Extracts from


Report of a WHO Study Group

World Health Organization Technical Report Series 797

A WHO Study Group on Diet, Nutrition and Prevention of noncommunicable Diseases met in Ceneva from 6 to 13 March 1989. The meeting was opened by Dr Hu Ching-Li, on behalf of the Director-General, Dr Hiroshi Nakajima. He said that the amount and type of food eaten were fundamental determinants of human health. Since health was a fundamental determinant of the quality of each individual's life, good health should be a primary social goal. Improvements in the collective good health of a population- particularly the avoidance of chronic diseases in adult life-also decreased the costs associated with both health care and lost economic productivity. Good health was therefore an important economic asset.

Changes in dietary habits towards the "affluent" diet that prevailed in many developed countries had been followed by increases in the incidence of various chronic diseases of middle and later adult life. Initially, those chronic diseases coexisted with the long-standing and persistent problems associated with nutritional deficiencies, which could affect all age groups. The continuing public health importance of such deficiency disorders was recognized. However, the task of the Study Group was to provide recommendations that would help to prevent the chronic diseases that were related to the newly emerging dietary changes in developing countries, and to help in reducing the impact of these diseases in developed countries.

The Study Group's report would describe recent changes in the dietary and health patterns of countries, define the relationship between the "affluent" diet that typically accompanied economic development and the subsequent emergence of chronic diseases, and explore the need for national food and nutrition policies to prevent or minimize costly health problems in both developing and developed countries.

3.3 Cancer

The relationships between specific dietary components and cancer are much less well established than those between diet and cardiovascular diseases. This is reflected in reviews of diet and cancer (43-45). However, the overall impact of diet on cancer: rates throughout the world appears to be significant. For populations in developed countries, where cancer rates are highest and account for approximately one-quarter of all deaths, some epidemiologists estimate that 30-40% of cancers in men and up to 60% of cancers in women are attributable to diet (46).

The evidence for the influence of diet on cancer risk is derived from several sources. Correlations between national and regional food consumption data and cancer rates, and studies of the changing rates of cancer in populations as they migrate from a region or country of one dietary culture to another, have led to many important hypotheses. Case-control studies of the dietary habits of cancer patients and comparison subjects, and prospective studies of populations with known dietary habits, provide stronger evidence for the effects of diet in relation to major cancers. Many of these observations from human populations have been supported by animal experimental data.

Studies of diet in relation to some cancers have been confined to relatively homogeneous populations and have not been replicated across a range of cultural and dietary settings; for other cancers, the research has been pursued over a wider range of dietary intakes included among the cancers that have been linked repeatedly to dietary factors in different populations are cancers of the oral cavity. pharynx, larynx, oesophagus, stomach, large bowel, liver, pancreas, lung, breast, endometrium, and prostate.

3.3.1 Cancers of the oral cavity, pharynx, larynx, and oesophagus

In developed countries, epidemiological studies clearly indicate that drinking alcoholic beverages is causally related to cancers of the mouth, pharynx, oesophagus, and upper part of the larynx (47). There is no indication that the effect is related to the type of beverage. Smoking also causes cancers at these sites. There is also some evidence that cancers of the mouth and throat are increased by poor oral and dental hygiene.

In correlation studies conducted in different parts of the world. investigators have found positive associations betweem oesophageal cancer and several dietary factors, including (a) low intakes of lentils, green vegetables, fresh fruits, animal protein, vitamins A and C, riboflavin, nicotinic acid, magnesium, calcium, zinc, and molybdenum; (b) high intakes of pickles, including salt-pickled vegetables, and mouldy foods containing N-nitroso compounds; and (c) consumption of foods and beverages at very high temperatures. The reported associations are consistent with the general hypothesis that certain nutrient deficiencies, such as are found in many high-risk populations, including heavy alcohol drinkers, might increase the susceptibility of the oesophageal epithelium to neoplastic transformation. Case-control studies of oral and laryngeal cancers have also shown an increased risk associated with infrequent ingestion of fruit and vegetables.

