- Email: [email protected]
Energy cost of fever in children on total parenteral nutrition
Clinical Nutrition (1997) 16:251-255 © Harcourt Brace & Co Ltd 1997
Energy cost of fever in children on total parenteral nutrition M. BENHARIZ, O. GOULET, J. SALAS, V. COLOMB, C. RICOUR Department of Pediatric Gastroenterology and Nutrition, Necker-Enfants Malades University Hospital, 149 rue de S#vres, 75743 Paris Cedex 15, France (Correspondence to OG) Abstract--The aim of the study was to measure the changes in energy expenditure (EE) and respiratory
quotient (RQ) induced by fever in children receiving total parenteral nutrition (TPN). Ten children aged 1-16 years (weight for height: 104.5 _ 13%) were included. They underwent two EE recordings of 3 h duration using indirect calorimetry, during a febrile (38.9 _ 0.5°C) and afebrile conditions, after parameters (nutritional intake, ambient temperature) being equal. The EEl (febrile phase) was significantly higher than the EE2 (afebrile) (2.13 __ 0.48 vs 1.78 _ 0.42 kcal/kg/h; P< 0.01). Increase in EE was significantly correlated with body temperature (r = 0.92, P< 0.001). The slope of the regression line indicates an increase of 16.2% in EE per degree of fever. By dividing the individual increase in EE by the individual increase in temperature, the energy expenditure during fever is 11.3% per unit rise (°C). During the febrile phase urinary nitrogen excretion was significantly higher (280 _+90 vs 210 _+70 mg/kg fat free mass/day, P< 0.02). The RQ during the febrile phase (0.90 _+ 0.13) did not differ significantly from mean RQ during the afebrile phase (0.95 +_ 0.07). Fever due to infection or inflammatory process is partly responsible for an hypermetabolic state including increased EE.
Introduction
tional intake during the study were also excluded. Informed consent was obtain from parents. Ten children (7 boys) aged 1 - 16 years were included in the study. All were on TPN, (infused over 12 h) for an average of 35 months (range 2-67) for small bowel resection (n = 5), intractable diarrhea (n = 4) and chronic intestinal pseudo-obstruction syndrome (n = 1). The nutritional intakes were stable for at least 1 week before entering the study (Table 1). The parenteral nutrition was carried out on the basis of 5-7 perfusions per week of a binary solution (Glucose, amino acids) and lipid as a piggy back. The mean non-protein energy intake was 57 kcal/kg/day (range 43-83). The mean rate of glucose infusion was 15.3 mg/kg/min (range 7.5-20.2). Only three children received a lipid infusion (Intralipid® 20% Pharmacia, Limoges, France) at the time of the study representing respectively 30, 38.5 and 43% of their total non-protein energy intake. The mean nitrogen intake (Vaminolact ® Pharmacia, Limoges, France) was 300 mg/kg/j (range 206-495).
Changes in body temperature lead to variations in energy expenditure (EE) (1-3). During an infectious episode, fever may increase EE and, if prolonged, be partly responsible for a negative energy balance. Few data on changes in EE related to modifications of body temperature are available in children (1, 2, 4, 5), whereas analysis of consequences of infection on EE gives controversial results (4-6). Meal induced thermogenesis represents a factor which should be taken into account for analysis of results (2, 7). Thus children on total parenteral nutrition (TPN), whose nutritional intake can be easily controlled, provide an interesting population in which to study the precise effect of fever on EE. To our knowledge, no studies have been performed in children receiving the same nutritional intake between febrile and afebrile phases. The aim of the current study was to assess the energy cost of fever. EE was measured by indirect calorimetry under a ventilated hood during the febrile phase, and compared to the EE measured a week later under the same environmental and nutritional conditions while the children were afebrile.
Measurements
Anthropometry Population and methods
Body weight and height were measured and referred to Sempe et al growth curves (8). The estimation of fat body mass was performed by the same examiner by measuring skinfold thickness at four different sites (biceps, triceps, subscapular and suprailiac) with a skin fold caliper (Holtain Ltd, Grymych, UK). The mean of three consecutive skinfold thickness measurements was recorded. The fat body mass (FBM) was calculated according to age by using the equations established by Brook or Dumin (9, 10). The fat
Subjects Hospitalized children on TPN, with fever defined by an axillary temperature of > 38°C, were eligible for the study. Children with cancer, inflammatory or respiratory diseases and children receiving corticosteroids, as well as those having hemodynamic disorders during the febrile phase were excluded. Patients with any modifications in nutri251
252 ENERGYEXPENDITUREINHYPERTHERMICCHILDREN Table 1 Clinicaldata and nutritionalintakes Patients
Sex
Age yrs, mths
Pathology PreviousTPN dmation (months)
Causeof fever
C-reactive Nonprotein Glucose Lipid Nitrogen protein energy infusionrate infusionrate intake mg/1 kcal/kg/day mg/kg/mn mg/kg/mn mg/kg/day
1 2 3 4 5 6 7 8 9 10
M M M F F M M M M F
1,0 2,2 2,11 3,5 4,0 5,6 5,8 8,6 12,10 16,11
SBS IDI SBS SBS SBS IDI IDI SBS SBS CIPOS
viral illness viral illness viral illness viral illness CRS chest infection CRS viral illness CRS viral illness
26 17 26 23 19 31 75 6 19 18
12 2 35 41 30 65 67 48 10 39
43 83 62 81 48 57 61 45 52 43
7.5 16.5 20.2 16.2 19.0 18.5 15.0 17.4 12.8 9.5
2.92 4.8 3.0
206 322 331 495 264 302 323 239 275 230
SBS: Short bowel syndrome;IDI: Intractablediarrhea of infancy;CIPOS:Chronicintestinalpseudo-obstructionsyndrome;TPN: Total parenteral nutrition;CRS: Catheterrelated sepsis. Table 2 Anthropometricdata Patients Weight1 (kg) 1
2 3 4 5 6 7 8 9 10
9.3 8.9 11.6 14.5 14.5 21.2 19.8 20.1 29.1 48.8
FFM 1 (kg)
Height (cm)
W/H (%)
FBM/W Weight2 (%) (kg)
FFM 2 (kg)
7.8 7.4 10 12.8 13 17.4 16.7 18.2 24 37.5
72 75 89 87 98 104 105 117 134 153
100 90 90 121 97 125 116 95 98 113
16.5 17.5 13.7 12 10 18 16 9.3 17.4 23.2
7.7 7.3 10 13 13.2 17 16.3 18 24 37.2
9.1 8.7 11.8 14.4 14.5 21 19.8 20.6 29.1 48.7
FBM: Fat Body Mass; Weight 1: Weightat phase 1; FFMI: Fat free mass at phase 1; Weight 2: Weightat phase 2; FFM2: Fat free mass at phase 2.
