Jump to ContentJump to Main Navigation
Nutrition and Poverty$

S. R. Osmani

Print publication date: 1993

Print ISBN-13: 9780198283966

Published to Oxford Scholarship Online: October 2011

DOI: 10.1093/acprof:oso/9780198283966.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2019. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use.  Subscriber: null; date: 26 August 2019

Undernutrition: Measurement and Implications

Undernutrition: Measurement and Implications

(p.17) 2 Undernutrition: Measurement and Implications
Nutrition and Poverty

C. Gopalan

Oxford University Press

Abstract and Keywords

This chapter examines the measurement of undernutrition in population groups. It describes the biological processes involved in the genesis and development of undernutrition and reviews the conceptual issues in the two major approaches to the measurement of undernutrition. It discusses some of the practical problems in measuring undernutrition and argues that the measurement of undernutrition in large populations should be based on dietary surveys supplemented by weight-for-age measurements of children under the age of five.

Keywords:   undernutrition, biological processes, population survey, dietary survey, nutrition evaluation

2.1 Introduction

In this chapter it is the measurement of undernutrition in population groups rather than individuals that is under consideration. It is important to make this distinction at the very outset, because yardsticks and procedures that may be adequate for evaluation of the nutritional status of whole communities may not be suitable for the assessment of the nutritional status of a given individual. Individual genetic variations with respect to requirements of nutrients and response to their deprivations could get largely neutralized when large population groups in nearly similar socioeconomic and environmental status are considered.

Economists and planners, understandably, look for tidy methods of quantifying undernutrition in population groups. Biologists, however, would readily recognize the inherent limitations and pitfalls of exercises that seek to ‘measure’ undernutrition with mathematical precision. These limitations stem from the very nature of the undernutrition process—the multiplicity of interacting, often mutually reinforcing, factors involved in its causation; its evolution, often so insidious that it is hard to decide where normalcy has ended and subnormality (or abnormality) has set in; and its multiple clinical dimensions. It is not the purpose of this paper to discuss these limitations in detail, but a broad appreciation of them is essential for any meaningful discussion of the problem of measuring undernutrition.

The paper is organized as follows. Section 2.2 describes the biological processes involved in the genesis and development of undernutrition. Section 2.3 reviews the conceptual issues involved in the two major approaches to the measurement of undernutrition. These issues have been the subject of intense controversy in recent times, leading to sharply contrasting views on how undernutrition should be measured. These controversies are reviewed briefly, and my own views on the subject are presented. Section 2.4 is concerned with some of the practical problems of measuring undernutrition, bearing in mind the limitations of data on the one hand and the multi-faceted nature of undernutrition on the other. Here I argue that the measurement of undernutrition in large populations (p.18) should be based on dietary surveys supplemented by weight-for-age measurements of children under the age of five.

My emphasis on the weight-for-age criterion is in sharp opposition to a recent trend which favours the alternative weight-for-height criterion. As I explain, the choice between these alternative criteria has a lot to do with the views one holds on the conceptual issues discussed in Section 2.3. It is, however, important to point out that my choice of the weight-for-age criterion for the under-fives in no way implies a negative judgement on the usefulness of height-for-age measurements: it is just that there are certain special problems in the height-for-age measurement of very young children. However, as is argued in Section 2.5, monitoring the height of older children is an exceedingly useful exercise; indeed, it is perhaps the best possible way of assessing the long-term changes in the nutritional status of a community. Finally, my major conclusions are summarized in Section 2.6.

2.2 The Biology of Undernutrition

The aetiology

Undernutrition, widely prevalent among socially and economically deprived population groups around the world, is associated with a cluster of related, often coexistent, factors which together constitute what may be termed the ‘poverty syndrome’, the major attributes of which are (1) income levels that are inadequate to meet basic needs of food, clothing, and shelter; (2) diets that are quantitatively and often qualitatively deficient; (3) poor environment, poor access to safe water, and poor sanitation; (4) poor access to health care; and (5) large family size and high levels of illiteracy—especially female illiteracy. Among most undernourished population groups, these factors often tend to coexist, though their relative severity and extent may vary in different locations. In the evalution of undernutrition, and indeed in its progression and perpetuation, these factors often act synergistically.

The effects of all these factors, both socioeconomic and environmental, are, however, ultimately mediated through a final common pathway. The ultimate determinant of nutritional status is the availability at the cellular level—in adequate amounts, in proper combinations, and at appropriate times—of all the essential nutrients required for normal growth, development, maintenance, repair, and functioning of the organism. This, in turn, is determined by two broad sets of factors: (1) the diet, which must provide adequate amounts of the essential nutrients, and (2) factors that condition the requirement, absorption, assimilation, and utilization of the nutrients of the diet. These latter include the activity level and environmental factors, particularly infections and stress situations.

(p.19) 2.2.1 Correlation between dietary intake and nutritional status

The correlation between the levels of dietary inadequacy prevailing in households and communities on the one hand and the degree of severity of undernutrition (as assessed by anthropometric criteria and clinical signs) obtaining among them on the other is not always strict. Three major reasons for this may be mentioned.

1 The severity of effects of primary dietary inadequacy in a population can be aggravated by superadded conditioning factors, such as infections and parasitic diseases, the extent of such aggravation being determined by the nature of such infections, their duration, frequency, and severity, and the promptness and efficiency with which they are prevented and treated in the community. Infections can increase requirements of nutrients and inhibit their absorption and assimilation.

For example, in communities subject to the same order of dietary deficiency right through the year, clinical manifestation of undernutrition could be more pronounced in seasons characterized by a high prevalence of infections than at other times. Thus, in Coonoor in India, in poor communities subject to the same monotonous dietary deficiency throughout the year, the peak incidence of ‘kwashiorkor’ (a disease caused by severe protein-calorie deficiency) in children in successive years was noticed in May–June, following the peak incidence of diarrhoeal diseases in the ‘fly season’ of April–May (Gopalan 1955). In parts of Kerala in India, nearly three decades ago, when health services were less adequate than at present, the peak prevalence of ‘kwashiorkor’ was noticed in the weeks following the monsoon when, again, diarrhoeal diseases attained their peak.

While the level of dietary inadequacy is undoubtedly the dominant determinant of undernutrition, the level of primary health care in the community can significantly modify the severity of its clinical manifestations.

2 Where diets of entire households rather than of individuals within the family are being used as yardsticks in the assessment of community nutritional status, differences in the nature of intrafamilial distribution of food, and in particular in infant feeding and child-rearing practices, between the families and between communities can result in important differences with respect to nutritional status (especially of children) between households, and between communities with nearly similar overall levels of dietary inadequacy.

Differences with respect to nutritional status of infants and young children between households with nearly similar dietary and socioeconomic status can arise from differences with respect to duration and intensity of breast-feeding, the time of introduction of supplements, and the nature and amount of such supplements. Relatively small proportions (p.20) of the overall family diet can make a significant difference to the level of adequacy or inadequacy of the diet of the preschool child. The level of female literacy in the household is often a major determinant of child-rearing practices and, therefore, of the level of child nutrition in poor households.

3 Furthermore, except in acute famine situations, the current nutritional status of a community is often a reflection of its erstwhile rather than (necessarily) its present dietary status. There is a variable time-lag between dietary deprivation and the onset of clinical undernutrition. This consideration, however, may not matter in the case of communities wherein no significant or striking changes in dietaries have taken place, and where seasonal fluctuations in dietaries are not marked. Current dietaries may then well reflect the situation responsible for the prevailing nutritional state.