3.3.2 Stomach cancer

A high incidence of stomach cancer is found in Japan and other parts of Asia and in South America, but not in North America or Western Europe where the rates are low and still decreasing. In the United States of America, stomach cancer rates are now among the lowest in the world, whereas in 1930, this was the leading cause of cancer death for men and the second leading cause in women. Gastric cancer incidence is decreasing in Japan, and a gradual decline in incidence over several generations has been noted among Japanese migrants to Hawaii. It seems most likely that these trends are related to changes in food consumption patterns, since several dietary factors have been implicated in gastric cancer risk. Stomach cancer is associated with diets comprising large amounts of smoked and salt-preserved foods (which may contain precursors of nitrosamines) and low levels of fresh fiuits and vegetables (acting as possible inhibitors of nitrosaminc formation). Dietary shifts away from this pattern could explain the declines in stomach cancer mortality in industrialized nations over the past 50 years, but the evidence is not conclusive.

3.3.3 Colorectal cancer

International comparisons indicate that diets low in fibre- containing foods and high in fat increase the risk of colon cancer. the initial suggestion that a lack of dietary fibre might increase the occurrence of large bowel cancer came from observations of the virtual absence of this cancer in southern Africa. The indigenous populations were known to eat a lot of plant foods, and to have much higher faecal weights than populations fron industrialized countries.

Several studies also demonstrate positive assoaations between the risk for colorectal (primarily colon) cancer and dietary fat. In general, the data suggest that saturated rather than unsaturated fatty acids may be responsible for this effect. In other studies, positive associations have been found between meat consumption and this cancer, but many studies have also shown no relationship between fat or meat intake and colorectal cancer. Several case-control and cohort studies provide suggestive but inconclusive evidence that drinking alcoholic beverages, in particular beer, has a causal role in the development of rectal cancer.

The data relating dietary fibre per se to colorectal cancer are equivocal. Although several studies have shown inverse relationships between the intake of high-fibre foods and colon cancer risk, these foods (vegetables to a large extent) are rich sources of other nutritive and non-nutritive constituents with potential cancer- inhibiting properties. Lower rates of colorectal cancer in Californian Seventh-Day Adventists, half of whom are vegetarians, support protective effect of a vegetarian diet, although this group also abstains from alcohol.

In summary, an increased risk of colorectal caner appears to be associated with high fat intake (particularly saturated fats) and low vegetable intake. It is not clear whether dietary fibre pev se is protective or whether the apparent,ffect is due to other food constituents . Rectal cancer risk may be increased by the consumption of beer.

3.3.4 Liver cancer

Primary liver cancer is relatively rare in North Alerica and most developed countries, but it is common in sub-Sahran Africa and south-east Asia, where it is associated primarily "the exposure to hepatitis B virus infection. Liver cancer incidence and mortality, by geographical area,or among different population groups, have been correlated with aflatoxin contamination of foodstuffs in Africa. On the basis of evidence in developed, countries, consumption of alcoholic beverages is causally related to liver cane, (47).

3.3.5 Lung cancer

In most industrialized countries, lung cancer is the leading cause of cancer deaths among men, and it is rapidly approaching this status among women. The most important causal factor is cigarette smoking. Luug cancer risk in males is clearly increased by certain occupational exposures (e.g., to asbestos, nickel, chromate, or gamma-radiation) several of which have been shown to interact synergistically with smoking.

Studies in several different populations have shown an interactive effect between smoking and a low frequency of intake or green and yellow vegetahles rich in beta-carotene. These findings are consistent with experimental data showing tumour inhibition by vitamin A and synthetic analogues. In prospective studies, the frequency of consumption of beta-carotene-containing foods and the concentration of beta-carotene in serum have been inversely associated with the risk of lung cancer, but earlv reports of a similar inverse association for erum retinol(vitamin A) have not been confirmed in subsequent studies. Detary fats and dietary cholesterol have also been positively associated with lung cancer risk.

