Previous Page Table of Contents Next Page


Although this report, like its predecessors, is primarily concerned with the requirements of healthy people, in developing countries the prevalence of malnutrition and common infections, particularly of the gastrointestinal and respiratory tracts, is such that they can be regarded as part of ordinary life. As was pointed out in 1975 by the joint FAO/WHO informal gathering of experts (1), when a public health problem is of such a magnitude that it affects the energy and protein requirements of a significant part of the population, it cannot be ignored in the assessment of such requirements.

9.1 Catch-up growth

Previous sections have been concerned with the energy and protein requirements of children and adults who are in the acceptable range of weight for age or weight for height. As discussed in section 3.5.1, their requirements can be estimated either on the basis of actual weight or median weight for age or height. The latter will have what has been termed a normative effect: those who are somewhat under- or overweight will tend to revert towards the median if the normative estimates of their requirements are fulfilled.

This section is concerned with those who are outside the acceptable ranges. If by definition, requirements are the amounts needed to maintain health, then the estimates for those who are outside the acceptable ranges must be normative and designed to bring them back to health. The problem is not one of principle, but of degree.

9.1.1 Adults

Since no change in height is to be expected with changes in intake, the objective will simply be to bring body weights into the normal range for actual height. The amount by which the requirement should be increased or decreased to achieve this objective will depend upon two things: the rate at which a desirable weight is to be achieved, and the extent to which the deficit or excess consists of fat and lean tissue.

Being overweight is predominantly due to excess fat, and even the milder degrees of obesity are regarded as a public health problem in many societies. The low energy diets used for the management of obesity, if maintained for long periods, must supply adequate amounts of protein and other nutrients. Therefore, as in other conditions in which the energy intake is low, such as the physically inactive and the elderly, the concentration of protein in the diet has to be increased (see section 10). Further discussion of the management of obesity is outside the scope of this report (2). However, it seems reasonable to suggest that the safe level of protein for those who are overweight should be based on the median acceptable weight for height and not on the actual weight. Since the adipose tissue that is lost does not consist solely of fat, this may involve a small negative nitrogen balance for a time.

In adults who are seriously underweight for their height, there will generally be a loss of both fat stores and lean body mass, particularly muscle (3). Therefore, to bring their weight into the normal range requires additional amounts of both energy and protein. Clinical experience suggests that underweight adults who are free from disease can be rehabilitated fairly rapidly if given food ad libitum, although not as rapidly as children.

The extra daily requirements of both energy and protein will depend not only on the initial deficits, but also on the desired rate of rehabilitation. Even when the deficits are known, there is no simple relationship between intake and the time needed to reach a desired weight. As weight is gained, the maintenance protein requirement and the energy expenditure, even at a fixed level of physical activity, will both increase. Therefore, if the intake is fixed, progressively decreasing amounts of energy and protein will be available for storage as new tissue. The simplified example in Annex 8 shows that if the intakes of energy and protein are simply set at the levels appropriate for a person of median weight for height, restoration of the deficits will be slow. Where the deficits are large, as during famines, and relatively rapid rehabilitation is needed, substantial amounts of extra protein and energy must be provided. Appetite is probably the best practical guide.

9.1.2 Children

To make quantitative estimates of the requirements for catch-up growth in children is even more difficult than for adults for two reasons: first, the target body weight is not fixed, but increases with time in a growing child, so that the longer the period of rehabilitation, the greater the gap to be filled; secondly, two components may contribute to the weight deficit: (a) weight may be reduced below the acceptable range of weight for height; (b) body size may be small because the increase in height has fallen below accepted standards. This impairment, if severe, has been described as “stunting” (4). In many, if not most stunted children, body weight is within the acceptable range for height (5,6). Low weight for height (wasting). Deficits in weight for height, particularly in young children, are associated with an increased risk of mortality (7) and morbidity (8). Except in famines, this type of severe weight deficit is generally found in only a small proportion of children.

