Chapter two - Scientific considerations in the development of food-based dietary guidelines (FBDG)
There is good scientific evidence that dietary patterns, i.e. a daily combination of foods and beverages, have specific health or disease outcomes. For example, a diet may be apparently adequate in all other ways but still be deficient in vitamin A or iron, and this may lead to xerophthalmia or anaemia. Conversely a diet high in saturated fat and energy is known, on a population basis, to lead to an increased likelihood of coronary heart disease.
The reasons for developing and using FBDG are many, and often surprisingly self-evident.
1. Foods make up diets; foods are more than just a collection of nutrients.
2. Nutrients interact differently when presented as foods.
3. Methods of food processing, preparation and cooking influence the nutritional value of foods.
4. There is already good evidence from studies (animal, clinical and epidemiological) that specific dietary patterns are associated with a reduced risk of specific diseases. Diets rich in fruits and vegetables are associated with various positive outcomes such as reduced incidence of lung cancer (1). Science has not been able to identify completely the specific nutrients involved. The protective effect could be due to a single nutrient, a combination of nutrients or non-nutrients, or the replacement of some other food in the diet when fruits and vegetables are consumed in large amounts (2).
5. Scientific evidence has not yet identified the potential health outcomes of some non-nutrient food components, hence the increasing attention being paid to flavonoids, phyto-oestrogens etc. If the focus is on a single nutrient, people may not appreciate the benefits of the ingestion of these compounds in foods.
6. Some food components may have biological functions that science has not yet identified.
7. Foods and diets have cultural, ethnic, social and family aspects that individual nutrients themselves do not have.
8. For certain micronutrients, evidence suggests that intake higher than present nutritional recommendations may help to lower the risk of noncommunicable diseases. FBDG can encourage dietary patterns that include these nutrients.
FBDG need to take into account: food/health patterns; the relative comprehensiveness of the foodbased versus the nutrient-based approach; the practicality of suggested goals; nutrition labelling directing the consumer to an over-simplified view of foods; and shifting paradigms of the nutritional basis of disease and health (the biological effects of food and food patterns can be greater than the sum of its parts).
FBDG need to be based on sound scientific principles and knowledge of local conditions. A summary of the four main areas of science relevant to the development of FBDG is provided below, with a more detailed description in Annex 3.
The study of energy balance and metabolism, and of macro- and micronutrient physiology, is well-established and has now evolved to take account of non-nutrient components of food of biological significance.
As far as energy is concerned, the next generation of FBDG can take account of both the lower limits of energy intake and the ways that energy balance can be achieved without recourse to food restriction, but with the emphasis placed on strategies to achieve energy balance either by decreasing the energy density of the diet or increasing energy expenditure. With this approach, concerns about both chronic energy deficit and excess body fat can be addressed through FBDG by encouraging the use of foods of appropriate energy density.
Micronutrients are now recognized as important in communicable disease (improving host defences), classical deficiency disorders and NCD. Therefore, the formulation of FBDG in all food cultures with varying health patterns merits consideration of non-energy nutrients.
Food chemistry is extraordinarily complex, and it is becoming clearer that health states previously not regarded as nutrition-related may actually be determined by the intake of non-nutrient substances contained in foods, for example, certain flavonoids from plants and phyto-oestrogens from certain legumes. Therefore, it is essential that dietary guidelines be based on foods and not simply on nutrients, taking into account that sound scientific evidence about foods should be the foundation of recommendations.
Food science and technology is creating a new framework for FBDG. This is principally in the areas of:
Methods of food storage and preservation
Changes in food preparation
Opportunities for use of formula foods where energy intake is low or regular food cannot be eaten
Nutrient restoration and fortification of foods
The development of health-focused designer and functional foods.
Both the practicality of FBDG and their consequences require analysis through these sciences. However, even though a substantial knowledge base exists in these areas, it is not commonly utilized by nutrition and food policy-makers. For example, FBDG that neglect basic methods of promoting behaviour change may be scientifically sound but totally ineffective for improving dietary intakes. Some of the major areas of concern are set out in Annex 3.
The scientific basis of FBDG is inclusive of these disciplines. There is evidence that even existing dietary guidelines in developed countries, such as Australia and the USA, cannot be easily achieved in the short term without increased fruit and vegetable production. Even more difficult may be the pressure on world fish supply because of increasing appreciation of the importance of long chain omega-3 fatty acids from fish in the human diet. The need for sustainable food production using existing natural resources is becoming more and more apparent.
