6.3.1 Energy requirements
Although, in principle, it would be desirable to determine the requirements of children, in the same way as for adults from measurements of energy expenditure, this approach involves many difficulties in practice. Information is indeed available on the BMRs of children of all ages for which prediction equations are given in Table 5. However, in young infants, in whom the requirement for growth is a substantial component of the total requirement for energy, there are large variations within the normal range, in the rate of growth, and probably also in the composition of the tissue laid down. Moreover, for both infants and children, it is not possible to specify with any confidence the allowance that should be made for a desirable level of physical activity. We have therefore followed the example of the 1971 Committee (la) and estimated the energy requirements from birth to 10 years from the observed intakes of healthy children growing normally.
18.104.22.168 Infants (birth to 12 months). Up to 6 months of age the 1971 Committee (la) used the results collected by Fomon et al. (55) for the intakes of infants fed breast milk by bottle. For older children, they used figures for the intakes of children in the United States of America and the United Kingdom presented by Leitch & Widdowson to the second FAO Committee on Calorie Requirements (56). A much larger collection of information is now available on the intakes of infants, children, and adolescents compiled from studies in Canada, Sweden, UK, and USA (57). Results from developing countries were not included in this analysis to ensure that the intakes represent those of groups of children who, on average, were growing along the 50th centile of the WHO reference standard. For the first 12 months there are some 4000 data points available. The means of the measured intakes at each month for the first year are given in Table 21. That table shows a fall in energy intake per kg of body weight between 3 and 6 months which is maintained until 9 months, and then rises again towards 1 year. We believe this reduction to be a real phenomenon, representing a period when the very high growth rate characteristic of the first 3 months of life has declined but is not yet balanced by increased physical activity.
|Median body weightc||Total requirement|
a Observed intakes at ages indicated, from data of Whitehead et al. (57), omitting studies 7 and Fb on technical grounds. Average intake predicted from equation (age in months): I (kcalth/kg) = 123 - 8.9 age + 0.59 age2.
b Requirement over interval indicated, calculated as predicted intake +5% (see text).
c NCHS median weights at mid-point of month.
|Average during first year||103||112||430||470|
The intakes of the breast-fed infants in these studies were measured by test-weighing. Recent measurements of breast-milk consumption by the deuterium oxide method suggest that test-weighing underestimates actual milk consumption by about 5% (58). Estimates of the intake of other foods are likely to have a similar bias. The Consultation therefore accepted that the estimates of the energy requirements of infants should be set at 5% above the average observed intakes. The figures that result are still 10–15% lower than those proposed by the 1971 Committee (1a) (Table 22).
22.214.171.124 Children 1–10 years. In order to calculate the energy requirements of children over 1 year of age from measurements of energy expenditure, both the time and cost of all types of physical activity need to be known. Unfortunately, this information is not available for children in this age group, and it is therefore necessary, as for infants, to evaluate their energy requirements from data on dietary intake. Table 23 shows the mean intakes of boys and girls from 1 to 10 years. These values are derived from a critical review of the recent literature,1 and are based on studies in developed countries and in the more affluent groups of developing countries. The regression equations were calculated from 6500 data points for girls and 6000 for boys.
1 Ferro-Luzzi, A. & Durnin, J.V.G.A. The assessment of human energy intake and expenditure: a critical review of the recent literature. Rome, FAO, 1981 (Document ESN: FAO/WHO/UNU/EPR/81/9).
Table 23 shows that after 3–4 years the estimated intakes fall below the requirements for children proposed by the report of the 1971 Committee (la). The meaning of this may be that physical activity, and thus the energy requirement of children, has declined in recent times, reflecting increasingly sedentary life-styles in the large cities of industrialized countries. There is some evidence of this in older children (60). If true, such a reduction may well be considered undesirable from the point of view of optimal health and function, as maintenance of adequate levels of physical activity is thought to be necessary in the formative years of the growing child. Therefore the Consultation considered that recommendations for this age group should be based on observed intakes, but that they should be increased by 5% to allow for a desirable level of physical activity.
