Calculation of short-term changes in body fat from measurement of respiratory gas exchange

Calculation

of Short-Term Changes in Body Fat From Measurement of Respiratory Gas Exchange

R.F. G.

J. King, D. J. Almond,

C. B. Oxby, J. H. M. Holmfield,

and M. J. McMahon

A method is presented where the gain in body fat of patients receiving intravenous nutrition (IVN) may be computed using indirect calorimetry. When patients are administered their nutritional requirements soley in the form of glucose and amino acids, metabolism of these will result in changes in the exchange of carbon dioxide and oxygen. Since net fat synthesis results in the production of excess carbon dioxide whereas degradation results in a net reduction in the production of carbon dioxide, changes in RQ will reflect underlying changes in net fat balance. A correction may be applied to these measurements to allow for net protein balance. The basic equations and stoichiometry for the synthesis of tripalmitin from glucose are shown together with the derivation of the formulae that enable the calculation of fat changes to be made. A group of 13 patients was used to assess the technique and given energy at 1.5 times their resting metabolic expenditure together with adequate nitrogen as amino acids to maintain protein balance. The patients were monitored daily in the resting state and fed at a constant rate so that errors due to discontinuous intake or exercise were avoided. The group was found to gain 874 + 243 g fat during the 14 day study. Although individual changes of fat in patients ranged from -819 to f2487 g, net synthesis of fat was found in ten of them. It is suggested that indirect calorimetry is useful not only in the assessment of actual expenditure in patients but can also be used to quantify the fate of administered fuels during intravenous nutrition.

I

NTRAVENOUS nutrition (IVN) is now firmly established as a technique for the maintenance or repletion of lean body mass in patients who are malnourished or catabolic.‘” However the choice of an intravenous regimen remains somewhat arbitrary, and the proportional composition of the nutritional solution and the amounts of energy and nitrogen that are optimal remain to be determined.7-9 The utilization of energy and amino acids in the maintenance of lean body mass has been shown to be related to changes in the storage and utilization of both glycogen and fat.lO-‘* Clearly, any assessment of the full effect of a nutritional regimen should consider such interplay, although in many studies only nitrogen balance has received attention. When patients receive IVN containing glucose as the source of energy most of the gain in weight that results is due to an accumulation of fat and water,13 changes in lean body mass being relatively insignificant. The repletion of the patient to original body mass may be a goal of IVN but is seldom realized, at least in the short-term, and may be unnecessary to achieve optimal clinical benefit. Although it is usual to place most emphasis upon these quantitatively small but physiologically important changes in lean body mass that are achieved by IVN, measurement of changes in body fat might provide valuable information concerning the efficiency and utilization of the nutritional

From the University Departments of Surgery and Medical Physics, The General Infirmary, Leeds, England. Address reprint requests to R. F. G. J. King, University Department of Surgery, The General Infirmary. Leeds, West Yorks, England. 01984 by Grune & Stratton, Inc. 0026-0495/84/3309-0008$03.00/0

826

substrates which are infused into the patient. Moreover, as changes in body fat are of considerably greater magnitude than changes of lean body mass, their measurement may enable more precise interpretation of the fate of infused nutrients. In this paper we outline the theoretical basis for calculating fat balance from measurement of carbon dioxide production and oxygen consumption, assuming an accurate knowledge of the provision of amino acids and energy substrates. These measurements are compared with data obtained from measurement of skinfold thickness, which is the more traditional method of assessment of change in body fat. MATERIALS

AND

METHODS

Patients The study was carried out with 13 patients who were under the care of the University Department of Surgery. IVN was provided for conventional clinical indications. Details of the patients are given in Table 1. None of the patients developed significant sepsis (sublingual temperature > 38 “C or white cell count > 15,000 . 109/L) or complications of IVN during this investigation. The study was carried out during the initial 14 days of IVN, none of the patients having received this therapy previously. The study was approved by the Research Ethics Committee of the General Infirmary at Leeds and informed consent was obtained from each patient. Intravenous