The time-lag between the onset of nutrient deprivation and the appearance of clinical (or functional) manifestation can vary, depending on the nutrient and the clinical sign. Thus, for example, it could take much longer for eye lesions to appear following a vitamin A deprivation than for growth retardation to occur following calorie-protein undernutrition. In the case of growth retardation consequent on calorie-protein undernutrition itself, retardation in linear growth (stunting) is generally the outcome of a more longstanding dietary deprivation than retardation in bodyweight increment (wasting).

These considerations will underscore the limitations with respect to the measurement of nutritional status of communities on the basis of the level of dietary inadequacy alone, and will highlight the need for additional yardsticks. This is not to minimize the importance of diet surveys in the assessment of nutritional status of population groups, but only to explain some of the seeming incongruities such as the lack of strict parallelism between dietary intakes and nutritional status. It will also explain the reasons why nutrition scientists rely not only on diet survey data but also on nutrition surveys (actual examination of human subjects—both adults and children) for the assessment of nutritional status of population groups.

Growth retardation

In children of poor communities, habitually subsisting on inadequate diets, there is a continuous and insidious transition from the stage of normalcy usually obtaining up to about the fourth or sixth month (many infants being small-for-date may never start from normalcy) to that of fully fledged, clinically manifest undernutrition which generally supervenes before the third year. The speed of this downward slide from normalcy to fully fledged disease will depend on the extent of the dietary inadequacy, its duration, and the presence or (p.21) absence of superadded aggravating factors such as infection. In poor communities we may expect to see children in different stages of this transition. Not all will go through the entire transition: the downward slide may be arrested at different stages, or it may be so slow that the child may manage to cross the critical age period of four to five years before the ‘end-point’ is reached. It is necessary to emphasize that, unlike many infectious diseases, in the case of undernutrition there is no point of striking or dramatic onset and no easily (visually) discernible dividing line between normalcy and the commencing of ‘disease’. In children whose growth is carefully monitored, a faltering in the growth rate and the point at which the growth curve begins to flatten and deviate away from the normal standard could provide the earliest indication of undernutrition; however, few children in poor communities of the developing world enjoy the benefit of such close and careful growth-monitoring. Biochemical tests could reveal sub-clinical undernutrition, but these are hardly feasible in large-scale community surveys.

Retardation of growth and the downward deviation from normalcy becomes progressively more pronounced with the passage of time, and children pass insidiously from the so-called ‘mild’ to the ‘severe’ grades of growth retardation. A considerable proportion of children presently in the ‘mild’ grades of growth retardation are potential candidates for the ‘moderate’ and ‘severe’ grades; those presently in the ‘severe’ grades were probably in the ‘mild’ and ‘moderate’ categories a few weeks or months earlier. A fortunate small proportion may even reverse their direction.

In order to arrive at a given level of growth retardation, not all children need follow the same route in the growth chart. The shape of the growth curve could vary. The speed and intensity of growth retardation and the consequent duration over which a given order of growth retardation results will differ depending on the nature and extent of dietary inadequacy and of superadded infections. Under these circumstances, as important as the child's current position in the growth chart will be the route which that child took in order to arrive at the point—whether the child is the victim of an acute fairly severe deficiency over a short duration, or of a chronic less severe deficiency spread out over a longer period. For the same low weight-for-age, the child in the latter category could be more ‘stunted’ (less height for age) than the one in the first, and might require a much longer duration of more intensive nutritional rehabilitation for recovery. Quantification of undernutrition purely on the basis of degree of deficit of weight-for-age thus has its complexities and limitations. These have been discussed in detail in an earlier communication (Gopalan 1984).

Multiple nutrient deficiencies

Children in poor communities suffer not merely from calorie deficiency but from other nutrient deficiencies as (p.22) well. Thus, Indian children in poor rural communities often suffer from moderate and severe iron deficiency, anaemia (63 per cent of children below three years belonging to poor rural communities were found to suffer from such anaemia, according to one study of the National Institute of Nutrition), vitamin A deficiency, and less frequently from deficiencies of vitamins of the Β group. Iodine deficiency resulting in endemic goitre is a massive problem of a special kind, and may be treated as a separate category; it lends itself to a simple technological solution capable of successful implementation even within the prevailing context of poverty.

The severity of deficiency of the different nutrients does not necessarily run parallel, possibly because of differences in the composition of diets and in the efficiency of absorption of different nutrients. The severity of iron deficiency or vitamin A deficiency may show a much lower positive correlation with the degree of weight deficit than the severity of calorie deficiency. Under the circumstances, different combinations of multiple nutrient deficiencies of varying orders of severity are seen in poor children. To quantify undernutrition under these circumstances, we need to be able to give a ‘value’ to different specific nutritional deficiencies in the total composite of undernutrition.

A recognition of these complexities inherent in the biology of the process of undernutrition will help us understand the problems and difficulties involved in the measurement of undernutrition in a community.

2.2. 2 Practical approaches to measurement

Despite these difficulties and limitations, a fairly reliable estimate of the quantum of undernutrition in a community may be made through two approaches which are practicable under the real-life conditions obtaining in the field, and which will largely serve the needs of the public health scientist and developmental economist:

  1. 1 a survey of the diets of representative households (supported by a survey of diets of individual members of the family in a sub-sample of households) in a community, in order to derive information on nutrient intake—especially the calorie intake;

  2. 2 anthropometric and clinical examination of children—especially the under-fives (who constitute the ‘most sensitive’ segment of the population from the point of view of nutritional vulnerability).

In the conventional procedure employed by nutrition scientists for the assessment of nutritional status of communities, these two approaches are combined with a broad survey of the socioeconomic and environmental status of the community, which not only could facilitate the (p.23) interpretation of data but also could provide valuable practical leads for combating undernutrition in the community.

The approaches described above have been widely used, and, subject to their inherent limitations, they are valid. The major controversies with respect to measurement of undernutrition pertain not so much to the choice of the above two approaches, as to the interpretation and evaluation of the data derived from them.

2.3 Some Basic Issues

In any discussion about the measurement of undernutrition, there are some basic issues which need to be considered and some crucial questions which need to be answered.

If we are going to use the two practical approaches described above, the following questions arise: What are the normal standards against which the prevailing level of calorie intake and the observed growth performance should be compared in order to determine adequacy or otherwise? Are the widely used standards that are recommended by international agencies and adopted with slight variations by most countries valid for both the rich and the poor? More specifically, with respect to calorie intake, is it necessary to use the recommended mean requirement level (M), or will it be more appropriate, especially in the case of poor populations, to use a level equivalent to the mean minus two standard deviations (M – 2 SD) as the standard yardstick? With respect to growth, how appropriate are the ‘international standards’ for developing countries? Even if we accept the position that the genetic potential for growth between populations is nearly similar, can we not accept some levels of growth retardation (as revealed by comparisons with international or the ‘best indigenous’ standard) as acceptable for poor children consistent with their ‘economy and ecology’ Should small body size cause any concern, and is it of any functional significance? Does ‘stunting’, so widely seen in developing countries, really matter?

We may consider some recent hypotheses which touch on these questions.

Sukhatme's hypothesis

With respect to calorie intake, a sharp debate as to whether, for the purpose of assessing the adequacy of a given level of calorie intake, the recommended mean energy requirement level (M) need be used as the standard for comparison, or whether a level corresponding to the mean minus two standard deviations (M – 2 SD) would suffice has been ongoing for some time. This debate was touched off largely by the postulate of Sukhatme (1978, 1981a) that a human subject can permanently ‘adapt’ himself to a low calorie intake level representing the lowest limit of his (p.24) ‘intra-individual variation’ and that, therefore, Μ – 2 SD rather than Μ would be the appropriate yardstick for comparison. Sukhatme's postulate has been dealt with in earlier publications (Gopalan 1983 a), and it is therefore proposed to touch on it only briefly here.