3.3.6 Female breast cancer

Several lines of evidence support the importance of dietary factors in the causation of breast cancer. The first derives from animal experimental studies, which have demonstrated that, both with and without the presence of known mammary carcinogens, the incidence of mammary tumours in rats increases substantially with diets high in total and saturated fat, provided that the diet contains a small amount of polyunsaturated fat. A role for fat and other dietary factors is also supported by descriptive epidemiological studies, correlation studies, case-control and cohort studies, and evaluations of nutrition-mediated biological risk factors.

Correlation studies provide evidence of a direct association between breast cancer mortality and the intake of energy, fats, and specific sources of dietary fats, such as milk and beef (see, for example. Fig. 9, page 68). Several case-control studies have associated breast cancer risk with dietary constituents, especially fats. However, not all studies show these relationships. There is epidemiological evidence--not fully consistent relating alcohol consumption to the risk of breast cancer in women. It is, at present, unclear whether this association is causal.

3.3.7 Endometrial cancer

A strong association between endometrial cancer risk and excess weight has been reported in several studies, and a hormonal mechanism has been postulated for this association. Specific dietary factors other than obesity have not been identitied for this disease.

3.3.8 Prostate cancer

Internationa1 incidence and mortality data generally show a positive correlation of prostate cancer with the incidence of other diet-related cancers, including cancers of the breast, corpus uteri. and colon. Inter- and intra-country analyses show positive correlations between mortality from prostate cancer and per caput intake of total fat. These findings have been supported in analytical studies showing an association of prostate cancer with the intake of high-fat foods.

Although studies of certain other cancers suggest that vitamin A and, in particular, beta-carotene may be protective factors, some case-control studies indicate that beta-carotene may be a risk factor for prostate cancer, especially among men aged 70 years and older. Increased weight or obesity has also been positively associated with the risk of prostate cancer.

3.3.9 Summary and conclusions: major associations between diet and cancer

Table 11 summarizes the strength of association between dietary components and cancers at various sites. A review of the evidence indicates that a high intake of total fat- and in some case-studies also saturated fat--is associated with an increased risk of cancers of the colon, prostate, and breast. The evidence is strongest for cancer of the colon, and weakest for breast cancer. The epidemiological evidence is not totally consistent, but is generally supported by laboralory data from studies in animals. The experimental data, however, also point to an adverse effect of very high intakes of polyunsaturated fats, at levels that are considerably above current intakes in human populations.

Diets high in plant foods, especially green and yellow vegetables and citrus fruits, are associated with a lower occurrence of cancers of the lung, colon, oesophagus, and stomach. Although the mechanisms underlying these effects are not fully understood, such diets are usually low in saturated fat anf high in starches and fibre and several vitamins and minerals, including beta-carotene and vitamin A. There is no conclusive evidence that these beneficial effects are due to the high fibre content of such foods.

Sustained heavy alcohol consumption appears to be causally linked to cancer of the upper alimentary tract and liver. Excessive hody weight is clearly a risk factor for endometrial and postmenopausal breast cancers, but the association of these cancers with excessive energy intake per se is less well established.

High fat intake is associated with cancer at several sites. Certainty about the optimum intake of fat in relation to cancer must await future research, such as controlled trials. In the meantime, international correlation analysis (Fig. 9) and other epidemiological data indicate that fat Intakes of less than 30% of total energy will be needed to attain a low risk of fat-related cancers. A reduction in risk is also likely when fat intake is reduced towards 30%, especially if this dietary change is combined with a change in other dietary components (Table 11).

