AJCN North Carolina Research Campus
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Faisy, C.
Right arrow Articles by Fagon, J.-Y.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Faisy, C.
Right arrow Articles by Fagon, J.-Y.
Agricola
Right arrow Articles by Faisy, C.
Right arrow Articles by Fagon, J.-Y.
American Journal of Clinical Nutrition, Vol. 78, No. 2, 241-249, August 2003
© 2003 American Society for Clinical Nutrition

ORIGINAL RESEARCH COMMUNICATION

Assessment of resting energy expenditure in mechanically ventilated patients1,2,3

Christophe Faisy, Emmanuel Guerot, Jean-Luc Diehl, Jacques Labrousse and Jean-Yves Fagon

1 From the Service de Réanimation Médicale, Hôpital Européen Georges Pompidou, Paris (CF, EG, J-LD, JL, and J-YF), and the Service de Pneumologie et Réanimation, Hôtel-Dieu de Paris, Paris (CF).

2 Supported by Service de Réanimation Médicale, Hôpital Européen Georges Pompidou and Service de Pneumologie et Réanimation, Hôtel-Dieu, Assistance Publique-Hôpitaux de Paris, Paris. This study was not sponsored by gifts or fellowships.

3 Address reprint requests to C Faisy, Service de Réanimation Médicale, Hôpital Européen Georges Pompidou, 20, rue Leblanc, 75908 Cedex 15 Paris, France. E-mail: christophe.faisy{at}wanadoo.fr.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Background: Usual equations for predicting resting energy expenditure (REE) are not appropriate for critically ill patients, and indirect calorimetry criteria render its routine use difficult.

Objective: Variables that might influence the REE of mechanically ventilated patients were evaluated to establish a predictive relation between these variables and REE.

Design: The REE of 70 metabolically stable, mechanically ventilated patients was prospectively measured by indirect calorimetry and calculated with the use of standard predictive models (Harris and Benedict’s equations corrected for hypermetabolism factors). Patient data that might influence REE were assessed, and multivariate analysis was conducted to determine the relations between measured REE and these data. Measured and calculated REE were compared by using the Bland-Altman method.

Results: Multivariate analysis retained 4 independent variables defining REE: body weight (r2 = 0.14, P < 0.0001), height (r2 = 0.11, P = 0.0002), minute ventilation (r2 = 0.04, P = 0.01), and body temperature (r2 = 0.07, P = 0.002): REE (kcal/d) = 8 x body weight + 14 x height + 32 x minute ventilation + 94 x body temperature - 4834. REE calculated with this equation was well correlated with measured REE (r2 = 0.61, P < 0.0001). Bland-Altman plots showed a mean bias approaching zero, and the limits of agreement between measured and predicted REE were clinically acceptable.

Conclusion: Our results suggest that REE estimated on the basis of body weight, height, minute ventilation, and body temperature is clinically more relevant than are the usual predictive equations for metabolically stable, mechanically ventilated patients.

Key Words: Resting energy expenditure • mechanical ventilation • indirect calorimetry • nutrition • metabolism • intensive care


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Metabolism is acutely modified by any form of severe disease. Resting energy expenditure (REE) is also influenced by malnutrition (1, 2). Results of one study showed that malnutrition affected 40–50% of the patients in the intensive care unit (ICU) (3). Malnutrition may be linked to higher morbidity and mortality rates and increased length of stay (3). Thus, caloric requirements and specific metabolism are essential components of the care of these patients.

Long et al (4) emphasized that variables such as fever or type of the injury or illness influence the REE of patients who have undergone surgery without respiratory assistance. Other variables affecting the REE of ICU patients include the following: medications used, treatment procedures (5–7), modalities of mechanical ventilation (8, 9), weaning of respiratory support (10, 11), type of nutrition (12, 13), and body composition (1, 14). Few studies have compared REE measured by indirect calorimetry or REE calculated by using Harris-Benedict predictive equations (15) for adult patients requiring respiratory assistance. Moreover, these studies mainly included patients who had undergone minor surgery (5–7, 16–21) and showed that the difference between measured and calculated REE is substantial (from -30% to 49%). However, most of these investigations did not consider situations that modify REE or the limits of accuracy of indirect calorimetry in the ICU. Therefore, we prospectively studied the REE of mechanically ventilated patients in the ICU with the objective of identifying factors that might influence REE to propose an accurate method for calculating REE in patients receiving respiratory support.