Sukhatme's contention that the energy requirement of normal individuals is not static and that there is intra-individual variation in calorie intake may not be disputed. In fact, it is to be expected that calorie intake of individuals not subject to socioeconomic constraint will show daily variation depending on the level of activity and the presence or absence of stress.

What is unacceptable, however, is the second postulate of Sukhatme, that subjects permanently obliged to subsist on calorie intakes representing the lowest levels of what he terms ‘intra-individual variation’ (equivalent to recommended mean level minus the standard deviations) can permanently adapt their requirement to this low intake without any functional impairment. In short, Sukhatme suggests that the pendulum of daily calorie intake in an individual, which normally oscillates between two points on either side of the mean, can be safely and permanently arrested at the lowest end of its oscillation. If intra-individual variation in energy intake is a physiological mechanism providing for daily variation in energy requirements, Sukhatme's postulate would imply that the human organism could do without this physiological adjustment and that that requirement will somehow get adjusted to a lower level in keeping with the lowered intake. A healthy subject responds to alterations in energy intake by burning body fat when dietary energy is deficient or by storing body fat when dietary energy is in excess, resulting in a continuous process of breakdown and synthesis of body energy reserves. Individuals subsisting permanently on low-energy intakes have no scope for this and lose the advantage of an important regulatory mechanism. We have no evidence that fluctuations in efficiency of energy utilization in the absence of variations in levels of activity or stress contribute significantly to prevailing intra-individual variation even over a short period. We have no evidence to believe that populations engaged in their expected levels of occupational activity and obliged permanently to subsist at nearly 70 per cent of the recommended mean energy intake can ‘adapt’ themselves to such a low level of calorie intake without suffering loss of body weight and consequent impairment of function.

Seckler's and Payne's hypotheses

Seckler (1982) had gone so far as to suggest that ‘smallness’ is an appropriate and welcome attribute of poor people, consistent with their good health. He advised Indian nutrition scientists not only not to use ‘international standards’ of growth (as these would yield an ‘overestimation’ of undernutrition) but also not to use even the ‘best indigenous standard’ of the Indian high socioeconomic (p.25) groups, because even these will be ‘abnormally large’ for the majority of Indians who are poor. This subject has been dealt with by the author in an earlier publication (Gopalan 1983 b).

More recently, Payne has argued that, even if children of developing countries have the same genetic potential for growth as those of the more fortunate countries of Europe and the USA, they will settle for a lower level of growth in keeping with their ‘economy and ecology’ (Pacey and Payne 1985). It is not surprising that this plea for acquiescence in growth retardation has been sharply rejected as an exercise in ‘perpetuation of undernutrition’ (Jaya Rao 1986).

The three hypotheses referred to above have one thing in common: they have all relied on the body's ability to permanently ‘adapt’ itself to environmental and dietary stress without any detriment whatsoever to functional competence. Unfortunately, ‘adaptation’ has apparently been loosely interpreted in the debate to signify an acceptable state of normalcy instead of being viewed as no more than a ‘strategic retreat from normalcy’ on the part of the organism (‘a contraction of its metabolic frontiers’) in order to face the stress and escape with minimal permanent damage to its vital tissues. ‘Adaptation’ thus represents a state of siege. A population that is permanently reduced to this state cannot be normal. I shall discuss this central issue at some length in the next section and shall try to highlight the ‘cost’ and functional implications involved in the so-called ‘adaptation’ process.

2.3.1 ‘Adaptation’ meaning and implications

When an organism is subject to the stress of dietary inadequacy, it responds to the stress in a number of ways in order to minimize permanent tissue damage. A reduction of physical activity to conserve energy and a retardation of growth to minimize nutrient requirement are well-known responses. However, these responses are not without their inevitable functional consequences and costs. Individuals responding to stress in this way should not be considered normal and indeed are not normal, as several functional studies have shown. They generally function at a substandard level.

Thus, an adult can (within limits) successfully ‘adapt’ himself to low calorie intake through a corresponding reduction in work output to reduce energy expenditure. Such adaptation will, however, result in limiting his productivity and earning capacity and thus could only serve to perpetuate his poverty. Payne would perhaps argue that the individual is ‘adapted’ to function as well as he needs to in his ‘economy and ecology’, meaning his poverty situation, in which he is often either unemployed or underemployed. This would imply that we accept a situation in which (p.26) the economic status of a poor country decides the level of development of its human resources, and that efforts at improvement of the quality of human resources of a country could wait until economic improvement has been registered. This would be contrary to the present strategies, whereby vigorous efforts at promoting the quality of human resources go hand in hand with efforts at economic improvement.

To take another example, children can adapt to energy deficiency by reducing play and other physical activities, but such restriction could impair their mental and physical development. The fascinating work of Torun et al. (1975) has in fact shown the importance of physical activity in children in promoting linear growth and ensuring an efficient pattern of energy utilization. A restriction of physical activity in children can, by reducing opportunities for stimulation and learning experiences, retard mental development as well. Adjustment to low energy inputs through restriction of necessary physical activity cannot, therefore, be considered an acceptable form of adaptation.

Rutishauser and Whitehead (1972) observed that the level of physical activity of Ugandan children was less than that of European counterparts of comparable age and that their caloric intake was also correspondingly low (80 kcal per kilogram of body weight as against 100 kcal for European children). We should avoid drawing the conclusion that the low physical activity of the African children is perhaps a ‘cultural attribute’ and that the low energy intake is the result rather than the cause of the decreased activity. We may safely predict that, if the European children were to subsist on the habitual diets of Ugandan children, their activity pattern would be no different. Several ‘natural’ and controlled studies have clearly demonstrated that the response of human beings to calorie undernutrition is identical, irrespective of their race and nationality. The picture of semi-starvation was the same in Belsen and Bengal, in Madras and Minnesota. There is no scientific evidence to show that people of different cultures will show different physiological responses when faced with the same order of restriction in calorie intake.

Implications with respect to protein nutrition

An important point that seems to have been totally lost sight of in discussions of Sukhatme's hypothesis of ‘adaptation’ to low calorie intake is its serious implications with respect to protein nutrition. Policy-makers, planners, and some economists apparently labour under the mistaken impression that, with the acceptance of that hypothesis, the current estimates of undernutrition among populations in developing countries will be reduced to ‘manageable proportions’. Far from this being the case, the acceptance of this hypothesis would actually imply that the nutrition situation in many developing countries is far worse than what the present estimates indicate—indeed, that the solution of the problem will be possible only (p.27) through a drastic qualitative upgrading of the current dietaries, which would be clearly well beyond the economic resources of these countries. Let me explain why this is so.

At the height of the great protein debate, which finally led to the ‘protein fiasco’ and the winding up of the PAG (Protein Advisory Group of the UN), we in India had argued that the problem of protein-calorie malnutrition was essentially a problem of calorie deficiency, and that such protein deficiency as existed was an incidental secondary byproduct of primary calorie deficiency. Sukhatme himself was very much in the forefront in this endeavour. We had shown that, if the habitual cereal-legume dietaries of poor Asian population groups were consumed at levels adequate to meet the full caloric needs (and here we were talking of caloric needs as conforming to present international recommended mean levels of intake, and not of Μ – 2 SD levels), then protein needs would be automatically met. A study carried out at the National Institute of Nutrition (Gopalan et al. 1973), which had then attracted international attention, actually helped to demonstrate how, in an undernourished community of children, an additional provision of 310 calories (even when such calories were derived from food sources practically devoid of proteins—‘empty calories’) could bring about a significant improvement in their nutritional status and a reduction in protein-calorie malnutrition. We therefore took the position that, under the circumstances, the right and feasible strategy was to bridge the calorie gap with existing habitual dietaries and not to go in for expensive protein concentrates or for drastic dietary changes which in any case were well beyond the economic resources of these countries. Our policy was based on the principle that children needed to be fed the cereal-legume diets, which alone they could afford, at levels that would at least provide their full calorie needs as per internationally accepted recommendations-levels at which they would not need to restrict play and physical activity in order to conserve energy. We sought to overcome the problem of low calorie density (‘bulk factor’) of cooked cereal diets through invoking traditional practices like malting, so that children could be fed such predominantly cereal-based diets in quantities that would provide them their full caloric needs as per current recommendations (and therefore also, incidentally, their protein needs).

