Contents - Previous - Next

Protein-energy interrelationships during rapid growth


* 15, Hillgate Street, London W8 7SP, U.K.

It is interesting to examine protein-energy interrelationships during rapid growth, when the effects of imbalances are likely to be shown most clearly.

1. Efficiency of protein deposition
2. Protein turnover during rapid growth
3. Energy cost of protein synthesis

1. Efficiency of protein deposition

Table 1 shows results from a selection of studies. In the two reports on prematures the absolute rates of gain of body weight and body protein are almost identical with those calculated from the data of SHEPHARD (1991) for the body composition of the fetus in utero at 30-35 weeks:

Efficiency has been calculated as N retained/(N intake - obligatory loss). I do not know whether this is the correct approach: to deduct the maintenance requirement rather than the obligatory loss would be to assume a fixed value for the efficiency of maintenance. The figure taken for the obligatory loss is 84 mg N or 0.52 g protein/kg/d, as proposed by FAO/WHO/UNU (1985). It may be that this is too high. In the study of CATZEFLIS et al. (1985) regression of N retained on intake gave an obligatory loss of 48 mg N/kg/d. If that value is nearer the mark, the efficiencies in Table 1 will be overestimated.

Be that as it may, it is clear that the premature and newborn deposit protein quite efficiently, while the older children are less efficient. Thus, in the study of FJELD et al. (1989), which is probably the most detailed and comprehensive that has been published on catch-up growth, the children showing moderate rates of gain had energy and protein intakes slightly higher than the prematures, but they only deposited about half as much protein, in spite of the stimulus of being severely malnourished (average Wt/Ht Z score -2.8). From the dietary data given in this paper there can be no question of deficiency of any known micronutrient. It looks as if older children may have an intrinsically lower capacity for growth, so that larger intakes of protein and energy are needed to achieve the same result as in prematures.

Table 1. Efficiency of protein deposition during rapid growth. All values per kg per day.

Subjects and Authors

Metabolizable energy intake kcal/d

Weight gain g/d

Protein intake g/d

Protein gain g/d

Net efficiencyX %

A. In utero







B. Prematures

1. DE BENOlST et al., 1984






2. CATZEFLIS et al., 1985






C. Newborn

FOMON, 1986+






D. Catch-up growth *

1. Rapid:

FJELD et al., 1989






2. Moderate:

FJELD et al., 1989






3. JACKSON et al., 1983**






X Calculated as: gain/(intake - obligatory loss).
Obligatory loss taken as the equivalent of 0.525 g protein/d.
+ Breastfed children 0-1 month.
* Children aged about I year.
** Children receiving 102 kcal/kg/d.

One could speculate that this reflects the transition from the infant to the child phase of growth, as described by KARLBERG (1989), a transition that presumably depends on hormonal factors. We might call this, in Millward's terms, a lower anabolic drive.

2. Protein turnover during rapid growth

It has been recognized for a long time that during periods of rapid growth rates of protein synthesis and breakdown are increased above the non-growing level. In a recent study in which labelled leucine was infused just before elective Caesarian section, the rate of protein synthesis in the fetus in utero was seven times the rate of accretion (CHIEN et al., 1992). Similar results have been obtained in prematures (Table 2A).

The uniformity of the synthesis rates shown in Table 2A disguises the fact that the total synthesis is made up of two components, the basal or maintenance synthesis and that associated with growth.

These components vary at different ages. When measurements have been made at different levels of protein intake and gain, it is possible to estimate separately the basal rate of protein synthesis at zero gain and the increase in synthesis per gram gain, according to the simple relationship y = A + Bx, where y = synthesis and x = gain. Some results obtained in this way are shown in Table 2B. Comparison of the prematures with the older children suggests that the basal rate of protein synthesis (A), just like the basal rate of oxygen uptake, is much lower in the prematures, whereas the slope (B) is greater.

Admittedly, doubts may be expressed about the validity of these values for protein synthesis, all of which except for those of DE BENOIST et al. (1984), were obtained with 15N-glycine, with urea as end-product. However, similar results have been recorded in growing animals studied with 14C-labelled amino acids (see e.g., REEDS and HARRIS, 1981), and I think it is clear that, during growth, rates of protein synthesis are substantially higher than rates of protein gain. Obviously there are implications for energy balance.

Table 2. Relationship of protein synthesis to protein gain

A. Total protein synthesis (g)1 per g protein gained

Fetus in utero

CHIEN et al., 1992



DE BENOIT et al., 1984


CATZEFLIS et al., 1985


Catch-up growth

GOLDEN et al., 1983


B. Basal rate of protein synthesis and net synthesis per g protein gained2

Basal synthesis

Synthesis, above basal


g/g protein gain


CATZEFLIS et al., 1985



Catch-up growth

GOLDEN et al., 1977



Children recovered from malnutrition

JACKSON et al., 1983



1 In all cases rates of gain were 1.8-2.0 g protein/kg/d.
2 The relation between synthesis (S) and gain (G) is of the form S = A + BG, where A, the intercept, is the basal rate of synthesis at zero gain, and B. the slope, is the synthesis above basal per g gained. Because of the intercept, calculations of total synthesis, as in A above, must be made at similar rates of gain.

