Contents - Previous - Next

This is the old United Nations University website. Visit the new site at

Quality versus quantity of infant diets: translating research into action

Judith McGuire


This paper discusses briefly the difficulty of translating scientific research findings in nutrition into operational intervention programmes and concludes that the conceptualization, conduct, and dissemination of results of research must be oriented toward the intended uses of the results.

Individuals who work in development agencies often have difficulty using research results. It is not just a question of language or jargon. It is often difficult to see exactly why some findings would make any difference to nutrition programmes. In the typical project-development process, the scientific justification for choosing an intervention is a minor concern to a task manager. He or she must be more concerned with institutional capacity, financing, training, supervision, and, of course, getting the loan approved on time. In other words, the process is more important than the content, within an acceptable range. Nonetheless, there are ways to improve the likelihood that scientific research will influence policies and programmes.


Clinical studies have demonstrated that dietary quality can have significant effects on infant nutrition. Yet we know very little about the global prevalence of deficiencies, even for vitamin A, iodine, and iron, which are the best documented micronutrient deficiencies.

National food consumption surveys have generally been analysed only for calories and perhaps protein. Statisticians and economists need guidance on how to measure nutritional quality and express it. It is not reasonable or valid to express the results of a survey based on household food purchases in terms of availability of a dozen or so nutrients. The results are also impossible to interpret if some nutrients are high and others are low. A simple way of measuring and expressing dietary quality is required to assure that it will be taken into account in the same fashion as quantity (which is measured as calories by and large). The methodological problems in measuring dietary quality have to be solved before better data can be collected. Those data will be necessary to convince policy makers that poor dietary quality is a problem.

An associated problem is the growing availability of anthropometric data. Without surveys of nutrition status with respect to specific nutrients, policy makers are led to believe that growth is the only indicator they should use, and that growth is affected only by dietary quantity (or calories). Efforts should be made to link anthropometric surveys with assessments of micronutrient status and to increase communications between nutritionists and policy makers. It is not even known what proportion of the population suffers from deficiencies of single or multiple micronutrients. Nonetheless, for many micronutrient and nutritional quality problems we have no good survey assessment tool. Food consumption is not as good an indicator as body stores or serum levels, but data on the former are a lot easier to collect. The population prevalence of deficiencies cannot be assessed by evaluating responsiveness to supplementation as is recommended for zinc. More effort has to be put into assessment methods if nutritional quality is to compete with quantity for policy and programme priority.

Dietary quality and development

Little documentation is available on the relationship between economic development (or household income) and dietary quality. One might assume that quality improves as household income increases. Well-designed studies have shown that significant increases in household income may not result in increased calorie intake [1; 2]. One hypothesis is that, as income increases, families buy more quality than quantity in their food. That hypothesis remains unproved. If the relationship between poverty and nutrition quality were strong, arguments to development institutions to address the latter would be much more persuasive.

The two-way linkage between dietary quality and development must be clarified further. It would be extremely useful to know, for instance, how much improvement in dietary quality is necessary to have a perceptible effect on economic development, the mechanisms through which it works, and the costs of effecting that magnitude of change. On the other hand, we have to know how much improvement in dietary quality we can expect over the normal course of socio-economic development with changes in the source, composition, and processing level of foodstuffs. Once again, national data are not available even for a handful of countries, and, if these data exist, they have not been systematically compiled and compared. Such data are necessary to make a convincing argument for investments to improve dietary quality.

The interactions among nutrients with respect to functional outcomes are also largely unknown but have major implications for programmes. A zinc- and iron-deficient child may be twice as likely to be disabled or sick as a child with only one deficiency, but perhaps addressing one deficiency is good enough. Another issue that comes up with micronutrients is the relative contribution of protein-calorie and micronutrient deficiency to growth failure and dysfunction. Sorting out the causes is vital to choosing the correct intervention. Another issue is that of cutoffs. Does a continuous relationship exist between deficiency and dysfunction, or does the dysfunction occur only above or below a certain threshold? Having the answer to this would help us determine whether targeting is reasonable and cost-effective.


Even if the prevalence and functional effects of sing]e and multiple micronutrient deficiencies are known, one needs to know the aetiology of the problem to design effective solutions. We know that micronutrient malnutrition results from food factors, health factors, nutrition behaviours, and intergenerational transmission due to maternal malnutrition. A task manager must know what bang to expect for the project buck: whether eliminating hookworm has greater impact than food fortification, how much behaviour change can compensate for lack of purchasing power, what differential in outcomes can be expected from monofocal versus comprehensive approaches, and whether to target the mother or the child or both.

