Even in the steady state, body proteins constantly undergo breakdown and resynthesis. When growth is occurring, not only is there a net deposition of protein, but the rates of both synthesis and breakdown are increased (1–3). The principles underlying this process of protein turnover have been described in detail elsewhere (4).
The rates of turnover vary from tissue to tissue, and the relative contributions of different tissues to total protein turnover change with age and adaptation to various levels of protein intake (4–7). The amino acids released by breakdown are reused for protein synthesis. However, because this process of reutilization is not completely efficient, some amino acids being lost by oxidative catabolism, both essential amino acids and a dietary source of nitrogen are needed. The daily turnover of body proteins is, in fact, several-fold greater than the amino acid intake, showing that the reutilization of amino acids is a major contributory factor to the economy of protein metabolism (4, 5).
This process of recycling, which includes interchange of amino acids between tissues as well as intracellular reutilization, depends on various metabolic and hormonal factors and is influenced by the physiological status of the host. Thus, reutilization of amino acids is highly efficient during rapid catch-up growth and in convalescence from a catabolic episode resulting from injury or infection (4). The increased efficiency of reutilization under these circumstances results in the improved use of dietary amino acids, giving the dietary protein an apparently improved biological value (8).
This principle extends to the adaptation of protein metabolism under circumstances of restriction or excess of dietary protein or amino acid supply. Adaptation to a submaintenance level of protein intake leads to a diminished turnover of tissue protein and a reduced rate of catabolism of the amino acids liberated by protein breakdown (9, 10). In this way, within limits, the tissue protein pool can reach a new steady state appropriate to the diminished intake of protein.
Under the experimental conditions of a protein-free diet, protein synthesis and breakdown continue via the reutilization of amino acids. This process becomes very efficient, but a small proportion of amino acids are still catabolized to urinary nitrogenous compounds and there is some nitrogen loss in the faeces. These represent what has been called the obligatory loss (11–13). The extent of this loss and its various components is discussed in section 5.5.
It has become clear since the 1971 Committee's report (12) that a key question in relation to the assessment of protein requirements is the extent to which people living on low protein intakes can adapt by increasing the efficiency of recycling and reducing the extent to which amino acids “escape” from the system and are catabolized. One objective, therefore, in determining protein requirements is to define the point at which adaptation is exceeded; beyond this point there will be progressive loss of body protein and deterioration of tissue function.
As has been fully discussed in previous reports, the protein requirement has two components—that for total nitrogen and that for essential amino acids—so that a diet may be deficient in quantity or quality of protein. Both aspects may, in theory, be affected by adaptive processes, but almost nothing is known about adaptation of essential amino acid requirements.
All healthy individuals are able to adjust total nitrogen (N) excretion to balance their intake over a certain range. For a time, as the lower limit of this range is approached, body N loss exceeds N intake, and there is a reduction in the mass of body protein leading to a new steady state. This is clearly a form of adaptation. At even lower intakes the limits of adaptation are exceeded, and there will be a continued depletion of body protein resulting ultimately in death. Two questions have to be considered:
(a) First, in passing from one steady state to another there is a loss of body protein; is this of any functional significance? When an adult man is transferred from a higher than customary N intake to one that is close to the physiological minimum, e.g., from 14 to 4 g N per day, a level of excretion close to a new steady state is reached in about 7 days (14), and over this period there is a cumulative N loss equivalent to about 1.5% of total body N. In similar studies on children the new steady state was reached more rapidly, with a total loss estimated at 1% of body N (15). In experimental animals the losses on passing from a customary to a lower protein intake are relatively greater. From such experiments it is known that initially most of the N is lost preferentially from the liver and gut (16, 17). Later, because of the recycling discussed above, it is the N content of muscle and skin that is mainly reduced (18).
It seems doubtful whether a loss of 1–2% of body N in man could represent a significant degree of depletion rather than adaptation. Experimentally, the capacity of animals maintained on low protein intakes to respond to stresses of various kinds does not seem to be impaired (19), except for an increase in perinatal mortality (20). Systematic evidence on this question is not available in man, except in children who are demonstrably malnourished (21), because of the difficulty in matching compounding variables other than nutritional state. Nevertheless, it is not justifiable, given the present state of knowledge, to assert that there are no functional differences between steady states at lower and higher protein intakes. For example, albumin synthesis and breakdown rates, albumin pool size, and perhaps plasma concentration, are somewhat lower at low intakes (22, 23).