free mass (FFM) was calculated as the difference between the body weight and the FBM (Table 2).
Experimental design Each child was studied successively during the febrile phase (phase 1) and after the fever had resolved (phase 2). Axillary temperature was measured using mercury thermometer in all children. Fever was defined as an axillary temperature higher than 38°C measured within 60 min of time, Onset of fever occurred during parenteral nutrition infusion in seven children and during the non-infusion period of cyclic TPN in three children. For the seven children in whom fever occurred during infusion, the indirect calorimetry recording started 4 - 8 h after the beginning of infusion. For the other three children, fever started 2 - 4 h after the end of the infusion period and the indirect calorimetry recording started 2 - 4 h later.
Methods Oxygen consumption (go2) and carbon dioxide production (VCO2) were measured by using indirect calorimetry (MMC Horizon, Beckman ®, Sensor Medics ®, Anaheim, CA, USA) with a ventilated hood system. Calibration was performed before each test, once with a mixture of 02 and CO2 of known concentration and secondly with pure nitrogen. In addition, quality control of the apparatus was
achieved by infusion of pure nitrogen, simulating 02 consumption and combustion of ethyl alcohol at 99.7% _ 0.2 (Carlo Erba®-Milan) as the control of respiratory quotient. Each indirect calorimetry recording was performed over 3 h in a relatively draught-free room with a stable environmental temperature of 23-24°C. Clothing was reduced and an acrylic leakproof hood was placed over the child's head. Expired air flow was adapted to maintain the CO2 level in the calorimeter around 0.7%. In 98% of cases, the recording was carried out during sleep. Only the sleeping periods were taken into account for the calculations. The axillary temperature was taken every half hour during the indirect calorimetry recording. In eight children 24 h urine collections were performed from the beginning of indirect calorimetry for urinary nitrogen excretion on the same day using the Kjeldahl method. This parameter was then used to estimate the nonprotein respiratory quotient (npRQ). Mean energy expenditure (EE) was calculated from the repeated measurements of VO2 and VCO2 during the 3 h recording and on the basis of the Weir equation (11). Substrate oxidation rate were calculated using published methods (12, 13). Constants used for this calculation were as follows: 1 g urinary nitrogen is produced by oxidation of 6.25 g protein; oxidation of 1 g protein consumes 966 ml 02 and produces 782 ml CO2; while oxidation requires 746 ml oxygen per gram of glucose (RQ -- 1) and 2.019 ml oxygen per gram of fat (RQ = 0.696).
CLINICAL NUTRITION
A repeat measurement (test 2) was carried out when the child was afebrile after a mean interval of 8 days after the first recording (range 6-14 days). The same relationships to parenteral nutrition were maintained from the first to the second phase in each child. Recording was performed in the same room under the same environmental conditions (room temperature, clothing) and included the same measurements: indirect calorimetry and 2 4 h urinary nitrogen excretion.
253
2000
A
1000 0
v
Statistical analysis The results are expressed as mean + standard deviation (SD). Comparisons between the two phases were made using Student' s paired t test. Correlations between variables were calculated using linear regression analysis and expressed as the correlation coefficient r. The level of significance was set at P < 0.05.
o 1
2
3
4
5
6
7
8
9
10
Subjects Fig. 1 Individual energy expenditure values at the two phases of the study: phase h Febrile phase; phase 2: Afebrile phase.
Results
respectively in phases 1 and 2 (P < 0.01). EE during phase 1 was, for each child, significantly higher than that measured during phase 2 (Fig. 1 and Table 3). The mean percentage of increase in EE was 20.8 _+ 11%. Expressed as a function of body weight, mean EE increase between febrile and afebrile phases was 20.4 _ 12% and 20.1 _+ 11% expressed as a function of FFM. The variable rise in EE during the febrile phase was not correlated with the duration of fever before starting the study. A positive correlation was found (r--0.92, P < 0 . 0 0 1 ) between the relative EE increase (EEI-EE2)/EE2) and the temperature variation (AT °) between the two phases (Fig. 2). The slope of regression line (16.2%) represents the estimated rise of EE by degree of fever. The intercept of the slope of the regression line on the AT ° axis is about 0.5°C. Another way to assess the EE increase related to fever would be to calculate, for each child, the increase of EE by degree of fever (AEE/AT °) between the two phases. This approach provides an average rise of EE of 11.3 _+ 2.7% per degree of fever.