Intervention options

Because the concept of dietary quality is fairly new, the average policy maker may not know the options for addressing the problem. Uncommon interventions may be called for, such as food technology, genetic improvement of food sources, fortification. Policy makers have just become used to targeted feeding programmes, growth monitoring, and nutrition education. They have to know the complexity and costs of alternative interventions.


Tied in with cost-effectiveness is the targeting question. There are real administrative costs to delivering a service only to those who are at highest risk. Policy makers must have advice on the relative costs and benefits of prevention versus treatment of malnutrition. They have to know what groups are at highest risk. The administrative costs of targeting may outweigh the increased cost-effectiveness, yet negative health consequences may result if the non-needy receive a mass-dose supplement, for instance. Researchers must think about these alpha and beta errors of targeting when they are advocating for action.

What researchers can do

Although this discussion is not exhaustive, it is meant to suggest a number a practical problems one is likely to encounter when trying to translate the research discussed in the preceding papers into benefit for malnourished people. If researchers do not think through these implications, either their ideas will remain interesting academic hypotheses, or programme designers will make a lot of shaky assumptions and undertake highly speculative interventions. The success rate of such interventions is likely to be low, and nutritionists will lose credibility.

To improve the likelihood that any research will actually make a difference to policies and programmes, I suggest that the following actions should be taken:

- Validate findings in a community setting. Only by examining the social ecology of a problem does one understand the true causality and feasibility of interventions.

- Compare data across regions and social and racial groups to provide some estimate of generalizability and some estimate of the global importance of the problem.

- Think of the meaning of the research within the context of national economics, household structure, economics, lifestyles, and political environment. Policy makers must be motivated to adopt a new idea, and these contextual factors can be used for advocacy.

- Disseminate the findings widely in the community and country where the research was done. Scientific colonialism is not only politically unpopular; it also weakens local experts' credibility at the same time that their participation is needed.

- Make recommendations for practical solutions and assist in implementing those recommendations.

- Make reasonable leaps of faith. If the weight of the evidence suggests an important relationship, that is sufficient reason to advocate action.

- Obtain experience in implementation. Approach the appropriate government bodies, donors, and international organizations to help design, implement, and evaluate intervention programmes. The obvious solutions may not be feasible or sustainable. Only by experiencing the difficulty and complexity of changing people's behaviour can researchers appreciate the difficulty of translating research into action.

Some exciting findings on the dietary quality of foods for infants are presented in this issue. Those that stand up to closer scrutiny and are corroborated by other research must be translated into action, even if it is only a pilot test. Most people working in international nutrition are in the field because they have a personal commitment to addressing global malnutrition. The challenge is to translate research into action as effectively and efficiently as possible.


1. Bouis HE, Haddad LJ. Effects of agricultural commercialization on land tenure, household resource allocation, and nutrition in the Philippines. IFPRI Report 79. Washington, DC: International Food Policy Research Institute, 1990.

2. Kennedy ET, Cogill B. Income and nutritional effects of the commercialization of agriculture in southwestern Kenya. IFPRI Report 63. Washington, DC: International Food Policy Research Institute, 1987.

Nutrition and the elderly

Editorial introduction

The Committee on Nutrition and Aging of the International Union of Nutritional Sciences (IUNS), in cooperation with the World Health Organization and the United Nations University, has organized comparative studies of food habits and health in later life in a number of countries. These studies use a combination of quantitative and qualitative methodologies to examine relationships between diet and such factors as social and economic status, beliefs about foods, and actual food practices. The following three papers, based on studies in Sweden, Greece, and Australia, were presented at the UNU-UNICEF International Conference on Rapid Assessment Methodologies for Planning and Evaluation of Health-Related Programmes (RAP) held in Washington, D.C., in November 1990.

The first of these papers describes the dietary habits of elderly people in Sweden and suggests the importance of supplementing such survey data with qualitative information on social, psychological, and medical circumstances in order to explain the basis for the findings. The second presents similar information for elderly Greeks living in Greece and adds qualitative information on beliefs and practices obtained by adapting rapid assessment procedures (RAP) developed for programmes of nutrition and primary health care. The paper illustrates the value of complementary qualitative data obtained through open-ended and informal interviews and observations. The third paper demonstrates the advantages of combining quantitative surveys with RAP-type methodology in studying the dietary patterns and health of elderly Australian Aboriginals.

Guidelines for the application of RAP procedures to the studies of the elderly are in preparation. In the meantime those published in English, Spanish, French, and Portuguese for programmes of nutrition and primary health care and in English for the study of AIDS-related behaviour continue to be available at nominal cost from the UCLA Latin American Center, University of California, Los Angeles, CA 900241447, USA. A 16-minute video describing RAP, with either an English or a Spanish sound track, is available from the UNU Food and Nutrition Programme office at the Harvard Center for Population and Development Studies, 9 Bow St., Cambridge, MA 02138, USA.