No single food other than breast milk for infants provides all the required nutrients. A range of nutrients is needed in amounts that change throughout life for optimal growth, health and avoidance of disease. The varied diets that people eat made up of many different foods have proved over the ages to be able to provide adequate nutrients in a variety of national patterns and combinations.
Many countries have adopted RDI, which are customarily defined as the intake of energy and specific nutrients necessary to satisfy the basic requirements of a group of healthy individuals. While the criteria to estimate requirements have changed over time, the three main methods of investigation still involve a clinical approach, studying the typical consumption of healthy populations, and looking at functional indicators of nutritional status. A detailed analysis of the relative merits of each is beyond the scope of this document, but additional information may be found in the references (3,4).
More recently the concept of "optimal nutrient intake" has evolved and is influencing both scientists and the public alike. The question of ''optimal for what?" is usually answered by the suggestion that diet or specific nutrients can lay the basis for improved physical and mental performance or for longer and healthier life. This concept is too broad. The preferred approach is to define clearly the function that is of interest in relation to the intake of a specific nutrient or a given food. The selected function should be of relevance to health or disease prevention.
A growing body of scientific evidence supports the relationships between food patterns and health. These relationships are derived from several types of studies:
Epidemiological which examine population-based data to associate disease patterns or health outcomes with food patterns. These studies can be prospective or retrospective. Several factors determine the strength of the associations concluded from epidemiology, including the consistency, strength, specificity and biological plausibility of the association. Epidemiological observations are, in principle, not sufficient to confirm cause and effect relationships between individual nutrients and health outcomes; however, they do provide valuable information about the food patterns to be considered in FBDG.
Clinical research which is designed to determine whether dietary modification or supplementation with selected nutrients can intervene in the disease process or modify one or more risk factors for disease. These studies confirm that the foods or nutrients in question do contribute to the health or disease outcomes of interest. The research is typically conducted in individuals who are known to be at risk for disease and, although the findings may be relevant to the general population, it is important to keep the limitations in mind when extrapolating from these studies to the general population (for example, trying to draw general conclusions from very low-fat diets in atherosclerotic regression studies).
Experimental studies which are designed to determine the effect of foods and food components on cell metabolism or physiology, and to understand the basic mechanism by which specific nutrients modify disease. These studies are essential to establish the plausibility of the diet-health relationship; however, they are often conducted in animal models and it can be difficult to extrapolate from one species to another.
An example of the need for integrating epidemiological, clinical and experimental data is the recent evidence associating a lower relative risk of age-related blindness due to macular degeneration with the consumption of dark-green leafy vegetables, particularly spinach and collard (5). While it may be that zeaxanthin and lutein components in these foods are protective against macular degeneration, establishing this relationship will require extensive experimental and clinical research. In the meantime an increased intake of dark-green leafy vegetables is a justifiable food-based dietary guideline, especially in view of other support for this recommendation.
Dietary patterns vary in different geographical areas and socioeconomic groups, and also over time. Such variation depends on agricultural practices, climatic, ecological, cultural and socioeconomic factors which determine the foods that people demand or expect and that are generally available.
Given the variety of food combinations that can provide a healthful diet, it is impossible to define the ranges of intake for all foods that could be combined to provide a nutritionally adequate diet. Although a large set of food combinations that are compatible with nutritional adequacy could be identified, they could hardly be extrapolated to cover every different ecological and social setting. An alternate approach to defining the nutritional adequacy of diets has therefore evolved, based on a scientific understanding of the biochemical and physiological basis of human nutritional requirements in health and disease. This approach has permitted the definition of essential nutrients and the establishment of RNI.
FBDG as an instrument and expression of food and nutrition policy should be based directly upon diet and disease relationships of particular relevance to the individual country. The priorities in addressing the dietary guidelines will depend on whether the relevant public health concerns are related to dietary insufficiency or excess.