(b) The second question is whether, on a habitually low protein intake, it is possible to reduce the lower limit of the range of protein intake at which N equilibrium can normally be maintained. Obligatory urinary N losses in subjects in different countries, presumably with different habitual protein intakes, are remarkably uniform (Table 3). Thus, the way to reduce the minimum requirement for nitrogen balance is probably to increase the efficiency of utilization of amino acids, i.e., an increase in the recapture of amino acids for protein synthesis, as discussed above, and a decrease in amino acid oxidation.
In some of the long-term balances summarized in section 6 (Table 17), in which protein was fed at about the same level as the estimated requirement, some subjects showed a trend towards a lower urinary N output at a fixed intake, but in others there were no such trends. Moreover, in spite of an earlier observation (38) that Nigerian farmers could achieve balance with an intake that was apparently inadequate for North American men, several short-term balances that have been carried out since 1971 reveal no striking differences in the estimates of maintenance requirement in relation to body cell mass obtained in studies of well-nourished subjects in different countries. However, the question of long-term adaptation cannot be answered definitively without further research.
Although it has been shown in both animals and man (39, 40) that enzymes of amino acid metabolism adapt to changing levels of protein intake, there is no evidence that their activity can be reduced to zero, allowing for 100% reutilization of amino acids. It is possible that some essential amino acids may be more efficiently conserved than others(e.g., lysine). From these considerations it is evident that at the maintenance level the limiting amino acid will be the one that is least efficiently conserved.
The question has been raised as to how far recycling of urea through the gut could contribute to amino acid economy. The ammonia liberated by hydrolysis in the colon is available for the formation of nonessential amino acids (41). However, this process cannot result in any net increase in protein synthesis unless there is a parallel increase in the availability of the essential amino acids or their carbon skeletons—a situation that is unlikely to exist under natural conditions (42).
Given the evidence currently available, it must be concluded that there is probably only limited scope for metabolic adaptation to N intakes below the physiological minimum for N balance found in subjects with “normal” intakes. The important question that remains is whether the degree of adaptation that does occur, consistent with the maintenance of nitrogen equilibrium, represents a metabolic adjustment without functional significance, or whether it is detrimental to health and long-term survival. The answers to these questions will obviously influence the nutrition of populations subsisting on diets providing low levels of protein. They are also relevant to the determination of protein requirements in healthy, well-nourished individuals.
Subjects and country | Reference | No. of subjects | Age | Body weight | BMR/kg | Excretion (mg N/kg) | Total N | ||||
(kg) | (kcalth) | (kJ) | Urine | Faeces | Totala | (mg/kcalth) | (mg/kJ) | ||||
Infants and children | |||||||||||
(24) | 7 (f) | 4–6 months | 7.3b | 54 | 226d | 37 | 20 | 67 | 1.24 | 0.30 | |
(25) | 11 (m) | 9–15 months | 10.2b | 56 | 234d | 54 | 22 | 86 | 1.54 | 0.37 | |
(26) | 5 (m) | 17–31 months | 12.6b | 57 | 238d | 34 | 20 | 64 | 1.12 | 0.27 | |
Young adults | |||||||||||
Women | |||||||||||
USA | (27) | 20 | 23 years | 59 | 23 | 96d | 25 | 8 | 41 | 1.78 | 0.42 |
(28) | 11 | 22 years | 60 | 23 | 96d | 31 | 8 | 47 | 2.04 | 0.49 | |
Men | |||||||||||
USA | (29) | 13 | 20 years | 71 | 26 | 109 | 38 | 14 | 60 | 2.31 | 0.55 |
USA | (30) | 83 | 21 years | 73.5 | 21 | 88 | 37 | 9 | 54 | 2.57 | 0.61 |
China | |||||||||||
(Province of Taiwan) | (31) | 50 | 23 years | 55 | 26 | 109 | 33 | 13 | 54 | 2.08 | 0.50 |
India | (32) | 4 | 27 years | 46 | 27 | 113 | 38 | 23 | 69 | 2.56 | 0.61 |
Nigeria | (33) | 9 | 26 years | 54 | 26 | 109 | 34 | 23 | 65 | 2.56 | 0.60 |
Japan | (34) | 9 | 63 | 26 | 109d | 33 | 13 | 54 | 2.08 | 0.50 | |
Elderly persons | |||||||||||
Women | |||||||||||
USA | (35) | 11 | 77 years | 63.5 | 17 | 71 | 24 | 10 | 42 | 2.47 | 0.59 |
Men | |||||||||||
USA | (36) | 6 | 68 years | 83c | 19 | 79 | 27 | 10 | 45 | 2.37 | 0.57 |
USA | (37) | 8 | 70 years | 71.6 | 22 | 92 | 34 | 12 | 54 | 2.45 | 0.59 |
b Assumed median weight (NCHS) at mid-point of age range.