If, in consonance with Sukhatme's hypothesis, we now take the position that what children (and adults) need is no more than 70 per cent of the recommended mean levels of calorie intake, then on such restricted levels, clearly, they cannot meet their protein requirements with their present habitual cereal-legume diets. As the intake of energy is restricted, protein requirement increases—by as much as 20–30 per cent. Individuals will need to include in their dietaries other, relatively expensive, items of protein-rich foods. Several studies carried out at the National Institute of (p.28) Nutrition in Hyderabad provide experimental support for this in both children and adults, as shown in Tables 2.1 and 2.2. Not only is the protein concentration in the habitual poor Asian dietaries too low to provide the protein requirement at the M–2SD levels of calorie intake, but the utilization of protein at this lower level is also relatively poor.

Table 2.1. Recommended dietary intakes of protein and energy for Indians, 1981

NetCalories (kcal)



Netcaloriesat 70% level

Protein-calorie % at low level of energy requirement


Sedentary work






Moderate work






Heavy work







Sedentary work






Moderate work






Heavy work






During pregnancy













1–3 years






4–6 years






7–9 years






Boys 10–12 years






Girls 10–12 years






Boys 13–15 years






Girls 13–15 years






Boys 16–18 years






Girls 16–18 years






Note: This table presents the RDA of calories and proteins for adults (at different physiological levels) and children. Also, it provides the protein-calorie percentage at requirement level and at the low level of energy intake. This shows that at low level of energy (M – 2SD) intake the protein-calorie percentage of the dietaries should be at higher levels to meet the recommended allowance of protein. The protein-calories percentage of the habitual rice/legume-based Indian dietaries is around 8–9 per cent. Also, since the protein requirement is known to increase when calories are restricted because of poorer utilization, the actual protein-calorie percentage required at the low energy will be higher by another 20–50 per cent.

Source: Recommended Dietary Intakes for Indians 1981 (Indian Council of Medical Research)


Table 2.2. Effect of energy restriction on protein requirements in india

Preschool children

Energy intake (kcal/kg/day)

Protein requirementa(g/kg)





Adults 1

Protein intake (g/day)

Energy required for nitrogen(N) balance (cal)





Adults 2: Heavy casual labourers

Energy intake (kcal/kg/day)

Protein intake (g/kg/day)

Mean Ν balance (g/day)







aA retention of 40 mg of nitrogen per kilogram was taken on meeting the requirement of preschool children.

Source: Provided at the author's request by Dr B. S. Narasinga Rao, Director, National Institute of Nutrition, Hyderabad, India.

Thus, a child of 3 years subsisting on a 1200 calorie diet could obtain 22 g protein daily with a diet that provides no more than just 7.2 per cent of its overall calorific value through protein. (Low-cost cereal-legume diets provide more than 8 per cent of protein calories.) Against this, if calorie intake using the same diet is reduced to 70 per cent of mean recommended levels (M– 2SD level), the protein-calorie percentage in the diet that will be needed to provide the same 22 g of protein daily would be 10.3 per cent (clearly, more than what is possible with existing poor Indian dietaries). Moreover, we have also to take into account the fact that with low levels of calorie intake the utilization of protein is poor. Thus, in a study of under-fives it was found that, while with a calorie intake of 100 kcal/kg per day the children need 1.35 g protein per kilogram to achieve a retention of 40 mg of nitrogen per day, they would need 1.64 g per' kilogram from the same protein source to achieve the same level of nitrogen retention when calorie intake is reduced to 80 kcal/kg per day. Thus, a decrease in calorie intake to 70 percent (p.30) will call for an increase of protein intake by an additional 20 per cent. In a study of adults it was found that, with a calorie intake of 2066 kcal daily, as much as 60 g protein was necessary to achieve nitrogen balance, while with a calorie intake of 2250 calories, the protein intake needed for achieving nitrogen balance was just 40 g daily.

In effect, then, it would appear that, in recommending lower levels of calorie intake for poor populations in the expectation that they will ‘adapt’ to such low levels, we are also implying that they substitute ‘cake’ for their usual ‘bread’.

2.3.2 Growth retardation: the minimal role of genetic factor

It has now been conclusively shown, on the basis of data from countries throughout the world, that differences currently observed with respect to growth patterns of children in the rich countries of Europe and North America on the one hand and in the poor countries of Asia, Africa, and Latin America on the other (and between the rich and poor within the developing countries themselves) are mostly attributable to differences in their socioeconomic status, and not to genetic differences (Habicht et al. 1974; Stephenson et al. 1983). The remarkable secular trend in heights of children and adults witnessed in post-war Japan underscores this fact. While a great majority of Indian children show varying degrees of growth retardation, Indian children who are not subject to dietary constraints have been shown to have growth levels that correspond closely to the international (Harvard) standards (Gopalan 1989). The genetic potential for growth and development is nearly similar among most peoples of the world.

The Lancet (1984), discussing the use in developing countries of international growth standards (particularly the WHO/NCHS standards), on the basis of data from different parts of the world, concluded in its editorial that ‘recent evidence suggests that the growth of privileged groups of children in developing countries does not differ importantly from these standards’ and that ‘the poorer growth so commonly observed in the underprivileged is due to social factors—among which malnutrition-infection complex is of primary importance—rather than to ethnic or geographic differences’.

Also, there are no known ethnic differences in human physiology with respect to metabolism of nutrients. Africans and Asians do not burn their dietary calories or use their dietary protein any differently from Europeans and Americans, It follows, then, that dietary requirements for normal growth, development, and function cannot vary widely between different races, unless we accept different standards between them with respect to normal growth and function.

(p.31) It is not so much the retardation of physical growth per se and the relatively small body size of the poor that need bother us: it is the fact that there is now mounting evidence, thanks to sophisticated functional tests which measure physical stamina and work capacity on the one hand and mental development and learning ability on the other, that impairment in physical growth (as assessed by the failure to achieve the full genetic potential for the attainment of physical stature) is accompanied by varying degrees of functional incompetence. The fascinating work of Spurr et al. (1982, 1983, 1984) in Colombia, Chavez and Martinez (1982) in Mexico, Viteri (1971) in Guatemala, and Satyanarayana et al (1977) in India has provided ample evidence of the functional implications of growth retardation. Indeed, there is often a linear relationship between the degree of growth retardation and the degree of physical and mental functional impairment. Measurement of the degree of growth retardation thus could serve as a proxy for the assessment of functional competence.

It will be difficult for any biologist to agree with Payne's strange suggestion that the term ‘undernutrition’ should be reserved only for children whose state of nutrition has deteriorated to the point where they are close to death, and that other children with less severe degrees of nutritional deprivation, who do not actually face the risk of imminent death, even if they happen to show clear evidence of functional impairment of various kinds, should not be included in the ‘undernourished’ category. This is almost like saying that a person should be considered ‘unhealthy’ only when he has reached the point of death.


