Contents - Previous - Next

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

5. Clinical methods for the evaluation of protein quality

Criteria for Evaluation

Clinical methods for the evaluation of protein quality are based on the same principles as the corresponding animal assays are, but require modification for application to humans. The principal procedures measure either growth or nitrogen balance, either alone or in combination with biochemical analyses for serum proteins and amino acids, haemoglobin, blood urea nitrogen, and the urinary excretion of creatinine, sulphur compounds, and hydroxyproline.

Depending on the objectives of the study and the intended dietary use of the test protein, either it may be the major or sole source of protein in the experimental diet, or it may be given as a supplement. Thus, amino acid and protein supplementation may be evaluated in this way. Ideally, a control diet based on milk, egg, or lean beef as the reference protein should be studied for comparison purposes (1), and the levels of dietary protein and calories should be identical for the experimental and control groups S.

For children, nitrogen balance and growth are alternative criteria giving generally similar results. Except in young infants, however, growth methods are too time consuming and too subject to environmental influences to be convenient. They may, nevertheless, be useful in demonstrating the value and acceptability of new protein foods designed for mass feeding of young children. A further important limitation to growth methods is that it is not ethical to feed protein to growing children for long periods of time at the low levels of intake necessary for sensitive detection of differences in protein quality among various food proteins. To do so might permanently retard the growth and/or development of the young child. Nitrogen-balance methods require more sophisticated facilities and personnel, but allow evaluation of relative protein quality in a shorter period of time and with fewer subjects.

In theory, a more refined growth criterion than overall increase in body weight or height is the estimation of body protein increased by the determination of body cell mass (2, 3), but the methods for estimation of body protein content are indirect, relatively imprecise, costly, and are only available in a few institutions. However, changes in body protein in the young can be approximated by measuring changes in 24-hour urinary creatinine excretion (4). Viteri has shown that the urinary creatinine/ height index (CHI) is a much better measure of protein repletion and biochemical recovery of a malnourished child than changes in weight, height, or weight-for-height (5). While urinary hydroxyproline excretion correlates reasonably well with skeletal growth (6), the changes in hydroxyproline output are usually too slow to be helpful in evaluating protein quality.

During recovery from protein malnutrition, serum albumin regeneration is slower in patients fed vegetable protein than in those given an equivalent level of milk protein (7). This has led to the suggestion that serum albumin regeneration might be used as a sensitive measure of protein quality (8). However, the amino acid composition of serum albumin is not representative of most other body proteins or of the pattern of essential amino acid requirements established for humans, and the metabolism of serum albumin does not necessarily reflect the overall response of the body to altered protein intake.

The most serious reservation to the use of serum albumin or haemoglobin values as a measure of the quality of protein fed to malnourished children is that both are distorted by changes in blood volume and lean body mass (9). Even total weight may be unchanged for two or three weeks, while lean body mass is increased rapidly along with blood volume. It has been suggested that plasma retinol-binding protein and thyroxine-binding prealbumin are more sensitive to nutritional deficiency (10).

The traditional measures of protein quality in experimental animals, protein efficiency ratio (PER) and net protein utilization (NPU), are conducted at levels of intake that are suboptimal for maximum growth. Comparable values to standard NPU in rats can be obtained in children or adults by measurement of nitrogen balance at a single, distinctly suboptimal level. It is now recognized, however, that protein is utilized more effectively at suboptimal levels than at levels in the near-maintenance range of intake. Accordingly, biological measures of protein quality conducted at suboptimal levels in either experimental animals or human subjects may overestimate protein value at maintenance levels.

At relatively high levels of protein intake, either milk or a vegetable mixture of good protein quality gives equally good nitrogen balance. As can be seen from figures 3 (11) and 4 (12), efficient utilization of egg protein in adult men begins to fall with intakes well below those sufficient for maintenance of nitrogen balance. The feeding of a test protein at a level sufficient to maintain nitrogen equilibrium gives a lower value than predicted from the classical biological value (BV) and NPU assays conducted at grossly inadequate levels of protein intake. A similar conclusion applies to studies in children (13). Moreover, these effects are more marked for protein of poor quality than for protein of high quality. To predict reliably their capacity to fulfil protein requirements, the values obtained at near maintenance levels of protein intake are the appropriate ones.

FIG. 3. Biological Value of Egg Protein at Different Levels of Protein Intake in Young Adult Men (From ref. 11)

Because the requirement level for an individual is not known precisely in advance, a multiple-protein-level feeding is necessary to obtain an estimate of the amount of the test protein required for maintenance. Digestibility is best calculated at this latter level, particularly if "apparent" rather than "true" digestibility is to be determined. Instead of NPU measured at a suboptimal level, an estimation conducted at the maintenance level of protein intake is recommended. Proteins can then be compared with one another at this level with egg or milk as reference proteins. It must be recognized, however, that this approach will not reveal differences in protein quality that are apparent on assays conducted at a single, suboptimal level. For example, milk and the vegetable mixture Incaparina give essentially similar protein values at maintenance levels, but milk is superior when determined at suboptimal protein intake levels.

