active dietary constituents, definition of the exact mechanism for the effect exerted by fat and other dietary components, and delineation of the precise diets capable of counteracting some of these effects. However, research on nutrition and carcinogenesis has not invari- ably followed the sequence described above. Of the many hypotheses generated by the results of early experiments in animals, only some have been followed up by epidemiological studies. For example, clues about the effect of the caloric content of the diet per _ on experimentally induced carcinogenesis have remained largely unexplored. Similarly, leads produced by international correlation studies of human popula- tions have not always been followed up by more controlled epidemiologi- cal and laboratory investigations. For example, the finding in the mid-1960's that low selenium intake may be associated with increased cancer incidence or mortality has been tested in well-controlled lab- oratory experiments, but no controlled epidemiological studies (i.e., case-control or cohort studies) could be conducted because of a lack of knowledge about the precise intake of selenium. It is never possible to predict exactly where major discoveries will be made, and any attempt to stipulate a particular sequence for research on diet and cancer would tend to stifle creativity. There- fore, the committee has been rather cautious in making suggestions. Nevertheless, it may be desirable to plan the research on diet and cancer in a logical but flexible conceptual framework that could encompass all the sources of data, i.e., surveys to monitor expo- sure, epidemiological studies, carcinogenesis bioassays in animals, short-term tests for genotoxicity, short-term in vivo bioassays to detect early biological indicators of carcinogenesis, and studies designed to elucidate metabolic pathways or pathogenic mechanisms. After completing an assessment of the literature in 1982, the Com- mittee on Diet, Nutrition, and Cancer concluded that 'the differences in the rates at which various cancers occur in different human popula- tions are often correlated with differences in diet. The likelihood that some of these correlations reflect causality is strengthened by laboratory evidence that similar dietary patterns and components of food also affect the incidence of certain cancers in animals. Thus, concordance between epidemiological and laboratory data served as the principal basis for the degree of certainty allotted to conclusions and as the basis for the interim dietary guidelines proposed in the first report. The selection of this criterion reflects the committee's con- viction that persistent interaction between epidemiologists and labora- tory investigators is necessary to provide a framework for future research that will lead to a more definitive understanding of diet and carcinogenesis.
Overview and Major Recommendations 3 STRATEGIC OBJECTIVES AND PRIORITIES FOR RESEARCH The committee ha s operated on the principle that research on diet and cancer should encompass the seven strategic objectives presented below. From the numerous suggestions for research made in this report, it wishes to call attention to certain general recommendations, which are listed following the strategic objective to which they apply. 1. Identification of the foods and of the dietary macro- and mirr~r~n.c~titilPntc th::~t alter the read of renter her] "lilr,;rl~t;~= of their mechanisms of action. In the first report, the assessment of the literature resulted in the preliminary identification of four categories of dietary constit- uents that are likely to affect the risk of cancer. These were satu- rated and unsaturated fat; certain fruits, vegetables, and whole grain cereals; smoked, cured, and pickled foods; and alcoholic beverages. The committee recommends that when the epidemiological and experimental evidence associating particular dietary components with cancer risk is sufficiently convincing, studies should be undertaken to identify the specific active constituents and their mechanisms of action. For example, attempts should be made to identify the consitituents of fruits and vegetables that are responsible for the observed reduction in risk associated with their frequent consumption and to define the mechanisms of action of those constituents (see Chapter 7~. Similarly, studies should be pursued to elucidate the mechanisms by which a high fat diet increases the incidence of certain cancers (see Chapter 6~. Information from such studies would be useful in refining the interim dietary guidelines recommended by the committee in its first report. 2. Improvement of the data base and the methodology for assess- ing human exposure to foods and dietary constituents that may alter the risk of cancer. Better epidemiological methods should be developed to monitor and quantify dietary exposures in human populations in order to establish more clearly the relationship of dietary constituents and dietary patterns to the occurrence of cancer. For example, innovative methods are needed to measure past dietary intake. Furthermore, better techniques should be sought to validate the data produced by all these methods. Regular nutrition surveys to monitor dietary intake would augment the data base for epidemiological studies of diet and cancer (see Chapter 4~. 3. Identification of markers of exposure and early indicators of the risk of cancer. The committee recommends that attempts be made to identify early biological or biochemical changes that reflect the ability of specific
4 DIET, NUTRITION, AND CANCER: DIRECTIONS FOR RESEARCH dietary constituents or dietary patterns to alter the risk of cancer in humans. For example, where neoplasia i s used as the sole end point, investigations are severely limited by the long latency period between "exposure" and "expression. " Therefore, one of the most pressing needs is the development of short-ter~ test s that could identify early bio- logical indicators of exposure to dietary constituents that affect carcinogenesis (see Chapters 4, 5, and 7~. 4. Determination and quantif ication of the adverse or beneficial effects of the foods and of the dietary macro- and microconstituents . . that affect the risk of cancer. The committee recommends that efforts be continued to evaluate the impact of potentially carcinogenic or inhibitory dietary constituents on cancer risk. These studies should include a focus on substances that can damage macromolecules, especially DNA; on those that can enhance experimentally induced carcinogenesis, i.e., promoters and cocarcinogens; and on those that can inhibit experimentally induced carcinogenesis (see Chapters 5, 6, 7, and 8~. 5. Determination of the ranges of optimal intake of dietary macro- and microconstituents. Attention should be given to determining ranges of dietary macro- and microconstituents that are optimal not merely for the prevention of deficiency diseases but also for the promotion of other aspects of health, including the reduction of the risk of cancer. For example, it would be useful to establish a dose-response curve for selenium and to define the optimal range of selenium intake, giving special attention to the levels that might be needed to achieve a reduction in the risk of certain cancers (see also Chapters 5, 6, and 7~. 6. Intervention to reduce the risk of cancer. Intervention studies should be conducted using foods or food con- stituents believed to be associated with a lower cancer risk, but only when a substantial body of data indicates a high likelihood of benefit without discernible risk. For example, attention might be given to reducing the consumption of fat and/or adding specific fiber compo- nents to the diet (see Chapter 6), and to the ingestion of different levels of certain microconstituents or of foods containing potential inhibitors (see Chapters 4 and 7~. 7. Application of knowledge about diet and cancer to programs in public health. To maximize the potential impact of public health programs to reduce the risk of cancer, studies should be pursued to elucidate factors that motivate people to modify their food habits. For example, it would be useful to analyze bodies of longitudinal data to learn what they reveal about factors that determine consumption patterns in dif- ferent populations ~ see Chapter 9) .