The extra amounts of energy and protein needed for catch-up growth have been calculated in some detail in studies on the rehabilitation of malnourished children (9,10). Representative values for the energy cost, in round figures, are 5 kcalth (21 kJ) per g tissue laid down (see Annex 6) and for the protein cost 0.23 g per gram [0.16 g deposited at 70% efficiency (section 6.3)]. The daily increments needed will depend on the rate at which catch-up growth is to be achieved, but it is clear that (as in normal growth, pregnancy, and lactation) at all rates of catch-up there will be some increase in the ratio of protein to energy requirement, over and above that appropriate to the age of the child. The increased requirement for protein is likely to be relatively greater than that needed for rehabilitation in adults, because except in extreme starvation the tissue lost by adults will generally include a high proportion of fat. Low height for age (stunting). Small size due to deficits in linear growth is extremely common. In some countries, almost 50% of children have been classified as stunted1 (11), so that its significance cannot be ignored from the point of view of public health (see section 3.2.1).

The Consultation has accepted the view (section 3) that children of different ethnic groups have basically the same growth potential, and that the cause of widespread growth retardation is principally environmental. There is clear evidence that catch-up in height is possible (12), although the process is much slower than catch-up in weight. There is little precise information on the age up until which the capacity for catch-up is retained, although it is probably up to the end of the adolescent growth spurt.

1 The cut-off point commonly taken for the diagnosis of stunting is 90% of the NCHS height for age.

The 1975 informal gathering of experts (1) proposed that the energy and protein requirements per kg of a stunted child should be taken as those appropriate for a normal child of the same age (height) (i.e., biological age based on height as opposed to chronological age). They did not specify to which weight these requirements per kg should be applied i.e., the actual weight or the median weight for height. Examples are shown in Table 52 and Annex 8. The increments based on age (height) and median weight for height initially amount to about 13% for energy, and 32% for protein, over and above the requirements based on actual age and weight. These percentage increments will decrease as the target weight and height are approached, although the absolute requirements will increase. As already stated, it is evident that, for catch-up to occur, there is a relatively greater need for protein than for energy. The implications of this for determining a safe level of protein energy ratio are discussed in section 10.

Table 52. Examples of different ways of calculating the energy and protein requirements of an undernourished child
 Child aged 2.0 years
 Wt 9.0 kg = 71% of median Wt for age
 = 92% of median Wt for Ht
 Ht 74 cm = 86% of median Ht for age
 Median Wt for age = 12.6 kg
 Age (height) = 12.0 months
 Median Wt for Ht = 9.8 kg

 Requirement per kgTotal requirement (× weight)
 Weight used
Age used ActualMedian for heightMedian for age
g/kgActual agea1.15   10.411.314.5
 Height/agea1.37   12.313.4c17.3
kcalth/kgActual ageb101   910   9901 270
kJ/kg 4233 8104 1405 310
kcalth/kgHeight/ageb105   9501 030c1 320
kJ/kg 4393 7755 5205 520

a Safe intakes (g/kg) from Table 33.
b Energy requirements, from Tables 21 and 23.
c Values produced by approach recommended by 1975 informal gathering of experts (1).

An alternative approach would be to estimate the requirements of the malnourished child as those appropriate if he or she were of normal weight for age. Table 52 shows that this would involve increments of 45% for energy and 70% for protein over and above those based on actual age and weight. It is likely that this approach, if fulfilled, would lead to obesity, which is, in fact, quite often observed in stunted children.

It can be predicted from short-term studies of catch-up in malnourished children (13) that increments of the order shown in Table 52, over and above the requirements based on actual weight, will allow the normal weight for height to be reached. What is much more difficult to predict is the rate of catch-up in height that will occur when these extra amounts are provided. In the approach proposed, height is the main determinant of the total requirement because age (height) determines the requirements per kg, and height determines the median weight by which those requirements should be multiplied. The deficit in height given in the example in Table 52 represents almost 4 SD below the median. It has been shown in studies of children severely malnourished as a result of untreated coeliac disease that, when treatment is given and the children are living under good conditions, a height deficit of this order can be corrected in a year (12). The extra provision shown in the example should allow for catch-up in weight within a year, but it is not known whether these extra amounts will provide for a corresponding catch-up in height.

