Previous PageTable of ContentsNext Page


D. M. Hegsted
Department of Nutrition
School of Public Health
Harvard University

During complete starvation relatively large amounts of nitrogen are excreted into the urine and are obviously derived from the catabolism of the tissue proteins. When calories are provided which contain no protein, the urinary nitrogen falls to much lower levels. These basic facts which demonstrate that "calories spare body protein" have long been known. The quantitative aspects of the relationship between calories and protein remain, however, relatively obscure. The field was critically reviewed by Munro in 1951 (1).

The evidence is convincing that at very low levels of calorie intake which are incapable of maintaining the body tissue over long periods, dietary protein serves little useful purpose. For example, in the studies of Gamble (2) on life boat rations during World War II in which provision of both food and water was severely limited, the conclusion was reached that 400 calories of sugar was more effective than diets containing protein at the same caloric level. The provision of dietary protein at these low caloric intakes simply elevated the urinary nitrogen an equivalent amount and this increased the water requirement.

At the other end of the spectrum, Munro cites numerous experiments in which the provision of a surfeit of calories as carbohydrate and fat, over and above an ordinarily adequate diet, lowered the urinary nitrogen excretion and increased nitrogen balance. These experiments are of short duration and it would seem clear that in the long run excessive intakes of calories will not sustain a continual increase in the body nitrogen other than the amount required to expand the body's component of adipose tissue.

Finally, as a third type of conclusion that has been drawn, are the studies of Mitchell and Beadles (3) in which growing rats were fed different levels of protein and calories. They conclude that "In the growing rat, the protein percentage in diets of equal caloric density necessary for maximum growth at any caloric intake level is independent of the caloric intake level within a considerable range." This is to say that the protein requirement is independent of the calorie level provided and vice versa.

The confusion in this area is demonstrated by the two common and contradictory statements that one often hears. On the one hand it is stated that since proteins are metabolized to yield calories when calories are limited, increasing the protein intake will be of no benefit; it is also commonly stated that since calorie limitation increases the breakdown of body proteins, protein needs are high when calories are limited.

The confusion probably is due in large part to the difficulties in the design of appropriate experiments. A major one is that the consequences of modifying the calorie or protein intake will probably depend largely upon the calorie (adipose tissue) and protein (labile protein) reserves of the individuals studied. The limitation of dietary calories in a person with a reasonable component of adipose tissue probably imposes little restriction upon the total calories available for the metabolic machinery since the adipose tissue constitutes an available supply of calories which is readily mobilized. The major change may be simply a shift from the catabolism of the usual dietary supply of carbohydrate and fat to the utilization of fat as the primary source of calories. Since some carbohydrate is presumably required for metabolism, this will have to be supplied by gluconeogenesis from body protein if the diet completely lacks carbohydrate and this would result in some breakdown of additional body protein. However, the effects of calorie limitation may be quite different in normal or obese individuals as compared to very lean individuals.

Similarly, the effects of protein or calorie limitation may depend upon whether the animal was previously fed a very low or a high protein diet. Since the so-called "labile protein reserves" are easily mobilized, short-term balance studies may yield results quite different from long-term studies. In any event, it should be clear that if either calories or protein are limiting so that body tissue is being lost or gained, these reserves will change with time. Thus, the ultimate effects may be quite different from those observable immediately after the dietary regimen is imposed.

Similarly in the types of studies reported by Mitchell and Beadles referred to previously, if one imposes defined levels of calories or protein on young growing animals such as young rats, the animals either gain or lose weight and continually modify their protein and calorie needs. Thus, within a few days the degree of deficiency or excess imposed may be quite different from that originally selected and the comparability of the groups may be lost. The design of an ideal experiment may be impossible.

Obviously, these are not the only problems involved in interpreting the data available. Different, more efficient or less efficient, metabolic pathways may be activated, the proportions of carbohydrate, fat and protein metabolized may have subtle effects, and so forth as discussed by Munro.

In many ways it is a contradiction of terms to speak of "protein requirements" when calories are limiting or "calorie requirements" when adequate protein is not provided. However, it also is clear that various populations do, in fact, exist at quite different levels of calorie intake, and these are most often quite low in the developing countries. It is necessary to consider the protein needs of such populations.

As Munro noted, the studies of Allison on dogs which received intakes of calories of 80 or 100 cal/kg of body weight (4, 5) or only 50% of their normal requirements, the nitrogen balance index of the protein was not affected. His data apparently show that this degree of restriction was essentially without effect upon the urinary nitrogen or the degree of nitrogen balance at comparable levels of casein intake. On the other hand, when the calorie intake was dropped to only 25% of requirements (5), there was an immediate rise in the urinary nitrogen. Even at this low caloric level, there was apparently some slight increase in nitrogen retention as the level of dietary casein was raised. Thus, Munro concluded that "On the whole, it seems that the capacity of the body to take full advantage of an increase in nitrogen intake is restricted only at very low levels of energy intake." However, Rosenthal and Allison found with both rats (6) and dogs (5) that at the lower levels of calorie intakes the nitrogen balance tended to drift downward and, as might be expected, eventually a marked negative balance was obtained. In the experiment on dogs, they tested animals with different "protein stores", keeping the case in intake constant and feeding about 20% of the normal calorie intake. It is instructive to note that the animals with presumably adequate stores went into a moderate negative balance immediately while the animal which was protein depleted was in positive balance. Thus, the well known ability of depleted animals to retain nitrogen more efficiently was demonstrated even at restricted calorie levels. However, the animal with adequate stores continued in moderate negative balance for a long period and only after approximately 70 days did it drift into a more marked negative balance. On the other hand, the depleted dog went into a negative balance after 20 days. A dog with moderate stores was intermediate between the two extremes.