Four possible approaches are useful in assessing the nutritional quality of diets for the development and evaluation of FBDG:
Assessing adherence to a particular food pattern usually associated with a favourable health outcome is one way of evaluating the nutritional soundness of an envisaged dietary guidelines approach. This reference is most likely to concern a traditional food pattern of people with longevity, low morbidity and low perinatal and infant mortality rates (e.g. Scandinavian, Japanese, Mediterranean) through tradition or through cultural adaptation. In most of these populations, other factors such as health care, educational system, safe water and socioeconomic development also play important roles in the favourable health outcomes. Negative effects following changes in dietary patterns might also indicate food patterns to be avoided.
Tracking health indices in populations in accordance with food intake patterns has, so far, been the most available evidence on which to base FBDG. Some of the examples are given in Annex 3.
While the value of increased food variety in either ensuring essential nutrient adequacy or decreasing the risk of food toxicity (health-adverse factors in food are generally diluted where the foods eaten are varied) has been understood for some time (6), using food variety as a predictor of health outcome is a relatively recent approach (7,8). However, enough evidence is available to justify promoting food variety through FBDG as a technique to reduce morbidity and mortality while awaiting further scientific studies on how exactly it operates.
In deriving indices of food variety, decisions are required about both the categories of foods and the time over which variety in food choices is achieved.
These have been the subject of both FAO/WHO technical reports (1, 9-16) which are currently being updated, and many countries have established their own recommended requirements. The numbers vary somewhat for the different nutrients although the implications of these differences for the establishment of dietary guidelines are small. Countries that have the technical capacity may develop their own RN I, otherwise the review of existing recommendations can be used to define which are most suited to the given national reality. FBDG should be structured to enable the population to meet RNI that are critical for diet-related public health problems.
The meeting recommended that FAO/WHO undertake the review of existing reports for all relevant nutrients (since the last time this task was completed was in 1974).
Using nutrient densities to evaluate dietary quality involves expressing existing RNI values per unit of energy (usually 1000 kcal) provided by the diet. The conditions for this model are that if a diet provides for the energy needs of individuals, it will also satisfy the RNI for all essential nutrients. This approach permits the simplification of age- and sex-related RNI figures since the values differ minimally when these figures are expressed per 1000 kcal. Therefore, for the purpose of establishing dietary guidelines for the general population, precise sex- and age-specific RNI are not needed. The figures presented should provide a way to assess dietary quality.
Individuals within a family group usually form the basic unit for food consumption. Thus, if there is enough food at the family or household level, all members can consume a diet with the recommended nutrient densities and meet their specific RNI. The problem of intrafamily distribution needs to be considered, since children and women may not receive an adequate proportion of foods with higher nutrient density. This should be considered in establishing both general dietary guidelines and those specifically addressing the needs of vulnerable groups in the community.
Relevant non-energy nutrients that represent public health concerns globally are included in Annex 3, where their need is not expressed as an absolute requirement but as nutrient density per 1000 kcal. This should not be interpreted as a physiological relationship between the specific nutrients and energy requirements but as a way of defining the adequacy of a given diet to meet the needs for specific nutrients if sufficient energy is consumed.
If intake for adolescents or adults is less than 2000 kcal per day, it will be difficult to meet their vitamin and mineral needs. Thus, adequate energy intake is a prerequisite for maintaining nutrient status, and a sedentary lifestyle should be discouraged. Children below 2 years of age need a different diet and require specific dietary guidelines.
In using nutrient densities to establish dietary goals and guidelines, the quantitative and qualitative aspects of the food supply should be considered. The quantitative aspects include the estimation of the amount of nutrients and the relative proportion of nutrient in the food sources that will meet nutritional needs in practice. The qualitative aspects relate to the biological quality of the nutrients in the food source and the potential for interaction of nutrients and non-nutrient components which may enhance or inhibit the biological quality of a given source if both are ingested simultaneously. The recommended nutrient densities should refer to the total diet available and not to individual foods.
Energy RNI should be treated separately, since age-and-sex as well as activity-specific RNI are necessary to define adequacy of energy intake and to validate proposed dietary guidelines.