Little information is available in the literature on the energy intakes of children in the developing countries, who may be expected to lead a more active life, having to walk long distances, undertaking hard physical work, sometimes being less tied to sedentary activities by strict school schedules, etc. Their energy intakes are smaller than those of their counterparts of the same age in developed countries, but the difference largely disappears when the intakes are recalculated on a body-weight basis. One might postulate that in this case, externally imposed limitations may restrict both energy intakes and energy expenditure. The extra 5% added to observed intakes may be considered a realistic estimate of the requirement if enough energy is to be available for a desirable level of physical activity.
126.96.36.199 Children 10–17 years. After 10 years of age it becomes feasible to base requirements on estimates of energy expenditure built up by the factorial method. The approach is basically the same as for adults.
As in adults, the BMR in children is generally the largest component of the requirement. Because of differences in the timing of the pubertal growth spurt, both weights and heights at any given age are rather variable. For example, in 12-year-old boys, weight may vary from 28 kg (3rd centile) to 59 kg (97th centile), and height from 1.36 to 1.64 m. The best predictor of the BMR is weight and, as in adults, the BMR per kg varies with body weight. At any given weight, variations in height make no difference in boys and only a small difference in girls (at a given weight the predicted BMR changes by less than 5% as one moves from the median height to the 3rd or 97th centiles, Annex 1). As discussed in section 3.5.1, the BMR may be estimated either from the actual weight or from the median weight for height and age, as shown in Baldwin's table (Annex 2(B)). Values for the BMR of adolescents at median weight for height and age are given in Table 24.
|Age||Boys||Girls||Requirement by weightc|
|(years)||(kcalth/day)||(MJ/day)||(kcalth/day)||(MJ/day)||(kcalth/day)||(MJ/day)||(kcalth/ day)||(MJ/day)||(kcalth/ day)||(MJ/day)||(kcalth/ day)||(MJ/day)||(kcalth/kg per day)||(kJ/kg per day)||(kcalth/ kg per day)||(kJ/kg) per day)|
|1–2||1 140||4.76||1 200||5.02||1 180||4.93||1 090||4.56||1 140||4.76||1 180||4.93||104||435||108||452|
|2–3||1 340||5.60||1 410||5.89||1 360||5.69||1 250||5.23||1 310||5.48||1 350||5.64||104||410||102||427|
|3–4||1 490||6.23||1 560||6.52||1 560||6.52||1 370||5.73||1 440||6.02||1 520||6.35||99||414||95||397|
|4–5||1 610||6.73||1 690||7.07||1 720||7.19||1 465||6.12||1 540||6.44||1 670||6.98||95||397||92||385|
|5–6||1 720||7.19||1 810||7.57||1 870||7.82||1 550||6.48||1 630||6.81||1 790||7.48||92||385||88||368|
|6–7||1 810||7.57||1 900||7.94||2 010||8.40||1 620||6.77||1 700||7.11||1 900||7.94||88||368||83||347|
|7–8||1 895||7.92||1 990||8.32||2 140||8.95||1 685||7.05||1 770||7.40||2 010||8.40||83||347||76||318|
|8–9||1 970||8.24||2 070||8.66||2 260||9.45||1 740||7.28||1 830||7.65||2 110||8.82||77||322||69||268|
|9–10||2 045||8.55||2 150||8.99||2 380||9.95||1 795||7.51||1 880||7.86||2 110||8.82||72||301||62||259|
a From data of Ferro-Luzzi & Durnin (see footnote 1 on page 92).
b Intakes +5% (see text).
c From NCHS median weights at mid-year.
a Median height for age from NCHS standards.
b Median weight for height and age from Baldwin's standards (Annex 2(B)).
c Boys:BMR = 17.5 W+651 kcalth/day (2.72 MJ/day). Girls:BMR = 12.2 W+746 kcalth/day (3.12 MJ/day).
