Contents - Previous - Next

Amino acid scoring in health and disease

D. J. MILT-WARD (Chairman), E.A. NEWSHOLME, P.L. PELLETT and R. UAUY

1. Introduction
2. Amino acid scoring in health
3. Amino acid scoring in special cases and disease
References

1. Introduction

The paradigm which has dominated attitudes to dietary protein quality is that the amino acid pattern of the dietary protein source has a marked influence on the utilisation of food protein by the organisms. The basis of this paradigm is that the need for protein is explained in terms of a need for individual amino acids, some of which are indispensable and must be supplied in the diet, the rest requiring only a dietary source of nitrogen and carbon skeletons from which cellular synthesis can occur. Furthermore, it has been implicit in this paradigm that the magnitude of the need for the indispensable amino acids (IAA) is such that different proteins fed at isonitrogenous intakes may vary in their ability to satisfy the dietary need for protein. Thus, dietary proteins can be classified in terms of their quality, a function of their amino acid pattern. The consequence of this approach is that nutritional adequacy is likely to be met by food in which the overall amino acid pattern is optimised by complementation, i.e., balancing inadequacies in one protein with abundance in others, to achieve an overall pattern which matches needs. In this respect traditional food intake practices, that provide a satisfactory intake of most micronutrients, also do so for dietary protein.

The two most recent international reports which have considered human IAA requirement patterns for healthy individuals of all ages are the Expert Consultations published in 1985 (FAO/WHO/UNU, 1985) and 1991 (FAO/WHO, 1991). The main feature of the 1985 report was the acceptance of data on IAA requirements for different age groups which indicated a marked fall with age, with especially low values for adults. This meant that the scoring pattern with which protein quality was assessed for adults would indicate that all diets, even those based on minimally supplemented cereals, are equally able to satisfy IAA requirements, apart from possible increases to account for low digestibility. In many ways this was a remarkable conclusion in the light of the all-prevailing view that protein quality was an important factor, but nevertheless logical if the magnitude of adult IAA requirements summarised in the 1985 report was correct.

The most recent report (FAO/WHO, 1991) arose out of an Expert Consultation convened to deal with the methodology of protein quality assessment and particularly to endorse the protein digestibility corrected amino acid score method as an alternative to the rat growth assay (PER). To do this, required among other things a review of the scoring patterns, an opportunity to reassess the values presented in the 1985 report for each age group. As discussed below, there remain areas of disagreement and uncertainty. It was agreed at the 1991 Consultation that the IAA requirement values for older children and adults in the 1985 report were inappropriate, and that alternative values remain to be established. While the Consultation fully accepted the infant and preschool child scoring patterns, the data on which the latter pattern is based have never been published in full and cannot be scrutinised. In particular, with the data deriving from balance studies on previously malnourished children, the extent to which the amino acid requirement values determined in these studies were influenced by continuing tissue repletion cannot be examined. This is of particular importance given that the preschool child pattern was proposed as the basis for scoring protein quality in older children and adults in the interim recommendations of the FAO/WHO 1991 Consultation which are discussed below.

Thus, until new acceptable values for IAA requirements for older children and adults have been determined, and the preschool child values published for scrutiny, assessment of protein quality for all age groups remains problematic. In this section no attempt will be made to resolve the present controversies regarding the criteria to define optimal amino acid intakes. Our purpose is to summarise existing views of amino acid requirement patterns, as reviewed by the 1991 Consultation, and the way these are used in the evaluation of protein quality, a necessary step in determining the amount of dietary protein needed to satisfy the individual's IAA requirement. Thus, the following section dealing with the methodology for protein quality evaluation in health derives mainly from the recent FAO/WHO report (FAO/WHO, 1991).