The starting point to devising FBDG is the relevance to a public health issue and not a gap which may exist between prevailing nutrient intake and a numerical recommended intake for a nutrient. This can be illustrated by comparing calcium intakes in the USA and China. In the USA, calcium intakes are about 600 mg/d, while in China calcium intakes are about 400 mg/d. The international range for calcium RDA is from 500 mg per day to 1200 mg per day. On that basis China has a bigger calcium nutrient intake gap than the USA. In reality, whereas osteoporosis in the elderly is a major public health issue in the USA, apparently it is not a current public health problem in China (17). Therefore, the starting point for devising FBDG should be the relevant public health issue. Even when a public health issue exists (for example, high rates of cardiovascular mortality) and a difference also exists between the numerical dietary goal (for example, 10% energy from saturated fatty acids) and prevailing intake of the target nutrient (for example, 16-18% energy from saturates as in Northern Europe), the FBDG should be based on what can be realistically achieved in the socioeconomic context rather than trying to eliminate in one step the whole difference between the desired and actual intakes.
Once the public health issue is identified, the first step should be to ascertain the extent to which it may be attributable to non-nutritional factors. Those non-nutritional factors then have to be addressed in tandem with, or even in advance of, any nutritional intervention. Specifically infection, safe food and water, physical activity or smoking may have to be addressed in order for nutrition strategies to be fully successful.
Since all FBDG are issued in a specific context, it is essential next to consider the likelihood of an intervention's success based on food-specific guidelines within the social, economic, agricultural, supply and cultural context. For example, the traditional Cretan diet contains about 50 g of olive oil per day, and some 8 million tonnes of olive oil would be needed to allow all citizens of the EU to enjoy this amount. The agricultural capacity of the EU to produce the required quantity of olive oil and the attendant impact on other aspects of agriculture production suggest that this is not feasible.
If the public health information indicates that nutrients are relevant to addressing a public health issue, a transition is needed from a nutrient orientation to focusing on food-based strategies likely to be successful. The following steps are useful in this transition. (See Figure 1 for a summary).
Step one: Consideration of associated nutrients prior to intervention
Having identified the nutrient relevant to the public health issue in question, try to determine whether any other nutrient or non-nutrient might be confounding the targeted nutrient. For example, if the objective is to increase carotenoid intake, one needs to know if the level of fat in the diet is adequate to ensure carotenoid absorption; and even though the intake of fat appears adequate for the purpose of absorbing carotenoids, are the fat and carotene usually consumed at the same time?
Step two: Identifying the foods to be included in the food-based dietary guideline to modify the intake of a specified nutrient
Several approaches should be used here, depending on available data.
Determine which foods have high levels of the target nutrient. This information may be sufficient to formulate FBDG, but one should be cautious about assuming that a food rich in a nutrient is one that is likely to make a significant contribution to the dietary intake of a given population group.
Determine the main dietary sources of the target nutrient. In some instances it will be found that the nutrient has a very wide distribution across food groups, in which case altering intake of the nutrient might be difficult. In such cases one might decide which of these equally important food groups, in terms of their supply of the targeted nutrient, should be selected to define a FBDG. The basis for such selection should include price, likely consumer acceptability, coincident benefit for other FBDG, or other factors critical to implementation. In other instances, it may be found that the nutrient is provided largely by one food source, in which case one should consider whether the intakes of non-targeted nutrients might be detrimentally effected were that food to be strongly promoted.
Where food and nutrient intake data are available, it is possible to develop FBDG based on the food consumption patterns of population subgroups that achieve a particular nutritional goal. This allows one to discriminate between subgroups with high and low intakes of a target food or nutrient. Such foods or nutrients are often not those which would be predicted on the basis of their contribution to actual average intake of the target nutrient. Thus, it is important to determine patterns of food intake not only on the basis of the population average but also for subgroups, or even individual consumers. It is possible for two sub-populations to have equal average intakes but to differ in terms of the proportion of consumers and their different intake levels.
In effect, any or all of these, and still other approaches, may be used to ensure that the foods and derived ingredients are likely to be effective and will not create confounding adverse effects.
Step three: Consideration of possible sociocultural and economic factors
Whereas steps 1 and 2 above may help us to identify, in nutritional terms, the ideal set of FBDG, many other factors may work against their being heeded and may mitigate their impact. These are considered in Chapter 4.
Step four: Possible nutritional consequences of implementing a food-based dietary guideline
If, for some public health issue, a given dietary guideline were to advise an increased intake of pulses, one should ask what the possible consequences might be. One consequence might be that pulses replace meat, with subsequent nutritional implications. Another consequence might be that pulses do not replace meat but replaced vegetables, again with subsequent nutritional consequences. Consideration of a variety of possible scenarios may help to tailor the food-based dietary guideline so as to minimize adverse effects or, more importantly, may help those charged with delivering messages about FBDG to provide a clearer message.