   SUBJECTS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Subjects
This prospective, observational study was conducted over a 1-y period in a university teaching hospital. We evaluated adult patients who were intubated and mechanically ventilated for > 24 h. The local institutional review board approved the study design, and the procedures followed were in accordance with the Helsinki Declaration of 1975 as revised in 1983. Indirect calorimetry is a noninvasive technique that assesses the heat liberated during metabolic oxidative processes by measuring oxygen consumption ( ·VO2) and carbon dioxide production ( ·VCO2). However, many clinical conditions common to ICU patients are responsible for erroneous values measured by indirect calorimetry (6, 13, 22,23).

Thus, to obtain accurate REE measurements, patients were excluded for the following reasons.

  1. Hemodynamic instability: introduction, dose modification, and withdrawal of inotropic drugs (catecholamines, ß2-mimetic, and theophylline) during the preceding 2 h.
  2. Respiratory instability: pulse oxygen saturation <90%, modification of ventilator settings during the preceding 2 h, and signs of hyperventilation (respiratory rate > 35/min) or respiratory weakness during pressure-support ventilation.
  3. Variations in the carbon dioxide pool: a constant carbon dioxide pool is a good indicator of respiratory stability and allows a close approximation of ·VCO2 and the respiratory quotient (RQ = ·VO2/( ·VCO2) (24). Also, bicarbonate infusion, digestive or renal losses of bicarbonate (diarrhea, ureterosigmoidostomy, use of acetazolamide) or the use of extracorporeal blood lines for dialysis were reasons for exclusion.
  4. An intravenous carbohydrate load >15 kcal·kg-1·d-1 (ie, 5000 mL of 5% dextrose/d): a relation exists between carbon dioxide production and carbohydrate or caloric intake (12, 25), especially in hypermetabolic and mechanically ventilated patients (5, 24, 26). Moreover, continuous enteral feeding increases thermogenesis from nutrient intake and could affect REE (13). Thus, continuous enteral feeding was stopped 6 h before calorimetry.
  5. Air leaks around the respiratory circuit (aerosolization, lines for nitric oxide or heliox, and tracheal gas insufflation) or visible air leaks in the chest-drainage system.
  6. Accumulation of intermediate metabolites [indirect calorimetry fails to assess the energy produced by metabolic interconversion processes (13)], ketogenesis, or alcoholic intoxication.
  7. High level of inspired oxygen (FiO2 expressed as a percentage): errors in ·VO2 measurement increase exponentially with rising FiO2 concentrations (27). As determined by Weissman et al (28), the limits of accuracy in measuring FiO2 concentrations with our metabolic monitor ranged from 0.21 to 0.80. Patients who required an FiO2 > 80% to obtain a pulse oxygen saturation > 90% were also excluded.

Measurements and instrumentation
After inclusion in the study, simple acute physiology scores 24 h after admission to the ICU and for the 24-h-period before indirect calorimetry were calculated (29). REEs were calculated with the use of Harris-Benedict equations (15) as follows:


(1)


(2)

where body weight is in kilograms, height is in centimeters, and age is in years. The values obtained with these equations were corrected according to Long et al (4) as follows:


(3)

Hypermetabolism factors were 1.13 per °C over 37°C; 1.2 for minor surgery, 1.35 for major trauma or surgery, and 1.6 for severe infection.

Each patient’s REE was then measured by indirect calorimetry. The Puritan-Bennett 7250 metabolic monitor (Puritan-Bennett, Carlsbad, CA) was used for these measurements. This device is an open-circuit calorimeter that interfaces with the 7200 Puritan-Bennett ventilator. The 7250 monitor measures the fractions of FiO2, expired oxygen (FeO2), expired carbon dioxide (FeCO2), and the expiratory flow rate (·VE) in breath-by-breath intervals. ·VO2 is calculated from the Haldane equation:


(4)

REE was calculated on the basis of Haldane’s hypothesis and with De Weir’s equations (30). The measured values were corrected according to the standard temperature pressure dry conditions (0°C, 760 mm Hg, no water steam) before being displayed. The monitor was calibrated before each use with a known gas mixture (5% CO2 and 95% O2). At each breath cycle, sensors were automatically calibrated. All metabolic measurements were made between 1600 and 1800. The eligible patients were at rest and noise was avoided. Nonessential and nonemergency treatments were strictly limited. REE was measured over 3 consecutive 30-min periods, which allowed the elimination of measurements obtained during a possible phase of instability. The 90-min duration is compatible with routine practice in the ICU. The value of each metabolic variable ( ·VO2, ·VCO2, and REE) was recorded every 5 min to obtain a mean of the 6 values collected over each 30-min period, ie, 3 mean values for ·VO2 and ·VCO2. The final mean values of ·VO2, ·VCO2, and REE were calculated from them.