Nutrition

The regimen of parenteral nutrition utilized a mixture of hypertonic glucose, crystalline amino acids (Fre-amine II, The Boots Co, Nottingham, UK), minerals, and vitamins. In each patient the energy value of the feed provided was 1.5 times the measured resting energy expenditure (REE) at a ratio of kJ:g of nitrogen, of 657: 1. NO oral intake other than small amounts of water was permitted during the course of this study. The intravenous (IV) regime was dispensed in a 3-L bag and the infusion rate was regulated by pump at a steady rate of 2.1 mL/min. The bags were changed at 9.00 AM each day.

Metabolism, Vol 33, No 9 (September),

1984

827

BODY FAT CHANGE FROM RCI

Table 1. Patient Data Weight Patient

sex

Weight Loss

(kg)

(kg)

EnergyIntake (kJ/d-‘)

Diagnosis

5.2

9840

55.5

4.5

7995

38.1

11.9

7380

F

46.5

2.9

5535

F

45.1

6.9

11070

Pancreatic pseudocyst

30

F

48.8

2.1

11070

Pancreatic pseudocyst

73

F

68.0

5.0

12300

8.

53

F

50.0

10.0

8405

9.

55

M

54.0

14.5

8405

Gastric lymphoma, preoperative weight loss

10.

53

M

54.5

19.0

8405

Rectal neoplasm. Anastomotic leak

1.

78

M

60.8

2.

76

M

3.

66

F

4.

30

5.

81

6. 7.

Rectal neoplasm; postoperative vomiting Vesicocolic fistula; diverticular disease *Weight loss following ulcer surgery Delayed gastric emptying following ulcer surgery

Esophageal neoplasm, postoperative IVN Vomiting following ulcer surgery

11.

69

M

49.7

13.8

7688

lPYloric stenosis,

12.

67

F

42

12.8

9225

Ulcerative colitis

13.

67

M

53.4

10.0

7688

duodenal ulceration

lColonic neoplasm. preoperative weight loss

*Preoperative IVN

Actual intake was calculated by measuring the volume of nutritional solution which had not been administered. Although the energy and nitrogen were given continuously at the same rate each day throughout the 14 days of the study, electrolytes were tailored to the patient’s individual needs on a daily basis.

Nitrogen Balance Total body nitrogen (TBN) was measured by in vivo neutron activation analysis at the beginning and end of the f-week period of the study.14 Nitrogen balance was calculated from the difference between the two values.

Measurement

of Skinfold

Skinfold thickness was measured at the triceps, biceps, subscapular, and suprailiac sites on the nondominant side using Holtain calipers, (Holtain Ltd. Crymych, UK). Three measurements were made at each site and their mean values recorded. To determine the precision of the method ten patients were measured on three separate occasions of one day by one observer. The precision of the sum of the four-skinfold thicknesses was 0.73 mm.

Measurement

of Expired Air From Patients

A modification of the technique of indirect calorimetry described by Kappagoda” was used in this study (Fig 1). Patients were measured daily in a supine position in a specially adapted investiga-

Fig 1. Diagram of the instrumentation used. and the direction of gas flow in the measurement of carbon dioxide and oxygen exchanges in patients receiving IVN.

tion room adjoining the ward. On certain occasions it was only possible to measure the patients every two days but this was exceptional. All patients were acclimatized to the procedure and were allowed to rest for 15 minutes in the canopy before any determinations were made. Measurements were then made continuously during a 30-minute period when all expired gas was collected by the use of the system involving a ventilated canopy in which the subject was able to lie comfortably. The flexible canopy allowed considerable freedom for the patient, and was transparent yet permitted complete gas collection. A stream of room air was drawn through the canopy which was placed loosely over the head of the resting, supine patient. Not shown in Fig 1 are the inlet ports for air at the rear of the canopy and the loose towel used as a seal around the neck. In this way it was extremely unlikely that any exhaled air was lost to detection, because the canopy was evacuated into the analyzer at a flow rate of 150 L/min. The oxygen content of the mixed gas was measured continuously by a paramagnetic transducer (TaylorServomex OA137, Crowborough, Sussex, UK) and carbon dioxide by an infrared gas analyzer (PK Morgan, Chatham, Kent, UK). Output from both instruments was recorded on a dual-pen chart recorder. The instruments were calibrated daily using certified gas checked against our own gravimetric gas standard (Gold Star, British Oxygen Company, London, UK) at concentrations similar to those obtained from patients and at a flow rate of 150 L/min. Gas from the standard was introduced into the system in a manner analogous to a respiring patient but using a calibrated flow valve (Fisher-Porter, Workington, Cumberland, UK). Temperature and