FBDG offer signal opportunities to alter the dietary habits of populations in a direction which reduces the risk of NCD or prevents nutritional deficiencies. The above steps and other pertinent considerations should ensure that the right foods are targeted and that no adverse effects arise.
Figure 1 - Reorienting from nutrients to foods
1. Diet, nutrition, and the prevention of chronic diseases. Report of a WHO Study Group. Geneva, World Health Organization, 1990 (WHO Technical Report Series, No. 797).
2. Ziegler RG, Subar AF, Craft NE, Ursin G, Patterson BH, Graubard Bl. Does b-Carotene explain why reduced cancer risk is associated with vegetable and fruit intake? Cancer Research 1992, 52:(Suppl):2060S-2066S.
3. Truswell AS, Irwin T, Beaton GH, Suzue R, Haenel H, Hejda S, Hou X-C, Leveille G, Morava E, Pedersen J, Stephen JML. Recommended Dietary Intakes Around the World: a report by Committee 1/5 of the International Union of Nutritional Sciences, Nutrition Abstracts and Reviews, 1983,53:939-1015.
4. Truswell AS, Irwin T, Beaton GH, Suzue R, Haenel H, Hejda S, Hou X-C, Leveille G, Morava E, Pedersen J, Stephen JML. Recommended Dietary Intakes Around the World: a report by Committee 1/5 of the International Union of Nutritional Sciences, Nutrition Abstracts and Reviews, 1983, 53: 1175-1119.
5. Seddon JM, Ajani UA, Sperduto RD, Hiller R, Blair N, Burton TC, Farber MD, Gragoudas ES, Haller J,
Miller DT, Yannuzzi LA, Willett W. Dietary Carotenoids, Vitamins A, C, and E, and Advanced Age-Related Macular Degeneration. Journal of the American Medical Association, 1994,272:1423-1420.
6. Smiciklas-Wright H, Krebs-Smith SM, Krebs-Smith J. Variety in foods. What is America eating?
Proceedings of a Symposium. Washington DC: National Academy Press, 1986.
7. Kant AK, Schatzkin A, Harris TB, Ziegler RG, Block G. Dietary diversity and subsequent mortality in the First National Health and Nutrition Examination Survey: Epidemiologic Follow-up Study. American journal of clinical nutrition, 1993, 57:434-440.
8. Hodgson JM, Hsu-Hage BH-H, Wahlquist ML. Food variety as a quantitative descriptor of food intake. Ecology of food and nutrition, 1994,32: 137148.
9. Energy and protein requirements. Report of a Joint FAO/WHO/UNU Expert Consultation. Geneva,
World Health Organization, 1985 (WHO Technical Report Series, No. 724).
10. Requirements of Vitamin A, Iron, Folate and Vitamin B12. Report of a Joint FAO/WHO Expert Consultation. FAO Rome 1988.
11. Calcium Requirements Report of a FAO/WHO Expert Committee. Geneva, World Health Organization (WHO Technical Report Series, No 230).
12. Passmore R. Nicol BM, Narayana Rao M et al. Handbook on Human Nutrition Requirements. WHO Monograph Series, No 61. WHO, Geneva, 1974.
13. Requirements of Ascorbic Acid, Vitamin D, Vitamin B12, Folate, and Iron. Report of a Joint FAO/WHO Expert Group. Geneva, World Health Organization, 1970 (WHO Technical Report Series, No 452).
14. Requirements of Vitamin A, Thiamine, Riboflavine and Niacin. Report of a Joint FAO/WHO Expert Group. Geneva, World Health Organization, 1967 (WHO Technical Report Series, No 362).
15. Trace Elements in Human Nutrition and Health. World Health Organization, Geneva 1996.
16. Fats and oils in human nutrition. Report of a Joint FAO/WHO Expert Consultation. Rome, Food and Agriculture Organization (FAO Food and Nutrition paper No. 57).
17. Avioli LV. Calcium metabolism and osteoporosis in China, Japan and Pacific Rim communities. Calcified Tissues International, 1991,48:149150.