The following 4 quality criteria used to validate the simultaneous measurement of ·VO2, ·VCO2, and REE obtained for each 5-min period were 1) variations in FiO2 < ±0.5% (limits of accuracy of our monitor) between each 5-min period because FiO2 instability can falsify REE measurements (23); 2) variations in FeCO2 < ±10% (limits of accuracy of the 7250 monitor) between each 5-min period because FeCO2 stability is a good indicator of the carbon dioxide pool steadiness; 3) an RQ between 0.7 and 1, because it characterizes metabolic stability (6); and 4) no tracheal aspiration during each 5-min period. If one or more of these quality criteria were not met, the REE value obtained during the 5-min period was excluded from the analysis.

Events leading to the cessation of calorimetric measurements were as follows: occurrence of shock, accidental extubation, leaks of gas around the respiratory circuit, patient agitation, and need for urgent diagnostic or therapeutic procedures.

Patient data
Before calorimetry began, patient characteristics that might influence REE were recorded, including immune deficiencies (AIDS and transplants); treatment for cancer; use of catecholamines, vasopressors, ß2-mimetics, theophylline, curare, sedatives, or morphine or its derivatives; documented infection at the time of the study (blood culture showing positive results, uroculture or other identification of microorganisms grown from bacteriologic samples compatible with severe infection, protected specimen brush with >103 colony forming units/mL or tracheal secretions with >106 colony forming units/mL, >25 leukocytes/field, or <25 epithelial cells/field); caloric intake during the previous 24 h; enteral or parenteral feeding; renal failure (defined as a creatinine clearance <50 mL/min); liver failure (defined as any 2 of the following conditions: serum bilirubin > 150 µmol/L, serum aspartate aminotransferase > 500 U/L, serum albumin < 41 g/L, and clinical signs and symptoms of hepatic coma); body temperature measured electronically in the ear; heart rate; type of respiratory support (pressure-support ventilation or volume-assisted or volume-controlled ventilation); respiratory rate; minute ventilation; level of positive end-expiratory pressure; weight measured electronically; height measured with a tape measure while the patient was in a supine position; body mass index (weight/height2); and height2/Z1 and height2/Z2 estimated by bioelectrical impedance. Z1 and Z2 are body impedances ({Omega}) measured at 1 and at 5 kHz, respectively. To estimate Z1 and Z2 by bioelectrical impedance, we used the 2-electrode, 2-frequency method (Impedance Analyser Analycor 3; Eugedia, Chambly, France). Height2/Z2 is a linear function of extracellular water volume, and height2/Z1 is correlated with fat-free mass and active cell mass in healthy adults (31–33). Active cell mass affects REE in healthy subjects (2, 34).

Statistical analysis
Results are expressed as numbers and percentages, means ± SDs, or medians and ranges (for data with nonnormal distribution). The statistics were calculated in part with STATVIEW 4.5 software (Abacus Concepts Inc, Berkeley, CA). Measured and calculated REE were compared by using 2-way Student’s t test, correlation coefficients, and Bland-Altman analysis (35). The sample size was calculated a priori based on the assumption that an expected minimal difference of 250–500 kcal/d between groups has clinical relevance. To detect such differences [with the use of a ß risk of 0.20, an {alpha} error of 0.05, and an REE SD of 250 kcal/d (36)], >=34 patients (17/group) for a difference of 250 kcal/d or 12 patients (6/group) for a difference of 500 kcal/d are required. The correlation coefficient r was calculated by using linear regression analysis. The difference between the regression line and the identity line is a measure of the accuracy of the predictive equations.

In addition, we conducted a Bland-Altman analysis to determine the limits of agreements between measured REE (reference method) and calculated REE. The mean bias, which represents the difference between measured and calculated REE, is calculated by adding the differences between paired measurements and dividing the sum by the mean of paired measurements. A bias of zero represents a perfect agreement between methods (measured compared with calculated methods). The SD of the bias represents the variability between the methods. The limits of agreements between methods were defined as the mean difference ± 2 SDs.

To identify variables significantly associated with measured REE, we first performed a univariate analysis with the nonparametric Mann-Whitney and Kruskal-Wallis tests for categorical variables and Spearman’s correlation coefficient for quantitative variables because all of the variables were not normally distributed. We considered a difference to be significant when the {alpha} risk was <5% (P < 0.05).