828

KING ET AL

pressure were noted daily and all gas exchanges were corrected to 273 K at 760 mm pressure. The precision of repeated measurements of standard gas mixture when mixed into the main stream flow of I50 L/min was 23 mL/min for oxygen and 17 mL/min for carbon dioxide at flow rates experienced with patients.

C. Protein and Amino Acid Oxidation: Due to the variation in the oxidative mechanism and pathways for the individual amino acids, an average RQ of .80 was assumed. Seventy grams of protein produces 2.4 mol of CO, and consumes 3.0 mol O2 (RQ = .8)

Statistics Statistical comparisons for all determinations were made by the two-sample t test and in patients before and after IVN by the paired t test.16 THE CALCULATION OF FAT SYNTHESIS

FROM THE

RESPIRATORY QUOTIENT

We have taken the synthesis of tripalmitin (RMM = 806) as a standard, but other unique triglycerides will differ only slightly from this. The following four reactions were taken as essential for the de novo synthesis of tripalmitin from glucose. A. Production of Acetyl CoA:

The respiratory quotient (RQ) is

12 C6H,*06 + 24 ATP = 24 Acetyl CoA

CO2 produced

+ 24 CO, + 48 ATP + 48 NADH

Eq 1

O2 consumed

Eq 6

B. Production of NADPH:

In a given situation the RQ will be the ratio of all the carbon dioxide produced to all the oxygen consumed, and principally determined by glucose oxidation, protein turnover, and fat metabolism. It can be shown that other reactions involved in the exchanges of these gases are insignificant. Bicarbonate losses in urine and feces are ignored, as are oxygenase reactions such as hemoxygenase and reactions into “terminal” products. Then: RQ=

The Synthesis of Fat

CQk) + COZ(P) + CO*(fo) + CO,(fs) O,(g) + O*(P) + O*(fo) + O*(fs)

Eq2

3.5 G6P + 42 NADP + 24.5 HZ0 = 21 CO2 + 42 NADPH + 42 H+ + 3.5 Pi

Eq 7

C. Production of Palmitic Acid: 24 Acetyl CoA + 42 NADPH + 42 H+ + 21 ATP + 3 Hz0 = 3 Palmitic acid + 24 COA +42NADP+21ADP+21Pi

Eq8

D. Production of Glycerol: 0.5 Glucose + ATP = DHAP

CO,(g) and O,(g) represent the gas exchanges seen purely in the oxidation of glucose and are identical as shown in section A below (Equation 3). CO,(p) and O,(p) represent the gas exchanges seen purely in the oxidation of protein with the stoichiometry as shown in section C below. CO,(fo) and O,(fo) represent the gas exchanges seen purely in the oxidation of fat and are shown stoichiometrically in section B below (Equation 4). CO,(fs) and O,(fs) represent the gas exchanges seen purely in the synthesis of fat from glucose as outlined in the section entitled “The Synthesis of Fat” and shown by equations 6 through 12 below. Protein synthesis is not included in this equation because all the patients in this study oxidized almost all the amino acids supplied intravenously, maintaining their nitrogen balance without significant net protein synthesis. In this situation, therefore, there is net oxidation of amino acids provided in the feed. During the study the quantities of glucose and amino acids administered and the nitrogen balance were accurately known. The overall change in nitrogen was averaged out to a daily figure for the purposes of computation. Thus each individual’s daily gas exchange was expressed as a nonprotein respiratory quotient and from this the rates of lipogenesis were calculated using the following equations and their stoichiometries.