Complete recordings were made successfully in the 10 children. Fever was related to viral illness (n -- 6), catheter related sepsis (n = 3), or chest infection (n = 1). During the febrile phase (phase 1), axillary temperature measured every 30 min during the indirect calorimetry recording was 38.9 _+ 0.5°C (Table 3). The axillary temperature was stable during the recording (average coefficient of variation: 0.3%) and significantly greater than during the afebrile phase (37.0 +_ 0.3°C), the mean difference being + 1.98°C (P < 0.0001). There was no significant difference between the febrile and afebrile phases for weight, FBM and FFM (Table 2).
EE, RQ and nutrient utilization EE was positively correlated with body weight (r = 0.86, P < 0.001 and r = 0.93, P < 0.001 respectively in phases 1 and 2) and with FFM (r = 0.90, P < 0.001 and r = 0.93, P < 0.001 respectively in phases 1 and 2). Mean EE was 2.13kcal + 0.48/kg/hour and 1.78kcal + 0.42/kg/hour
Table 3 Body temperature (T), energy expenditure (EE), respiratory quotient (RQ), net oxidation rate (NOR) of glucose and lipid, and urinary nitrogen (UN) during febrile (1) and afebrile phase (2) Patient
T1 mean _+SD (°C)
T2 mean (°C)
EEl kcal/kgFFM/h (mean +_SD)
EE2 kcal/kgFFM/h (mean + SD)
RQ1
RQ2
Glucose NOR1 mg/kg/h
1 2 3 4 5 6 7 8 9 10
39.7 39.2 38.5 38.8 39.0 38.0 38.6 38.5 39.0 39.4
36.5 37.4 36.5 37.5 37.5 37.0 37.0 37.2 37.1 37.2
2.88 2.08 2.98 3.0 2.51 2.43 2.33 3.09 2.03 1.70
2.00 1.77 2.37 2.52 2.11 2.24 2.06 2.80 1.79 1.31
0.82 0.81 0.71 0.95 1.06 0.78 1.01 1.14 0.95 0.82
0.83 1.00 0.89 0.95 1.07 1.00 0.93 1.01 0.98 0.89
. 135 -17 . 623 122 466 918 327 125
+ 0.1 -+ 0.1 -+ 0.2 + 0.2 -+ 0.2 -+ 0.1 -+ 0.1 + 0.1 + 0.1 _+0.1
+ 0.29 + 0.10 + 0.29 + 0.18 -+ 0.10 -+ 0.20 _+0.24 -+ 0.19 -+ 0.09 _+0.18
*Difference between EEl and EE2: P < 0.01.
+ 0.34* + 0.09* -+ 0.29* + 0.16" -+ 0.17" - 0.13" _+0.11" -+ 0.25* -+ 0.15" +- 0.16"
Glucose NOR2 mg/kg/h .
Lipid NOR1 mg/kg/h .
345 289 .
. 557 444 283 584 313 147
Lipid NOR2 mg/kg/h
. 89 247 . -68 138 -23 -61 -4 69
Daily UN1 mg/kgFFM
Daily UN2 mg/kgFFM
340 260
130 260
280 210 230 410 400 150
150 240 200 300 280 100
. -9 56 . 57 -18 23 -30 -11 31
.
254
ENERGY EXPENDITURE IN HYPERTHERMIC CHILDREN
/
50-
40 o~
.~
r
= 0.92
P
< 0.01
3o
m
[]
El
[]
El
[]
[]
c, N
10'
0
i 1
De l ta
2
temperature
3
4
(°C)
Fig. 2 Correlation between the difference in temperature between p h a s e 1 and 2 (DT ° = T °] - T°2) and the relative increase in EE (DEE = EEI -
EE2/EE~). The mean respiratory quotients (RQ) in phase 1 (0.90 -+ 0.13) and phase 2 (0.95 + 0.07) were not different. Mean nonprotein RQ calculated for eight children did not differ significantly between phase 1 and 2 being respectively 0.95 _+0.19 and 1.00 _+0.07 (Table 3). Twenty four hours urinary nitrogen excretion measured in eight children was significantly higher during the febrile phase: 280 + 90 mg/kg/FFM/24 h versus 210 _+ 70 mg/kg/ FFM/24 h (P < 0.02) (Table 3). Net glucose oxidation calculated for eight children was not different between the two tests (337 _+ l l 4 m g / k g / h versus 370 _+ 148 mg/kg/h) (Table 3).