Table 11. Associations between selected dietary components and cancer (a)

Site of cancerFatBody
and vegetables
salted and
pickled foods
Lung -
Breast + + +/-
Colon + + - -
Prostate + +
Bladder -
Rectum+ - +
Endometrium + +
Oral cavity - +(b)
Stomach - + +
Cervix -
Oesophagus - + +(b) +
+ = Positive association; increased intake with increased cancer.
- = Negative association; increased intake with decreased cancer.
a Adapted and extended from reference 44.
b Synergistic with smoking.

In conclusion, although several lines of evidence indicate that dietary factors are important in the causation of cancer at many sites and that dietary modifications may reduce cancer risk, the contribution of diet to total cancer incidence and mortality cannot be quantified on the basis of present knowledge. Nevertheless, evidence indicates that a diet that is low in total and saturated fat, high in plant foods, especiaIly green and yeIlow vegetables and citrus fruits, and low in alcohol, salt-nickled, smoked, and salt-preserved loods is consistent with a low risk of many of the current, major cancers, including cancer of the colon, prostate, breast, stomach lung, and oesophagus.

5.3 Derivation of population nutrient goals

The evidence is clear that the risk of certain types of cancer is directly associated with the level of total fat in the diet. At present the range over which the relationship holds is not clear. It may extend across a range of total fat intakes from > 400% of energy down to 20% or lower. There is suggestive evidence that obesity may he associated with high total fat intake, but again there is no specific lower level that marks the end of the relationship.

The Study Group noted increasing evidence of specific beneficial effects of complex carbohydrates on intestinal function, on the chemistry of the gut and the physiology of the gut wall with a possible relationship to cancer, and on the absorption and metabolism of carbohydrates and other energy sources (including short-chain fatty acids formed by fermentation in the lower intestine) in relation to the amelioration of diabetes and other metabolic diseases. At present, the available evidence does not lead to the definition of specific population nutrient goals. Rather, it argues in favour of maximizing the intake of this class of carbohydrate, taking into account the energy provided by fat and protein. The lower and upper limits were set on that basis.

For the purpose of this recommendation, "fruits and vegetables" does not include potatoes, other tubers, or cassava. The recommendation is made recognizing: the epidemiological evidence of an increased risk of cancer with low intakes of certain fruits and vegetables; the continuing problems of vitamin A deficiency and low availability of dietary iron, which would be ameliorated with increased intakes of vitamin A (or beta-carotene) and ascorbic acid, respectively, and the specific contribution that these groups of foods make to micronutrient and protein intakes. The population nutrient goals have bean set judgementally rather than on the basis of specific evidence as to the necessary level of intake. The recommended lower limit is higher than current intake in many populations and much higher than current intake in some of the developed countries. No upper limit has been suggested. The carbohydrate, lipid, and protein contributionsof the fruits and vegetables are included within, not in addition to, the previously stated goals.


8.1 Recommendations to WHO

The Study Group recommended that WHO should consider:

1.Developing a coherent nutrition and health policy with targets for promoting the concept of healthy nutrition. This concept, which supplements the long-standing concern with nutritional deficiencies and food shortages, takes account of the increasecd risks of chronic diseases in later adulthood caused by the "affluent" dietary pattern (i.e., high in fats and free sugars, and low in starchy foods) that tends to accompany economic development.

2.Establishing a mechanism to ensure that the impact of diet upon a range of health problems is recognized in all relevant WHO programmes.

3.Discussing with other United Nations and international agencies the major implications for agriculture and trade of the nutritional needs ofboth developing and developed countries, as described in this report. Commodity analyses and the projections of FAO should be considered in relation to the nutritional quality of the diet.

4.Communicating to the major trading organizations the consequences of trading developments that might lead to adverse effects on health. The export of subsidized saturated fat is an example of one potential problem.

5.Strengthening its programme to provide expert guidance the national governments seeking to establish a nutrition and food policy, for example by means of seminars for national governments and nongovernmental organizations.

6.Developing a basis for monitoring the progress of the initiatives described in this report on diet and health, and setting targets for this work. In this regard the development of surveillance programnes at country level should be promoted.