DHAP + NADH + H+ = Glycerol-3-phosphate + NAD

Eq 9

Thus, in the production of 1 mol of tripalmitin 45 mol of carbon dioxide is produced and 23.5 mol of oxygen consumed. (RQ = 1.91). Equation 2 can now be written in terms of moles of glucose oxidized (G), protein/amino acid oxidized (P), fat synthesized (FS), and fat oxidized (FO), ie: RQ=

CO,(G) + CO,(P) + CO,(FS) + CO,(FO) O,(G) + O,(P) + O,(FS) + O,(FO)

Because CO,(FS) = 45(FS) (each mole of fat synthesized produces 45 mol of carbon dioxide) and O,(FS) = 23.5(FS) (each mole of fat synthesized uses 23.5 mol of oxygen) and CO,(FO) = 48(FO) (each mole of fat oxidized produces 48 mol of carbon dioxide) and O,(FO) = 69.5(FO) (each mole of fat oxidized uses 69.5 mol of oxygen) then equation 2 can be written in two parts, (10 and 1 I) replacing CO,(FS), O,(FS), CO,(FO), and O,(FO) by their known reaction stoichiometries. Total CO1 produced = CO,(G) + CO*(P) + 45(S) + 48(O)

Eq IO

= O,(G) + O,(P) + 23.5(S) + 69.5(O)

Eq 11

Total O2 consumed A. Glucose oxidation: C,H,,06 + 6 0, = 6 CO, + 6 H,O

(RQ = 1.0)

Eq3

B. Fat oxidation: Due to the fatty acids of various chain length the RQ was taken as 0.702 although the oxidation of individual fatty acids have unique RQ values, eg, Palmitic acid + 23 O2 = 16 CO2 + 145 HZ0 (RQ = ,695)

Eq 4

Subtracting equation 1I from equation 10 and since CO,(G) = O,(G): CO,(T) - O,(T) = CO,(P) - O,(P) + 21.5(S) - 21.5(O) For purposes of clarity it is not necessary to substitute the reaction stoichiometry for glucose oxidation because these cancel out by subtraction as seen. After rearrangement substitutions for CO,(P) and O,(P) may be made using the stated reaction stoichiometry for protein or amino acid oxidation. In essence, the final equation

BODY FAT CHANGE FROM RQ

829

quantifiesthe changesin total fat storesthat may be seen given the measured carbon dioxide and oxygen exchanges corrected for nitrogen balance. This relationship is true for both fat synthesis and breakdown. On rearrangement: 21.5(S - 0) = CO,(T) - O,(T) - CO*(P) + O,(P) Because net moles of fat synthesized = (S - 0): net fat synthesis (moles) =

CO,(T) - W-U - CW’)

+ O,(P)

21.5

If this mole of fat is tripalmitin (RMM = 806) then grams of fat synthesized

= CO,(T) - OAT) - CO,(P) + O,(P) . 806 21.5 Since 1 g protein produces .0343 mol CO2 and consumes .0429 mol O1 then: g fat synthesized = (CO,(T) - O,(T) + (g protein oxidized . 37.488)) .0086

E?q12

RESULTS

The consumption of oxygen and the production of carbon dioxide (mean A 1SEM) for the patients during their course of IVN is shown in Fig 2. The results are presented as mL/min-‘W-.” ” and it can be seen that before the commencement of feeding that the production of carbon dioxide (9 + .5) was much lower than the consumption of oxygen (12.3 + .7) but that the introduction of feeding was associated with changes in both of these rates. During feeding the production of carbon dioxide increased steadily to 14.8 + 0.3 by day