Discussion This study, carried out in children on TPN, demonstrates a linear relationship between changes in body temperature and energy expenditure as measured by indirect calorimetry. The children studied had a normal nutritional status at the time of the study, weight for height being greater than 90%. The short period of time between the two parts of the study allowed a stable nutritional status during the whole study period to be maintained. Thus the variations of EE were not related to modifications of the body composition (2, 14). The total parenteral protein-energy intakes during the study corresponded to the requirements according to the different ages of the children. Standardization of nutrient intake according to a strict protocol in individual children obviated the need for controls in the study. Thus, the children were studied as their own controls. The constant level of fever during the indirect calorimetry measurements allowed a relative stability of these measurements, and a longer period of recording than those performed during other studies in children (4-6). For all children, REE recorded during the febrile phase
was higher than that in the afebrile phase (20.8 _+ 11% on average). The increase of EE between phase 1 and 2 is lower than that observed by Stettler et al in children with malaria (29.5 _+ 2.6%) (4). This difference could be explained by the lesser degree of severity of the infection in our children. The variable amplitude of the rise in EE cannot be explained by a different duration of fever before the study (no correlation), nor by a difference in physical activity at the time of the study (recording during sleep). In this study, the rise of EE and that of the axillary temperature are correlated. The slope of regression line (16%/1°C) represents the fraction of the rise in EE related to fever. In the Stettler et al study (4) this fraction is equal to 6.9%, while it is 13% in the Dubois study (1). This difference could be due to the different populations in each study. The small number of children in our study might also account for this difference. The intercept on the AT ° axis (for AEE = 0) should be considered as the minimal variation level of the temperature from which we detect a difference in EE. It is equal to approximately 0.5°C. The rise in EE is related to fever, and to other phenomena such as rise in protein catabolism as suggested by the urinary nitrogen loss (15, 16). By dividing the individual rise of EE by the individual rise of temperature, EE increase is equal to 11.3% for I°C of temperature. This result is close to those reported by Dubois et al (1) and Stettler et al (4) but differs from those of a recent study performed in sick infants less than one year of age (5). The average RQ during the two phases were no different. However six out of the 10 children studied had a lower RQ during the febrile phase. In some studies carried out in children or adults during an infectious episode, the RQ measured in the acute period was lower than that measured during recovery period (2, 4, 17, 18). This decrease in RQ during the febrile phase might be explained by a rise in lipid oxidation. However, in three cases the RQ was either identical or lower during the afebrile phase. Moreover, as suggested by the absence of difference in RQ, net glucose and lipid oxidation were not different between phases 1 and 2. The heterogeneity of the studied population, in terms of nutritional intakes (seven fed and three fasting), and the relatively moderate infectious state observed in this study might account for these results in RQ and net substrate oxidation. On the basis of 24 h urinary-nitrogen excretion, protein catabolism rose significantly during the febrile phase (P < 0.02). These results are in agreement with other studies using identical or different methodologies (17, 20). More recently Berclaz et al have demonstrated a substantial increase (by a factor of two) in whole body protein turnover in Gambian children suffering from an acute episode of malaria, by using [tSN] glycine administration (21). In conclusion, in this pediatric population parenterally fed, fever is responsible for an EE rise of 16% per degree of fever. Protracted fever due to infection or inflammatory process is partly responsible for an hypermetabolic state including increased energy expenditure and protein breakdown.
CLINICAL NUTRITION 255
References 1. Du Bois E F. The basal metabolism in fever. JAMA 1921; 77:352-355 2. Eccles M P, Cole T J, Whitehead R G. Factors influencing sleeping metabolic rate in infants. Eur J Clin Nutr 1989; 43: 485-492. 3. Kluger M J. Body temperature changes during inflammation: their mediation and nutritional significance. Proc Nutr Soc 1989; 48:337-345 4. Stettler N, Shutz Y, Whitehead R, Jequier E. Effect of malaria and fever on energy metabolism in gambian children. Pediatr Res 1992; 31: 102-106. 5. Mclntyre J, Hull D. Metabolic rate in febrile infants. Arch Dis Child 1996; 74:206-209 6. Duggan M B, Alwar J, Milner R D G. The nutritional cost of measles in Africa. Arch Dis Child 1986; 61:61-66 7. James W P T. From SDA to DIT to TEF. In: Kinney J M, Tucker H N, (eds). Energy metabolism. Tissue determinants and cellular corollaries. New York: Raven Press 1992: pp 163-183 8. Sempe M, Pedron G, Roy-Pernot M P. Auxologie: m6thodes et sequences. Edition Theraplix Paris, 1979: pp 138-162 9. Brook C G. Determination of body composition of children from skinfold measurements. Arch Dis Child 1971; 46:182-184 10. Dumin J V, Rahaman M M. The assessment of amount of fat in the human body from measurements of skinfold thickness. Br J Nutr 1967; 21:681-689 11. Weir J B. New methods for calculating metabolic rate with special reference to protein metabolism. J Physiol 1949; 1 0 9 : i - 9 12. Jequier E. Measurement of energy expenditure in clinical nutritional
Submission date: 26 May 1997 Accepted: 29 July 1997
assessment. J Parenter Enteral Nutr 1987; 11: 86S-89S. 13. Ferrannini E. The theoritical bases of indirect calorimetry: a review. Metabolism 1988; 37:287-301 14. Hill J R, Rahimtula K A. Heat balance and the metabolic rate of newborn babies in relation to environmental temperature, and the effect of age and weight on basal metabolic rate. J Physiol 1965; 180:239-265 15. Long C L, Jeevanandam M, Kim B M, Kinney M. Whole body protein synthesis and catabolism in septic man. Am J Clin Nntr 1977; 30:1340-1344 16. Beisel W R. Metabolic response to infection. Ann Rev Med 1975; 26:9-20 17. Askanazi J, Carpentier Y A, Elwyn D H et al. Influence of total parenteral nutrition on fuel utilization in injury and sepsis. Ann Surg 1980; 191:40-46 18. Stoner H B, Little R A, Frayn K N, Elebute A E, Tresadern J, Gross E. The effect of sepsis on oxidation of carbohydrate and fat. Br J Surg 1983; 70:32-35 19. Long C L, Schiller W R, Blakmore W S, Geiger J W, O'Dell M, Henderson K. Muscle protein catabolism in the septic patient as measured by 3-methylhistidine excretion. A m J Clin Nutr 1977; 30:1349-1352 20. Tomkins A M, Garlick P J, Schofield W N, Wateflow J C. The combined effects of infection and malnutrition on protein metabolism in children. Clin Sci 1983; 65:313-324. 21. Berclaz P Y, Benedek C, J6qnier E, Schutz Y. Changes in protein turnover and resting energy expenditure after treatment of malaria in Gambian children. Pediatr Res 1996; 39: 401-409.