For growth, an addition of 5 kcalth (21 kJ) per g is allowed for the average daily cost of weight gain (61). It is recognized that growth does not occur at a regular rate from day to day (section 3.2). However, even during the pubertal spurt the requirement for growth is so small compared with the total energy requirement that no extra allowance in the energy intake need be made for this variability.
In order to illustrate quantitatively the energy requirements for different patterns of activity, an attempt was made to identify the time that might be spent by each sex and age group sleeping, going to school (including homework), and undertaking light, moderate, and heavy physical activity. There is little information in the literature on the amount of time children and adolescents spend on different types of activity, but these are bound to be highly variable. The time allocations were estimated as the daily means throughout the year, assuming that the average child still goes to school at the age of 10. It was also assumed that the child begins secondary school at the age of 13. In societies where children begin work rather than go to secondary school, patterns of activity should be calculated accordingly.
The values shown in Table 25 represent the best estimates of levels of activity at different ages that would be compatible with a good rate of growth and optimal development and health in children of appropriate weight and height for their age.
In calculating energy expenditure, it has been assumed that the energy cost of sleep is equal to the BMR. Estimates of the additional energy costs of other activities, over and above the BMR, were based upon the principles described in section 4. The gross energy costs were assessed as follows in terms of BMR units:
|Going to school and light activity||1.6 BMR||1.5 BMR|
|Moderate activity||2.5 BMR||2.2 BMR|
|High activity||6.0 BMR||6.0 BMR|
The somewhat lower values given for girls assume that their intensity of activity would decline to the level found in adult women (see section 6.2).
Table 26 shows an example of how energy expenditure was calculated from timed activities of a 10½-year-old boy in a developing country. Table 27 shows a similar but less detailed calculation for a girl in an industrialized country. The calculations for both sexes and for ages 10–17 years are shown in Annex 7. Table 28 shows, for each age group, values for energy needs based in this way on an estimate of energy expenditure plus the increment for growth, together with the relationship to BMR. The BMR factor varies over a rather narrow range, from about 1.6 to 1.75 in boys and from 1.5 to 1.65 in girls.
The new estimates of energy requirements, based on calculated energy expenditure, are compared in the table with observed energy intakes of adolescents. The 1971 Committee's recommendations (1a) are also shown to serve as a reference. It is obvious that the new estimates are appreciably and consistently lower than those proposed by that Committee and that the observed values of energy intakes are lower still. Between 10 and 18 years the new estimates of energy requirements of boys exceed the actual figures of observed energy intakes by an amount that corresponds almost exactly with the amount of energy thought desirable for children to spend in high activity (½ hour at 6.0 × BMR). In girls the discrepancy is greater, presumably reflecting a low level of physical activity in the sample whose intakes were measured. The Consultation considered that the estimated requirements for this age group should not be decreased to match the observed intakes in affluent countries. Fulfilment of the requirements as proposed is likely to be beneficial if physical activity is increased, and in developing countries will provide a margin of safety.
a Average over whole year.
|Sleep at 1.0 × BMRa||9||455||1900|
|School at 1.6 × BMR||2.5||200||840|
|Light activity at 1.6 × BMR:|
|sitting, standing, moving around||6.5||525||2200|
|social activities, washing, play||2||160||670|
|Moderate activity at 2.5 × BMR:|
|walking, household tasks, agricultural tasks, play||3||380||1590|
|Heavy activity at 6.0 × BMR:|
|fetching wood and water, agricultural tasks||1||300||1260|
|Total requirement per 24 hours||2080||8710|
|=1.71 × BMR|
a BMR estimated to be 1215 kcalth/day (5080kj/day).
|Sleep at 1.0 × BMRa||9||435||1820|
|School at 1.5 × BMR||4||290||1210|
|Light activity at 1.5 × BMR||4||290||1210|
|Moderate activity at 2.2 × BMR||6.5||690||2890|
|High activity at 6.0 × BMR||0.5||145||610|
|Total expenditure||1 850||7740|
|Total requirement per 24 hours||1 915||3010|
|= 1.65 × BMR|
a BMR estimated to be 1160 kcalth/day (4850kJ/day).
a Expenditure calculated as in Tables 26 and 27 Annex 7. BMR from equations in Table 5.
b Intakes from reference (62).
c Reference (1a).