8 and after this varied from 13.7 to 17. However, the consumption of oxygen, despite some variation, did not show such a large increase and the mean consumption by day 8 was 14.0 + 1.3. ‘Variation after this ranged from 12.0 to 16.3. The data in Fig 2 can be used to calculate the RQ. The data expressed in this way are shown in Fig 3, where it can be seen that the mean RQ was .76 at the commencement of feeding but increased steadily during feeding so that by day 3, it was greater than unity. After day 5, the mean RQ remained in the range 1.03 to 1.28. Fig 4 shows the mean calculated daily synthesis of fat for this group of patients during their course of feeding. Before feeding there was a net mean breakdown of 6.6 g fat (W-.‘*day-‘) but by day 3 this was reversed to a net fat balance. By day 4 fat synthesis had increased to 7 g and net synthesis was then maintained during the remaining period of the study. Although the mean rate of fat synthesis never fell below 2 g per day, there was considerable variation. The mean change in fat for the 13 patients was 874 + 243 g during the course of the 1Cday study. Individual patients ranged from - 6 19 g to 2487 g, but net synthesis was found in ten of them. The mean change in total body nitrogen was 26.3 t 144.8 g. Individual patients range from - 133 to + 337 g of nitrogen gained. CONCLUSIONS

The observation that RQ rises to values above unity during high glucose-infusion rates had been noted previously.“~‘2~‘8Although it has been concluded that fat synthesis must have occurred, no attempts have been made to quantify the gain in fat on a daily basis

16-

8L 0

I

2

3

4

5

6

1

8

9

10 II

I2

I3

W

IS

DAYS

Fig 2. Oxygen consumption and carbon dioxide production in 13 patients receiving IVN over a period of 15 days. Patients were supplied with nutrition at an energy level of 1.5 times their measured REE. Data points are daily means (+ / - 1 SEM).

Fig 3. Respiratory quotient in 13 Patients receiving IVN over a period of 15 days. Patients were supplied with feed at an energy level of 1.5 times their measured REE. Data points represent daily mean R Cl. I+ / - 1 SEW The lines on the graph at RQ = 1 and RQ = 0.7 represent the values for pure carbohydrate and pure fat oxidation respectively.

830

Fig. 4. Fat synthesis in thirteen patients receiving IVN over a period of 15 days. Patients were supplied with feed at an energy level of 1.5 times their measured REE. Data points represent the mean daily rate of fat synthesis (+ / - 1 SEMI calculated from RQ by the method described in this paper.

from changes in RQ. We were able to do this, using equation 12, because we knew the total intake of both energy and amino acids and because the net change in body nitrogen was measured. The method is suited to both ambulatory and recumbant patients who are fed intravenously. It could be adapted to subjects receiving oral nutrition but indirect calorimetry would have to be employed for longer periods in order to correct for changes in RQ that follow meals.” In the present study calorimetry was only used for short periods of time, but we have no reason to doubt that the instrument could be used continuously and patients could then be monitored overnight. The real value of indirect calorimetry is that it can be applied with moderate ease in the surgical ward and does not involve any undue discomfort for the patient. The data we have recorded show the potential of this instrument for the assessment of the fate of infused substrates in patients who are given IVN. Our daily measurements were taken in the resting state and thus give a measure of resting metabolic expenditure. The patients in this study were relatively inactive physically, and active metabolic expenditure would probably result in only a small increase of overall energy expenditure.20 Any exercise on the part of the patient would tend to reduce RQ and hence our computed fat balance would then be an overestimate. Conversely, if RQ rose at night (due to a more “basal state”) then our estimate of fat balance would have been too low. In