Energy cost of fever in children on total parenteral nutrition M. BENHARIZ, O. GOULET, J. SALAS, V. COLOMB, C. RICOUR Department of Pediatric Gastroenterology and Nutrition, Necker-Enfants Malades University Hospital, 149 rue de S#vres, 75743 Paris Cedex 15, France (Correspondence to OG) Abstract--The aim of the study was to measure the changes in energy expenditure (EE) and respiratory
quotient (RQ) induced by fever in children receiving total parenteral nutrition (TPN). Ten children aged 1-16 years (weight for height: 104.5 _ 13%) were included. They underwent two EE recordings of 3 h duration using indirect calorimetry, during a febrile (38.9 _ 0.5°C) and afebrile conditions, after parameters (nutritional intake, ambient temperature) being equal. The EEl (febrile phase) was significantly higher than the EE2 (afebrile) (2.13 __ 0.48 vs 1.78 _ 0.42 kcal/kg/h; P< 0.01). Increase in EE was significantly correlated with body temperature (r = 0.92, P< 0.001). The slope of the regression line indicates an increase of 16.2% in EE per degree of fever. By dividing the individual increase in EE by the individual increase in temperature, the energy expenditure during fever is 11.3% per unit rise (°C). During the febrile phase urinary nitrogen excretion was significantly higher (280 _+90 vs 210 _+70 mg/kg fat free mass/day, P< 0.02). The RQ during the febrile phase (0.90 _+ 0.13) did not differ significantly from mean RQ during the afebrile phase (0.95 +_ 0.07). Fever due to infection or inflammatory process is partly responsible for an hypermetabolic state including increased EE.
Introduction
tional intake during the study were also excluded. Informed consent was obtain from parents. Ten children (7 boys) aged 1 - 16 years were included in the study. All were on TPN, (infused over 12 h) for an average of 35 months (range 2-67) for small bowel resection (n = 5), intractable diarrhea (n = 4) and chronic intestinal pseudo-obstruction syndrome (n = 1). The nutritional intakes were stable for at least 1 week before entering the study (Table 1). The parenteral nutrition was carried out on the basis of 5-7 perfusions per week of a binary solution (Glucose, amino acids) and lipid as a piggy back. The mean non-protein energy intake was 57 kcal/kg/day (range 43-83). The mean rate of glucose infusion was 15.3 mg/kg/min (range 7.5-20.2). Only three children received a lipid infusion (Intralipid® 20% Pharmacia, Limoges, France) at the time of the study representing respectively 30, 38.5 and 43% of their total non-protein energy intake. The mean nitrogen intake (Vaminolact ® Pharmacia, Limoges, France) was 300 mg/kg/j (range 206-495).
Changes in body temperature lead to variations in energy expenditure (EE) (1-3). During an infectious episode, fever may increase EE and, if prolonged, be partly responsible for a negative energy balance. Few data on changes in EE related to modifications of body temperature are available in children (1, 2, 4, 5), whereas analysis of consequences of infection on EE gives controversial results (4-6). Meal induced thermogenesis represents a factor which should be taken into account for analysis of results (2, 7). Thus children on total parenteral nutrition (TPN), whose nutritional intake can be easily controlled, provide an interesting population in which to study the precise effect of fever on EE. To our knowledge, no studies have been performed in children receiving the same nutritional intake between febrile and afebrile phases. The aim of the current study was to assess the energy cost of fever. EE was measured by indirect calorimetry under a ventilated hood during the febrile phase, and compared to the EE measured a week later under the same environmental and nutritional conditions while the children were afebrile.
Measurements
Anthropometry Population and methods
Body weight and height were measured and referred to Sempe et al growth curves (8). The estimation of fat body mass was performed by the same examiner by measuring skinfold thickness at four different sites (biceps, triceps, subscapular and suprailiac) with a skin fold caliper (Holtain Ltd, Grymych, UK). The mean of three consecutive skinfold thickness measurements was recorded. The fat body mass (FBM) was calculated according to age by using the equations established by Brook or Dumin (9, 10). The fat
Subjects Hospitalized children on TPN, with fever defined by an axillary temperature of > 38°C, were eligible for the study. Children with cancer, inflammatory or respiratory diseases and children receiving corticosteroids, as well as those having hemodynamic disorders during the febrile phase were excluded. Patients with any modifications in nutri251
252 ENERGYEXPENDITUREINHYPERTHERMICCHILDREN Table 1 Clinicaldata and nutritionalintakes Patients
Sex
Age yrs, mths
Pathology PreviousTPN dmation (months)
Causeof fever
C-reactive Nonprotein Glucose Lipid Nitrogen protein energy infusionrate infusionrate intake mg/1 kcal/kg/day mg/kg/mn mg/kg/mn mg/kg/day
1 2 3 4 5 6 7 8 9 10
M M M F F M M M M F
1,0 2,2 2,11 3,5 4,0 5,6 5,8 8,6 12,10 16,11
SBS IDI SBS SBS SBS IDI IDI SBS SBS CIPOS
viral illness viral illness viral illness viral illness CRS chest infection CRS viral illness CRS viral illness
26 17 26 23 19 31 75 6 19 18
12 2 35 41 30 65 67 48 10 39
43 83 62 81 48 57 61 45 52 43
7.5 16.5 20.2 16.2 19.0 18.5 15.0 17.4 12.8 9.5
2.92 4.8 3.0
206 322 331 495 264 302 323 239 275 230
SBS: Short bowel syndrome;IDI: Intractablediarrhea of infancy;CIPOS:Chronicintestinalpseudo-obstructionsyndrome;TPN: Total parenteral nutrition;CRS: Catheterrelated sepsis. Table 2 Anthropometricdata Patients Weight1 (kg) 1
2 3 4 5 6 7 8 9 10
9.3 8.9 11.6 14.5 14.5 21.2 19.8 20.1 29.1 48.8
FFM 1 (kg)
Height (cm)
W/H (%)
FBM/W Weight2 (%) (kg)
FFM 2 (kg)
7.8 7.4 10 12.8 13 17.4 16.7 18.2 24 37.5
72 75 89 87 98 104 105 117 134 153
100 90 90 121 97 125 116 95 98 113
16.5 17.5 13.7 12 10 18 16 9.3 17.4 23.2
7.7 7.3 10 13 13.2 17 16.3 18 24 37.2
9.1 8.7 11.8 14.4 14.5 21 19.8 20.6 29.1 48.7
FBM: Fat Body Mass; Weight 1: Weightat phase 1; FFMI: Fat free mass at phase 1; Weight 2: Weightat phase 2; FFM2: Fat free mass at phase 2.