FIG.4. Biological Value of Egg Protein and Wheat Gluten at Different Levels of Nitrogen Absorption in Young Adult Men (From ref. 12)

Although evaluation for practical feeding should be based on comparisons of nitrogen retention at maintenance levels of test protein intake, sensitive detection of processing effects or screening of possible impairments in the mixture would be most effectively done by measurement of NPU at a single, suboptimal level. The effects of amino acid supplementation of cereals are generally undetectable in normal human subjects at intakes adequate for maintenance and growth, but measurable at slightly lower levels (14). Final choice of method will depend, therefore, on the practical objectives of the evaluation.

Criteria for Evaluation

Nitrogen-Balance Techniques

Nitrogen-balance data are obtained by measuring dietary nitrogen intake and urinary and faecal nitrogen output by direct analysis. In clinical trials, nitrogen-balance results are commonly expressed simply as the percentage of nitrogen intake retained without taking into account integumental, obligatory urinary, and obligatory faecal nitrogen losses because these are relatively constant and hard to measure. Generally, sweat and other miscellaneous losses are also not taken into account because of the problems of measurement and variation with activity, clothing, and environmental conditions. Any significant infection or other cause of stress will invalidate the results because of the resulting increased loss of nitrogen in the urine.

The percentage of ingested nitrogen retained in humans measures the same thing as apparent net protein utilization (apparent NPU) in the rat, and agreement is generally good ( 1). The calculation is as follows:*

Apparent NPU = (I - F - U) / I

NPU is the fraction of ingested nitrogen retained; BV is the fraction of absorbed nitrogen retained; and digestibility is the proportion of ingested nitrogen that is absorbed. The classical BV calculation corrects both for digestibility and for obligatory faecal and urinary nitrogen. The latter are determined by feeding the subject a diet that is free, or nearly free, of nitrogen. Thus:

BV = I - (F - FK) - (U - UK) / I - (F - FK )

If uncorrected for obligatory losses, the calculation gives apparent BV. Thus:

Apparent BV = (I - F - U) / I - F

Subtracting faecal nitrogen from ingested nitrogen gives a measure of the apparent digestibility of a protein meal, and correcting this for endogenous faecal nitrogen is supposed to give "true" digestibility. Because digestibility is an important component of the nutritive value of food protein sources, NPU is more appropriate than BV for practical purposes. NPU is influenced, however, by factors other than the inherent amino acid composition of the protein. For example, reduced digestibility caused by overheating will lower protein value by decreasing the availability of several of the essential amino acids, particularly Iysine, which is usually limiting in cereal-based diets (chap. 2).

Considerable variability and error in faecal nitrogen measurements are introduced by the difficulty of separating faecal samples corresponding to different diet periods. The common practice is to feed a small amount of faecal marker (e.g., carmine, FD & C Blue No. 2, or charcoal) at the beginning and end of each period, but these often do not give sharp divisions in faecal outputs for respective periods. This is more of a I = nitrogen intake; F = faecal nitrogen: U = urinary nitrogen; FK = obligatory faecal nitrogen; UK = obligatory urinary nitrogen. problem with low-residue, liquid formula diets because stools are frequently less bulky and softer. More consistent results can be obtained by the use of chromium sesquioxide or polyethylene glycol as markers (15, 16).

Several investigators ( 17-20) have called attention to the cumulative errors inherent in long-term balance studies in humans. Because of the difficulty of getting all of the material out of the containers used, nitrogen intake will tend to be overestimated and nitrogen excretion underestimated. These and other errors common in nitrogen-quibalance procedures have been extensively discussed (21).

Nitrogen losses from the skin are neglected in all of the above formulae because it is usually impractical to measure them. It should be obvious, however, that any appreciable integumental nitrogen losses will affect the interpretation of the above values. The 1973 FAO/WHO report Energy and Protein Requirements ( 1 ) proposes a value of 3 mg N for minimum sweat losses and 2 mg N for miscellaneous losses for a total of 5 mg N per kg of body weight, based on studies in adults. However, the discrepancy between nitrogen intake and urinary and faecal nitrogen in long-term balance studies under temperate conditions approximates 15 mg/kg (16, 22). Because it is unlikely that this large and consistent difference is due to methodological error, it is thought that it reflects higher integumental nitrogen losses than can be measured directly by available techniques. Sweat nitrogen losses with elevated environmental temperature and humidity are easily more than doubled (23, 24). But this may be at least partially compensated for by a reduction in urinary nitrogen excretion (25, 26).

The classical studies of nitrogen balance in humans and other mammals have shown that the response of body protein metabolism to changes in the quantity and quality of ingested nitrogen takes time to stabilize. With abrupt change from a normal protein intake to a nitrogen-free diet, five to seven days are required for stabilization of urinary nitrogen excretion in young adults, and four to five days in young children (B. Torun, personal communication). Similarly, with a sudden increase in nitrogen intake, it may take a week or more to reach a higher steady-state level. This means that in nitrogen-balance studies there must be a suitable adaptation period to each dietary change, which can be quite short if the change is small, longer if it is large.

