Contents - Previous - Next

This is the old United Nations University website. Visit the new site at

Human amino acid requirements: A re-evaluation

Vernon R. Young and Antoine E. El-Khoury



The most recent internationally stated estimates of the amino acid requirements in adult humans are those given in the 1985 report of the Joint FAO/WHO/UNU Expert Consultation on Energy and Protein Requirements. In this review we present, in brief a number of scientific concerns and problems that lead us to conclude that these current recommendations for amino acid requirements are not valid and that the appropriate values are likely to be considerably higher. Following a short review of the C-labelled amino acid tracer studies carried out at the Massachusetts of Technology (MIT) and designed to reassess the requirements for specific indispensable amino acids, we focus particular attention on the lysine requirement in adults. When various criteria and methods are used to estimate this requirement, it appears that a cohesive body of data indicates the mean requirement value for lysine in healthy adults to be about 30 mg/kg/day or 50 mg/g protein. Although this value contrasts with the FAO/WHO/UNU value of 12 mg/kg/day or 16 ma/g protein, this new, tentative requirement value is consistent with findings from studies carried out earlier at MIT on the nutritional quality of wheat proteins. We propose that it would be prudent to apply the MIT amino acid requirement pattern (see Food and Nutrition Bulletin 1990;12:298-300), rather than the 1985 FAO/WHO/UNU adult amino acid requirement pattern, in the design and implementation of sound nutrition policies and programmes that include considerations of the amount and quality of the protein component of national and regional diets.


A brief assessment is made of the strengths and weaknesses of current estimates of the dietary requirements for the indispensable (essential) amino acids in humans throughout the life cycle, with particular emphasis on the requirements in pre-school children and adults. This topic has been reviewed earlier [15]. Particular focus is given to the lysine requirements of adults, since this indispensable amino acid is most likely to be limiting in the cereal-based diets characteristic of populations in large areas of the developing world [6, 7]. We begin with a restatement of the problem, as previously outlined [8]. We then survey the data generated from a continuing series of studies in our laboratories at the Massachusetts Institute of Technology (MIT) and provide a tentative new estimate of the amino acid requirement pattern based on these studies. We discuss the limitations of these studies and the support that our conclusions have received from our other studies and from those of other investigators. Finally, the implications of our tentative new estimates for food and nutrition policy are considered.

The problem

The current, authoritative, internationally stated estimates of the amino acid requirements in humans at various ages are those in the 1985 report of the Joint FAO/WHO/UNU Expert Consultation on Energy and Protein Requirements [9]. These estimates, for preschool and school-age children and for adults, are summarized in table 1. Except for the requirements for pre-school children, these estimates are taken from a 1973 FAO/WHO report [11] or are similar to those summarized by others [10].

TABLE 1. FAO/WHO/UNU (1985) estimates of amino acid requirements in pre-school and school-aged children and in adultsa

Amino acid Pre-school children (2-5 yr) Schoolchildrenb (10-12 yr) Adults (18+ yr)
mg/kg/day mg/g protein mg/kg/day mg/g protein mg/kg/day mg/g protein
Histidine ?   ?   8-10 16
Isoleucine 31 28 28 28 10 13
Leucine 73 66 44 44 14 19
Lysine 64 58 44 44 12 16
Methionine and cystic 27 25 22 22 13 17
Phenylalanine and tyrosine 69 63 22 22 14 19
Threonine 37 34 28 28 7 9
Tryptophan 12.5 11 3.3 9 3.5 5
Valine 38 35 25 25 10 13
Total (except for histidine) 352 350 216 222 84 113

a. From tables 4 and 39 in FAO/WHO/UNU [9].
b. Also based on a summary by Williams et al. [10].
c. Sulphur amino acids.

We will not discuss dietary amino acid requirements for infants, because the 1985 FAO/WHO/UNU report [9] recommended that these requirements be met by breastmilk. However, it is worth noting that breastmilk provides higher amounts of all the indispensable amino acids than the estimated requirements for infants in the 1973 FAO/UNU report [11], except perhaps for threonine. Moreover, in a paper presented at the 1994 meeting of the International Dietary Energy Consultative Group (IDECG), Dewey et al. [12] suggested, based on a factorial approach, that the needs for indispensable amino acids in infants three to six months of age were much lower than those in the 1985 FAO/WHO/UNU report [9]. For example, the requirements for lysine were estimated at 63 mg/kg/day by Dewey et al. and at 103 mg/kg/day by FAO/WHO/UNU. For the sulphur amino acids (methionine and cystine), the values were 27 mg/kg/day for Dewey et al. and 58 mg/kg/day for FAO/WHO/UNU. These are very large differences and they demand more careful scrutiny and experimental verification. It is apparent, therefore, that the quantitative needs for indispensable amino acids in the first year of life are still a matter of considerable uncertainty, as has been pointed out earlier [13].