free mass (FFM) was calculated as the difference between the body weight and the FBM (Table 2).
Experimental design Each child was studied successively during the febrile phase (phase 1) and after the fever had resolved (phase 2). Axillary temperature was measured using mercury thermometer in all children. Fever was defined as an axillary temperature higher than 38°C measured within 60 min of time, Onset of fever occurred during parenteral nutrition infusion in seven children and during the non-infusion period of cyclic TPN in three children. For the seven children in whom fever occurred during infusion, the indirect calorimetry recording started 4 - 8 h after the beginning of infusion. For the other three children, fever started 2 - 4 h after the end of the infusion period and the indirect calorimetry recording started 2 - 4 h later.
Methods Oxygen consumption (go2) and carbon dioxide production (VCO2) were measured by using indirect calorimetry (MMC Horizon, Beckman ®, Sensor Medics ®, Anaheim, CA, USA) with a ventilated hood system. Calibration was performed before each test, once with a mixture of 02 and CO2 of known concentration and secondly with pure nitrogen. In addition, quality control of the apparatus was
achieved by infusion of pure nitrogen, simulating 02 consumption and combustion of ethyl alcohol at 99.7% _ 0.2 (Carlo Erba®-Milan) as the control of respiratory quotient. Each indirect calorimetry recording was performed over 3 h in a relatively draught-free room with a stable environmental temperature of 23-24°C. Clothing was reduced and an acrylic leakproof hood was placed over the child's head. Expired air flow was adapted to maintain the CO2 level in the calorimeter around 0.7%. In 98% of cases, the recording was carried out during sleep. Only the sleeping periods were taken into account for the calculations. The axillary temperature was taken every half hour during the indirect calorimetry recording. In eight children 24 h urine collections were performed from the beginning of indirect calorimetry for urinary nitrogen excretion on the same day using the Kjeldahl method. This parameter was then used to estimate the nonprotein respiratory quotient (npRQ). Mean energy expenditure (EE) was calculated from the repeated measurements of VO2 and VCO2 during the 3 h recording and on the basis of the Weir equation (11). Substrate oxidation rate were calculated using published methods (12, 13). Constants used for this calculation were as follows: 1 g urinary nitrogen is produced by oxidation of 6.25 g protein; oxidation of 1 g protein consumes 966 ml 02 and produces 782 ml CO2; while oxidation requires 746 ml oxygen per gram of glucose (RQ -- 1) and 2.019 ml oxygen per gram of fat (RQ = 0.696).
CLINICAL NUTRITION
A repeat measurement (test 2) was carried out when the child was afebrile after a mean interval of 8 days after the first recording (range 6-14 days). The same relationships to parenteral nutrition were maintained from the first to the second phase in each child. Recording was performed in the same room under the same environmental conditions (room temperature, clothing) and included the same measurements: indirect calorimetry and 2 4 h urinary nitrogen excretion.
253
2000
A
1000 0
v
Statistical analysis The results are expressed as mean + standard deviation (SD). Comparisons between the two phases were made using Student' s paired t test. Correlations between variables were calculated using linear regression analysis and expressed as the correlation coefficient r. The level of significance was set at P < 0.05.
o 1
2
3
4
5
6
7
8
9
10
Subjects Fig. 1 Individual energy expenditure values at the two phases of the study: phase h Febrile phase; phase 2: Afebrile phase.
Results
respectively in phases 1 and 2 (P < 0.01). EE during phase 1 was, for each child, significantly higher than that measured during phase 2 (Fig. 1 and Table 3). The mean percentage of increase in EE was 20.8 _+ 11%. Expressed as a function of body weight, mean EE increase between febrile and afebrile phases was 20.4 _ 12% and 20.1 _+ 11% expressed as a function of FFM. The variable rise in EE during the febrile phase was not correlated with the duration of fever before starting the study. A positive correlation was found (r--0.92, P < 0 . 0 0 1 ) between the relative EE increase (EEI-EE2)/EE2) and the temperature variation (AT °) between the two phases (Fig. 2). The slope of regression line (16.2%) represents the estimated rise of EE by degree of fever. The intercept of the slope of the regression line on the AT ° axis is about 0.5°C. Another way to assess the EE increase related to fever would be to calculate, for each child, the increase of EE by degree of fever (AEE/AT °) between the two phases. This approach provides an average rise of EE of 11.3 _+ 2.7% per degree of fever.