c Body weights were 8–19 kg above the expected weight for height.
d BMRs estimated from equations in Table 5. Others were measured by the investigators.
At the other end of the scale, it is necessary to define the limits of successful adaptation to high protein intakes. As pointed out in section 3.1, it would not be justified to assume that high intakes are automatically optimal. It is known that excessive protein intakes are accompanied by modest elevations in blood urea nitrogen (43), which facilitates urea excretion, and by an increase in urinary calcium content (44, 45). Low-birth-weight infants fed very high levels of protein (5–6 g/kg per day) have in some instances experienced a reduction in growth rate, urine abnormalities, lethargy, and fever (46, 47), and some studies suggest an impairment in neurological development (48).
The processes of protein synthesis and possibly of breakdown (turnover) require sources of dietary energy and are thus sensitive to energy deprivation. Consequently, the energy balance of the body becomes an important factor in determining nitrogen balance and influences the utilization of dietary protein.
The magnitude of the basal energy needs and of the total amount of protein turned over in a day are both related to active tissue mass (4, 49, 50). Moreover, in young animals and growing children both rates per unit of active mass are increased compared with those observed in adults (51). Nevertheless, as discussed in the next section, it has not proved possible to establish a constant numerical relationship, covering all age ranges, between BMR and either protein requirement or obligatory nitrogen loss, although such a relationship has been assumed by previous committees (11, 12).
There are, however, other ways in which the interactions between energy and protein metabolism are important in relation to protein requirements. It has been known for some time that the utilization of dietary protein is influenced by energy intake and notably by energy balance (13, 52–54). It has been demonstrated (54–57) that, at any given level of dietary protein, addition of energy improves N balance until the response reaches a plateau, which represents the limitations imposed by the dietary protein level. This effect of energy balance can be extended further by raising the protein intake. Studies on animals (13) and on man (58) suggest that increasing the plane of energy intake enhances protein synthesis and reduces amino acid oxidation.
The influence of energy balance on N balance extends from suboptimal up to excess levels of energy intake, so that any change in energy intake above or below the subject's needs is likely to influence his N balance, the effect being of the order of 1–2 mg of N retained per kcalth added (0.24–0.48 mg of N per kJ) (59, 60). This has important implications for the determination of protein requirements when N balance is used as the criterion of adequacy (59–61). In view of the difficulty of determining the energy needs of individual experimental subjects, this effect of energy intake must be carefully considered when assessing estimates of protein requirements obtained by the N-balance method.
A less well defined relationship appears to exist between protein intake and the efficiency of utilization of dietary energy. A limited number of studies (60, 62) suggest that changes in intake or utilization of protein produce changes in the rates of weight gain of children and adults under isoenergetic conditions of intake and expenditure. This may be an aspect of the general principle that the improvement in N balance caused by adding energy to the diet can be inhibited when the protein content of the diet is too low (52, 53).
In previous reports (11, 12) a factorial method was used as the basis for predicting the protein requirements of various age groups. This method involved measuring obligatory nitrogen losses (i.e., the amount of nitrogen present in urine, faeces, sweat, etc.) when the diet consumed contained no protein but was otherwise adequate. The requirement for dietary protein was considered to be the amount needed to replace this loss, after adjustments for the inefficiency of dietary protein utilization and the quality of the dietary protein based on its amino acid pattern. For children and pregnant and lactating women, an additional amount of protein (required to support tissue growth and milk formation) was incorporated into this factorial estimate of requirements.