In clinical protein deficiency, nitrogen-balance results vary widely with the condition of the individual. However, children who are in a late stage of recovery from proteincalorie malnutrition and still mildly depleted of nitrogen and lean body mass will retain nitrogen maximally over a wide range of intakes. This situation can be partially simulated in healthy, well-nourished adults by placing them on a nitrogen-free diet before the study is begun. The problem here is that the individual may be so depleted that he will temporarily retain extra nitrogen for that intake level even when the protein is of poor quality. The converse of this is that the well-nourished individual shows a less than maximal retention during the initial days of such a study even when the experimental diet is inadequate.

In adults, a one-day nitrogen-free period at the start helps to reduce the necessary length of the adaptation period when the test protein is fed either at or below requirement level. Once adaptation to the test level of nitrogen intake has taken place, the adaptation period necessary for different proteins depends on how different the test protein value is. This ranges from none at all, if they are of nearly equal protein value, to a number of days if they are quite different. After a one-day nitrogen-free period, five days of feeding the test protein are generally sufficient for adaptation, and the apparent biological value can be estimated from the nitrogen balance for the next five days. In children, the periods can be shorter and the level of protein fed higher, depending on age and nutritional status. The limiting factor is the number of days necessary to obtain reliable faecal samples through the use of orally administered markers.

For adults, it has been found effective to measure nitrogen balance at four levels, ranging from the highest approximating the requirement level to a low level suitable for obtaining conventional NPU determination values. For good-quality protein, levels of 0.6, 0.5, 0.4, and 0.3 9 test protein/kg/day have been found suitable (27). Balance periods are usually ten days, with values for the last five days of each period used for urinary nitrogen estimates.

Some workers favour randomization of diet presentation, using a Latin-square design. In this case, each period is preceded by two or more days on an adequate ad libitum diet followed by a one-day isocaloric nitrogenfree period. Consistent results are also obtained by giving the test protein levels consecutively in a step-wise reduction after an isocaloric, protein-free day only at the start of the experiment. Faecal nitrogen values are based on the full ten-day period, and some investigators use an average digestibility figure for all four diet periods. When randomization of levels is not employed, it is desirable to study half of the subjects given the levels of test protein in descending order and half in ascending order, combining the data used for calculations of protein quality.

Analysts of Nitrogen-Balance Results

The slope ratio technique of analysing and presenting nitrogen-balance responses to different protein sources has the limitation that it considers only the slope of the response curve and ignores its position. As an alternative, on the basis of the results of Bressani (28, 291 and Hegsted (30) and their co-workers, Young et al. (31) examined dietary protein quality in adults in reference to an estimation of the intake necessary to achieve zero balance (fig. 5). This method is termed relative nitrogen retention (RNR). With this procedure, the nitrogen-balance response is analysed for each individual studied at several intake levels. The linear regression line representing each individual's response is determined by standard statistical techniques (31) as discussed in chapter 10. The intersection of the regression line with zero balance in the adult (including estimates of integumental nitrogen losses) or an acceptable positive balance in infants and children that is sufficient to promote adequate growth is defined as the minimum need for the test protein source. The mean requirement estimates are then expressed as a fraction of the mean estimates obtained under comparable experimental conditions with an appropriate reference protein source such as egg or milk. A more critical statistical appraisal of the differences obtained with various test protein sources obviously can be carried out if allowance is made for the variability among subjects.

FIG. 5. Two Nitrogen-Balance Response Criteria and the Three Methods of Estimation for the Assessment of Dietary Protein Quality in Human Subjects(From ref. 31)

TABLE 11. Hypothetical Nitrogen-Balance Data and Individual Regressions of Intake on Nitrogen Balance for Young Men Receiving Two Protein Sources at Various Levels within the Submaintenance to Near-Maintenance Range of Intake


Nitrogen balance in mg N/kg/day (Y) at each level of nitrogen intake in mg N/kg/day (X)

Regression equations.


Zero intercept










mg N/kg/day


Receiving milk (reference protein source)
1 - 46.0 - 24.6 - 21.2 -8.2 -5.1 0.45 - 39.8 3.1 89 0.56
2 - 46.0 - 33.1 - 25.2 - 16.2 - 10.5 0.48 - 48.1 1.0 100 0.63
3 -46.0 -21.6 -11.7 - 9.2 + 5.7 0.53 -38.7 3.6 73 0.46
4 -46.0 -39.7 -13.7 - 5.2 + 6.7 0.92 -64.7 6.0 70 0.44
Mean -46.0 - 29.8 - 17.9 - 9.7 - 0.8 0.595 - 47.9   83 0.52
Receiving test protein source
1 -46.0 -24.8 -10.2 - 9.0 -7.0 0.34 -31.9 5.0 93 0.58
2 - 46.0 - 22.4 - 17.8 - 14.9 + 0.1 0.44 - 38.4 4.3 87 0.55
3 - 46.0 - 21.6 - 15.2 + 1.6 - 4.7 0.42 - 33.6 6.9 80 0.50
4 -46.0 -26.9 -22.9 - 9.7 - 3.7 0.52 -44.8 2.7 87 0.5
Mean -46.0 - 23.9 - 16.5 -8.0 -3.8 0.43 - 37.1   87 0.54

The following calculations can be made for NPU, RPV, and RNR for a test protein and milk, using hypothetical nitrogen-balance data (mg N/kg/day) from table 11 for four young men fed diets in the sub-maintenance to maintenance range of intakes. These approaches have been discussed in detail (31), and it is recommend' ed that both RPV and RNR be calculated.