In young children recovering from malnutrition very high caloric levels have been found to be necessary for maximum growth and nitrogen retention, i.e., intakes of the order of 150 cal/kg/day (7, 8, 9). In the infants studied by Ashworth and Waterlow (8, 9) the calorie intakes during the rapid growth phase were 160 cal/kg and fell to 117 after recovery; growth rates were 10.1 gm/kg/day and fell to 3.3. Now it is perfectly clear that recovery from kwashiorkor does occur when the total food intake is substantially below these levels of calorie supply. It would appear that in these rapidly growing infants the situation is generally similar to that indicated by Mitchell and Beadles in the young rat. Protein retention obviously occurs at any calorie intake which permits growth. As far as I am aware the rate of growth is primarily a function of the calorie intake if reasonable levels of protein are supplied. That is, no advantage can be shown for excessive levels of protein. Thus, these types of data apparently support the conclusion of Mitchell and Beadles that the protein requirement as percentage of calories is independent of the level of caloric intake. It should also be noted that satisfactory results are obtained with a variety of proteins of reasonable quality when these are fed to infants at levels supplying 5 to 7% of calories (10, 11), levels of the same order of magnitude as those in breast milk.

Ashworth and Waterlow found a large increase in metabolic rate after meals in the children during "catch up growth". These are interpreted as the "cost of protein deposition" rather than as specific dynamic action since no comparable increase was found after the rapid growth phase terminated. These costs were estimated to range from 50 to 80 calories per 10 gm of new tissue - 100 to 160 per gm of protein deposited.

Data on the relative roles of protein and calories during recovery from severe malnutrition in adults, such as that of Beattie et al. (12) appear to be generally similar. That is, they emphasize the role of calories once the protein requirement is met rather than the advantage of excessively high levels of protein.

The conclusion that I draw on calorie-protein relationships are (1) that more metabolic experiments in which caloric restriction has been imposed are likely to be misleading as difficult to interpret since the total calories available to the body are unknown and will vary with time; (2) that at levels of caloric intake which are compatible with a reasonable existence over long periods of time, there is little evidence that the total protein requirement is raised or that the efficiency of utilization of various proteins is markedly changed; (3) that calorie needs eventually dominate all other needs. The addition of protein to the diet is not useful in overcoming caloric deficits except as an expensive source of calories. Protein needs are a percentage of the calories are not increased by caloric restriction.


1. Munro, H.N. Carbohydrate and fat as factors in protein utilization and metabolism. Physiol. Rev., 31: 449, 1951.

2. Gamble, J.L. The water requirement of castaways. Proc. Am. Phil. Soc., 88: 151, 1944.

3. Mitchell, H.H. and J.R. Beadles. The determination of the protein requirement of the rat for maximum growth under conditions of restricted consumption of food. J. Nutr., 47: 133, 1952.

4. Allison, J.B. Calories and protein nutrition. Ann. N.Y. Acad. Sci., 69: 1009, 1958.

5. Rosenthal, H.L. and J.B. Allison. Some effects of caloric intake on nitrogen balance in dogs. J. Nutr., 44: 423, 1951.

6. Rosenthal, H.L. and J.B. Allison. Effects of caloric intake on nitrogen balance and organ composition of adult rats. Agri. Food Chem., 4: 792, 1956.

7. Graham, G.G., A. Cordano and J.M. Baertl. Studies in infantile malnutrition. III. Effect of protein and calorie intake on nitrogen retention. J. Nutr., 84: 71, 1964.

8. Ashworth, A. Malnutrition and metabolic rates. Nutr. Rev., 28: 279, 1970.

9. Ashworth, A. and J. Waterlow. Calorie and protein intakes and growth. Lancet, 1: 776, 1969.

10. Graham, G.G. The significance of the first limiting amino acid in human infant diets. In: Amino Acid Metabolism and Genetic Variation (W.L. Nyhan, ed). McGraw Hill Book Co., New York, 1967.

11. Fomon, S.J. and L.J. Filer, Jr. Amino acid requirements for normal growth. In: Amino Acid Metabolism and Genetic Variation (W.L. Nyhan, ed.). McGraw Hill Book Co., New York, 1967.

12. Beattie, J., P.H. Herbert and D.J. Bell. Nitrogen balances during recovery from severe undernutrition. Brit. J. Nutr., 1: 202, 1947.

Previous Page Top of Page Next Page