- N.S.S.

Qualitative factors as possible indicators of inadequate dietary habits in the elderly: Findings from recent population studies

Bertil Steen, Anna Sjögren, Ulla Sonn, and Tor Österberg


Two studies of elderly populations in Sweden which included surveys of their dietary intake are discussed: a study of 621 men in Malmö who were born in 1914, and a gerontological and geriatric population study of approximately 1,000 men and women in Göteborg who were 70 years old at the start of the study in 1971 and have been reexamined several times in subsequent years. The average dietary intakes were high, but there was considerable variation. This paper points out that quantitative data such as the intake of energy and nutrients alone are not adequate for identifying people at risk. Significant relations exist between such quantitative factors and qualitative aspects of people's life circumstances. Qualitative data are required in planning interventions for the elderly.

Knowledge of the nutritional situation of a group of individuals is necessary when evaluating their present and future state of health, in planning social and health activities, and in tracing population subgroups in the risk zone for malnutrition. The increasing number of elderly people in industrial as well as developing countries makes the elderly population an especially interesting one in this respect. The population pyramid of the 1890s is no longer a pyramid in countries like Sweden but a broad pillar. Life expectancy at birth is increasing, being at the moment 80.6 years in women and 74.8 years in men. In the 1960s for women and in the 1980s for men something historically new happened in Sweden: more than 50% of the increase of life expectancy at birth could be explained by the life expectancy at age 65. In the year 2000 the average 65-year-old woman can expect to reach 86.4 years of age! At present only 11% of women in Sweden die before age 65, compared to 20% of men. Death in early life or in middle age is thus uncommon. These changes are happening or will happen in most countries - both developed and developing - although at very different levels and rates of increase.

Possible changed nutritional demands in old age due to altered physical activity, the ageing processes per se, and the frequent occurrence of disease make these demographic facts even more pertinent.

The nutritional status of a population cannot be described by dietary surveys alone. There is also a need for social, psychological, and medical data, along with assessment of clinical signs of malnutrition, anthropology, biochemical analysis, and biophysical tests.

During the last decade numerous dietary surveys in elderly populations have been performed in various countries. In Sweden, the dietary habits of elderly people are well known from epidemiological studies in the northern, western, and southern parts of the country [1-4]. One starting point for the dietary parts of these studies is that quantitative data have to be combined with qualitative ones in order to obtain not only average intake values and variation but also explanatory qualitative data to help in intervention procedures and public health programmes. Two of these population studies will be discussed in this paper - a study of men born in 1914 conducted in Malmö [3], and gerontological and geriatric population studies in Göteborg [1]; results from the dietary parts of these studies were published during the 1980s [5-7].

The Malmö study of men born in 1914 is a prospective study, which began in 1969 with the primary purpose of studying risk factors for cardiovascular disease. Malmö is a city in southern Sweden with about 230,000 inhabitants. The study population comprised all men in Malmö who were born in even months in 1914 - a total of 621 men.

The gerontological and geriatric population studies in Göteborg started in 1971 with a cohort of about 1,000 men and women 70 years of age. This cohort has been re-examined several times, most recently at age 88. Subsequently other cohorts of similar size of other generations have also been studied at various ages.

The results of the sampling procedure can be said to include qualitative factors of importance to the results of the dietary substudies within the study. Such factors include the non-response and participation rates. In the Malmö study of 68-year-old men, more non-responders lived alone, had a worse self-perceived health, were more often hospitalized during a certain period before the study, and were more often alcohol-dependent [3]. Obviously, such qualitative factors could distort the representativity of the results from a study if the non-response rate were high.

A large range in the participation rates in different city areas was noted; in Malmö the rates varied from 27% to 100%. Areas of low participation were characterized by more single-person households, more households on social welfare, and higher rates of unemployment [3].

Generally, in the two studies the average dietary intake values were high, but there was considerable variation. This does not help much in identifying people at risk; we also need to include qualitative data in planning interventions.

In the Malmö study of 68-year-old men, the proportion with "inadequate dietary habits" - using a specific and arbitrary definition [8] - differed in groups of high, medium, and low social status (12%, 18%, and 29% respectively) and in corresponding groups with different degrees of social participation. The same was true regarding physical activity (10%, 21%, and 30% respectively). The proportion of men with inadequate dietary habits was higher among those with low or high alcohol consumption than in the in-between group [7].