NPU * = ( balance + obligatory loss) / intake

Milk NPU = ( - 17.9 +46 ) / 48 = 0.59

Test NPU = ( -16.5 +46 )/ 48 = 0.61

RPV pooled regression:

Milk = 0.595x - 47.9

Test = 0.430x - 37.1

RPV = 0.430/0.595 = 0.72

RNR mean intercepts at zero balance:

Milk 83

Test 87

RNR= 83/87 = 0.95

While it is not fully verified, it is probable that for both adults and children comparable results can be obtained using considerably shorter consecutive diet periods (32, 33). In this case, faeces might be pooled for the entire experimental period, and a single digestibility figure calculated. Balances are calculated on the basis of 24-hour urine collections in the same manner as for longer studies. After five days at the initial level, even one day at successive levels may prove sufficient (32).

Necessary Precautions in Nitrogen-Balance Studies

Nitrogen balance in both children and adults has given variable and unsatisfactory results in the hands of some investigators because of failure to control the large number of factors that can influence it and reduce or destroy its validity as a measure of protein quality. The following precautions are essential:

  1. Caloric intake must be adequate and constant. If caloric intake is too low, the test protein will be used to provide energy and a falsely low protein value will be obtained. Moreover, nitrogen balance is sensitive to increased caloric intake above requirements (22, 34-37) so that the level of intake must be the same for control and test diets and should neither be excessive nor inadequate. This, however, is difficult to achieve with precision because calorie requirements vary considerably among apparently similar individuals, and the energy needs of individuals cannot be readily predicted. The best estimate possible should be obtained of normal caloric needs by a combination of dietary history and weight-height nomogram. Activity level must be maintained reasonably constant. If the subject begins losing weight, this should be noted, but the duration of such studies is too short to know this with certainty in most cases, and an adjustment would confound the data more than it would help.
    As fat and carbohydrate sources have differential effects on nitrogen balance (34), it is also important to standardize the ratio of calories provided by these sources.
  2. Protein intake must be constant and appropriate. Variations in protein intake confound comparisons between experimental and control groups. For well-nourished adult subjects, the level of protein fed should be within the submaintenance to requirement level, but extremely low levels should be avoided. For childrer, protein intake should be at or near requirement level or slightly higher.
  3. Water intake must be controlled. A large fluid intake increases urinary nitrogen excretion (38, 39) so that variation in fluid intake causes greater variability in nitrogen balance results.
  4. No infection, even of seemingly mild degree, may be present. All systemic infections, even as mild as vaccination against smallpox and immunization against yellow fever, produce a stress reaction with a resulting increase in urinary nitrogen excretion (40). Some upper respiratory infections may also have this effect (41). Fever need not be present (42). Some intestinal parasitic infections are also sufficiently severe to interfere with nitrogen absorption (43).
  5. Excessive integumenta/ nitrogen losses must be avoided. Nitrogen-balance studies to determine protein quality should be conducted under conditions of minimum sweat loss. Excessive losses associated with physical activity and high environmental temperatures will bias the results.
  6. Subjects should be reasonably content and not psychologically disturbed. The stress reaction produced by anxiety and fear can increase urinary nitrogen excretion. In some university students, examinations have this effect (44). Even anxiety about the experiment itself can produce negative nitrogen balance in a susceptible subject (45). Much of the variation encountered in studies with children has probably been the result of stress reactions.
  7. Degree of depletion of the subject should be taken into account in the experimental/ design. Considerable adaptation time may be required to obtain a stable value if either the subject is severely depleted or the prior diet has been high in protein.
  8. Meal times should be standardized both between and within treatments. "Nibbling" studies have established that protein is better utilized in animals if given in divided amounts rather than concentrated in one or two meals (46). Dividing the protein to be fed among three or four meals is a convenient practice. In any event, the test and reference protein should be fed in the same manner.
  9. Periods on the same diets must be long enough to determine trend as well as initial response to a dietary change. The initial response will be heavily influenced by the preceding period, and considerable time may be required to obtain a stable value. In general, the greater the change in protein value, the longer the adjustment time.
  10. Measurements of food intake and/or urine and faecal collections must be accurate. An inherent limitation to the balance period is that the dietary intake tends to be overestimated and the urinary and faecal nitrogen underestimated because of the mechanical problems of scraping or rinsing out ail of the contents of any container.
  11. Determinations of nitrogen in food, urine, and faecal collections must be accurate. A common error is that introduced by poor homogenization of the dietary and/or faecal samples before taking aliquots for analysis.
  12. Other dietary constituents. In addition to the mineral and vitamin supplements, all other components added to the diet should be stated and identified because certain components, for example, wheat bran, might supply supplementary protein. Intakes of all nutrients, other than protein, should be adequate and therefore equal to or exceeding the levels of intake regarded as being sufficient to meet the needs of all individuals. These levels would be those recommended by WHO/FAO or by national groups such as the US National Research Council (47).