KING

ET AL

many respects the most worthwhile aspects of this technique may be the dynamic assessment of fat synthesis, ie, the potential to measure fat synthesis at different times during the infusion of substrate at a constant rate. Our data clearly show a change in RQ from an initial value of 0.8 to levels in excess of unity as feeding progressed thus we were able to calculate not only the net fat synthesis but also the rates of synthesis on individual days. Although we have chosen to measure these patients for only 30 minutes during each day, ten patients were measured on both the morning and afternoon of the same day. The mean difference (-t 1 SD) of 7.73 f 21 mL/min for oxygen consumption and 3.23 * 28.5 mL/min for carbon dioxide production was not significant (paired t test) compared to the day-to-day variations in the group. Despite this day-to-day variation, significant increases in carbon dioxide production were found over the initial ten days of the study. There was a small rise in oxygen consumption during fat synthesis. This was expected due to the specific dynamic action of fat (and protein) synthesis and constitutes the oxygen used for the energy cost of synthesising fat from glucose. The positive synthesis of fat in these patients was not unexpected because the mean daily energy input exceeded mean daily energy expenditure. It is very probable that the fat gains seen in our patients during the study (874 + 243 g) were in part a repletion of the patients’ body mass, however, it is unclear whether repletion of fat without repletion of the lean body is of any nutritional advantage. The method described in this paper enables a daily calculation of fat balance to be made. Comparison with fat change calculated from the measurements of skinfold (1.14 + 1.12 kg (P < .Ol)) showed a high degree of agreement between the two techniques. Whereas indirect calorimetry is probably ideal for a dynamic assessment of fat breakdown or synthesis but costly in terms of apparatus, skinfold measurement is cheap but only useful to measure changes which might be expected after 2 weeks or more of IVN. We feel that the method outlined in this paper is of potential value in the design of IV regimens. Clearly, the smaller the calorie supply, consistent with the provision of adequate nitrogen retention, the lower the probability of metabolic complications of IVN. Our calculations for the synthesis of fat are based on the equations given in the text of this paper. It has been well documented that the synthesis of fatty acids occurs in the cytosol of the adipocyte or other fatsynthesizing cells and this process uses NADPH as a source of energy for both of the enzyme-catalyzed reduction steps (/3 ketoacyl-acylcarrier protein reductase and enoyl-acylcarrier protein reductase). It is

BODY FAT CHANGE FROM RO

831

possible to approach the energetics of fat synthesis and assume that any surplus NADH derived from mitochondrial oxidation of pyruvate to produce acetyl CoA (equation 6) can be equated 1: 1 to produce NADPH in the cytoplasm. However, there is no basis on which such a stoichiometry can be founded. There have been two significant approaches which address this point albeit with a different intention. Woolfe et al’* have taken the basic equation and used only that stoichiometry shown in equation 13 13.5 C,H,,O, + 30, = CSsH,,06 + 26 CO2 + 29 HZ0

Eq 13

in their calculation, for the rate of fat synthesis in patients receiving short-term glucose infusions at high rates. In this work it was shown by use of infusions of 13Cglucose that 1.41 mg/kg/min of glucose administered at a rate of 9 mg/kg/min could be accounted for in net fat synthesis. This is no different in value to the mean rate between days 4 and 10 in the present study calculated solely from indirect calorimetry of 1.43 mg/kg/min. It would appear from this that approaches from either isotopes or calorimetry are comparable and that minor differences in the theoretical derivation of equations suitable for the calculation of fat synthesis, viz, the generation of NADPH, produce no real differences in the practical application of such methods on patients, thus strengthening the different individual approaches. The same point has been addressed by Elwyn et al4 although their main aim was to consider changes in nitrogen balance of patients

infused with glucose. Nevertheless, due allowance was made for NADPH in their calculations used solely for the effects of glucose on resting energy expenditure. Under the conditions of their study 22% of the starting energy was utilized as the cost of lipogenesis as measured by increases of oxygen consumption. RQ values greater than 1, which were observed at high glucoseinfusion rates, were taken to indicate net lipogenesis. Our own equations used to calculate fat synthesis from RQ have been set up so that the cost of NADPH synthesis is met de novo from glucose units. It may well be true that humans can utilize one half of the NADH produced mitochondrially, like the rat,” but we have no evidence for this. The error caused by this assumption is not great and would produce only a small change in the computed rates of fat synthesis. We have also taken the triglyceride synthesized to be tripalmitin because palmitic acid is the first product of fat synthesis. All other fatty acids are modifications of this key product but involve extramitochondrial or cytoplasmic reactions. If the major product of excess glucose utilization is not palmitic acid but one of the modifications of it, then the effect on fat balance would be only marginal because the COJO, differences this would involve are insignificant compared to the primary synthesis of palmitic acid. ACKNOWLEDGMENTS We are grateful to the Department of Medical Physics for measurements of total body nitrogen, and to Mrs Margaret Richardson for the preparation of the manuscript.