The method adopted by the 1971 Committee (12) has not always been clearly understood. It may be useful to retrace the steps of that Committee's argument and to fill in some gaps where the basis on which estimates were determined is not entirely clear.
Direct measurements on young adult males show that the sum of obligatory losses in urine and faeces is approximately 49 mg of nitrogen per kg of body weight. This figure was derived from the literature and from other studies (30, 63–65) available to the Committee, which had not been published at the time it met. To this figure was added 5 mg of N/kg to allow for miscellaneous unmeasured losses (sweat, etc.). In order to extrapolate these data to other age and sex groups for which direct results were not available, the Committee, following the example of its predecessor (11), made use of the general relationship that has been observed in animals of different species between obligatory or endogenous nitrogen loss and basal metabolic rate (BMR) (66).
The BMR of adult males was estimated as 25–27 kcalth/kg per day (105–113 kJ/kg per day), so that the obligatory loss was taken as approximately 2 mg of N/basal kcalth (0.48 mg of N/basal kJ). The Committee used this value to estimate the obligatory losses in other groups. Thus, for adult women, the estimate of obligatory loss was reduced by about 10% on the basis of the known sex differences in the BMR of adults. In children and adolescents the obligatory loss was calculated from the figures for BMR given in Annex 4 of the Committee's report (12). These BMR figures related to children of reference weight for age and height up to the age of 18 years, and were not used for estimating the BMR in adults. The Committee recommended that, after the BMR had been determined for any age and sex group, further adjustments within the group should be related to body weight.
Since 1971 additional studies have been published on obligatory losses in 2-year-old children, young men, and elderly men and women. There is no further information about losses in older children and adolescents. The results of these studies are summarized in Table 3.
The agreement among the values for urinary loss per kg of adult body weight is remarkable, particularly when one considers the difficulty in defining the exact period of time needed for the urinary loss to reach a stable level. The length of the study will probably have less effect on the faecal loss, yet it is the faecal loss that is more variable. It was higher in Indians and Nigerians, presumably because of the nature of their diet.
From evidence available to the 1971 Committee it was apparent that the ratio of obligatory losses to BMR might be significantly less in infants and young children than in adults. The new data from preschool children confirm this difference.
The 1971 Committee was the first to consider the important question of whether the obligatory nitrogen loss can be used to predict the amount of dietary protein needed to meet the minimum physiological requirement. Minimal protein needs had previously (11) been expressed in terms of a hypothetical reference protein that could be used with 100% efficiency; i.e., when fed at the level of the obligatory nitrogen excretion there would be no increase in the urinary and faecal nitrogen losses in excess of those found on a protein-free diet. An ideal amino acid pattern was proposed for the hypothetical reference protein that would provide a standard for determining the quality or number of other proteins.
The 1971 Committee examined the results of balance studies in which egg and milk proteins had been fed to infants, children, and adults at levels below or close to the requirement. This was defined as the amount needed to achieve nitrogen balance in adults and adequate retention in children. It was evident that even these high-quality proteins were not utilized with the efficiency previously assumed, on the basis of animal studies and a few studies in man which had involved very low protein intakes. To meet the minimum requirement, the dietary intake of various age groups had to be 25–50% above that expected from the obligatory losses plus a growth increment. It was proposed that the average requirement for egg and milk protein should be taken as 30% greater than that given by the factorial method for all ages. Re-examination of the data now suggests that the addition should have been of the order of 45%.
The present Consultation has adopted a modified approach, based on that of its predecessor. It was useful to establish the constancy of the obligatory loss, as shown in Table 3. However, once it became clear that N balance could not be achieved simply by replacing the obligatory loss, even with the best quality proteins, the magnitude of that loss becomes of little relevance. The important variable is the efficiency of utilization of dietary protein at the maintenance level. The starting point in determining the present estimates is therefore the direct measurement of the N needed for zero balance in short-term or long-term studies.
Most of the nitrogen balance studies that have been made in recent years have been either on young adult men or on young children. These provide two relatively fixed points, so that the protein requirements of other age groups can be obtained by interpolation, with an allowance, where appropriate, for the growth increment. As discussed in section 6.3, the BMR was not found to be a satisfactory basis for estimating either the total protein requirement or the maintenance requirement.
Since all the estimates of protein requirements are obtained either directly or indirectly from measurements of N balance, the limitations of balance studies need to be considered in some detail.