2. Amino acid scoring in health

2.1. Protein quality evaluation: The protein digestibility-corrected amino acid score method
2.2. Protein digestibility
2.3. Amino acid scoring patterns

2.1. Protein quality evaluation: The protein digestibility-corrected amino acid score method

It has been widely recognised that clinical studies which measure growth and/or other metabolic indicators, including nitrogen balance, provide the most accurate assessment of protein quality. For reasons of both cost and ethics, it is considered inappropriate to routinely measure protein quality through the use of such techniques. Consequently, in the past, assay techniques designed to measure the effectiveness of a protein in promoting animal growth have been utilised and the Protein Efficiency Ratio (PER) method, which measures the ability of a protein to support growth in young, rapidly growing rats, was used. However, after decades of research with this method, it is now accepted that this was inadequate. Because of the rapid growth of rats, the assay measured mainly the requirement for tissue deposition. With the considerably slower rate of human growth, amino acid requirements are much more related to metabolic needs for maintenance, and these needs may be quite different from needs for growth. Thus, PER overestimates the value of some animal proteins for human needs while underestimating the value of some vegetable proteins for that purpose.

Theoretically, the nutritive value of a protein depends upon its capacity to provide nitrogen and amino acids in adequate amounts to meet the requirements of an organism. This directly reflects two factors, the amino acid content and the digestibility of the protein which limits the bioavailability of the amino acids in food proteins. Consequently, both amino acid composition and digestibility measurements are considered necessary to accurately predict the protein quality of foods for human diets. For some time the use of an amino acid score on its own has been advocated as an alternative to the PER. However, when there is poor bioavailability the quality of proteins will not be adequately assessed. This has led to the adoption of the protein digestibility-corrected amino acid score method.

2.2. Protein digestibility

Differences in protein digestibility may arise from inherent differences in the nature of food protein (protein configuration, amino acid bonding), from the presence of non-protein constituents which modify digestion (dietary fibre, tannins and phytates), from the presence of antiphysiological factors, or from processing conditions that alter the release of amino acids from proteins by enzymatic processes. Thus, amino acid scores must be adjusted for 'true' protein digestibility.

After reviewing the methodology of determining protein digestibility, the 1991 FAO/WHO report concluded that (a) the rat balance method is the most suitable practical method for predicting protein digestibility by humans, (b) the true digestibility of crude protein is a reasonable approximation of the true digestibility of most amino acids, (c) for use in the amino acid scoring procedure, established digestibility values of well-defined foods (composition and processing parameters) may be taken from a published data base for routine analysis, assuming all safety and toxicological criteria have been met, and (d) for new or novel products or processes, digestibility values must be determined.

2.3. Amino acid scoring patterns

The evolution of amino acid scoring has involved transition from a pattern based on the amino acid composition of food proteins judged to be high quality like egg (BLOCK and MITCHELL, 1946), to patterns which are assumed to reflect actual human needs for each age group (FAO/WHO/UNU, 1985).

Thus, the 1985 FAO/WHO/UNU report was able to formulate scoring patterns from IAA requirement values, expressed as mg per kg body weight per day, for infants, preschool children, school children and adults. These values were then divided by the recommended safe level of protein intake (g protein per kg body weight per day) for each age group to calculate the corresponding amino acid scoring pattern (mg/g protein). For infants the amino acid composition of human milk was proposed to calculate the amino acid scoring pattern.

The calculation of scoring patterns for the four separate age groups (Table 1) is a logical consequence of deriving the pattern from estimates of requirements and hence explicitly implies that protein quality is not an unchanging attribute of protein but varies with the age of the individual consuming it. The reduction of the magnitude of each amino acid in the pattern with age implies that proteins and diets with an IAA content and pattern that effectively meet the needs of young children are also adequate for older children and adults, whereas the reverse need not be true.

There have been a number of criticisms raised about the accuracy of the estimates of human IAA requirements and the scoring patterns which derive from them. Short-term balance studies in adults (WELLER, CALLOWAY and MARGEN, 1971) failed to confirm the requirement values suggested by ROSE (1957), which were the major basis for the 1985 FAO/WHO/UNU adult values.

It has been pointed out (PINEDA et al., 1980a; b) that there is considerable uncertainty about the IAA requirements that had been established for school-age children (NAKAGAWA et al., 1963). Problems with those studies include the excessive amount of dietary nitrogen used, the short N-balance periods that did not allow for adaptation to new levels of amino acid intake, the lack of allowance for integumental and miscellaneous N losses in estimating N balance and the modification in dietary amino acid composition from one experiment to another, which influenced the outcome and interpretation of the nitrogen balance studies.