When these 12 requirements are all observed, nitrogen-balance methods can produce biologically significant and consistent results for most protein sources in human subjects, both children and adults.

Use of Nitrogen-Balance Technique for Determination of Minimum Essential Amino Acid to Total Nitrogen Ratio

As listed in table 12, milk and other animal sources of protein have higher E/TN ratios (grams essential amino acids per gram total N) than those of vegetable origin; but except for cassava, which is recognized to be a very poor protein source, all have ratios that are similar to or greater than that of the 1973 FAD/WHO reference "Essential Amino Acid Pattern" (1). Because sources of non-specific nitrogen are readily available, it should be useful to know the extent to which various food proteins can be extended by adding nitrogen from these non-specific sources without impairing the capacity of food proteins to supply the "protein" requirements of children and adults.

Evaluation of E/TN ratio by examining amino acid patterns of protein as analysed is affected by the same factors limiting the value of amino acid scores-uncertainty as to the biological availability of the amino acids measured by the analytical procedures. For some purposes, therefore, it may be desirable to determine minimum E/TN ratios directly by the nitrogen-balance technique. Once urinary nitrogen excretion is constant, the nitrogen of the test protein can be isonitrogenously replaced at varying levels by a suitable non-specific nitrogen source until a level is reached at which urinary nitrogen excretion increases (48-51). The relationship that must exist between the E/TN ratio of a protein supplement and its ability to improve a diet of poor protein value has not been explored experimentally.

TABLE 12. Ratio of Total Essential Amino Acids to Total Nitrogen in Selected Foodstuffs and in the FAO Provisional Patterns

Protein source E/TN ratio Protein source E/TN ratio
  (g/gN)   (g/gN)
Casein 3.25 Sesame seed 2.47
Hen's egg 3.22 Oats 2.30
Cow's milk 3.20    
Human milk 3.13 FAO/WHO (1973)  
Beef liver 2.94 Pattern 2.25
Beef heart 2.85    
Beef muscle 2 79 Rye 2.17
Cotton seed 2.15    
Navy beans 2.79 Sunflower seed 2.11
Corn meal 2.78    
Peanut flour 2.08    
Millet 2.75 White wheat flour 2.02
Sweet potato 2.70    
Pork tenderloin 2.67 FAO 1957  
Fish 2.66 Pattern 2.02
Rice 2.61    
Peas 2.59 Wheat gluten 1.99
soy flour 2.58 Cassava 1.31
Spinach 2.50 Gelatin 1.05

FAO/WH01973 (1).

Plasma Amino Acids

Plasma amino acid changes seen after feeding a standard protein meal are influenced by the protein quality of the diet the subject has been consuming (52). Similarly, subjects fed a uniform diet before the test feeding of protein show that the serum amino acid response is influenced by the amino acid pattern of the test meal (531. In either case, qualitative differences in amino acid availability from different proteins can be deduced. Although there have been a number of studies on the use of plasma amino acid changes as a basis of estimation of protein quality, the conditions required for uniform response and quantitative interpretation of the data have not been established (53-57).


Premature and young infants have been favoured as subjects because of their normally rapid growth rate, but they should not be given diets that might significantly impair their growth or development. Long-term feeding studies in preschool children who are in orphanages, day-care centres, and other institutions are useful for demonstrating the value of an improved or supplemented dietary regimen but are too time-consuming to be feasible for studying individual protein values.

As has been indicated earlier, even under these relatively controlled circumstances, growth studies are subject to confounding by diarrhoea! disease, epidemics of one or more of the common communicable diseases of childhood, and other infections. Any appreciable differences between groups in the frequency and severity of infections will bias growth results. Under the usual conditions of village populations, it is virtually impossible to control sufficiently the many variables and to obtain regular enough participation in the feeding programme for results to be reliable. The members of the family may be too occupied with other duties to bring the child to the feeding centre; the family may leave the village for a short or long period; the child may be kept away because of illness; or a variety of other reasons may result in irregular participation.

It is tempting to think that some problems of working with preschool children can be avoided by conducting clinical trials with school children, who are generally regimented and more resistant to infectious disease. School children are usually not suitable subjects, however, because for most of them food intake is no longer the factor limiting their growth. For children of school age, factors reducing the sensitivity of their response to supplementary feeding include less body protein depletion following infection, their lower protein requirement per unit of body weight, the fewer foods that are withheld from them, and their improved ability to obtain food for themselves.

Growth trials will be a test of protein value only if other dietary elements are not limiting. The diet must therefore supply adequate calories, vitamins, and minerals. It is usually difficult to be certain of this when measurements are made on children living at home, a further limitation to the value of growth studies not under rigidly controlled institutional conditions. For both ethical and practical reasons, the total protein intake should be sufficient to support growth.