A view that is now being widely propagated is that, irrespective of deficits in height-for-age or weight-for-age, children with weights ‘appropriate’ to their heights (‘normal’ weight-height ratio) could be considered to have successfully adapted themselves to their dietary deprivation, and to be practically ‘normal’ and free from undernutrition! According to this postulate, it is not so much height-for-age or weight-for-age that matters but the weight-height ratio that is the crucial indicator of normalcy. This convenient hypothesis could lend legitimacy to the ‘small but healthy’ hypothesis. However, there is not an iota of hard scientific evidence to justify this sweeping postulate. Indeed, such evidence as is available points to a near-linear relationship between deficits in height-for-age or weight-for-age and functional impairment, irrespective of weight-for-height.

The above postulate is an unwarranted distortion of Waterlow's morphological classification of growth retardation into stunting, wasting, and stunting + wasting. Waterlow clearly did not invest this classification with functional significance. All that he implied was that, in the case of such stunted children with a normal weight-height ratio, it could be (p.32) argued that their current level of calorie intake is probably adequate to sustain them in the context of their stunted stature. He did not claim that stunted children with appropriate weights for their heights were functionally normal or that stunting, even if associated with appropriate body weight-for-height, was an acceptable state.

Satyanarayana has shown a direct correlation between productivity and body weight in industrial workers drawn from the poor socioeconomic groups, even with respect to operations in which body weight may not be expected to make a difference. In a longitudinal study on undernourished boys in India, Satyanarayana and colleagues (1979) showed that the wages earned by adolescent boys employed by farmers in rural areas were significantly related to body weight and height. Men and women with better nutritional anthropometry earned 30–50 per cent additional incentive money (over and above the uniform basic pay) in factories where an individual incentive system based on work output was in operation.

Agarwal et al. (1987) have provided convincing evidence, on the basis of an intensive study of over 1300 children in rural areas of Uttar Pradesh in India, that stunted children with a ‘normal’ weight-height ratio show the same order of functional (physical and mental) impairment as equally stunted children with poorer weight-height ratios. The view that a stunted child may be considered ‘adapted’ if it happens to have a weight appropriate to its stunted height is apparently untenable.

A considerable proportion of girls in developing countries who are stunted and of low body size because of undernutrition during the crucial years of their growth and development end up with heights of below 145 cm when they enter motherhood. It is now known that there is a direct relationship between stunting of mothers and the occurrence of low birth weights in their offspring. According to the recommendations of international agencies, maternal heights below 145 cm may be considered indicative of risk of obstetric complications and low birth weight. It will be seen from the data presented in Table 2.3 that a distinctly higher proportion of offspring of mothers with heights of less than 145 cm were of low birth weight. In India, as in many other developing countries, more than one-third of all infants born alive have birth weights below 2500 g. It is now known that, with respect to both height and weight, infants who start with the initial handicap of a low birth weight apparently never fully recover from it. Thus, low birth weights in full-term infants make a lasting contribution to stunting.

Stunting is the outstanding feature of so-called ‘adaptation’. It is the feature that ensures that not only this generation, but also the next, does not escape from the poverty trap. Stunted children with impaired learning abilities and schooling end up as stunted adults with low levels of productivity, educational attainment, and resourcefulness, earning low (p.33) incomes and thus continuing to be enmeshed in the poverty trap, and so proving unable to feed their children adequately. Stunted women beget offspring with low birth weights who start their lives with an initial handicap from which they never fully recover. Thus, stunting and the poverty with which it is invariably associated continue from one generation to another. To view this scenario as ‘acceptable adaptation’ is cruel irony!

Table 2.3. Maternal height and incidence of low birth weight (LBW) in offspring, India

Maternal height (cm)

Income group (Rs/he ad/month)

Incidence of LBW (%)

〈 145

〈 50


〉 145

〈 50


〉 145

〉 200


Source: Shanti Ghosh et al. (unpublished research)

A country or community in which large segments of the population suffer from growth retardation is one in which the quality and calibre of human resources is eroded and of substandard quality. A level of dietary intake and nutrition which can permit only substandard growth must inevitably lead to an erosion of the quality of human resources in developing countries or a perpetuation of such erosion where it already exists.

A sensible developmental policy of any country must obviously aim at providing for a level of calorie intake that will permit the full productivity and work output from its labour force, and a level of growth and development for its children that represents the fullest expression of their genetic potential. It is possible that many poor countries may be in no position to achieve these targets in the near future, and may have to settle for a policy that will enable them to reach these goals in a phased manner. It will, however, be an act of self-deception and political expediency to tailor standards of growth, physical and mental function, and dietary requirements in order to minimize the problem of undernutrition and win the war against poverty on paper. Standards are meant to determine the magnitude of a problem: it will be perverse to let this magnitude frighten us or tempt us to tailor the standard so that the ‘problem’ is reduced to ‘manageable proportions’.

There is no scientific justification for double standards between rich and poor countries in the matter of dietary requirements of their populations or growth levels of their children. We may, of course, argue as to whether current levels of body weight (not height) observed in European (p.34) and American infants and preschool children are not somewhat high and are representative of over mutrition and obesity. This is a different matter; it could call for a revision of standards for American and European children as well, Establishment by each country of a standard yardstick of its own in conformity with the pattern observed among affluent children of that country, who are not subject to dietary and environmental constraints, will be in order. On the basts of all available evidence, we may expect that the standards thus established by different countries (including developing countries) will not be widely different. What we should guard against, however, is an acceptance of the concept that levels of growth and of dietary intake known and accepted to be subnormal and substandard for the affluent are appropriate and good enough for the poor and consistent with their poverty status. In the measurement of undernutrition, we should be guided by these considerations.

2.4. Some Problems in Measurement

I shall now briefly consider the two practical approaches referred to earlier. There are inherent practical limitations involved in the use of these approaches as measures of undernutrition. For our present purpose, we will not consider the limitations that pertain to the actual collection of data and the possibilities of measurement errors, but merely the broader issues.

2.4.1 The calorie intake yardstick

Estimations of levels of dietary intake of calories have been widely used in the evaluation of nutritional status of population groups. The limitations of the calorie intake yardstick have been discussed in an earlier communication (Gopalan 1983a); these need only be recapitulated briefly here.

First, as pointed out earlier, poor diets are deficient not in calories alone but in several nutrients as well, though in dietaries based mostly on a single major staple there is a close correlation between calorie intake and intake of essential nutrients. Where the major dietary item is lacking in an important nutrient (as is the case with cassava and tapioca with respect to protein), the diet could be quite adequate with respect to calories while being highly deficient with respect to protein. This is not unusual in Africa. Under the circumstances, calorie intake levels may provide a flattering picture of the nutritional status. Therefore calorie intake measurements generally provide a quantitative and not necessarily a qualitative measure of the adequacy of diets.

(p.35) Second, measurements of calorie intake, especially under field conditions, lack precision; daily and seasonal fluctuations in dietary intake could add to this problem, which even seven-day weighment methods cannot entirely solve. Repeated diet surveys in different seasons may be necessary to obtain a reliable picture.

Third, where diets of entire households (and not of individual members therein) are estimated, as is generally the case, the actual calorie intake of the most vulnerable segment of the population, namely the children, is often indirectly derived through the application of certain arbitrary coefficients based on the assumption that the intrafamilial distribution of food conforms to relative physiological needs—an assumption not often valid. Actual estimations of individual dietary intakes within families in a representative sub-sample of households surveyed, and the application of necessary correction based on these data to the figures for individual intakes, may obviate the error to some extent.