The N-balance technique used for the assessment of IAA requirements has been criticised on a number of grounds (HEGSTED, 1976; YOUNG, 1986). Briefly, these concerns include the inadequate criteria used in earlier studies to estimate N balance, the difficulty faced in evaluating the nutritional and health significance of a given N balance under a particular diet and experimental condition, and the complicating effects of energy intake on N balances. It was suggested (YOUNG et al., 1989) that such problems would lead to underestimates of actual minimum physiological needs and, therefore, the relatively low requirement values proposed for the adult by the 1985 FAO/WHO/UNU Consultation must be questioned.

In support of these criticisms, metabolic isotopic studies had indicated higher requirement values for leucine, lysine, valine and threonine in adults (MEGUID et al., 1986a; b; MEREDITH et al., 1986; ZHAO et al., 1986; CORTIELLA et al., 1988). Further, in reviewing the metabolic basis of IAA and protein requirements, it has been suggested (MILLWARD and RIVERS, 1988; 1989) that the apparent age-related fall in the scoring patterns adopted by the 1985 FAO/WHO/UNU Consultation primarily reflected the different dietary designs of the various original balance studies. These experimental designs would have induced different rates of oxidative losses of amino acids and, therefore, inappropriate estimates of requirements. In particular, the amino acid mixtures used in the N-balance studies of ROSE (1957) and NAKAGAWA et al., (1963) included a disproportionate quantity of nonessential nitrogen in comparison to the composition of food proteins.

Table 1. Comparison of suggested patterns of amino acid requirements with the composition of high-quality animal proteins*

* Reproduced from FAO/WHO/UNU (1985), where references in this table can be found.
^a Amino acid composition of human milk.
^b Amino acid requirement/kg divided by safe level of reference protein/kg (Tables 4, 33, and 34). For adults, safe level taken as 0.75 g/kg; children (10-12 years), 0.99 g/kg; children (2-5 years), 1.10 g/kg. (This age range is chosen because it coincides with the age range of the subjects from whom the amino acid data were derived. The pattern of amino acid requirements of children between 1 and 2 years may be taken as intermediate between that of infants and preschool children).
^c Composition of cow's milk and beef or egg (LUNVEN et al., unpublished data, 1972).
^d Values in parentheses interpolated from smoothed curves of requirement vs age.

Thus, various authors (TORUN, 1990; MILLWARD and RIVERS, 1988) now agree that there is no justification for the continued use of the scoring patterns proposed by FAO/WHO/UNU for school-aged children and adults. The 1991 Consultation, however, faced a considerable dilemma in identifying a precise and practical alternative.

YOUNG and colleagues (PELLETT and YOUNG, 1988; BIER, YOUNG and PELLETT, 1989) had proposed, on theoretical grounds, a new amino acid scoring pattern to be employed for all ages except for the infant, and also provided some experimental support from their stable isotope studies for the valid use of this pattern in relation to adult protein nutrition. However, the Consultation recognised that this theoretical scoring pattern was subject to considerable controversy (MILLWARD et al., 1989; MILLWARD, 1990; YOUNG, BIER and PELLETT, 1990).

The 1991 FAO/WHO Consultation carefully considered the various arguments which had been raised in the light of current knowledge of the metabolic basis of IAA needs, and concluded that it was unlikely that there was a marked age-related fall in the IAA requirement as implied by the 1985 report. Given the slow rate of growth of the human, net accretion of proteins only accounts for a significant proportion of protein needs in the infant, and the maintenance component accounts for most of the requirement for all other age groups. As there is little evidence to suggest that maintenance nitrogen requirements substantially change with age, it is unlikely that IAA requirements change markedly with age.

Recognising the need for amino acid scoring patterns which can be used to assess quality of food protein sources and diets in all age groups, the Consultation decided that the scoring pattern proposed for the preschool child, which is based on various criteria of amino acid adequacy (PINEDA et al., 1981; TORUN et al., 1981), is robust and represents the best available estimates of IAA requirements for this age group. In the absence of sufficient new experimental data to determine more definitive scoring patterns for older children and adults, it was agreed that, in the interim, the preschool child scoring pattern should be employed for all ages, except for infants.

