Contents - Previous - Next


Quantitative relationships between protein and energy metabolism: Influence of body composition

C.J.K. HENRY*

* School of Biological and Molecular Science, Oxford Polytechnic, Oxford OX3 OBP U.K.


Abstract
1. Introduction
2. Theoretical basis
3. Constancy of tissue mobilisation
4. Tissue mobilisation in the obese
5. Allometric analysis
6. Conclusions
References


Abstract


Using the model for energy regulation proposed by PAYNE and DUGDALE (1977), a quantitative relationship between energy metabolism, i.e., basal metabolic rate (BMR) and protein metabolism, i.e., fasting urinary nitrogen loss (FUNL) is presented. An analysis of classical human studies on starvation shows that the fraction of total energy expenditure derived from protein (i.e., the P ratio) remains unchanged, confirming a close association between protein and energy metabolism. The propensity to use protein as a fuel appears to be fixed in an individual regardless of whether the synthesis of protein is limited by energy or protein intake. During starvation, obese individuals appear to mobilise about 5% of total energy from protein breakdown, while normal weight subjects about 20%. Body composition exerts a profound influence on protein metabolism. A quantitative relationship between BMR, FUNL and obligatory nitrogen loss (ONL) is also presented. It is thus proposed that body composition may significantly influence ONL and hence human protein requirements.

1. Introduction


The existence of a relationship between energy metabolism or, more specifically, basal metabolic rate (BMR) and protein metabolism, i.e., obligatory urinary nitrogen loss (OUNL) was first proposed by TERROINE and SORG-MATTER in 1927 and later elaborated by BRODY (1945) and SMUTS (1935). The main evidence used to justify this relationship were comparative animal studies and the allometric (inter-species) analysis of BMR and OUNL. Allometric analysis showed that both BMR and OUNL were related to weight raised to a power of 0.75 (W.75) and the values obtained were:

OUNL (ma N/d) = 140 W.75
BMR (kcal/d) = 70 W.75

Dividing one by the other leads to a mean OUNL of 2 mg N/basal kcal. Such a relationship is of theoretical and practical importance in the estimation of protein and energy requirements.

BMR, the metabolic rate of a resting, non-stressed subject in a thermoneutral environment, after a short fast, is relatively easy to measure. Obligatory urinary nitrogen loss measurements, i.e., the nitrogen lost on a non-protein diet for some days, are experimentally more difficult to obtain. Estimates of the magnitude of OUNL in a range of animals have been obtained by using BMR as a proxy to predict the nitrogen lost on a non-protein diet, using the value of 2 mg N/basal kcal. This method was also used by the 1973 FAO Expert Committee on protein requirements to estimate the obligatory nitrogen loss in children (FAO, 1973). Critics have rightly pointed out that, while this is an inter-species average, directly measured values of BMR and ONL in the same subject can change considerably. This is well illustrated by re-analysing the study conducted by DEUEL et al. (1928) shown in Figure 1. The ONL/BMR ratio varied from 2.8 to 1.2 during the 30-day period on a non-protein diet.

2. Theoretical basis


To date, no theoretical reasons have been advanced to explain this putative relationship between OUNL and BMR. This paper therefore aims to present a theoretical framework to describe the relationship between protein and energy metabolism. The two main questions addressed here are:

1. How close is the quantitative relationship between protein and energy metabolism in humans?
2. Does body composition influence protein and energy metabolism?



3. Constancy of tissue mobilisation


A possible theory to accommodate the observed relationship between OUNL and BMR can be derived from the model developed by PAYNE and DUGDALE (1977), initially proposed as a model for energy regulation. These authors suggested that the way in which individuals regulate energy balance depends on their P ratio, defined as:

Figure 1. Relationship between ONL and BMR over time. (After DEUEL et al.,1928)

Subjects with a high P ratio will deposit their excess food energy as protein which has a high maintenance energy requirement and thereby maintain energy balance more effectively than those with a low P ratio. Those with a low P ratio will tend to deposit excess food energy as fat.

A direct relationship between protein and energy is also suggested by the P theory. During tissue mobilisation, it predicts that a constant fraction of energy should be derived from the breakdown of protein, i.e., the P ratio should remain constant. Extending this model further, I propose that the P ratio may not only influence energy balance but also influence human protein requirements. In other words, the propensity to use protein as a fuel is fixed in an individual regardless of whether the net synthesis of protein is limited by energy intake or protein intake. If the assumption is made that body protein is 16% nitrogen and has a metabolisable energy value of 4 kcal (16.7 kJ)/g (HENRY, 1984), then the P ratio during fasting may be calculated, using the fasting urinary nitrogen loss and fasting metabolic rate as follows:

Figure 2 shows the urinary nitrogen excretion and fasting metabolic rate in the professional faster named Levanzin, studied by BENEDICT (1915). If energy expenditure is taken as approximately equal to the BMR, the P ratio for Levanzin averaged 0.187 0.013 and, as Figure 1 shows, after the first 3 days of the fast, during which time it rose, it exhibited no tendency to vary systematically over time. The value for the last 10 days of fasting (0.175) was not significantly different from that observed on days 3-13 (0.190). This constancy is in marked contrast to the changes in both numerator and denominator in equation 1, both of which changed by more than 25% during the fast. The interesting aspect of this re-analysis is that the fraction of protein-energy mobilised, i.e., the P ratio, remains constant during the duration of the fast, despite dramatic falls in both nitrogen loss and energy expenditure.

Figure 2. The constancy of P ratio during fasting.

The P ratio was 0.187 0.013. Thus, Levanzin obtained approximately 18.7% of this energy from protein metabolism. Benedict's study is exceptional, but not unique. Two other, less well-known human studies from Japan (TAKAHIRA, 1925) and Holland (VAN NOORDEN, 1907) were also used to determine the P ratio during fasting. The results are presented in Table 1. In every case the P ratio remained remarkably constant. I believe that this is the first direct evidence we have of a clear and, more importantly, constant quantitative relationship between energy and protein metabolism in humans. Put another way the FUNL/BMR ratio in an individual is constant during a fast, but varies widely among individuals (see Table 1).

Table 1. Urinary nitrogen loss, metabolic rate and P ratio in non-obese human subjects during fasting

Subject No.

Investigator

Initial body weight [kg]

Duration of fast [days]

Fasting urinary N loss [g/d]

Metabolic rate [kcal/d]

P ratio [mean SD]

CV [%]

1

BENEDICT, 1915

60

31

7.8

1130

0.19 0.013

6.8

2

TAKAHIRA, 1925

58

30

7.9

951

0.19 0.017

8.9

3

TAKAHIRA, 1925

79

26

7.5

1176

0.16 0.011

6.8

4

TAKAHIRA, 1925

51

17

7.9

950

0.19 0.011

5.7

5

TAKAHIRA, 1925

43

16

10.3

951

0.24 0.202

8.3

6

TAKAHIRA, 1925

50

12

10.7

1087

0.23 0.012

5.2

7

VAN NOORDEN, 1907

57

109.5

1508

0.17 0.003

5.6


8

VAN NOORDEN, 1907

59

7

9.7

1568

0.16 0.011

6.8

9

VAN NOORDEN, 1907

60

6

9.9

1292

0.19 0.011

5.7

10

VAN NOORDEN, 1907

70

5

11.5

1970

0.15 0.012

7.5


4. Tissue mobilisation in the obese


A closer examination of the pattern of tissue mobilization during starvation in 'obese' and 'normal' subjects reveals some interesting features. Figure 3 shows the pattern of urinary nitrogen loss during the 31-day fast in a normal weight subject (60.6 kg) studied by BENEDICT (1915) and compares it with an obese subject (154 kg) studied by GILDER et al. (1967) and starved for a comparable period. After the first week, the obese subject lost less fasting urinary nitrogen than the normal weight subject.

This difference in nitrogen loss is further highlighted if we compare the P ratio of normal subjects (Table 1) with that of obese subjects (Table 2). The obese have a lower P ratio and appear to mobilise about 5% of their energy from protein breakdown, in contrast to normal weight subjects, who mobilised approximately 20% of their energy from protein breakdown. It therefore appears that normal weight and obese subjects respond differently during starvation, and that the adipose tissue exerts a profound influence on protein mobilisation.

Although there are no data on P ratio and direct body composition measures in humans, weight and height measured in subjects cited in Tables 1 and 2 were supplemented with single point measurements of BMR and nitrogen loss during fasting on seven more subjects. These values are shown in Figure 4. It can be seen that a close agreement exists between the P ratio and the BMI at the start of the fast. The Figure also shows GARROW's (1983) proposed relationship between BMI and body fat. It is clear from the graph that the P ratio declines in a curvilinear fashion with increasing adiposity. It therefore appears that the tendency to use protein as a fuel in starvation is much lower in the obese than the non-obese human, as has already been suggested by experiments on laboratory animals (HENRY, 1984).

Table 2. P ratio in obese human subjects during fasting

Subject No.

Initial body weight [kg]

Duration of fast [d]

Fasting urinary nitrogen [g/d]

BMR [kcal/d]

P ratio

1

123

38

3.90

1985

0.049

2

132

38

4.00

1707

0.058

3

147

38

3.27

1773

0.046

4

154

21

4.40

1740

0.063

5

190

21

4.56

2480

0.046

6

166

14

3.41

2100

0.040

7

73

14

2.83

1200

0.059

8

124

21

5.70

2328

0.061

9

120

18

3.20

1984

0.040

10

103

4

4.50

1509

0.074





Mean

0.0536






0.011

Figure 3. Mean fasting urinary nitrogen loss in normal weight and obese subjects.