The nitrogen-balance technique involves the determination of the difference between the intake of nitrogen and the amount excreted in urine, faeces, and sweat, together with minor losses by other routes. In most experiments only the nitrogen content of the diet, urine, and faeces has been directly measured. Allowances are made for losses by other routes on the basis of a limited number of published studies. Thus, to allow for these other losses, any estimate of nitrogen balance limited to measurements of diet and excreta in non-growing individuals must be positive if overall body N maintenance is to be achieved.
In using this method to predict protein requirements, the usual procedure is to feed a series of different levels of dietary protein. The requirement is estimated by extrapolating or interpolating the N-balance data to the zero balance point (N equilibrium) for adults or for adequate growth (positive balance) in children. In early studies the levels fed often included one diet period without protein and other levels of intake far below the requirement. However, from studies on experimental animals and on man, it is known that the N-balance response is not linear throughout the entire submaintenance range; the slope decreases considerably as intakes producing zero balance are approached and slightly exceeded (34, 67).
Accordingly, recent studies have attempted to assess requirements by using several levels of intake that encompass the expected range of requirements. This is one of the reasons why most estimates of requirements based on contemporary studies are higher than those based on data reported in the past. In addition, other variations in experimental design contribute to the differences, such as the level of dietary energy intake and physical activity. In the earlier studies, energy intake was intentionally increased to ensure weight maintenance at low levels of protein intake. However, it is known that this results in more positive (less negative) N balances and therefore lowers the apparent protein requirement (53, 61) (section 5.3).
A second problem is the magnitude of the N losses other than from urine and faeces, the most important of which is via the skin. During heavy work in hot climates appreciable amounts of nitrogen are lost in sweat, mainly as urea, although the losses are lower in those who are acclimatized (68). There is some evidence of a compensatory reduction in urinary urea output (69). The N content of sweat is related to blood urea and both decrease with a low protein intake. Other forms of loss, e.g., in hair clippings and menstrual flow, have been measured in detail in some studies (Annex 6).
In the conventional balance study, these losses are not determined and an estimated allowance is made for them. On the basis of the available evidence this allowance has been set at 8 mg of N/kg per day for adults and 10 mg/kg for children up to the age of 12 years. It is unlikely that a single figure will be applicable under all conditions, but there is no realistic alternative to using this method of correction.
A further important consideration is the length of time needed to achieve a steady state at given levels of protein intake. Because adjustments in urinary N excretion do not occur immediately following a change in N intake, it is necessary to allow an adequate period of time for adjustment of N output to the new N-intake level. Most recent experiments concerned with the determination of N requirements have involved diet periods of 1, 2, or 3 weeks at each intake level. This approach characterizes the so-called “shortterm” N balance determinations of protein requirements in man.
The experimental design may be criticized on the ground that the time necessary to achieve a new steady state resulting in zero N balance may in fact be longer than that allowed for in the shortterm studies. If this is so, this method will result in overestimation of the requirements. Data from studies on obligatory N losses show that there is initially a sharp drop in urinary N followed by a long period of relatively stable but slowly declining excretion. The major adjustment appears to be complete by days 5–7 in most adults over a range of age and sex categories (13, 14) and somewhat sooner in children (15, 70). Because of day-to-day variation in urinary N excretion it is difficult to make accurate measurements of the subsequent slope or to prove that it is significantly different from zero. Most adult subjects receiving protein intakes close to their maintenance needs have not shown statistically significant differences in urinary N output when days 5–10 are compared with days 10–15 (70). However, in a long-term study of men with a low protein intake (0.36 g of protein/kg), it was shown that subjects required from 8 to 28 days to reach a new steady state.1 From the little evidence available, it appears that when subjects are fed a fixed protein intake over a long period of time, some do and some do not show a slow drift towards a lower rate of urinary N excretion.2 Data from a very long-term study of men fed 6.5 g of N/day (∼ 0.64 g of protein/kg) show that urinary N continued to fall for at least 90 days and perhaps much longer; the rate of decline after the first 2 weeks was about 0.01 g of N/day, which would be statistically undetectable in short-term studies (67).
It has also been claimed that autocorrelation between day-to-day rates of urinary N excretion will bias the interpretation of balance measurements (73). Statistically significant autocorrelation has been found in some studies but not in others (71).