It was recognised, however, that the use of this preschool amino acid scoring pattern means that there will be some uncertainty about the extent to which protein quality will be accurately predicted for older children and adults, and that there may be some chance of overestimating the protein needs and underestimating the protein quality for these age groups. However, the Consultation considered that, in this event, this would result in a smaller error when protein quality is evaluated, than when the currently accepted scoring pattern for adults (FAO/WHO/UNU, 1985) is used.

The Consultation recognised the urgent need for further research in older children and adults to supplement the existing information and ultimately define the needs for IAA in these age groups. This should include research to identify functional indicators of amino acid adequacy.

It also recognised the need and importance to confirm and reinforce the existing information on IAA requirements for infants and preschool-aged children, since they form the basis of this Consultation's recommendation for an amino acid scoring pattern to evaluate protein quality.

The 1991 Consultation concluded with the following recommendations:

1. The amino acid composition of human milk should be the basis of the scoring pattern to evaluate protein quality in foods for infants under 1 year of age.

2. The amino acid scoring pattern proposed by FAO/WHO/UNU (1985) for children of preschool age should be used to evaluate dietary protein quality for all age groups, except infants.

3. The recommendations for the two amino acid scoring patterns to be used for infants and for all other ages must be deemed as temporary until the results of further research either confirm their adequacy or demand a revision.

4. Further research must be carried out to confirm the currently accepted values of requirements of infants and preschool-aged children, which are the basis for the scoring patterns recommended by this Consultation.

5. Further research must be carried out to confirm the IAA requirements of school-aged or adolescent children and of adults.

6. Given the urgency of these research needs and the magnitude of the task required, it is recommended that a FAO/WHO-coordinated international research programme be immediately established to assist in the determination of human amino acid needs.

3. Amino acid scoring in special cases and disease

3.1. Amino acid essentiality
3.2. Glycine
3.3. Glutamine
3.4. Arginine
3.5. Cysteine/taurine
3.6. Branched-chain amino acids (BCAAs)

It might be predicted that metabolic alterations during growth or illness will induce changes in amino acid requirements for optimal organ function and hence modify nutritional needs. Nutritional rehabilitation involves repletion of lean body mass and will be expected to change the balance between obligatory metabolic needs to sustain organ mass and the specific needs for protein accumulation. At the present time there is insufficient information to formulate special recommendations for amino acid requirement patterns for disease or nutritional rehabilitation. We will review in this section current information about individual amino acids for which there is evidence of metabolic responses to dietary supplementation in particular circumstances or in disease which point towards benefit or improvement in organ function from this modified amino acid supply (see also Bistrian).

3.1. Amino acid essentiality

As reviewed elsewhere, there is a substantial body of evidence which demands that the adequacy of current concepts be challenged (JACKSON, 1982; VISEK, 1986; LAIDLAW and KOPPLE, 1987; MILLWARD et al., 1989). Early N-balance studies showing that nonessential nitrogen might be limiting for normal growth (SNYDERMAN et al., 1962; KIES, 1972), and ¹⁵N studies showing specific channelling of amino groups between amino acids (JACKSON and GOLDEN, 1980, JAHOOR et al. 1988), demonstrate the essentiality of the amino group. According to JACKSON (1982), only two amino acids can be considered to be absolutely essential: lysine and threonine; and a small number absolutely non-essential: alanine, glutamate and aspartate, which can be formed by simple transamination from readily available intermediates, (pyruvate, µ-ketoglutarate or oxaloacetate). All the other non-essential amino acids can be shown to be conditionally essential, in that either they derive from IAA or there are circumstances in which their demand exceeds the capacity of the organism for their synthesis. Particularly interesting conditionally essential amino acids are glycine, glutamine, arginine, cysteine, and taurine.

3.2. Glycine

SNYDERMAN et al., (1962) demonstrated that glycine deficiency could limit growth in infants, and the limited availability of glycine in the developing foetus and in infants and children during rapid growth can be demonstrated by the increased urinary excretion of 5-oxoproline (see JACKSON et al., 1987), indicating a drain on the glycine pool. This is also seen in children recovering from malnutrition (PERSAUD et al., 1987), during pregnancy (PERSAUD et al., 1989) and in preterm infants (JACKSON, 1989).