The clinical methods for evaluating protein quality are based on the same principles as the corresponding animal assay, with specific adaptations to human subjects. The most useful criterion currently available is nitrogen-balance, despite the many recognized limitations of this approach. I nadequate and excessive caloric intakes, too high a test level of protein, intercurrent infections, inadequate time for adaptation to a diet change, and failure to adhere strictly to the test diet and sample collection schedules are the most common sources of error. Classical NPU measurements, based on nitrogen balance at a single level and markedly suboptimal level of intake, overestimate the relative capacities of proteins to meet requirements, and the discrepancy increases with decreasing protein quality. For this reason, multiple assays using protein levels at points slightly above, approximate to, slightly below, and well below requirements are recommended. In reference to the slope-ratio procedures for animal assays, statistical evaluation of nitrogen-balance results suggests that the criteria for comparison of the quality of different proteins for human subjects be based on the intercept with zero balance in adults, after allowing for integumental and other nitrogen losses, or a stated positive balance in the young child, to measure RNR.

Finally, it should be emphasized that the critical interpretation of the nutritional significance of results obtained in clinical studies must take into account the nature of the diet of the target population in which the test protein will form a part. Many of the procedures discussed earlier in this chapter are further discussed in chapter 9, where more detailed methodology is described.


1. Joint FAD/WHO Ad Hoc Expert Committee, Energy and Protein Bequirements, WHO Technical Report Series, no. 522; FAO Nutrition Meetings Report Series, no. 52 (WHO, Geneva; FAO, Rome, 1973).

2. K.G McNeill, J.R. Mernagh, K.N. Jeejeebhoy, S.L. Wolman, and J.E. Harrison, "In Vivo Measurements of Body Protein Based on the Determination of Nitrogen by Prompt gama Analysis,'' Am. J. Clin. Nutr. , 32: 1 955-1969 1 1979) .

3. F D. Moore, K.H. Olesen, J.D. McMurray, H.V. Parker, M,R, Ball, and C.M. Boyden, The Body CellMass and lts Supporting environment lW.B. Saunders, Philadelphia, Penn., USA, 1963) .

4. J.E. Graystone, "Creatinine Excretion during Growth," in D.B. Cheek, ea., Human Growth: Body Composition, Cell Growth, Energv and Intelligence (Lea and Febiger, Philadelphia,Penn., USA, 1968), pp. 182-192.

5. F.E. Viteri, G. Arroyave, and M. Behar, in Seventh International Congress of Nutrition, Abstracts of Papers, Hamburg 3-10 Vlil, 1966 (Pergamos Druck, Hamburg, Germany, 1966), p.46.

6. R.G. Whitehead, "The Assessment of Nutritional Status in Protein-Malnourished Children," Proc. Nutr. Soc, 28: 1-16 (1969).

7. G.G. Graham, J.M. Baertl, and A. Cornado, "Studies in Infantile Malnutrition. V. The Effect of Dietary Protein Source on Serum Protein," Am. J. Clin. Nutr., 18: 16-19 ( 1966).

8. J.D.L. Hansen, W. Wittmann, A.D. Moodie, and S.A. Feiiingham, "Evaluating the Synergism of Infection and Nutrition in the Field," in N.S. Scrimshaw and J.E. Gordon, eds., Ma/nutri tion, Learning and Behavior (MIT Press, Cambridge, Mass,, USA, and London, 1968), pp. 438-455.

9. F.E. Viteri, J. Alvarado, D.G. Luthringer, and R.P. Wood, "Hematological Changes in Protein Calorie Malnutrition," in R.S. Harris, I.G. Wool, J.A. Loraine, and K.V. Thimann, eds., International Symposium on Vitamin-Related Anemias, Vitamins and Hormones (Academic Press, New York and London, 1968), pp. 573-615.

10. P.S. Shetty, K.E. Walrasiewicz, R.T. Jung, and W P.T. James, "Rapid Turnover Transport Proteins: An Index of Subclinical Protein-Energy Malnutrition," Lancet, 2: 230-232 (1979).

11. N.S. Scrimshaw and V.R, Young, "Application of Assays of Protein Quality to Human Requirements for Protein," in P.L. White and D.C. Fletcher, eds., Nutrients in Processed Foods-Proteins (published for the AMA by Publishing Sciences Group, Inc., Acton, Mass,, USA, 1974).

12. G. Inoue, Y. Fujita, K. Kishi, S. Yamamoto, and Y. Niiyama, "Nutritive Value of Egg Protein and Wheat Giuten in Young Men," Nutr. Rep. Int'l.,10: 201 (1974)

13. R. Bressani and F. Viteri, "Metabolic Studies in Human Subjects," in Proceedings SOS/70, Third International Congress on Food Science and Technology Institute of Food Technology, Chicago,1971), pp. 344357.

14. N.S. Scrimshaw, M. Behar, D. Wilson, F. Viteri, G. Arroyave, and R. Bressani, "All Vegetable Protein Mixtures for Human Feeding. V. Clinical Trials with INCAP Mixtures 8 and 9 and with Corn and Beans," Am J. Clin. Nutr., 9: 196-205 (1961).