6.3.2 Protein requirements
188.8.131.52 Infants from birth to 6 months. As in the case of energy, the 1971 Committee based its estimate of the protein requirements from birth to 6 months on intake data because of the difficulties of accurately allowing for growth and maturation. Many observations show that infants breast-fed by healthy well nourished mothers (62–65) or fed breast milk by bottle (55) can grow at a satisfactory rate for 4–6 months using the standards adopted in this report. It may therefore be concluded that for the first 6 months of life the protein needs of an infant will be met if its energy needs are met and the food providing the energy contains protein in quantity and quality equivalent to that of breast milk.
The average protein content of human milk, calculated as total N × 6.25, has been taken as 1.15 g per 100 ml after the first month of lactation1. It is recognized that human milk contains about 40 mg of non-amino nitrogen per 100 ml (approximately 20% of total N) (66, 68) but, following the usual convention, we have assumed that this nitrogen fraction is utilized. If the average energy content is taken as 70 kcalth (290 kJ) per 100 ml, the protein content is 1.64 g per 100 kcalth (6.85 g per MJ).
|(months)||(ml)||(g/day)||(kg)||(g/kg per day)||(g/kg per day)|
a From Table 20. In accordance with the findings of Wallgren (69) and Whitehead (63), the breast-milkconsumption by female infants is taken as 8% less than that of male infants.
b NCHS median, mid-point of months.
c From reference (1a), Table 15 (data of Fomon).
An estimate of the average protein intake per kg in breast-fed infants up to 4 months is shown in Table 29. Although the amount of milk consumed by infant girls is less than that by boys, on a body-weight basis the intakes are virtually the same. For comparison, the table also shows the average protein intakes of infants who were fed breast milk by bottle (55). These were the data on which the 1971 Committee based its estimates of protein requirements for infants up to 6 months old.
1 The values adopted, 1.15 g of total N and 70 kcalth (290 kJ) per 100 ml of human milk are based on the compilations of Macy (66), the Department of Health and Social Security, United Kingdom (52), and the WHO Collaborative Study on Breast-feeding (67).
Estimates of the protein intake of breast-fed infants are not shown after the age of 4 months, because from this age there is insufficient information on the intakes of exclusively breast-fed infants who are growing satisfactorily. It may be noted that if the estimates of energy requirements shown in Table 21 are correct, it would need 1040 ml of breast milk to fulfil the energy needs of a male infant at 5–6 months. This volume is larger than has usually been observed.
For the purpose of comparison with older infants, it may be noted that if the average daily energy intake at 6 months is 95 kcalth (400 kJ) per kg (Table 22), the corresponding average protein intake from breast milk will be 1.56 g per kg per day.
184.108.40.206 Children from 6 months onwards. The period from 6 to 12 months is clearly the most critical, because of rapid growth during this time and because the child increasingly relies on supplementary foods. The first priority must therefore be to establish as reliable an estimate as possible of the safe level of protein intake for children of this age. One can then with some confidence interpolate the safe level for older children from this estimate.
The protein requirements of children have been calculated in the first instance by a modified factorial method. As with adults, and with the same limitations, the maintenance requirements can in principle be estimated from measurements of nitrogen balance at several levels of intake. There is the additional problem of determining an appropriate value for the N retained during growth and the requirement for achieving this retention.
(a) Maintenance requirement. Several short-term N-balance studies have been carried out in older infants and young children to determine the protein requirement by the slope-ratio method (see section 5). These have been healthy children, usually recovered from malnutrition, short in stature but in the normal range of weight for height. The usual design follows that of adult studies: energy intake is kept constant at a level assumed to be adequate and protein is given at different levels, each for a period of several days. From these studies it is possible to calculate a maintenance requirement (no growth, N equilibrium) from the regression of N balance on N intake, allowing 10 mg of N per kg per day for sweat and miscellaneous N losses. In theory it is also possible from these studies to determine the amount of protein needed in the diet to meet any chosen value for N retention.