Taken together these data lend strong support to the proposition that glycine is of limited availability during the early months of life and could be a rate-limiting amino acid not only for growth but also for other important functions including the synthesis of nucleotides, porphyrin and haem, creatinine, bile salts and glutathione. These are likely to constitute the demands for glycine in the adult, and YU et al., (1985) have demonstrated that, in normal adult men on a low-protein diet, the endogenous synthesis of glycine may be inadequate to satisfy the normal metabolic demand.

3.3. Glutamine

Interest in glutamine has developed over the last five years subsequent to developments in two areas: (1) the understanding of the metabolic importance of glutamine as a fuel for lymphocytes and other rapidly dividing cells (e.g., NEWSHOLME et al., 1985; SOUBA et al., 1985), (2) new knowledge on the importance and regulation of the skeletal muscle glutamine pool (KAPADIA et al., 1985; RENNIE et al., 1986; MacLENNON et al., 1987; 1988; JEPSON et al., 1988). Concern has been expressed that inadequate provision of glutamine from skeletal muscle as a result of stress, such as caused by burns, may be potentially deleterious for the immune system (PARRY-BILLINGS et al., 1991). Concern has also been expressed that the ability of the gut mucosa to act as a barrier to pathogens in sepsis or trauma may be limited by inadequate mucosal cell metabolism, secondary to reduced availability of glutamine to the gut (SOUBA et al., 1990). Finally, the suggestion that glutamine in skeletal muscle may act to regulate protein balance in this tissue has also led to attempts to limit the negative N balance in trauma by reversing the observed fall in muscle glutamine (VINNARS et al., 1975, ASKANASI et al., 1980; FURST et al., 1989).

However, in parenteral nutrition solutions the instability of glutamine during autoclaving has resulted in its effective omission from such solutions. Several authors have suggested the inclusion of glutamine containing peptides in TPN solutions (ADIBI et al., 1987). The recommendation has been made that enteral provision of glutamine could also serve to provide this amino acid and be advantageous by reducing bacterial translocation across the gut mucosa (SOUBA et al., 1990). To date there is no quantitative information about the dose response of either immune function or gut mucosal function to enable recommendations of intake to be made, although the safety, pharmacokinetics and metabolic responses to 0.1 and 0.3 g/kg per day have been demonstrated (ZIEGLER et al., 1990).

3.4. Arginine

This basic, traditionally non-essential diamino acid is, in fact, essential for several rapidly growing mammalian species, and is now classified as conditionally essential in human nutrition (VISEK, 1986; LAIDLAW and KOPPLE, 1987). As documented by VISEK (1986), the role of arginine in priming the urea cycle and in the activation of the carbamyl phosphate synthetase pathway are important regulatory roles for this amino acid in the newborn, while at all ages its requirement as a precursor for creatine synthesis may exceed dietary supply. Another important role of arginine, especially its guanidino nitrogen, is as a potential precursor of nitric oxide, the potent endothelial releasing factor (MONCADA, HIGGS and PALMER, 1988), and this further extends its metabolic importance. Arginine serves as a potent stimulant of insulin (FAJANS and FLOYD, 1972; HENQUIN, 1987), and of growth hormone (e.g., SOLIMAN et al., 1986), so that it may play an important role as an enhancer of the anabolic drive.

Conditional essentiality can be postulated during sepsis, and multisystem organ failure (BAUE, 1991) and recovery from injury and surgery (DUDRICK and SOUBA, 1991), since arginine supplementation has been shown to enhance N retention, wound healing and the immune response in these circumstances. The evidence supporting a specific role of arginine in the nutritional modulation of the inflammatory and immune response is particularly convincing (CERRA, 1991; DALY et al., 1990). Additionally, it may also contribute to enhance the growth of newborn infants since accretion rates for this amino acid exceed the provision in human milk (LAIDLAW and KOPPLE, 1987).

Contents - Previous - Next