7.Discussing with nongovernmental organizations how best to ensure that communities throughout the world participate in activities to promote health by eating an appropriate diet. The variety of diets consistent with the nutritional goals should be recognized, and the nutritional goals translated into food needs at national level.

8.Monitoring research developments and promoting international collaboration in the study of the relationships between diet and health so that further refinements to the Organization's nutrition policy can be made.

9.Involving the Coder Alimentarius Commission (1) in establish- ing standards that will help to implement the proposals in this report.

(1)The Codex Alimentarius Commission is an intergovernmental body with a membership, in 1989, of 137 countries; it is charged with the implementation of the FAO/WHO Food Standards Programme.

8.2 Recommendations to national governments

1O.Governments are recommended to establish a national board for nutrition and food policy involving, in addition to the ministry of health, the many government ministries whose policies affect the production, distribution, and consumption of food.

11.Governments should ensure that experts are available to the ministry of health to monitor the nutritional and health status of the population, as assessed by a national surveillance system.

12.Ministries of health should initiate or strengthen professional training programmes, at both undergraduate and postgraduate levels, to ensure that the role of diet in the prevention of chronic diseases is understood by the medical profession and other health care workers.

13.Each ministry of health should, as part of its health promotion programme, establish regular contact with non governmental organizations, consumer representatives, and the media to develop jointly a community-based programme. This activity should be in addition to any government-sponsored healh promotion campaign.

14.Governments should ensure that adequate nutritional com- petence exists within the ministry of agriculture to allow full participation in a national board for nutrition and food policy.

15.Governments should consider their investment and subsidy policies in both agriculture and the food industryto ensure that they are consistent with the nutritional concepts contained in this report. Policies should be geared to promoting the growing of plant foods, including vegetables and fruits, and to limiting the promotion of fat-containing products.

16.As part of a national policy, each government should set its own goals and strategy for reducing the incidence of chronic diseases.

17.Governments are recommended to establish appropriate food standards, to ensure the nutritional quality of foods that are substantial contributors to the national diet.

18.Each government should consider all new legislation bearing on agriculture and food, to ensure that it is compatible with the prevention of chronic diseases.

19.Governments are recommended to establish, where possible, compulsory labelling of food products based on the Codex Standardsand Guidelines for the Labelling of Foods and Food Additives. (1) Labelling should be clear and consistent and, to be understandable as well as scientifically correct, information should be simply presented and expressed both graphically and numerically. One method is proposed in Annex 6.

20.Discussions should be encouraged between the government, the food indusry, and the consumers to ensure the development of food products that are low in fat, free sugars, and salt.

21.Each government should examine its animal production policies and incentes to ensure that they do not promote the production of excessive quantities of saturated fats.

22.Ministriesof education should ensure that, as part of nutrition and health education for teachers and children, due attention is given to ths prevention of diet-related chronic diseases.

(1) CODEX ALIMENTARIUS COMMISSION. Codex Alimentarius, Volume VI. Codex Standards and Guidelines for the Labelling of Foods and Food Additives, 2nd ed. Rome, Food and Agriculture Organization of the United Nations/World Health Organization, 1987.


The Study Group wishes to acknowledge the valuable contributions made to its work by: Dr K.V. Bailey, Regional Officer for Nutrition, WHO Regional Office for Africa, Brazzaville, Congo; Dr H.W. Heiss, University of Freiburg, Freiburg, Federal Republic of Germany; Dr E. Helsing, Regional Officer for Nutrition, WHO Regional Office for Europe, Copenhagen, Denmark; Dr F.K. Kaferstein, Chief, Food Safety, WHO, Geneva, Su;itzerland; Dr E. Nicholls, Regional Adviser, Noncommunicable Diseases, WHO Regional Oflice for the Americas, Washington, DC, USA; and Dr R. Saracci, Unit of Analytical Epidemiology, International Agency for Research on Cancer, Lyon, France.

Back to Media Page