The results in Table 30 indicate that in the critical age group from 6 to 20 months, with milk as a source of protein, the average maintenance requirement is approximately 115–120 mg of N per kg per day. Studies of a different design, in which N balance was measured only once or twice in each child over a range of intakes, gave a maintenance requirement of 110 mg of N per kg per day (75).
|6–20 months||17–31 months||38–62 months|
|Protein source||Milk||Egg||Milk||Soy||Rice, fish||Milk||Soy||Mixed|
|No. of children||24||10||10||10||7||7||7||3|
|Intake range (/kg/day)|
|• Energy (kcalth)||61–92||69–88||100||100||110||100||100||100|
|• N (mg)||16–173||71–150||80–320||80–320||120–320||120–240||120–240||160–320|
|Average maintenance requirementb|
|(mg N/kg per day)||117||120||80||98||130||91||137||164|
|Corrected for digestibilityc||117||117||80||97||98||91||100||118|
a Column 1 = reference (70); column 2 = reference (70); column 3 = reference (71); column 4 =reference (71); column 5 = reference (72); column 6 = reference (73); column 7 = reference (73); column8 = reference (74).
b Allowing 10 mg of N/kg for skin losses.
c Corrected for digestibility of milk, where appropriate.
The results of the balance studies in the older age groups are somewhat variable. The estimated maintenance requirement, corrected to the digestibility of cow's milk, ranged from 80 to 118 mg of N per kg per day. Since the average maintenance requirement in young adults studied in short-term balances was estimated at 98 mg of N per kg per day (section 6.1.2), it seems reasonable, for intermediate age groups, to interpolate between two well-established values. These, rounded off, would be 120 mg of N per kg per day at 1 year, falling to 100 mg of N per kg per day at 20 years.
In the balance measurements on children the range of inter-individual variation in the estimate of the maintenance requirement was similar to that found in the much more numerous short-term balances in adults, and the Consultation has therefore accepted the same value of 12.5% for the coefficient of variation of the maintenance requirement.
It is assumed that this represents variability in the efficiency of utilization—an assumption that is important when the requirement for growth is being considered.
(b) The requirement for growth. The mean rate of N accretion during growth can be estimated from the expected daily rate of weight gain (NCHS 50th centile) and the N concentration in the body. This is low at birth, and increases to the adult value by 5 years of age or earlier. The extent of the increase is important between 6 and 12 months, when growth is rapid. The 1971 Committee (1a) reported values for body N concentration at different ages obtained by three different methods. At some ages the values were not in good agreement.
More recent estimates of N accretion during growth have been provided by Fomon et al. (76). They are similar to those accepted by the 1971 Committee (la, Table 12) but lower than some earlier estimates (77).
However, as pointed out in section 3.2.1, it should not be assumed that growth always proceeds at exactly the same rate from day to day, even in apparently healthy children. The cause, extent, and significance of these fluctuations in growth rate are difficult to assess. Table 31 illustrates the extent of variation in weight gain that has been observed over a period of 4 weeks in healthy children aged 4–6 months (78; and S.J. Fomon, personal communication).
|Period (days of age)|
|Weight gain (g/day)|
|Weight gain g/100 kcalth (g/1000 kJ)|
|10th centile||1.7 (4.1)||1.3 (3.1)||0.8 (1.9)|
|50th centile||2.5 (6.0)||2.5 (6.0)||2.1 (50.2)|
|90th centile||3.8 (9.1)||3.4 (8.1)||3.0 (7.1)|
|mean||2.6 (6.2)||2.4 (5.7)||2.0 (4.8)|
|SD||1.0 (2.4)||1.0 (2.4)||0.9 (2.2)|
|CV (%)||38 (90.8)||42 (100)||45 (107.6)|
|Energy intake kcalth/day (kJ/day)|
|10th centile||78 (326)||74 (310)||72 (301)|
|50th centile||95 (397)||95 (398)||89 (372)|
|90th centile||115 (481)||114 (477)||109 (456)|
|mean||96 (402)||96 (407)||90 (377)|
|SD||14 (59)||15 (63)||15 (63)|
|CV (%)||15 (63)||16 (67)||17 (71)|
a Unpublished data of Fomon.