Complete recordings were made successfully in the 10 children. Fever was related to viral illness (n -- 6), catheter related sepsis (n = 3), or chest infection (n = 1). During the febrile phase (phase 1), axillary temperature measured every 30 min during the indirect calorimetry recording was 38.9 _+ 0.5°C (Table 3). The axillary temperature was stable during the recording (average coefficient of variation: 0.3%) and significantly greater than during the afebrile phase (37.0 +_ 0.3°C), the mean difference being + 1.98°C (P < 0.0001). There was no significant difference between the febrile and afebrile phases for weight, FBM and FFM (Table 2).
EE, RQ and nutrient utilization EE was positively correlated with body weight (r = 0.86, P < 0.001 and r = 0.93, P < 0.001 respectively in phases 1 and 2) and with FFM (r = 0.90, P < 0.001 and r = 0.93, P < 0.001 respectively in phases 1 and 2). Mean EE was 2.13kcal + 0.48/kg/hour and 1.78kcal + 0.42/kg/hour
Table 3 Body temperature (T), energy expenditure (EE), respiratory quotient (RQ), net oxidation rate (NOR) of glucose and lipid, and urinary nitrogen (UN) during febrile (1) and afebrile phase (2) Patient
T1 mean _+SD (°C)
T2 mean (°C)
EEl kcal/kgFFM/h (mean +_SD)
EE2 kcal/kgFFM/h (mean + SD)
RQ1
RQ2
Glucose NOR1 mg/kg/h
1 2 3 4 5 6 7 8 9 10
39.7 39.2 38.5 38.8 39.0 38.0 38.6 38.5 39.0 39.4
36.5 37.4 36.5 37.5 37.5 37.0 37.0 37.2 37.1 37.2
2.88 2.08 2.98 3.0 2.51 2.43 2.33 3.09 2.03 1.70
2.00 1.77 2.37 2.52 2.11 2.24 2.06 2.80 1.79 1.31
0.82 0.81 0.71 0.95 1.06 0.78 1.01 1.14 0.95 0.82
0.83 1.00 0.89 0.95 1.07 1.00 0.93 1.01 0.98 0.89
. 135 -17 . 623 122 466 918 327 125
+ 0.1 -+ 0.1 -+ 0.2 + 0.2 -+ 0.2 -+ 0.1 -+ 0.1 + 0.1 + 0.1 _+0.1
+ 0.29 + 0.10 + 0.29 + 0.18 -+ 0.10 -+ 0.20 _+0.24 -+ 0.19 -+ 0.09 _+0.18
*Difference between EEl and EE2: P < 0.01.
+ 0.34* + 0.09* -+ 0.29* + 0.16" -+ 0.17" - 0.13" _+0.11" -+ 0.25* -+ 0.15" +- 0.16"
Glucose NOR2 mg/kg/h .
Lipid NOR1 mg/kg/h .
345 289 .
. 557 444 283 584 313 147
Lipid NOR2 mg/kg/h
. 89 247 . -68 138 -23 -61 -4 69
Daily UN1 mg/kgFFM
Daily UN2 mg/kgFFM
340 260
130 260
280 210 230 410 400 150
150 240 200 300 280 100
. -9 56 . 57 -18 23 -30 -11 31
.
254
ENERGY EXPENDITURE IN HYPERTHERMIC CHILDREN
/
50-
40 o~
.~
r
= 0.92
P
< 0.01
3o
m
[]
El
[]
El
[]
[]
c, N
10'
0
i 1
De l ta
2
temperature
3
4
(°C)
Fig. 2 Correlation between the difference in temperature between p h a s e 1 and 2 (DT ° = T °] - T°2) and the relative increase in EE (DEE = EEI -
EE2/EE~). The mean respiratory quotients (RQ) in phase 1 (0.90 -+ 0.13) and phase 2 (0.95 + 0.07) were not different. Mean nonprotein RQ calculated for eight children did not differ significantly between phase 1 and 2 being respectively 0.95 _+0.19 and 1.00 _+0.07 (Table 3). Twenty four hours urinary nitrogen excretion measured in eight children was significantly higher during the febrile phase: 280 + 90 mg/kg/FFM/24 h versus 210 _+ 70 mg/kg/ FFM/24 h (P < 0.02) (Table 3). Net glucose oxidation calculated for eight children was not different between the two tests (337 _+ l l 4 m g / k g / h versus 370 _+ 148 mg/kg/h) (Table 3).