9.2 Effects of infection on energy and protein requirements

Children under 3 years of age, who have the highest protein and energy needs per kg of body weight, are the group most frequently affected by infections and most severely ill. The average number of days per year during which a child is ill varies from country to country and is very difficult to establish with any accuracy. Thus in one country it was estimated that the average child between 6 months and 3 years was ill with diarrhoea 13% of the time (14), and had some clinically recognizable illness for almost half of each month. The resulting anorexia and metabolic disturbances probably continued for longer. These episodes are well documented by several authors (15–17).

Respiratory and other infections are also very frequent in countries with a high prevalence of diarrhoeal disease. These infectious episodes frequently result in a negative protein and energy balance due to anorexia, catabolic processes, and decreased absorption in diarrhoeal illness. Such frequent episodes of infection are known to reduce the rate of increase in both weight and height, and probably account for the wide range in rates of weight gain found in longitudinal studies in developing countries (18). However, intestinal parasites, which are usually common under these environmental conditions, apparently have little or no effect on protein and energy requirements unless the infection is extensive (19, 20) or causes acute diarrhoea.

Two approaches have been used to estimate the extent of the nutritional deficits caused by infection. The first is to measure the reduction in food intake during the period of illness. Thus during an episode of diarrhoea, total energy intake can be reduced by as much as 20–40% (21). However, this method does not take account of nutrient losses caused by diarrhoea and vomiting, nor of metabolic losses attributable to the infection and the accompanying fever. It is virtually impossible to quantify these losses under field conditions.

An alternative approach is to estimate the amount of energy and protein needed to bring the child back to its normal growth channel. This method covers both the deficiencies of intake and the metabolic losses during the period of illness. A number of studies in different countries have attempted to quantify the proportion of the growth deficit that can be attributed to infections (14–17, 22, 23) and the extra requirements for recovery from them. Although such calculations may not be totally reliable, the contribution of infection to growth failure has sometimes appeared to be surprisingly small. For example, in a Mexican study (16) children with a low morbidity from infections had a slightly better growth pattern than their counterparts with a high morbidity, but their heights and weights were still well below the reference values.

9.3 Conclusion

It is impossible to generalize about the amounts of additional energy and protein needed for catch-up growth in children who have become malnourished, usually as a result of the combined effects of an inadequate intake and frequent infection. The relative contributions of these two factors, and their severity, will vary in different communities. In many countries there are also important seasonal effects, the time when food is in short supply coinciding with the highest prevalence of infection.

Since this report is primarily concerned with the requirements to maintain health, and therefore a good level of mental and physical performance, rather than with the treatment of those who are ill, it is appropriate to examine the requirements for preventing the development of a significant degree of malnutrition under the existing conditions in Third World countries.

There are two approaches that could be used to obtain an estimate of the requirements for prevention. Longitudinal studies in three countries have shown that children in their usual environment may, over a period, grow at 2–3 times the median reference rate (18). Often these growth spurts will follow a period of faltering. It might then be reasonable to ask: what would be the effect of doubling the growth components of the energy and protein requirements, to allow for twice the normal rate of growth during favourable periods? This doubling would be over and above the 50% addition that is made to allow for day-to-day fluctuations in growth in normal children (section 6.3.2). The consequent increases in energy and protein requirements at various ages are shown in Table 53. The increase in protein would be about 40% in the youngest age group, falling to 20% at 2 years.