The multivariate analysis was performed with the use of a multiple linear regression model. The variables significantly associated with measured REE according to univariate analysis (P < 0.05) were entered into the model. The variables associated independently and significantly with measured REE in this model were used to establish our predictive equation. Because we had fewer data points than needed for a proper cross-validation study, we used the jackknife method to test the robustness of the estimate of the ß coefficients and their SEs. These calculations were performed with STATA software (Stata Inc, College Station, TX).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Of the 132 patients intubated and mechanically ventilated for >24 h over a 1-y period, only 70 patients were eligible for metabolic monitoring. Sixty-two patients were excluded from the study: 32 because of hemodynamic instability, 13 for respiratory instability or a high FiO2 concentration, 7 for carbon dioxide pool variation, 4 because of inadequate nutrition, and 4 because of air leaks around the respiratory circuit or visible air leaks in the chest-drainage system. The demographic and descriptive data of the eligible patients are summarized in Tables 1 and 2GoGo. Data collected just before the calorimetric measurements are listed in Tables 3 and 4GoGo.


View this table:
[in this window]
[in a new window]
  		TABLE 1 Causes of acute respiratory failure in the patients whose resting energy expenditure was measured by indirect calorimetry1
 

View this table:
[in this window]
[in a new window]
  		TABLE 2 Demographic and descriptive characteristics of the patients eligible for metabolic monitoring1
 

View this table:
[in this window]
[in a new window]
  		TABLE 3 Medications used, organ dysfunction, nutrition, vital signs, and anthropometric and bioelectrical variables in patients eligible for metabolic monitoring1
 

View this table:
[in this window]
[in a new window]
  		TABLE 4 Modalities of respiratory support and respiratory variables for the patients eligible for metabolic monitoring1
 
We performed 1260 REE measurements over 5-min periods; 128 (10%) of these values were excluded from the statistical analysis because the quality criteria were not satisfied (RQ < 0.7 or tracheal aspiration was needed). The median of excluded 5-min periods was 1/patient (range: 0–12). Mean measured REE and calculated REE (Harris-Benedict equations) differed significantly (1890 ± 404 compared with 1399 ± 243 kcal/d, respectively; P < 0.001). Mean calculated REE corrected according to Long et al (4) was not significantly different from measured REE (1817 ± 528 compared with 1890 ± 404 kcal/d, respectively). The mean FeCO2 was 36 ± 9 mm Hg, and the mean RQ was 0.76 ± 0.05.

Univariate analysis
No significant relation was found between measured REE and the cause of acute respiratory failure or presence of organ dysfunction, documented infection, or immune deficiency (Table 5Go). Women had a significantly lower REE than did men (Table 5Go) and women were significantly shorter. Type of respiratory or nutritional support or the use of sedatives or vasoconstrictors, inotropic agents, morphine or its derivatives, or curare did not significantly change the measured REE in our patients (Table 6Go). These observations did not change when REE was adjusted for weight or height (data not shown). Measured REE did not differ between patients who died and those who survived (1883 ± 389 compared with 1898 ± 422 kcal/d, respectively; P = 0.99). A statistically significant relation was noted between measured REE and 7 variables: body weight, height, body mass index, height2/Z2 estimated by bioelectrical impedance, body temperature, arterial blood oxygen saturation, and minute ventilation (Table 7Go).


View this table:
[in this window]
[in a new window]
  		TABLE 5 Univariate analysis: association between resting energy expenditure (REE) measured by indirect calorimetry and various clinical characteristics of the patients eligible for metabolic monitoring1
 

View this table:
[in this window]
[in a new window]
  		TABLE 6 Univariate analysis: association between resting energy expenditure (REE) measured by indirect calorimetry and type of respiratory or nutritional support or medications used in patients eligible for metabolic monitoring1
 

View this table:
[in this window]
[in a new window]
  		TABLE 7 Univariate analysis: relation between resting energy expenditure measured by indirect calorimetry and quantitative variables in the patients eligible for metabolic monitoring1
 
Multivariate analysis
Multivariate linear regression analysis, performed with the abovementioned 7 variables and sex, identified 4 independent predictors of REE (body weight, height, minute ventilation, and body temperature) and yielded the following equation to predict REE from these independent variables for men and women (Table 8Go):


View this table:
[in this window]
[in a new window]
  		TABLE 8 Multivariate analysis: independent variables determining resting energy expenditure in the patients eligible for metabolic monitoring1
 

(5)

The estimates generated by the jackknife method (data not shown) did not differ by >6% from the coefficients given in Table 8Go, which indicated no major bias.