When the same definition of inadequate dietary habits was used in the gerontological and geriatric population studies in Göteborg for the subjects at age 70, some qualitative factors showed a clear relation to the dietary habits. The proportion of men and women with inadequate dietary habits was significantly higher among people living in apartments than among those living in single-family houses. Among men, the proportion was significantly higher among those who were not able to use their ovens than among those who could. Forty-seven per cent of the men and 58% of the women with walking difficulties were considered to have inadequate dietary habits by the definition used.

In the same study the proportion of subjects at age 76 with a low intake of thiamine, according to current standards, was considerably higher among unmarried people and blue collar workers. For people living together, the proportion was higher among those who had been together for less than a year.

Also at age 76, the proportions of subjects with low intakes of thiamine and protein were considered in relation to activities of daily life function. Significant differences existed between those who needed and those who did not need different kinds of support for such activities. For example, in the group who needed personal help, 34% of the subjects had a low intake of thiamine, compared to 11% of those who did not need help. The same tendency could be seen with respect to low protein intakes.

In the Göteborg studies feelings of loneliness were examined in relation to disease and social factors. Loneliness was a problem to 24% of the women and 12% of the men. Fatigue, negative self-assessment of health, high consumption of hypnotics and sedatives, and high frequency of seeking medical advice were all significantly more common in the lonely group [9].

An important phenomenon to be taken into consideration is the dynamic changes in the medical, psychological, and socio-economic circumstances under which the elderly live. Thus there are cohort or "generation" differences - which are significant even in short perspectives such as a decade [10; 11]. Important information from recent population studies indicates that functional variation between individuals is very large at the same chronological age, and.that chronological age seems, in fact, to become increasingly less valuable as an information source about an individual's functional ability with advancing age. For example, in the studies in Göteborg of cohort differences between 70-yearolds in 1971-1972 and in 1981-1982 respectively, it was noted that the later cohort felt healthier (both sexes), smoked less (males), and had their own teeth to a markedly higher degree [12].

Health and vitality seem to be increasing among the younger elderly people in countries like Sweden [10; 11]. On the other hand, increasingly more individuals will for that same reason reach ages where a so-called uncompensated ageing is common, and the number of cases of certain diseases will, therefore, increase with age.


Information about dietary habits should be collected from broad epidemiological and multidisciplinary studies that examine psychological, social, and medical as well as nutritional data. This paper has tried to indicate that it is obvious that the analysis of quantitative data alone, such as intake of energy and nutrients, is not sufficient to describe and explain the dietary situation of an elderly population. Significant relations exist between qualitative and quantitative factors, and we suggest that qualitative factors could be used as indicators of adequate dietary habits in nutritional studies in the elderly and in planning intervention and public health programmes.


1. Rinder L, Roupe S, Steen B, Svanborg A. 70-year-old people in Gothenburg - a population study in an industrialized Swedish city: 1. General presentation of the study. Acta Med Scand 1975;198:397-407.

2. Österlind P-O, Löfgren A-C, Sandman P-O, Steen B, Winblad B. Health disorders and drug consumption in an elderly population in northern Sweden. Gerontology 1984;32:52-59.

3. Janzon L, Hanson BS, Isacsson S-O, Lindell S-E, Steen B. Factors influencing participation in health surveys: results from the prospective population study "Men born in 1914" in Malmö, Sweden. J Epidem Comm Health 1986;40: 174-77.

4. Steen B, Björn A-L, Küller R et al. Age retirement in women: I. General presentation. Compr Gerontol 1988;A2:7177.

5. Steen B, Isaksson B, Svanborg A. Intake of energy and nutrients and meal habits in 70-year-old males and females in Gothenburg, Sweden: a population study. Acta Med Scand 1977;suppl.611:39-86.

6. Lundgren BK, Steen B, Isaksson B. Dietary habits in 70- and 75-year-old males and females: longitudinal and cohort data from a population study. Naringsforskning 1987;31:53-56.

7. Hanson BS, Mattisson I, Steen B. Dietary intake and psychosocial factors in 68-year-old men: a population study. Compr Gerontol 1987;BI:62-67.

8. Österbert T, Steen B. Relation between dental state and dietary intake in 70-year-old males and females in Göteborg, Sweden: a population study. J Oral Rehab 1982; 9:509-21.

9. Berg S, Mellström D, Persson G, Svanborg A. Loneliness in the Swedish aged. J Gerontol 1981;36:342-49.

10. Berg S. Psychological functioning in 70- and 75-year-old people: a study in an industrialized city. Acta Psychiatr Scand 1980;62(suppl.288) :24-25.

11. Steen B. The elderly in the changing world. In: Barac B, Lechner H, eds. Neurologija 1990;39(suppl.2):19-20.

12. Steen B. Environment and the elderly: the European perspective. Austral J Ageing (in press).

Contents - Previous - Next