The variability of gain is much greater than the variability of intake, so that presumably it results to a large extent from variation in the efficiency of utilization of food for growth.
Longitudinal studies in Jamaica on reasonably healthy and well nourished children showed that over successive months a period of faltering would usually be followed by a period of weight gain at 2–3 times the normal rate (79). Very little is known about variations over shorter periods. Daily measurements of children's weights even under standardized conditions in a metabolic ward show fluctuations, which no doubt resulted partly from differences in the amounts of urine and stool retained at the moment of weighing.
They may also represent differences from day to day in the proportions of fat and lean tissue laid down. It is known that individual children, at least when recovering from malnutrition, may differ widely in the composition of weight gain (80).
If it is accepted that different amounts of protein may be laid down from day to day, as part of the normal process of growth, the question then arises, what is the effect of this on the child's daily protein requirement? In order to maintain a satisfactory overall rate of growth, any failure to lay down protein on one day must be compensated for on a subsequent day. Studies such as those cited in section 5 suggest that the body has a very limited capacity for storing amino acids or for drawing on the free amino acid pool for protein synthesis. Even during short periods such as 12 hours without food, nitrogen balance becomes negative (81). Therefore, in accordance with classical teaching, it seems very unlikely that amino acids provided on a day when there was no growth could be held “in stock” to be utilized for growth later on. It follows that since it is impossible to foretell on which days the growth rate will be low or high, it is necessary to provide enough every day for a possible extra demand.
There is no evidence available on which to base a realistic estimate of the extra requirement for protein that might arise in this way. In this situation a judgement has to be made. A reasonable judgement must lead to estimates that are similar to values established independently, such as the intakes of healthy breast-fed children. It was found that if, in the factorial calculation the growth component of the protein requirement is set at 50% above that based on the theoretical daily amount of N laid down, the calculated average requirement at 4 months comes close to the average intake of breast-fed infants (Table 32). To provide a physiological margin of safety, it was therefore decided to increase the theoretical growth requirement by a factor of 50%.
|(mg N/kg per day)||(g protein/kg per day)|
a Although body weights differ between male and female infants, it is not considered that requirements per kg will differ.
b From reference (76).
c See text.
d Efficiency of utilization assumed to be 70%.
e See text.
f From Table 29.
The amount of dietary nitrogen needed to allow for a given amount of N deposited can be derived from the same slope as the requirement for maintenance. It is assumed that dietary protein is used with the same efficiency for growth as it is for maintenance, and on theoretical grounds there is no reason to suppose that this assumption is not valid (section 5.4). The appropriate slope for diets based on milk or egg has been taken as 0.7 at all ages.
Table 32 shows the detailed calculation of the average requirement of protein for infants up to 1 year, over the period when growth makes a significant contribution. Although the factorial method is not, in fact, used for infants below the age of 6 months, the calculations have been made in order to compare the results with the estimated protein intakes from breast milk (Table 29). This comparison suggests that the proposed addition of 50% to the nitrogen increment does not unrealistically increase the estimate, and it may even be too small. This is a subject on which more research is urgently needed.
Finally, a correction has to be made for the inter-individual variability of growth, in order to arrive at a safe level of protein intake for virtually all healthy children. The coefficients of variation shown in Table 31 represent a mixture of inter-and intra-individual variability over periods of 4 weeks. This interval should be long enough to smooth out the effects of day-to-day fluctuations in the growth of each individual child, as discussed above. It is apparent that over a period of a month children do vary in their average daily rate of growth, with a CV of approximately 37%. The CV will be lower over longer periods and higher over shorter periods. If the one-month CV is accepted as a reasonable compromise, an overall CV can be calculated, as shown in the footnote to Table 33. This is 15% at 6–9 months, falling to 12.5% in the second year. The data do not justify more detailed estimates.