Because of these limitations of short-term balances, longer-term studies should provide a better basis for determining protein requirements. In principle they permit the measurement of other variables that respond more slowly to dietary inadequacy, such as alterations in lean body mass or growth rate in children. In the few long-term studies that have been reported to date the usefulness of various biochemical and other measurements has been explored, e.g., serum albumin and total body potassium (K), but no sensitive and reliable marker has been found (56, 74). There is undoubtedly a need for other criteria of protein nutritional status as a firm basis for estimating requirements.
1 Durkin, N., et al., unpublished data, 1981.
2 Rand, W.M. & Young, V.R., unpublished observations, 1981.
Clearly, it would be very difficult to carry out a study in which subjects received a range of different intakes, remaining on each one for several months rather than several weeks. Published long-term balances have so far been conducted at only one level of protein intake. However, longer-term studies under somewhat more realistic conditions, in which adequate growth and positive N balance in children and N equilibrium in adults are maintained, provide the best evidence that the level fed is adequate, whether or not it is a minimum figure.
The preceding discussion assumes that the aim of both short-term and long-term balances is to find the minimum protein intake that will maintain the status quo in terms of mass of body protein. This aim allows for only minimal changes in protein mass as the body moves from one steady state to another (see section 5.2). Quite large losses of body protein can be supported without loss of life (75), but the degree of loss that is compatible with optimal function remains unknown.
The requirements for essential amino acids have been assessed by nitrogen balance in adults, starting with the classical work of Rose (76), by determining the amounts needed for normal growth and N balance in infants and children, and for infants, by comparison with observed intakes of good quality proteins. The report of the 1971 Committee (12) contains estimates of the amino acid requirements of infants, older children, and adults. Since then, information has become available on the amino acid requirements of preschool children aged 2–4 years, and the requirements of infants and older children have been reassessed.
Infants. The 1971 Committee (12) estimated the amino acid requirements of infants by using the intakes of cow's milk formulae or breast milk that supported satisfactory growth (77–79). With the exception of the requirement for tryptophan, the values estimated from the amounts of protein consumed in formulae were lower than those determined by Holt & Snyderman from N balances and the growth of infants given amino acid mixtures (80). The lower values are given in Table 4.
Snyderman et al. were also able to maintain satisfactory rates of growth in a small group of infants who consumed a diluted cow's milk formula to which glycine and urea had been added (81). These infants were consuming amino acids in amounts slightly below those estimated to be adequate on the basis of the studies of Fomon et al. (78).
Amino acid | Infants | Children | Schoolboys | Adults | |
(3–4 months) | (2 years) | (10–12 years) | |||
(ref. 12) | (ref. 82, 83) | (ref. 12) | (ref. 86) | (ref. 12) | |
Histidine | 28 | ? | ? | ? | [8–12]b |
Isoleucine | 70 | 31 | 30 | 28 | 10 |
Leucine | 161 | 73 | 45 | 44 | 14 |
Lysine | 103 | 64 | 60 | 44 | 12 |
Methionine + cystine | 58 | 27 | 27 | 22 | 13 |
Phenylalanine + tyrosine | 125 | 69 | 27 | 22 | 14 |
Threonine | 87 | 37 | 35 | 28 | 7 |
Tryptophan | 17 | 12.5 | 4 | 3.3 | 3.5 |
Valine | 93 | 38 | 33 | 25 | 10 |
Total essential amino acids | 714 | 352 | 261 | 216 | 84 |
b Not given in reference 12 (see text).
Preschool children. Data on the amino acid requirements of preschool children are now available (82, 83). An experimental design similar to that of Holt & Snyderman was used (80). The children received diets consisting of 0.3 g of cow's milk protein/kg per day plus an amino acid mixture in proportions and amounts equal to 0.9 g milk-protein/kg per day. The diets provided 100 kcalth/ kg per day (418.4 kJ/kg per day) with proper vitamin and mineral supplements. The single essential amino acid under study was partially replaced in the diet by glycine at five different levels. Nitrogen balance was calculated with an allowance of 8 mg of N/kg per day for integumental losses. It was assumed that a retention of 16 mg of N/kg per day (i.e., 100 mg of protein or about 0.5 g of lean tissue gain/kg per day) allowed for normal growth, and results were validated by N-balance studies conducted on children fed milk or soy protein to assess protein needs. Table 4 summarizes the new requirements. As this table shows, the values for preschool children are closer to those of older children than to those of infants.