15. T. Hargreaves and GA. Rose, "Reproducibility of the Balance Method in Man as Applied to Calcium and Phosphorus," C/in. Sci., 28: 537-542 (1965).

16. B. Isaksson and B. Sjogren,"A Critical Evaluation of the Mineral and Nitrogen Balances in Man," Proc. Nutr. Soc., 26: 106-116 (1967) .

17. W.M. Wallace, "Nitrogen Content of the Body and Its Relation to Retention and Loss of Nitrogen," Fed. Proc., 18: 1125-1136 (1959).

18. E.C. Owen, "Nitrogen Balances," Proc. Nutr. Soc., 26: 116-124 (1967).

19. J.C. Waterlow, "The Assessment of Protein Nutrition in the Whole Animal with Special Reference to Man," in H.N. Mun ro, ea., Mammalian Protein Metabolism, vol. l l l (Academic Press, New York and London, 1969), pp. 325-390.

20. D.M Hegsted, "BalanceStudies," J.Nutr., 106:307-311 (1976).

21. D.H. Calloway, A.C.F. Odell, and S. Margen, "Sweat and Miscellaneous Nitrogen Losses in Human Balance Studies,"J. Nutr., 101: 775-786 (1971).

22. C. Garza, N.S. Scrimshaw, and V.R. Young, "Human Protein Requirements: The Effect of Variations in Energy Intake within the Maintenance Range," Am. J. Clin. Nutr., 29: 280-287 (1976).

23. C.F. Consolazio, H.L. Johnson, R.A. Nelson, J.G. Dramise, and J.H. Skala, "Protein Metabolism during Intensive Physical Training in the Young Adult," Am. J. Clin. Nutr., 28: 29-35 (1975) .

24. H.H. Mitchell and M. Edman, Nutrition and Climatic Stress (Charles C. Thomas, Springfield, 111., USA, 1951).

25. T.R.A. Davis, "The Influence of Climate on Nutritional Requirements," Am. J. Publ. Health, 54: 2051 -2067 (1964).

26. P.C. Huang, H.E. Chong, and W.M. Rand, "Obligatory Urinary and Fecal Nitrogen Losses in Young Chinese Men,"J. Nutr.,102: 1605-1614 (1972).

27. V.R. Young, N.S. Scrimshaw, B. Torun, and F, Viteri, "Soybean Protein in Human Nutrition: An Overview," J. Am. Oil Chem. Soc., 56: 110-120 (1979).

28. R. Bressani, F. Viteri, and L.G. Elias, "The Usefulness of Results with Human Beings in Standardizing Rat Assays," in J.W.G. Porter and B.A. Rolls, eds.,Proteins in Human Nutrition (Academic Press, London and New York, 1973), pp. 293-316.

29. R. Bressani, "Human Assays and Applications," in C.E. Bodwell, ea., Evaluation of Proteins for Humans (Avi Publishing Co., Westport, Conn., USA, 1977), pp. 81-118.

30. D.M. Hegsted, "Nutritional Research on the Value of Amino Acid Fortification-Experimental Studies in Animals," in N.S. Scrimshaw and A.M. Altschul, eds., Amino Acid Fortification of Protein Foods (MIT Press, Cambridge, Mass., USA, and London, 1971), pp.157-178.

31, V.R. Young, W.M. Rand, and N.S. Scrimshaw, "Measuring Protein Quality in Humans: A Review and Proposed Method," Cereal Chem., 54: 929-948 (1977).

32. D. Navarrete, V.A.L. de Daqui, L.G. Elias, P.A. Lachance, and R. Bressani, "The Nutritive Value of Egg Protein as Determined by the Nitrogen Balance Index (NBl)," Nutr. Repts. Intern., 16: 695-704 (1977).

33. R. Bressani, D.A. Navarrete, V.A.L. de Daqui, L.G. Elias,J. Olivares, and P.A. Lachance,

"Protein Quality of Spray-Dried Whole Milk and of Casein in Young Adult Humans Using a Short Term Nitrogen Balance I ndex Assay," J. Fd. Sci., 44: 1136- 1149 11979) .

34. H.N. Munro, "General Aspects of the Regulation of Protein Metabolism by Diet and by Hormones," in H.N. Munro and J.B. Allison, eds., Mammelian Protein Metabolism, vol.1 (Academic Press, New York and London, 1964), pp. 381-481.

35, D.H. Calloway and H. Spector, "Nitrogen Balance as Related to Caloric and Protein Intake in Young Active Men,"Am. J. Clin. Nutr., 2: 405412 (1954).

36. G. Inoue, Y. Fujita, and Y. Nilyama, "Studies on Protein Requirements of Young Men Fed Egg Protein and Rice Protein with Excess and Maintenance Energy," J. Nutr., 103: 16731687 (1973) .