Discussion This study, carried out in children on TPN, demonstrates a linear relationship between changes in body temperature and energy expenditure as measured by indirect calorimetry. The children studied had a normal nutritional status at the time of the study, weight for height being greater than 90%. The short period of time between the two parts of the study allowed a stable nutritional status during the whole study period to be maintained. Thus the variations of EE were not related to modifications of the body composition (2, 14). The total parenteral protein-energy intakes during the study corresponded to the requirements according to the different ages of the children. Standardization of nutrient intake according to a strict protocol in individual children obviated the need for controls in the study. Thus, the children were studied as their own controls. The constant level of fever during the indirect calorimetry measurements allowed a relative stability of these measurements, and a longer period of recording than those performed during other studies in children (4-6). For all children, REE recorded during the febrile phase
was higher than that in the afebrile phase (20.8 _+ 11% on average). The increase of EE between phase 1 and 2 is lower than that observed by Stettler et al in children with malaria (29.5 _+ 2.6%) (4). This difference could be explained by the lesser degree of severity of the infection in our children. The variable amplitude of the rise in EE cannot be explained by a different duration of fever before the study (no correlation), nor by a difference in physical activity at the time of the study (recording during sleep). In this study, the rise of EE and that of the axillary temperature are correlated. The slope of regression line (16%/1°C) represents the fraction of the rise in EE related to fever. In the Stettler et al study (4) this fraction is equal to 6.9%, while it is 13% in the Dubois study (1). This difference could be due to the different populations in each study. The small number of children in our study might also account for this difference. The intercept on the AT ° axis (for AEE = 0) should be considered as the minimal variation level of the temperature from which we detect a difference in EE. It is equal to approximately 0.5°C. The rise in EE is related to fever, and to other phenomena such as rise in protein catabolism as suggested by the urinary nitrogen loss (15, 16). By dividing the individual rise of EE by the individual rise of temperature, EE increase is equal to 11.3% for I°C of temperature. This result is close to those reported by Dubois et al (1) and Stettler et al (4) but differs from those of a recent study performed in sick infants less than one year of age (5). The average RQ during the two phases were no different. However six out of the 10 children studied had a lower RQ during the febrile phase. In some studies carried out in children or adults during an infectious episode, the RQ measured in the acute period was lower than that measured during recovery period (2, 4, 17, 18). This decrease in RQ during the febrile phase might be explained by a rise in lipid oxidation. However, in three cases the RQ was either identical or lower during the afebrile phase. Moreover, as suggested by the absence of difference in RQ, net glucose and lipid oxidation were not different between phases 1 and 2. The heterogeneity of the studied population, in terms of nutritional intakes (seven fed and three fasting), and the relatively moderate infectious state observed in this study might account for these results in RQ and net substrate oxidation. On the basis of 24 h urinary-nitrogen excretion, protein catabolism rose significantly during the febrile phase (P < 0.02). These results are in agreement with other studies using identical or different methodologies (17, 20). More recently Berclaz et al have demonstrated a substantial increase (by a factor of two) in whole body protein turnover in Gambian children suffering from an acute episode of malaria, by using [tSN] glycine administration (21). In conclusion, in this pediatric population parenterally fed, fever is responsible for an EE rise of 16% per degree of fever. Protracted fever due to infection or inflammatory process is partly responsible for an hypermetabolic state including increased energy expenditure and protein breakdown.
CLINICAL NUTRITION 255
References 1. Du Bois E F. The basal metabolism in fever. JAMA 1921; 77:352-355 2. Eccles M P, Cole T J, Whitehead R G. Factors influencing sleeping metabolic rate in infants. Eur J Clin Nutr 1989; 43: 485-492. 3. Kluger M J. Body temperature changes during inflammation: their mediation and nutritional significance. Proc Nutr Soc 1989; 48:337-345 4. Stettler N, Shutz Y, Whitehead R, Jequier E. Effect of malaria and fever on energy metabolism in gambian children. Pediatr Res 1992; 31: 102-106. 5. Mclntyre J, Hull D. Metabolic rate in febrile infants. Arch Dis Child 1996; 74:206-209 6. Duggan M B, Alwar J, Milner R D G. The nutritional cost of measles in Africa. Arch Dis Child 1986; 61:61-66 7. James W P T. From SDA to DIT to TEF. In: Kinney J M, Tucker H N, (eds). Energy metabolism. Tissue determinants and cellular corollaries. New York: Raven Press 1992: pp 163-183 8. Sempe M, Pedron G, Roy-Pernot M P. Auxologie: m6thodes et sequences. Edition Theraplix Paris, 1979: pp 138-162 9. Brook C G. Determination of body composition of children from skinfold measurements. Arch Dis Child 1971; 46:182-184 10. Dumin J V, Rahaman M M. The assessment of amount of fat in the human body from measurements of skinfold thickness. Br J Nutr 1967; 21:681-689 11. Weir J B. New methods for calculating metabolic rate with special reference to protein metabolism. J Physiol 1949; 1 0 9 : i - 9 12. Jequier E. Measurement of energy expenditure in clinical nutritional
Submission date: 26 May 1997 Accepted: 29 July 1997
assessment. J Parenter Enteral Nutr 1987; 11: 86S-89S. 13. Ferrannini E. The theoritical bases of indirect calorimetry: a review. Metabolism 1988; 37:287-301 14. Hill J R, Rahimtula K A. Heat balance and the metabolic rate of newborn babies in relation to environmental temperature, and the effect of age and weight on basal metabolic rate. J Physiol 1965; 180:239-265 15. Long C L, Jeevanandam M, Kim B M, Kinney M. Whole body protein synthesis and catabolism in septic man. Am J Clin Nntr 1977; 30:1340-1344 16. Beisel W R. Metabolic response to infection. Ann Rev Med 1975; 26:9-20 17. Askanazi J, Carpentier Y A, Elwyn D H et al. Influence of total parenteral nutrition on fuel utilization in injury and sepsis. Ann Surg 1980; 191:40-46 18. Stoner H B, Little R A, Frayn K N, Elebute A E, Tresadern J, Gross E. The effect of sepsis on oxidation of carbohydrate and fat. Br J Surg 1983; 70:32-35 19. Long C L, Schiller W R, Blakmore W S, Geiger J W, O'Dell M, Henderson K. Muscle protein catabolism in the septic patient as measured by 3-methylhistidine excretion. A m J Clin Nutr 1977; 30:1349-1352 20. Tomkins A M, Garlick P J, Schofield W N, Wateflow J C. The combined effects of infection and malnutrition on protein metabolism in children. Clin Sci 1983; 65:313-324. 21. Berclaz P Y, Benedek C, J6qnier E, Schutz Y. Changes in protein turnover and resting energy expenditure after treatment of malaria in Gambian children. Pediatr Res 1996; 39: 401-409.