Table 53. Percentage increases needed in energy and protein requirements of children to allow for twice the “normal” growth rate
AgeAverage weight gain% increase over “normal” requirementa
(g/kg per day)EnergybProteinc,d,e
0.5 –0.751.8314.550
1 –1.50.675   32
1.5 –20.513.525

a It is assumed that the requirements for “normal” growth are 1.5 × the theoretical estimates based on weight gain and N increment (see section
b “Normal” energy requirements from Tables 22 and 23.
c “Normal” safe level of protein from Table 33.
d If the same value is assumed for the coefficient of variation of weight gain as in Table 33 (i.e., 35%),doubling the growth rate causes substantial increases in the overall CV. The values for the successive age groups, calculated according to the formula shown in Table 33, are: 20.7, 19.0, 16.1, and 14.5%.
e It may be noted that the average protein requirement for twice normal growth is little greater than the normal safe level. The average requirements for the successive age groups are: 1.75, 1.55, 1.6, and 1.13 g of protein per kg per day (cf. Table 33). Thus with the normal safe level a large proportion of children would be able to grow at twice the normal rate.

Since the safe level of protein intake (Table 33) already allows for growth rates that are 70% above the average (CV of growth taken as 35% over a 1-month period), it follows that within the safe level of intake the child with average protein requirement could still show a substantial increase in growth. To provide for a 2-fold increase in the growth rate of virtually all children may be an unattainable ideal.

Whatever target is set for a desirable increase in growth rate and the provision that should be made for it, the point remains that the relative increase must be greatest in the youngest children. This is extremely important when one considers the high morbidity and mortality between 6 and 9 months (24–27) and the fact that the risk attached to growth deficits is greater in children of less than 1 year than in older children (27).

Although the calculated increases in energy requirement are less, in practice the limiting factor may be the energy intake which can be achieved, because of the problems of bulk and energy density discussed in section 7.2.

The second line of evidence is derived from the long-term balance studies summarized in section 6.3.2 and Table 35. As pointed out in that section, the Consultation considered the protein intake in study A to meet the average protein requirement, but not the safe level. If the average protein requirement is accepted as 1.35 g per kg per day for these children aged 8–12 months, living under relatively controlled conditions, but exposed to some degree of infection and parasitization, the safe level would be 1.69 g per kg per day (1.35 × 1.25). This is 14% above the requirement calculated by the factorial method for children aged 9–12 months (Table 33).

It is not the purpose of this report to make recommendations about the application of requirement estimates, but rather to provide the biological basis and framework for practical recommendations. On the particular subject of this section it is impossible to make firm estimates and only an indication of the order of magnitude of the increases that might be needed by children exposed to infection can be given. If it is accepted on the basis of the long-term balance studies that children at 12–18 months need an approximately 20% higher protein intake, younger children would need more and older children less. Thus the increased requirement averaged over a population of preschool children would not be large. More detailed calculations are not warranted. Moreover, in many countries infections have a seasonal prevalence so that increases of the same magnitude may not be needed throughout the year.

In practice children should be fed according to their appetite. The important point is that in periods when the child is recovering and the appetite is restored, food should be available to make good the deficits caused by the anorexia and metabolic losses which accompanied the infection. As mentioned earlier, the bulkiness of the food may be a major factor limiting the intakes of adequate amounts of protein and energy by young children.

It should also be recognized that supplying the child's increased requirement is only one of the measures needed to counteract the effects of infection. The primary need is for prevention; infectious diseases should be controlled by improved sanitation and public health measures.

Because these interrelationships are complex, it is hoped that in the future independent measures of health and function could be used in a field study to test our assumptions about requirements under actual environmental conditions. A subgroup of the population with an apparently inadequate intake and level of performance can be selected. While continuing to live and work as before, they can then be provided with graded levels of the nutrient being examined to determine the extra amount needed to achieve adequate performance. Examples are growth studies in children in Papua New Guinea (28) and Colombia (29), who were fed supplements of protein, and experiments in India (30), Guatemala (31), and the Gambia (32) in which pregnant women provided with various amounts of food supplement were studied for weight gain during pregnancy and birth weight of the infant. Another opportunity to obtain this type of information is by the objective study of calamities such as wars, droughts, and famines.

In the planning of community studies to define requirements, the problem of measuring the exact home intakes should not be minimized. Knowledge of individual intakes is essential and these must be assessed by a reliable method over a period long enough to account fully for day-to-day variations, as well as seasonal fluctuations in food availability and intake.