It is apparent that, at present, we do not have an adequate theoretical basis for calculating the variability of the protein requirement for growth in young children, a major problem being that this variability depends on the length of the period over which the growth is measured. It is hoped that this discussion of the problem will stimulate further research.
The estimated safe levels calculated in this way are shown for children in Table 33, and for adolescents in Table 34, and compared with those of the 1971 Committee (la). In adolescents the requirements per kg have to be given separately for the two sexes, because of differences in the timing of the growth spurt. In younger children there is little difference from the earlier figures, although the ways in which the values have been derived are different. From 6 years onwards the present estimates are somewhat higher than the earlier ones. This reflects the higher current estimate for adults, since the maintenance requirement has been calculated by linear interpolation between infants and adults. Much of the difference disappears when revised corrections are made for protein score and digestibility (section 7.3).
|Maintenancea||Growthb||Total||+2 SDc||CV%||Safe level||1971|
|(mg N/kg per day)||(g protein/kg per day)|
a See text.
b After the addition of 50% to theoretical increment, and correction for 70% efficiency of utilization (see Table 32).
c CV for maintenance taken as 12.5%; CV for growth taken as 35%. Combined CV (CVtotal) calculated as:
d Reference (1a).
|Maintenance||Growth||Total||+2 SD||Safe level||1971|
|(mg N/kg per day)||(g protein/kg per day)|
Calculations and notes as for Table 33.
The growth component in Table 32 has been derived from theoretical values for N increment, whereas nearly all balance studies in children on usual intakes, many of which were well above the requirement, have shown that they retain more than the theoretical amounts of N (82–84). Although these studies differ in type of subject and amount and kind of protein, they agree in showing apparent N balances greater than those expected from measurements of body composition in man or carcass nitrogen in animal experiments.
The discrepancies between observed and expected balance are smallest when N retention is high, such as in infants and also in children during catch-up growth (75). They are largest when intakes are high and retention is expected to be low, such as in older children during normal growth. This suggests an inherent problem of methodology. It is well known that the errors of the N balance method summate to exaggerate the apparent retention (85).1
1 An alternative approach examined by the Consultation was to estimate the protein requirement from the basal metabolic rate—a procedure analogous in principle to the use by the 1971 Committee of the factor 2 mg N per basal kcal to estimate the maintenance requirement. Since reliable estimates of the maintenance requirement have now been obtained from N-balance measurements, calculation from the BMR would only be useful to provide values for the total requirement. However, the BMR per kg changes little between 6 months of age and 3 years (Table 5), whereas the growth rate rapidly falls off. Thus changes in BMR with age in no way reflect the changes in growth rate, and hence in protein requirement, that occur in infancy and to a lesser extent in puberty. This approach, therefore, cannot be considered useful.
Longer-term balances. As emphasized earlier (section 5), although short-term N-balance studies provide valuable information, the conditions are artificial and the conclusions drawn depend heavily on a series of largely unverified assumptions. Long-term studies during which children grow normally would clearly inspire much greater confidence regarding the adequacy of the diets fed. Details of the few available long-term balances at a single level of protein intake are shown in Table 35. The diets fed in these long-term studies were controlled at a fixed level and were composed of foods commonly eaten by poor people in the countries represented. All except the oldest group were poor children and most had a previous history of malnutrition and stunting. Some were parasitized and all had minor febrile and afebrile illnesses during the studies. The weight gain was in general satisfactory, but some children did not gain weight at the expected rates at some periods. It is therefore questionable whether the diets fed should be thought of as meeting average requirements or constituting a safe level for groups of children growing at individually variable rates, including short intervals of little or no gain.
Interpretation of the long-term balance figures is difficult in view of the different age ranges covered by the various studies.