Older children. The amino acid requirements proposed by the 1971 Committee (12) for boys between 10 and 12 years were based on studies in Japan (84, 85). These values represented the lowest amounts of amino acid needed to bring all subjects into positive N balance. The data have been re-examined by a committee in the USA (86), which derived estimates of requirements from the information provided on individual subjects. Table 4 gives both sets of estimates. The 1971 Committee's figures are consistently higher than those calculated by the committee in the USA. There are no other studies of amino acid requirements of school-age children, but judging from measurements of N balance in girls of 7–9 years fed two levels of protein, the need for sulfur-containing amino acids is greater than 13 mg/kg and is met by 30 mg/kg (87). These discrepancies emphasize the limited and unsatisfactory state of knowledge concerning the amino acid requirements of children.
Adults. The figures for adults in Table 4 summarize the data of the 1971 Committee (12) for both sexes together. The figures for men were derived from studies in which the criterion of adequacy was the attainment of positive N balance. In the studies on women a positive balance was not necessarily attained; the authors accepted a balance of 0 ± 5% of intake as the criterion of adequacy. The data for the two sexes are not entirely consistent, but whether the differences are attributable to biological differences in the requirements for individual amino acids or to differences in methodology is not known. The requirement for sulfur-containing amino acids has been confirmed by subsequent N-balance studies (88). The report of the 1971 Committee (12) does not include a value for histidine. Evidence is now accumulating that histidine is essential even for adults, and that this requirement may be between 8 and 12 mg/kg (89). For other essential amino acids, present information is insufficient to provide more precise figures for adult requirements.
The adult requirement for essential amino acids falls more sharply from infancy than does the total protein requirement (section 6). Thus, the proportion of total amino acids (T) that must be supplied as essential amino acids (E) (the E/T ratio) falls with age. This implies that an evaluation of dietary protein quality based on the amino acid requirements or the E/T ratio for infants or young children may underestimate the effectiveness of that protein in meeting the requirements of older children and adults. How this problem is to be dealt with when evaluating local diets is discussed in section 7.
In determining protein requirements for maintenance or growth, the present Consultation has relied on direct N-balance studies, although they are not without problems, rather than on the earlier factorial method. The Consultation believes that evidence of N balance in long-term studies constitutes the most acceptable direct evidence. Regrettably, long-term N-balance data are lacking, and the few studies that have been done must be supplemented by results for short-term N balance.
None of the current evidence is entirely satisfactory because there is no method available for the independent validation of an optimal state of protein nutrition. The functional significance of larger and smaller total-body N pools and faster or slower protein turnover rates is unknown. Most biochemical markers (plasma retinolbinding protein, albumin, etc.) are either unchanged even after relatively long periods (30 days or more) of negative N balance or are not readily interpretable (e.g., enzyme changes). There are no functional indicators that can usefully be applied in experimental situations to detect protein inadequacy before clinically detectable changes occur. This area urgently requires further research.
In the final analysis, one would wish to set protein allowances in accordance with such characteristics as health, growth, development, and longevity. This was, in fact, the approach used by our predecessors at the end of the nineteenth and beginning of the twentieth centuries—Voit, Atwater, Benedict, and Cathcart (90, 91). The majority view, with Chittenden dissenting, appears to have been that protein intakes well in excess of physiologically determined requirements were associated with active and healthy lives.
However, the evidence on which these observations were based is of limited value. Firstly, most habitual diets derive 10–14% of their energy from protein. Thus, when energy intake rises, so does protein intake and also intakes of many of the nutrients associated with protein in foods, such as the B-group vitamins and trace elements. Secondly, it is obvious that many environmental factors will influence any selected measure of health. Populations that characteristically have higher levels of protein intake tend to live under healthier conditions, whereas those with habitually lower intakes are much more likely to be exposed to parasitic and infectious disease. These confounding factors make it extremely difficult to attempt to draw causal relationships. Thirdly, there are many different measures of health and wellbeing; the criteria are therefore complex and cannot easily be used to set physiological requirements for protein.