37. D.H. Calloway, "Nitrogen Balance of Men with Marginal Intakes of Protein and Energy," J. Nutr., 105: 914923 11975).

38. K.W. Keene, C.J. Smutko, C.H. Krieger, and H.E. Denton, 'Water Addition to Purified Diets. l l. Effects of Water and Carbohydrate on Protein Efficiency Ratio in Rats," J. Nutr., 81: 87-90 (1963).

39. R. Bressani and J.E. graham, "Effect of Water Intake on Nitrogen Metabolism in Dogs," J. Nutr.,82: 469474 (1964).

40. Y.R. Gandra and N.S. Scrimshaw, "Infection and Nutritional Status. ll. Effect of Mild Virus Infections Induced by 17-D Yellow Fever Vaccine on Nitrogen Metabolism in Children," Am. J. Clin. Nutr., 9: 159-163 (1961).

41. N.S. Scrimshaw, D. Wilson, and R.J. Bressani, "Infection and Kwashiorkor," Trop. Pediat and African Child Hlth., 6: 3743 11960).

42. W.R. Beisel, W.D. Sawyer, E.D. Ryll, and D. Crozier, "Metabolic Effects of Intracellular Infections in Man," Ann. Int. Med., 67: 744-779 (1967).

43. P.S. Venkatachalam and V.N. Patwardhan, "The Role of Ascaris lumbricoides in the Nutrition of the Host: Effect of Ascariasis on Digestion of Protein," Trans. Roy. Soc. Trop. Med. Hyg., 47: 169-175 (1953).

44. N.S. Scrimshaw, J,P. Habicht, M.L. Piche, B. Cholakos, and G. Arroyove, "Protein Metabolism of Young Men during University Examinations," Amer. J. Clin. Nutr., 18: 321 - 324 (1966).

45. N.S. Scrimshaw, "Discussion: Nutrition and Stresses of Short Term Space Flight," in Conference on Nutrition in Space and Related Waste Problems, University of South Florida, Tampa, Florida, April 27-30, 1964, NASA SP 70 (Scientific and Technical Information Division, National Aeronautics and Space Administration, Washington, D.C.,1964), PP. 79-84.

46. G. Leveille, "Meal Frequency and Time of Feeding," in P.L. White and D.C. Fletcher, eds., Nutrients in Processed Foods: Proteins (published for the AMA by Publishing Sciences Group, Acton, Mass., USA, 1974), pp. 61-66.

47. Food and Nutrition Board, National Research Council, Recommended Dietary Allovvances, 9th rev. ed. (National Academy of Sciences, Washington, D.C., 1980).

48. P.C. Huang, V.R. Young, B. Cholakos, and N.S. Scrimshaw, "Determination of the Minimum Dietary Essential Amino Acid-to-Total Nitrogen Ratio for Beef Protein Fed to Young Men," J. Nutr.,90: 416-466 11966).

49. N.S. Scrimshaw, V.R. Young, P.C. Huang, 0. Thanangkul, and B.V. Cholakos, "Partial Dietary Replacement of Milk Protein by Non-specific Nitrogen in Young Men," J. Nutr., 98: 9-17 (1969).

50. N.S. Scrimshaw, V.R. Young, R. Schwartz, M.L. Piche, and J.B. Das, "Minimum Essential Amino Acid-to-Total Nitrogen Ratio for Whole Egg Protein Fed to Young Men," J. Nutr., 89: 9-18 (1966).

51. C. Kies, "Non-specific Nitrogen in the Nutrition of Human Beings," Fed. Proc., 31: 11721177 (1972).

52. J.M. McLaughlan, "Nutritional Significance of Alterations in Plasma, Amino Acids, and Serum Proteins," in Committee on Amino Acids, Food and Nutrition Board, National Research Council, Improvement of Protein Nutriture (National Academy of Sciences, Washington, D.C., 1974), p.89.

53. G. Arroyave, D. Wilson, M. Behar, and F. Viteri, "The Development of INCAP Vegetable Mixtures. I I. Biochemical Testing," in Progress in Meeting Protein Needs of Infants and Preschool Children, NAS-NRC Publ.843 (National Academy of Sciences, Washington, D.C., 1961), pp. 49-56.

54. J.B. Longenecker and N.L. Hause, "Relationship between Plasma Amino Acids and Composition of the Ingested Protein. ll. A Shortened Procedure to Determine Plasma Amino Acid (PAA) Ratios,"Am. J. Clin. Nutr., 9: 356-362 (1961),

55. J.M. McLaughlan, "Relationship between Protein Quality and Plasma Amino Acid Levels," Fed. Proc., 22: 1122-1125 (1963).

56. V.R. Young and N.S. Scrimshaw, "Nutritional Significance of Plasma and Urinary Amino Acids," in The /nternational Encyclopaedia of Food and Nutrition, vol. 11: Protein and Amino Acid Functions, ed. E.J. Bigwood (Pergamon Press, Oxford and New York, 1972), pp. 541-568.

57. C.E. Bodwell, "Biochemical Indices in Humans," in C.E. Bodwell, ea., Evaluarion of Proteins for Humans (Avi Publishing Co., Westport, Conn., USA, 1977), pp. 119-148.


Contents - Previous - Next