In the long-term balance study on children aged 8–12 months (Table 35, study A) an intake of 1.35 g of protein per kg per day (after correction to the digestibility of cow's milk) supported satisfactory growth in almost all of the children for most of the time, but not in all the children at all times. It may, therefore, be reasonable to consider this as an average requirement for children at this age under the conditions of the study. This value is 8% higher than the average requirement of 1.25 g per kg per day at 6–9 months derived by the factorial method (Table 32). Addition of 2 SD (CV 12.5%) would give a safe level of 1.75 g of protein per kg per day. This may be regarded as a realistic estimate of the safe level of protein intake for children of 6–9 months in a developing country where the child is exposed to infections and perhaps periodic shortages of food. Under these conditions, it may be wise to adopt an estimate of the safe level of protein intake that has been derived from studies carried out in a comparable situation. The question of the additional demands imposed by these stresses is discussed in more detail in section 9.
|Study A||Study B||Study C||Study D||Study E||Study F|
|(ref. 86)||(ref. 87)||(ref. 83)||(ref. 84)||(ref. 88)||(ref. 88)|
|Age (x)||8–12(10) months||22–40(30) months||29–46(30) months||2–5 years||7–9 yearsa|
|No. of subjects, sex||6 M||6 M||11 M||20 F + M||13 F||6 F|
|Duration of study (days)||90||120||77||180||48||42|
|Weight gain (g/day)||9.4b||11c||7.2±8.2d||-e||NA||NA|
|95% beans + corn|
|82% beans +|
corn + veg.
|wheat or rice + veg.||mixed, 45% animal||all vegetable|
|Intake/kg per day energy, kcalth|x||88–93(90)||82–91(86)||79–93(85)||100||78||80|
|Apparent digestibility (%)f||73||59±9||72±5||66||80||77|
|Protein intake corrected|
to digestibility of cow's milk (g/kg)
|Crude N balanceg|
|(mg/kg per day)||71–100||68±7||90±22||100||25(13–43)||28(13–42)|
a Data for diets 8 and 11–12 of the published study (88).
b 1 child did not gain weight at an adequate rate.
c 4 children had intestinal parasites and all had mild upper respiratory tract infections during the study. All children gained weight and 3 showed catch-up linear growth.
d 4 children had febrile infections during the study, 5 had other afebrile illness, and a few vomited occasionally. 1 child did not gain weight and 2 gained at less than the expected rate. 5 children had normal linear growth and 1 showed catch-up growth.
e Children described as healthy and growing according to US standards.
g Intake - faecal N - urinary N.
In the long-term studies on preschool children the diets fed included a large proportion of plant food, so that digestibility was below that recorded for diets based on milk and eggs. Once the factor of digestibility is taken into account, the amount of protein that appears to support expected growth rates of 2–5-year-old children in the long-term studies (Table 35, studies B,C, and D) does not differ markedly from the requirement of high-quality protein predicted from the short-term studies. This observation indicates that the amino acid composition of practical diets is not necessarily a limiting factor for preschool children when consumed in the proportions used in the long-term studies. However, the habitual home diets of some populations may provide the same foods in different proportions, and their constituent proteins may not supply an adequate combination of essential amino acids. Under these circumstances the poor protein quality may require higher intakes. The margin of safety is obviously less for children than for adults, and the range of national diets needs to be examined with respect to amino acid content as well as digestibility, before concluding that no further adjustment for these factors is needed. How this is to be done is shown in section 7.3.
Long-term balance studies provide little information that might allow a firm estimate to be made of the safe levels of intake of school-age children and adolescents. A sample of data from studies of 7–9-year-old girls is included in Table 35 (studies E and F).
In the United States of America a single study of 14–15-year-old boys (89) indicated that an intake of 100–120 mg N per kg per day from a mixed diet was needed to produce consistently positive N balances. This represents an average intake, not corrected for digestibility, of 0.62–0.75 g of protein per kg per day in boys towards the end of the pubertal growth spurt.