3. Energy cost of protein synthesis

This is a well worn subject. Table 3 shows calculations in which we tried to assess the energy costs associated with protein synthesis, over and above the cost of peptide bond formation (WATERLOW and MILLWARD, 1989). The expenditure on messenger RNA formation can be little more than a guess without more knowledge about mRNA turnover. Some routes of protein degradation are ATP-dependent, such as that which requires coupling with ubiquitin (e.g., HERSHKO, 1985), but we do not know the quantitative importance of these pathways. The overall figure that we arrived at was about 2.2 kcal/g protein synthesized. If protein breakdown does have a significant cost, this figure will be an underestimate, but if messenger RNA is utilized more than once it will be an overestimate.

CATZEFLIS et al. (1985) plotted the regression of energy expenditure on protein gain. Combining this with their results for protein synthesis, we get a value of 2 kcal/g protein synthesized, which fits quite well with our theoretical estimate.

Table 4 summarizes the energy balance in these prematures. The energy cost of depositing protein (Table 4B) was twice as great as the energy content of the protein stored (Table 4A). When this cost is deducted from the energy expenditure (58 kcal/kg/d), there remain 38 kcal/kg/d for metabolic processes other than growth. In fact, CATZEFLIS et al., (1985) calculated from the regression of energy expenditure on growth that the basal metabolic rate at zero growth would be 40 kcal/kg/d, which shows good agreement with my calculations. The prematures also stored 30% of their energy intake as lipid.

It seems, therefore, that the premature infant is able to store large amounts of protein and fat on a modest energy intake, because it has a low basal metabolic rate compared with the normal newborn. This recalls the suggestion of WIESER (1989) that it may be incorrect to regard expenditures on maintenance and on growth as necessarily additive. Apparently in the biological realm there are organisms which are able to grow at the expense of their basal metabolism.

The results of JACKSON et al. (1983) shown in Table 5 provide in some ways an interesting contrast. This is the only study I know of in which protein turnover was measured with 15N-glycine while intakes of protein and energy were varied independently. The synthesis rate, based on urea as end-product, was well maintained on the low-protein (LP) diet except at the lowest level of energy intake. With this exception, energy intake had no effect on the rate of protein synthesis determined from the labelling of urea. The different diets had far more dramatic effects on protein deposition. On the LP diet, the children were just able to maintain protein balance provided that the energy supply was more than about 80 kcal/kg/d. On both protein intakes, retention decreased progressively as the energy intake fell. Deposition of protein on the HP diet involved very small energy costs, yet even this cost could not be maintained when the energy intake fell below 100 kcal/kg/d.

Table 3. Components of the energy cost of protein synthesis


kcal/g protein synthesized

Peptide bond formation

(4 P per bond)


RNA turnover






1.2 (1)

Amino acid transport

(0.3 P per amino acid)


Breakdown coupling with ubiquitin


(1 P per protein molecule)

Lysosomes - maintenance of acidity





(1) It is assumed that each molecule of RNA is utilized only once for transcription. For details of the calculation see WATERLOW and MILLWARD (1989).

Table 4. Energy balance in premature infants (CATZEFLIS et al., 1985)



A. Metabolizable energy intake




Energy stored in protein


Energy stored in lipid



B. Protein deposited (Table 1)

1.76 g/kg/d

Synthesis associated with protein deposition

at 5.2 g/g (Table 2)

9.15 g/kg/d

Energy cost of this synthesis at 2.2 kcal/g

(Table 3)

20 kcal/kg/d

Table 5. Rates of protein synthesis and deposition at different levels of protein and energy intake (from JACKSON et al., 1983). All values per kg per day


Metabolizable energy intake kcal

Protein synthesis g (1)

Protein deposited g

Synthesis associated with protein deposition g (2)

Energy cost of protein deposition kcal (3)

High protein






1.66 g/kg/d











Low protein




0.69 g/kg/d







(1) From results based on labelling of urea.
(2) At 2.7 g synthesized per g deposited (Table 2).
(3) At 2.2 kcal per g synthesis (Table 3).

We must conclude that, compared with the prematures, these children had a much higher threshold at which energy became limiting for growth in protein; other pathways of energy utilization clearly had a higher priority. The paper does not give information about rates of weight gain from which fat deposition could be calculated.

This study brings out another point of great interest. Rates of nitrogen flux were calculated from ammonia (QA) as well as from urea (QU). On the low protein intake QA was greatly reduced. There was no effect of the energy intake; combining results at all three energy levels, the mean QU/QA on the higher protein intakes was 1.31, on the lower intakes 2.13. This difference is highly significant p=0.001). We put forward the hypothesis some years ago (FERN et al., 1985) that QU predominantly reflects nitrogen metabolism in the visceral tissues, QA that in the periphery.

Along the same lines, SOARES et al. (1991) showed that QU/QA was higher in undernourished Indian labourers than in normal weight controls. According to this interpretation, Jackson's children on the low protein intake showed little change in overall protein turnover, but a shift in pattern, from periphery to viscera, which would fit the argument in the previous IDECG workshop (WATERLOW, 1990) that the body gives metabolic priority to the central tissues. This, of course, is not a new idea; it was well established long ago on the basis of animal experiments on the loss of protein from different tissues. It would be convenient if it turned out that QU/QA could be used in vivo as a measure of the impact of marginal protein intakes.

Contents - Previous - Next