With respect to the requirement figures given in table 1, some points should be made. First, the values for pre-school children are based on studies carried out by investigators at the Institute of Nutrition of Central America and Panama (INCAP), and only a summary of the findings from these studies is currently available in the literature [14, 15]. Thus, a detailed analysis of these important experiments cannot be made at this time. Second, the children who were studied in this series of investigations may not have fully recovered from earlier protein-energy malnutrition, as suggested by the magnitude of the positive nitrogen balances that were obtained in the experimental subjects. For example, at the estimated requirement value for lysine of 64 mg/kg/day, the children were found to be in a positive nitrogen balance of approximately 80 mg N/kg/day. However, the nitrogen increment in normal 18-month-old boys is only about 18 mg N/kg/day [16]; according to the data of Pineda et al. [14], a dietary intake of between 39 and 53 mg lysine/kg/day would have been sufficient to induce this lower rate of daily nitrogen retention. Of course, we do not know whether the children were using dietary lysine more efficiently than might be the case for a normal child. If this were the case, then the total daily amino acid requirements may not have been very different from those of normal and fully repleted children. However, the requirement figures given in table I for pre-school children may turn out to be somewhat higher than the true minimal physiological needs for fully repleted, normal, pre-school children. Additional studies in this age group would be highly desirable. The requirement figures for school-age children are based on the metabolic balance studies carried out at the Institute of Public Health in Tokyo [17] and are designed to help establish the minimal needs for the specific indispensable amino acids in this age group. Thus, in the absence of replication and confirmation of these earlier studies, the reliability of the values given in table 1 for the school-age child is still a question that future research will need to resolve.

The amino acid estimates for adults in the 1985 FAO/WHO/UNU report [9] were taken directly from the 1973 FAO/WHO report [11] and are based largely on the studies by Rose and co-workers in men [18] and on similar studies by various investigators in women [19]. There are major problems in the design and interpretation of these studies [2 4, 8]. These problems will not be discussed here, but it is our strong recommendation that the results of these earlier nitrogen balance experiments not be used to establish the minimum quantitative dietary requirements for specific indispensable amino acids that are needed to support long-term maintenance of protein nutrition status in adults.

From the foregoing, a case can be made that current knowledge about human amino acid requirements is highly unsatisfactory. There is an urgent need to close our wide gap of ignorance about the amino acid requirements of humans, if rational and safe nutrition policies and programmes, especially in reference to the nutritional well-being and development of populations in developing regions, are to be designed and implemented successfully. Therefore, we have sought new experimental data to support our arguments and to strengthen our previous conclusions [2, 8], and to make tentative new estimates of the amino acid requirements of humans from pre-school to adult [2, 4].

New approaches and tentative new amino acid requirement values

Our earlier studies [20--22] have already been reviewed, and the following discussion will focus on our work done during the 1990s. Our experimental approach has involved the application of C-labelled amino acid tracer techniques, the estimation of body amino acid carbon balance at varying test levels of amino acid intake [3, 4, 23], and the prediction of the obligatory amino acid losses and intakes necessary to just balance these, especially for the amino acids that we have not yet studied directly using C tracers [5].

Our working definition of the requirement for an indispensable amino acid in healthy individuals is "that minimal intake level which represents a single point on a dose-response curve and that is sufficient to maintain a specific criterion of nutritional adequacy (such as growth performance, body composition, body amino acid balance, or measure of organ [liver, muscle] or system [immune/defence, nervous] function)." For practical reasons we have chosen to use body amino acid balance, as determined by the difference between the intake of the test amino acid (e.g., leucine or lysine) and the whole-body oxidation of that amino acid, as measured by the appearance of the C label of the amino acid in expired carbon dioxide. Measures of organ or system function have not yet been sufficiently explored to permit them to be used for estimating the requirements for specific indispensable amino acids.

From our initial series of C-tracer studies, we estimated the requirements in adults for leucine, lysine threonine, and methionine (without dietary cystine) to be 30-40, 30, 15, and 13 mg/kg/day, respectively [4]. Except for the sulphur amino acid requirement, these estimates are far higher than the 1985 FAO/WHO/UNU values for adults shown in table 1. In addition, we attempted to predict the requirements for those indispensable amino acids that we had not yet studied with the aid of tracer techniques, based on considerations of obligatory amino acid loss (OAAL). This approach has been described in detail elsewhere [2, 5]. We appreciate that for some amino acids, and perhaps lysine in particular, the daily OAAL might be lower than the predicted value. This difference could be due to the possibility that some of the amino acid is retained in the free amino acid pool (possibly in muscle) during the fasting period and then reutilized for protein synthesis when amino acids are again supplied from the ingestion of protein-containing meals. Our preliminary C-lysine oxidation data suggest that possibly one-third of the lysine released via protein breakdown during the fasting period may be retained without undergoing terminal oxidation. However, the revised requirement for lysine discussed below, is based on the lower figures derived from metabolic data rather than on the somewhat higher predicted obligatory lysine loss.

Based on these earlier C-tracer studies and estimates of OAAL, revised values for the amino acid requirements of adults were proposed. These are presented in table 2 expressed in terms of milligrams of amino acid required per kilogram body weight per day. Table 2 also expresses these estimates in the form of an amino acid scoring pattern relating these estimates of mean requirements for specific amino acids to the mean requirement for total dietary protein [9]. The proposed new requirement pattern is quite different from that proposed by FAO/WHO/UNU [9]. Indeed, our proposed pattern for adults is similar to the FAO/WHO/UNU pattern for the pre-school child. The latter is also the amino acid requirement pattern that FAO/WHO [24] has recommended as an interim measure for use in the assessment of dietary protein quality for all humans over two years of age.

Contents - Previous - Next