REFERENCES 1. Elwyn DH: Nutritional requirements of adult surgical patients. Crit Care Med 8:9-20, 1980 2. Collins JP, McCarthy ID, Hill GL: Assessment of protein nutrition in surgical patients-the value of anthropometrics. Am J Clin Nutr 32:1527-1530, 1979 3. Dudrick SJ, Long JM. Steiger E, et al: Intravenous hyperalimentation. Med Clin North Am 54577-589, 1970 4. Elwyn DH, Gump FE, Munro HN, et al: Changes in nitrogen balance of depleted patients with increasing infusions of glucose. Am J Clin Nutr 32:1597-1611, 1979 5. Smith RC, Burkinshaw L, Hill GL: Optimal energy and nitrogen intake for gastroenterological patients requiring IVN. Gastroenterology 82:445-452, 1982 6. MacFie J, Smith RC, Hill GL: Glucose or fat as a non protein energy source? Gastroenterology 80:103-107, 1981 7. Hegsted DM: Energy needs and energy utilisation. Nutr Rev 32:33-38,1974 8. Rutten P, Blackburn GL, Fiatt JP, et al: Determination of optimal hypexalimentation infusion rate. J Surg Res 18:477-483, 1975 9. Jeejeebhouy RN, Anderson GH, Nathooda AF, et al: Metabolic studies in total parenteral nutrition with lipid in man. J Clin Invest 57:125-136, 1976 10. King RFGJ, MacFie J, Hill GL, et al: Effect of intravenous

nutrition with glucose as the only calorie source on muscle glycogen. J Parenter Enteral Nut 5:226-229, 1981

11. Askanazi J, Carpentier YA, Elwyn DH, et al: Influence of total parenteral nutrition on fuel utilisation in injury and sepis. Ann Surg 19 1:4&46, 1980 12. Wolfe R, O’Donnell TF, Stone MD, et al: Investigation of factors determining the optimal glucose infusion rate in total parenteral nutrition. Metabolism 29:892-900, 1980 13. Hill GL, Bradley JA, Smith RC, et al: Changes in body weight and body protein with intravenous nutrition J Parenter Enteral Nut 3:215-218, 1978 14. Hill GL, King RFGJ. Smith RC, et al: Multi-element analysis of the living body by neutron activation analysis-application to critically ill patients receiving intravenous nutrition. Br J Surg 66:868-872, 1979 15. Kappagoda CT, Stoker JB, Linden RJ: A method for the continuous measurement of oxygen consumption. J Appl Physiol 31:60&607,1974 16. Snedecor GW, Cochran WG: Statistical Methods, (ed 6). Ames, Iowa, The Iowa State University Press, 1967, pp 91-l 19 17. Stock M, Rothwell R: Obesity and Leanness. London, John Libbey, 1982, pp 39940 18. Macfie J, Holmtield JH, King RFGJ: Effect of the energy

832

source on change in energy expenditure and respiratory quotient during total parenteral nutrition. J Panter Enteral Nut 7:1-5, 1982 19. Garrow JS: The regulation of energy expenditure in man, in Bray GA (ed): Recent Advances in Obesity Research. 1978, pp 204-206

KING ET AL

20. Stock M, Rothwell R: Obesity and Leanness. London, John Libbey, 1982, pp 43-44 21. Flatt JP: Conversion of carbohydrate to fat in adipose tissue. An energy yielding and, therefore, self limiting process. J Lipid Res 11:131-143,197o