For the purpose of assessment of dietary/nutritional status, the actual observed calorie intake for a given age/sex/occupational group is compared with the mean energy requirement level for that group as recommended by international (and national) expert bodies on the basis of well conducted, fairly reliable intensive studies in several laboratories of the world. The Indian Council of Medical Research, for example, has recommended energy intake levels appropriate for different categories (age, sex, and occupation) based on Indian studies as well as on published observations from elsewhere and the recommendation of international agencies. Through such comparison, it will be possible to determine proportions of population groups—manual labourers (male and female), sedentary workers (male and female), children between 5 and 12 years of age, and children below 5 years of age—in the community surveyed, obtaining, say, 90, 80–90, 70–80, 60–70, and less than 60 per cent of the respective recommended intake energy level.

2.4.2 Growth retardation as a measure of undernutrition

Children under 5 years of age represent the most vulnerable segment of the population from the nutritional standpoint. For all practical purposes, the growth performance of children of this age group is a convenient measure of the nutritional status of the community. This procedure is now being widely employed in many developing countries. The pitfalls and limitations of this approach have been discussed in an earlier publication (Gopalan 1984).

Among the different anthropometric measurements—namely, weight-for-age, height-for-age, weight-for-height, and arm circumference—taking practical considerations into account, the balance of advantage rests heavily with weight-for-age measurements as far as infants and under-fives (p.36) are concerned. Heights (or lengths) for age in children are more difficult to measure accurately, and height measurements are also less sensitive to dietary deprivation. The significance of weight-for-height measurements, and of the classification of growth retardation into ‘wasting’ and ‘stunting’, are debatable. Arm circumference measurements are simple to carry out but there are doubts on the one hand about their being age-independent (a merit often claimed in their favour), and on the other hand about their reliability in comparison to weight measurements. For all practical purposes, therefore, as far as under-fives are concerned, it may be best to rely on weight-for-age surveys.

The procedure that is now being widely adopted is to compare the weight-for-age of the child with the weight-for-age in the international (Harvard or NCHS) standard. According to the Gomez scale, which is generally used (Gomez et al. 1956), children with weights within 90 per cent of the standard are considered ‘normal’, those between 90 and 75 per cent of the standard as being in ‘mild’ malnutrition, those between 75 and 60 per cent of the standard as being ‘moderately’ malnourished, and those below 60 per cent of the standard as being ‘severely’ malnourished.

The cut-off points of 90, 75, and 60 per cent, are admittedly arbitrary and have no real scientific basis. The use of the international standard rather than the ‘best indigenous’ standard as the yardstick for comparison has also been questioned. Even so, Gomez's classification has proved useful in enabling health/nutrition scientists to quantify undernutrition in the children of a community and to assess the impact of nutrition intervention programmes. However, in order that weight-for-age measurements in under-fives could be thus used for comparisons of nutritional status between population groups and between two time-points in the same population group, some precautions are absolutely necessary. These precautions are not being currently observed, with the result that the data from different locations and at different time-points are not truly comparable.

In all comparisons that use cut-off points that are percentages of the standard median, it is extremely important to ensure that the age and sex composition of the under-five groups are very similar. A given order of weight deficit in a child under 2 years of age carries a far greater significance than the same order of weight deficit in a child of 5 years. Also, where there are wide seasonal fluctuations with regard to food availability, comparisons of measurements between two populations in two such different seasons may prove fallacious. Mistakes in age assessment, especially in children under 3 years, can modify results significantly.

According to available data, generally less than 10 per cent of under-fives in South-Bast Asian countries suffer from ‘severe malnutrition’ (p.37) (weight-for-age less than 60 per cent of the standard), less than 15 per cent are normal (weight-for-age more than 90 per cent of standard), while the remaining belong to the ‘mild’ and ‘moderate’ grades of undernutrition.

If used with due precautions, weight-for-age surveys among under-fives and the quantification of the order of weight deficit observed among them provides the health/nutrition scientist with a convenient and practicable tool for the quantification of weight deficits in children and therefore, indirectly, for the quantification of undernutrition. This procedure will be specially useful in comparisons of different population groups within a country in order to identify the most depressed groups requiring special attention, and also to monitor changes in the profile of undernutrition (the quantum and pattern) following from nutrition intervention programmes.

The Gomez scale and other similar methods attempt to classify undernutrition in terms of various cut-off points defined as a stipulated percentage of the median of the reference population. The limitation of this approach is that it does not take into account the variability of the relative width of the distribution of weight-for-age across different age periods. For example, 60 per cent of median weight-for-age indicates a much more severe state of malnutrition for infants and young children than for older children. In order to overcome this problem, a new method has been developed which measures the deviation of the anthoropometric measurement from the reference median in terms of standard deviation units or ‘Z’ scores (Waterlow et al. 1977). This is coming into increasing use.

Unfortunately, weight-for-age measurements are now being put to the improper use of subserving a nutrition policy of brinkmanship in some developing countries, not for the promotion of child health/nutrition and prevention of undernutrition, but mainly to identify cases of so-called ‘severe’ malnutrition (that is, children with weight-for-age of less than 60 per cent of the standard) who could be chosen as beneficiaries for supplementary feeding programmes. In a longitudinal study in children of Bangladesh, Chen et al. (1980) observed that the risk of increased mortality was observed only in under-fives who suffered from ‘severe’ malnutrition. This has been unfortunately misinterpreted to imply that children suffering from ‘mild’ and ‘moderate’ malnutrition can somehow muddle through and that only the ‘severely’ malnourished need attention in nutrition intervention programmes. Though Chen tried to rebut this inference in a later publication (Chen et al. 1982), the impression conveyed by his original paper has continued to misguide health workers in developing countries.

Such exclusive attention to the severely malnourished to the point of neglect of the ‘moderately’ and ‘mildly’ malnourished (nutrition policy (p.38) of brinkmanship) can only result in reducing the size of the tip of the iceberg and in farther increasing the pool of moderate and mild malnutrition cases in the community. But children who are currently in the so-called ‘moderate’ stage could move into the ‘severe’ stage within a few weeks or months. To withhold action until they are actually at that end-stage would be poor strategy. The inputs needed to prevent children who are in the mild and moderate stages of malnutrition from passing into the severe stage are far less than those that would be needed to rehabilitate the severely undernourished ones. It may even be possible for mothers of poor households to provide the inputs of the former category in their own homes with their own resources, if they are properly assisted and educated by health workers in the course of domiciliary visits. The inputs needed for rehabilitation of the severely undernourished, on the other hand, will be clearly beyond the means of poor households and will need expensive institutional support. Apart from this practical consideration, there is now evidence that even the so-called moderately undernourished children show functional (physical and mental) impairment.

2.5 Height as an Indicator of Nutritional Status

It was earlier pointed out that, as far as infants and very young children are concerned, height measurements may not be feasible in large-scale field studies; moreover, height (length) is less sensitive to dietary fluctuations in the short run. For these reasons, it was recommended that weight-for-age measurements would suffice for all practical purposes as far as under-fives are concerned.

However, the merit of height measurements as an indication of socioeconomic status of a community deserves special emphasis. Serial height measurements of children of 6–7 years of age have an important place in national nutrition surveys. Cross-sectional measurement of heights in adult populations of different classes can provide valuable indicators of disparities with respect to nutritional status arising from socioeconomic inequalities. Tanner (1982), in his remarkable paper on ‘The Potential of Auxological Data for Measuring Economic and Social Well-being’, has provided a fascinating historical account which highlights the great value of height measurements as an instrument for monitoring progress with respect to the state of health, nutrition, and well-being of communities. Steckel (1983) found a close correlation between height and per capita income in a study based on the result of 56 height studies and per capita income estimates for 20 countries. When it is recognized that ‘socioeconomic’ factors and per capita income (p.39) could affect height only by mediating changes in nutritional inputs, the importance of height as a measure of nutritional status of a community becomes obvious.