  1. Energy and protein requirements: recommendations by a joint FAO/WHO informal gathering of experts. Food and Nutrition, 1(2): 11–19 (1975).
  2. Garrow, J.S. Treat obesity seriously. London, Churchill Livingstone, 1981.
  3. Barac-Nieto, M. ET AL. Am. J. Clin. Nutr., 31: 23–40 (1978).
  4. Waterlow, J.C. Lancet, 2: 87–89 (1973).
  5. Waterlow, J.C. & Rutishauser, I.H.E. In: Cravioto, J. et al., ed. Early malnutrition and mental development. Swedish Nutrition Foundation Symposia XII. Stockholm, Almqvist & Wiksell, 1974.
  6. Waterlow, J.C. Courrier, 38: 455–460 (1978).
  7. Chen, L.C. ET AL. Am. J. Clin. Nutr., 33: 1836–1845 (1980).
  8. Graitcer, P.L. ET AL. J. Trop. Pediat., 27: 292–298 (1981).
  9. Spady, D.W. ET AL. Am. J. Clin. Nutr., 29: 1073–1088 (1976).
  10. Jackson, A.A. ET AL. Am. J. Clin. Nutr., 30: 1514–1517 (1977).
  11. Waterlow, J.C. Proc. Nutr. Soc., 40: 195–207 (1981).
  12. Prader, A. Postgrad. Med. J., 54 Suppl: 133–146 (1978).
  13. Ashworth, A. Brit. J. Nutr., 23: 835–845 (1969).
  14. Rowland, M.G.M. ET AL. Brit. J. Nutr., 37: 441–450 (1977).
  15. Mata, L.J. ET AL. Am. J. Clin. Nutr., 25: 1267–1275 (1972).
  16. Condon-Paoloni, D. ET AL. Am. J. Pub. Hlth., 67: 651–656 (1977).
  17. Tomkins, A.M. ET AL. Trans. Roy. Soc. Trop. Med. & Hyg., 72: 239–243 (1978).
  18. United Nations University. Protein-energy requirements under conditions prevailing in developing countries: current knowledge and research needs. Tokyo, United Nations University, 1979, Report No. WHTR-1/UNUP-18.
  19. Brown, K.H. ET AL. Am. J. Clin. Nutr., 33: 1975–1982 (1980).
  20. Calloway, D.H. Rev. Infect. Dis., 4: 891–895 (1982).
  21. Martorell, R. ET AL. Am. J. Clin. Nutr., 33: 345–354 (1980).
  22. Cole, T.J. & Parkin, J.M. Trans. Roy. Soc. Trop. Med. & Hyg., 71: 196–198 (1977).
  23. Martorell, R. ET AL. Arch. Lat.-amer. Nutr., 27: 311–324 (1977).
  24. Mata, L.J. The children of Santa Maria Cauqué. Cambridge, MA, MIT Press, 1978.
  25. Puffer, R.R. & Serrano, C.V. Patterns of mortality in childhood. Washington, DC, Pan American Health Organization, 1975 (Sci. Publ. 262).
  26. Dyson, T. Wld. Hlth. Stat. Rep., 30: 282–311 (1977).
  27. Kielmann, A.A. & McCord, C. Lancet, 1: 1247–1250 (1978).
  28. Lampl, M. et al. Ann. Hum. Biol., 5: 219–227 (1978).
  29. Mora, J.O. ET AL. Nutr. Res., 1: 213–225 (1981).
  30. Belavady, B. Dietary supplementation and improvements in lactation performance of Indian women. In: Aebi, H. & Whitehead, R.G., ed. Maternal nutritional during pregnancy and lactation. Berne, Hans Huber, 1980, pp. 264–273.
  31. Lechtig, A. & Klein, R.E. Maternal food supplementation and infant health: results of a study in rural areas of Guatemala. In: Aebi, H. & Whitehead, R.G., ed. Maternal nutrition during pregnancy and lactation. Berne, Hans Huber, 1980, pp. 285–313.
  32. Prentice, A.M. Hum. Nutr. Clin. Nutr., 37: 53–64 (1983).

Previous Page Top of Page Next Page