Figure 4. Relationship between P ratio, W/H2 and percentage fat in humans.



5. Allometric analysis


Some years ago, using allometric analysis, I proposed a relationship between obligatory nitrogen loss (ONL) and fasting urinary nitrogen loss (FUNL) in homeotherms (HENRY, 1984). The allometric analysis of FUNL of 12 mammalian species in the weight range of 0.022 kg to 453 kg produced the following results:

FUNL = 418 W0.77 (mg N/d)
ONL = 272 W0.75 (mg N/d)
BMR = 70 W0.75 (kcal/d)

Since the weight exponent in all cases is equal, or close to 0.75, we can derive the following relationship between FUNL, ONL and BMR:

FUNL = 1.5 ONL
FUNL = 6 mg N/basal kcal

Using published values for the initial FUNL and ONL in infants, young men and old men after a 2- to 3-day fast or on an adequate protein diet, the ratio obtained (Table 3) is similar to that observed in animals. The suggested relationship between FUNL/BMR and ONL provides a possible theoretical basis for the enigmatic ONL/BMR relationship. The argument is as follows: When a subject is deprived of food, there is a constant relationship between FUNL and BMR which is determined by the P ratio of the individual. Thus, the FUNL/BMR ratio not only provides a measure of the catabolism of protein in relation to metabolic rate but of the propensity to catabolize free amino acids.

Since the ratio FUNL/BMR has been shown to be constant in an individual, the allometric relationship reported here between FUNL/BMR and ONL suggests that all three parameters are closely connected and bear a physiological relationship to one another. It is therefore proposed that subjects with a low P ratio (i.e., the obese) will not only have a lower FUNL but also a smaller ONL and hence a lower protein requirement. Although direct evidence to support this is lacking in man, rats made obese by cafeteria feeding were found to have a significantly smaller FUNL and ONL in comparison to controls (HENRY et al., 1985).

Table 3. Fasting urinary nitrogen loss (FUNL) and obligatory nitrogen loss (ONL) (mean SD) and the ratio FUNL/ONL in man

Subjects

Age [years]

FUNL [mg/kg/d]

Age [years]

ONL [mg/kg/d]

FUNL ONL

Infant

newborn

78 30a

4-6 months

57b

1.3

Young men

23.5

156

22.8

89c

1.7


3.7

26a

0.6



Old men

82.1

125

70.13

80d

1.5


3.9

14a

1.55



a McCANCE and STRANGEWAY, 1954
b FOMON, DE MAEYER and OWEN, 1965
c BODWELL et al., 1979
d UAUY et al., 1978

Most importantly, although the obese animals had a smaller FUNL and ONL, the ratio FUNL/ONL remained close to 1.5. Since in a population a continuum exists between those with a low P ratio and those with a high P ratio, it is proposed that the variation in P ratio dictates their FUNL and ONL values, and the observed variation in obligatory nitrogen loss in a group of subjects is a reflection of their pattern of tissue mobilisation when fed a protein-free diet.

6. Conclusions


The evidence presented in this paper suggests that there is a quantitative relationship between protein and energy metabolism. This is reflected in the association between FUNL and BMR in humans. It is suggested that the often reported relationship between ONL and BMR is an indirect consequence of the real relationship between FUNL and BMR. In other words, I propose that FUNL and BMR are the primary parameters of which the relationship between ONL and BMR is a derivative.

In estimating protein requirements, the factorial method has proved to be of great practical value. Central in this is the estimation of the obligatory nitrogen loss (ONL), which is the nitrogen loss (mainly in the urine and faeces) when a protein-free diet is consumed. The ONL value has been widely studied because of its significance as a measure of maintenance protein requirement which is the dominant component of protein requirements, estimated in humans to account for 70-100% of the total protein requirement depending upon age.

Body composition (notably adipose tissue) appears to determine protein metabolism as measured by FUNL and ONL. The results presented in this paper suggest that obesity is associated with a lowering of ONL and that such changes are secondary to the development of obesity itself. The implication of this - that the obese may have lower protein requirements - has not been fully explored. This is important, not only for our understanding of the biology of obesity, but also because many of the experiments conducted to determine human protein requirements have been done in developed nations on volunteers who, if not obese, were at least fatter than is usual in developing countries. Our study suggests the possibility that a previously unconsidered factor could lead to an underestimation of protein requirements in these experiments.


Contents - Previous - Next