The most convenient age group that could be captured for large-scale height surveys would be schoolchildren of the 6–7-year age group—i.e. those belonging to the first or second standard (the stage at which drop-outs are few). In fact, such surveys must constitute an important item of any national nutrition survey.

Height measurements and quantification of height deficit (in comparison with an international or national standard) will help to identify differences in nutritional status between different regions, population groups, and social classes in a country and to monitor changes over a period of time. Indeed, height deficits in a population group could be considered to provide an even more reliable indication of nutritional status than weight deficits. It may be argued that the standards against which weight measurements are being compared have been derived from relatively obese affluent American subjects and may not necessarily reflect optimal nutrition. Also, temporary diminution in weights occurring in communities of children or even adults can be caused by short-lived epidemics. These criticisms will not apply to height measurements. Overnutrition and resultant obesity can result in more than optimal weights, but not in more than optimal heights.

Height-for-age deficits can be quantified in the same way as weight-for-age deficits using international standards and applying the procedures of ‘Z’ scores. Where facilities exist, each country could develop its own weight-for-age and height-for-age standards based on measurements carried out on truly affluent sections of its own populations which are free from dietary and environmental constraints. Many developing countries do not at present have such standards of their own. Under the circumstances, in view of the mounting evidence that genetic differences with respect to growth potential between population groups are relatively minor, it will be quite in order for international standards (WHO/ NCHS/Harvard) to be used for all practical purposes for the assessment of weight and height deficits. It is possible that even standards developed by each developing country based on observations on its affluent sections could fall short of the widely used international standards to some extent. This is because of the strong likelihood that the secular trend with respect to growth has not as yet reached its maximum limit in many developing countries. It may probably need two or three generations of affluence for populations in many such countries to attain their fullest growth potential, which can be expected to be almost similar to the prevailing pattern in Europe and America, as has been revealed by experience in Japan during the last two to three decades.

(p.40) Biological significance of height measurement

Tanner (1982) quotes Villerme, who wrote as long ago as 1928 that ‘Human height becomes greater and growth takes place more rapidly, other things being equal, in proportion as the country is richer, comfort more general, houses, clothes and nourishment better, and labour, fatigue and privations during infancy and youth less; in other words, circumstances which accompany poverty delay the age at which complete stature is reached and stunt adult height.’

In Japan, between 1957 and 1977 average mature height increased by 4.3 cm in males and 2.7 cm in females; age at maximum increment dropped by 0.97 years in males and 0.53 years in females. Practically all the height increase was due to increase in leg length, not in sitting height, with the result that within 20 years of economic advancement the entire body proportions of the Japanese had changed. This is perhaps the most striking and spectacular evidence of the importance of height measurements as an index of the nutritional status of a population which parallels economic advancement, and it reveals that height measurement is clearly an indicator of as much importance to the developmental economist and planner as it is to the health/nutrition scientist.

There is a large body of evidence pointing to a relationship between height and mental function. Indeed, as early as 1893, William Porter (quoted by Tanner 1982) had shown in the schools of St Louis that pupils who were academically advanced for their age were also taller. There have been quite a few similar observations in recent years pointing to a correlation between height and IQ.

Tanner also quotes findings from a massive study in Norway, in which height measurements were recorded in 1.8 million subjects over 15 years of age: there it was found that mortality in those 185–9 cm tall was half the rate of that in those 150–5 cm tall. A similar lower mortality among taller children less than 5 years old in Ghana has also been reported by Billewicz and McGresor (1982).

Reviewing all the available evidence on height measurements and attempting an answer to the question, ‘Is being taller better?’, Tanner (1982) concludes: ‘It does look, therefore, as though height indeed can be a proxy for health and for the attainment of biological potential. This is true, of course, only when comparing groups, not in comparing isolated individuals, the variation between whom, is due overwhelmingly to genetic causes. But between social classes, urban and rural dwellers, educated and uneducated, height is a useful proxy for ‘aisance de vie’.

Height measurements will not only be helpful in monitoring secular trends in nutrition and economic status: they will also be useful in making interregional and interclass comparisons of nutritional status. Height measurements could help to bring out glaring socioeconomic inequalities and the consequent disparities in nutritional status among (p.41) classes within countries. Tanner refers to a report on Trinidad slaves in 1815 which showed that Trinidad foremen were on average an inch taller than the fieldhands. Bielicki et al (1981) showed that the sons of Polish peasants raised in villages in families containing four children had an average height of about 172 cm, while sons of professional men with small families working in large cities averaged 176 cm,. Goldstein (1971) reported a similar phenomenon in the UK on the basis of a national sample survey of heights of 7-year-olds in 1971.

The difference between the average heights of non-manual classes (class III) and labouring classes (classes IV and V) was also reported as being roughly 1 in. by Clements and Picket in 1957 (quoted by Tanner 1982); strangely enough, the data of the office of Population Censuses and Surveys of 1980 show that this difference still persists.

It is only with respect to Scandinavian countries—Sweden and Norway—that there is, today, convincing evidence of an absence of significant differences with respect to height between occupational classes. The attainment of such a situation of equity and distributive justice wherein there are no striking differences with respect to nutritional status between different occupational and income groups must be considered the hallmark of truly successful socioeconomic development; in such a situation, even the groups with the lowest income levels are apparently able to achieve an optimal level of nutrition. Unfortunately, most developing countries still appear to be far away from this goal. Not only is the general level of health and nutrition in their populations low, but there is also apparently far greater evidence of disparities among populations. Evidence of differences in height between different occupation groups in India are unfortunately quite striking, as the observations that follow will show.

Indian studies

Three Indian studies covering fairly large numbers of subjects indicate the value of height as a measure of nutritional status. In these studies, height measurements (along with weight measurements) have been carried out in subjects of different socioeconomic groups. It was presumed that the dietary intake would largely parallel the socioeconomic status; and in any case, as was pointed out earlier, there is no way in which socioeconomic or occupational status can exert a direct metabolic effect on the body in order to influence height except through its effect on nutritional inputs. So it will be justifiable to view the observed relationship in these studies between height and economic status as in fact a relationship between height and nutritional status.

1 Shanti Ghosh and her colleagues have carried out an extensive longitudinal study on growth and development of children of different socioeconomic groups, from birth to nearly 15 years of age. Nearly 8,200 children were covered in the study. The communities investigated (p.42)

                   Undernutrition: Measurement and Implications

Fig. 2.1 Weight and standing heights of Indian girls, by per capita income and age

Source: Shanti Ghosh et al., unpublished research

ranged from the poorest (less than Rs 50 per head per month—1969 level) to the fairly well-to-do upper middle class (more than Rs 200 per head per month—1969 level). The longitudinal data from children belonging to these two income groups (Figure 2.1) show a clear relationship between socioeconomic status and heights and weights of children.

2 Satyanarayana and colleagues (1980) have assembled data from longitudinal observations on the heights and weights of children of different socioeconomic groups in rural Hyderabad observed over a 15-year period from 5 to 20 years of age. The children belonging to their Group I, with heights between Μ and Μ – 2SD of Boston standard, mostly came from families of affluent landlords owning more than 5 acres of fertile land; those of their Group III were from the poorest rural households owning no land of their own, with adults being illiterate and eking out their living from seasonal agricultural wage-labour. In Table 2.4 some of their observations have been set out. The table also shows data on heights and weights of children of the most affluent Indian communities as observed and reported by Hanumantha Rao and Sastry (1977) on the basis of their cross-sectional studies. The striking differences between the different socioeconomic groups will again be obvious. (p.43)

Table 2.4. Longitudinal studies of growth of Indian children of different socioeconomic groups


Initial (aged 5)

Final (aged 20)

Ht (cm)

Wt (kg)

Ht (cm)

Wt (kg)

Mostly from families of well-to-do landlordsa (owning about 5 acres of fertile land)





Mostly from families of agricultural labourers on seasonal/daily wagesa





Highly affluentb





a Based on Satyanarayna (1986)

b Based on Hanumantha Rao and Sastry (1977)

3 The National Nutrition Monitoring Bureau (1980) has recently completed a study of the dietary, nutritional, and anthropometric status of 32,332 subjects (12,925 adults and the rest children) drawn from 15 major cities of India. The sample households were classified into five major socioeconomic categories. The high income group (HIG) and the slum labour (SL) represented the two extreme ends of the economic spectrum, with the other three groups lying in between. SL was the group subject to the greatest socioeconomic deprivation—poor, largely illiterate or semi-literate, living in highly overcrowded and unhygienic conditions, and having to depend mostly on unskilled manual labour to eke out a precarious livelihood. Their diets were decidedly lower in energy content, and their children showed a higher prevalence of signs of vitamin deficiencies. The heights and weights of children and adults faithfully reflected the socioeconomic gradient, with HIG at one end, SL at the other, and the remaining groups falling in between. For the sake of convenience, only part of the data from the two groups at the extreme ends (HIG and SL) have been set out in Tables 2.5 and 2.6.

The poverty trap

The outstanding finding in all the three Indian studies cited above is the striking relationship between income and occupational status on the one hand and physical stature on the other. It would appear that the more lowly (using the expression for the sake of convenience) the job that a community is engaged in, the greater the degree of stunting in its children and adults. The cart-pullers, scavengers, manual labourers (including those engaged in strenuous work), stone-cutters, porters having to carry heavy loads, and agricultural labourers are apparently the (p.44) ones who are most stunted and have the lowest body weights; unfortunately, these are precisely the occupation groups (rather than the business executives and academicians) who are in greatest need of a strong and sturdy body for optimal productivity and output and for earning a reasonable wage from their occupation.

Table 2.5. Heights and Weights of children in urban India, 1975–1979

Height (cm)

Weight (kg)





At 5 years











At 12 years











At 16 years











aHigh-income group

bSlum labour

Table 2.6. Heights and weights of adults in India,1975–1979

Height (cm)

Weight (kg)





At 20–25 years















At 40–45 years











aHigh-income group

bSlum labour

cFigures in parentheses in columns (2) and (4) are measurements of corresponding rural groups

Source: National Nutrition Monitoring Bureau (1980)

(p.45) As was pointed out earlier, my concern is not over small body size per se. Earlier in this chapter I pointed out the functional implications of stunting. Stunting in a community is but a proxy for current substandard function and for past malnutrition which must have involved a considerable cost to society.

A community in which a considerable part of the population is stunted is usually a community with high infant and child mortality, high levels of morbidity in children, a high rate of drop-outs from schools. This is also a community in which children have lost valuable time for learning skills, mothers have lost considerable part of their daily wages, and health services are so overburdened with curative work that preventive and promotive health programmes are relegated to the background.

2.6 Summary and Conclusions

The major propositions of this chapter can be briefly recapitulated.

  1. 1 There are two practical approaches to the measurement of undernutrition: (a) through a survey of diets of representative households (supported by surveys of diets of individual members of the family in a sub-sample of households), in order to derive information about the nutrient intake, especially calorie intake; and (b) through an anthropometric and clinical examination of children, especially the under-fives. These two procedures, combined with a broad survey of socioeconomic and environmental status of the community, will yield data which, when properly interpreted and evaluated, could provide valuable practical leads for combating undernutrition in the community.

  2. 2 The basic issue that arises in the interpretations of the data gathered through these above approaches is, what are the normal standards against which prevailing levels of calorie intake or observed growth, performance should be compared in order to determine adequacy or otherwise? The three hypotheses—of Sukhatme, Seckler, and Payne—that have been advanced in this connection have been critically examined. Sukhatme has argued that human subjects can permanently adapt themselves to a low calorie intake level representing the lowest limit of their intra-individual variations and that, therefore, Μ – 2SD (recommended mean energy requirement level Μ minus two standard deviations) rather than Μ would be the appropriate yardstick for an assessment of calorie intake. Sukhatme's hypothesis is unacceptable because there is no convincing evidence that populations engaged in their expected levels of occupational activity and obliged to subsist on only about 70 per cent of their recommended mean energy intake can permanently adapt themselves to such a low calorie intake through metabolic adjustment and increased efficiency of energy utilization without (p.46) suffering a loss of body weight or consequent impairment of function. With a calorie intake that is around 70 per cent of the currently recommended levels, it will be difficult for children to meet their protein requirements with cereal-based diets: a drastic qualitative upgrading of current cereal-based diets would be necessary. This implies a reversal to the discredited view that the answer to the problem of protein-calorie malnutrition in developing countries lies in increasing use of expensive protein concentrates, a view that has been convincingly rejected on the basis of extensive studies.

  3. 3 Seckler's hypothesis that moderate degrees of growth retardation are a welcome attribute of poor people consistent with their good health and an acceptable form of adaptation is untenable because of the evidence that even moderate degrees of growth retardation have been shown to be associated with an impairment of physical and mental function.

  4. 4 Payne's hypothesis that, even if children of developing countries have been shown to have the same genetic potential for growth as those of the more fortunate countries of Europe and the USA, their lower levels of growth should not be viewed as evidence of undernutrition but as an adjustment to their ‘economy and ecology’ is also untenable. Payne does not deny that such growth-retarded children suffer from impaired function; but he considers that, despite such impairment, the subjects can function as well as they need to in their economy and ecology, and therefore, he would not consider them undernourished. The acceptance of this postulate can lead to the perpetuation of the present substandard state of growth and development in populations of the poor countries and is therefore wholly unacceptable.

  5. 5 In all the above three hypotheses the word ‘adaptation’ has been loosely employed to signify an acceptable state of normalcy instead of being viewed as a strategic retreat from normalcy which involves compromise with respect to both physical and mental function.

  6. 6 The view propounded by Payne that only those children whose state of nutrition has deteriorated to the point where they face the risk of imminent death should be considered undernourished obviously cannot be accepted. This is almost like saying that a person is ‘unhealthy’ only when he is on the point of death. The so-called moderately malnourished children of today could gravitate into severe malnutrition in a few weeks or months. It is far easier and far less expensive to prevent them from sliding to such a severe stage of undernutrition than to rehabilitate them after they have reached that stage. Moreover, the so-called moderate degree of growth, retardation is also associated with impaired function. Populations subjected to such retardation represent a substandard human resource.

  7. 7 The view sometimes propagated that, irrespective of deficit in height-for-age or weight-for-age, children with weights appropriate to (p.47) their heights (normal weight-for-height ratio) could be considered adapted is also untenable. Stunting, irrespective of body weight, has been shown to be associated with impaired function. Stunted children with normal weight-for-height ratio also show evidence of impaired physical and mental function.

  8. 8 A level of dietary intake and nutrition which can permit only substandard growth must inevitably lead to an erosion of the quality of human resources of developing countries or a perpetuation of such erosion where it already exists. It must, therefore, be the policy of developing countries to achieve for their children a level of growth and development which represents the fullest expression of their genetic potential.