Accuracy of the equations for predicting REE
Calculated REE (Harris-Benedict equations) strongly correlated with measured REE (Figure 1AGo). Bland-Altman analysis showed a mean bias of 491 ± 282 kcal/d (95% CI: 425, 557 kcal/d) between measured REE and calculated REE (Harris-Benedict equations; Figure 2AGo). The limits of agreement between the 2 methods were -73 to 1055 kcal/d. The 95% CI for the lower and upper limits of agreement ranged from -139 to -7 kcal/d and from 989 to 1121 kcal/d, respectively.



View larger version (13K):
[in this window]
[in a new window]
  		FIGURE 1. Correlation between resting energy expenditure (REE) measured by indirect calorimetry and REE calculated with Harris-Benedict equations (A: r = 0.73; r2 = 0.53; 95% CI: 0.35, 0.67; P < 0.0001; y = 205 + 1.2x), with the Harris-Benedict equations corrected according to Long et al (4) (B: r = 0.44; r2 = 0.19; 95% CI: 0.05, 0.37; P = 0.0001; y = 1274 + 0.34x), and with the equation derived from our multivariate analysis (C: r = 0.79; r2 = 0.61; 95% CI: 0.45, 0.74; P < 0.0001; y = 1 + x. The solid lines are the regression lines, and the dashed lines are the identity lines.

 


View larger version (17K):
[in this window]
[in a new window]
  		FIGURE 2. Bland-Altman analysis between resting energy expenditure (REE) measured by indirect calorimetry and REE calculated with the Harris-Benedict equations (A), with the Harris-Benedict equations corrected according to Long et al (4) (B), and with the equation derived from our multivariate analysis (C). The solid lines represent the mean bias between measured and calculated REE, and the dashed lines represent the limits of agreement (± 2 SDs) between measured REE and calculated REE.

 
Measured REE and calculated REE [Harris-Benedict equations corrected according to Long et al (4)] were also correlated, but less well (Figure 1BGo). Bland-Altman analysis showed a mean bias of 73 ± 502 kcal/d (95% CI: -20, 166 kcal/d) between measured REE and calculated REE (corrected Harris-Benedict equations; Figure 2BGo). The limits of agreement between the 2 methods were -932 to 1078 kcal/d. The 95% CI for the lower and upper limits of agreement ranged from -1136 to -728 kcal/d and from 874 to 1282 kcal/d, respectively.

REE calculated with the equation derived from our multivariate analysis was strongly correlated with measured REE (Figure 1CGo). Bland-Altman analysis showed a mean bias of 2 ± 251 kcal/d (95% CI: -58, 62 kcal/d) between measured REE and REE calculated with our predictive model (Figure 2CGo). The limits of agreement between the 2 methods were-500 to 502 kcal/d. The 95% CI for the lower and upper limits of agreement ranged from -602 to -398 kcal/d and from 400 to 604 kcal/d, respectively.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
The results of this study show that the REE of 70 mechanically ventilated patients was affected by only a few determinant variables (weight, height, body temperature, and minute ventilation) and that usual predictive equations failed to accurately estimate REE in these patients. However, because calculations that use predictive equations appear easier to apply than does indirect calorimetric REE in routine practice in the ICU, we wanted to identify those variables that are most closely linked to REE to attempt to establish a relation between these variables and REE.

To the best of our knowledge, this is the first study that assessed indirect calorimetric measurements to develop a simple predictive formula for evaluating REE in clinical practice in the ICU. To remain within the scope of this pilot study, we were particularly cautious in selecting the patients by using multiple exclusion criteria to avoid the unsteady state; in addition, we used a rigorous protocol of indirect calorimetric measurements by using 4 different quality criteria. Indirect calorimetry does not seem applicable immediately after ICU admission, particularly in critically ill patients but could be after a delay that allows stabilization. This delay is probably not crucial, considering the possible usefulness of REE evaluation to routinely manage the nutrition of critically ill patients. As a consequence, the sickest patients could not be evaluated in our study because they had multiple criteria of exclusion; nonetheless, studied patients were indeed severely ill considering their simple acute physiology scores and ICU outcomes.

Measured REE was 25% higher than the calculated REE obtained with the Harris-Benedict equations. This finding agrees with previously reported values (5–7, 16–21). In our study, the difference between measured REE and calculated REE was statistically significant but not when the latter equations were corrected for the hypermetabolism factors proposed by Long et al (4). Although a good correlation was obtained between calculated REE (Harris-Benedict equations) and measured REE, the Bland-Altman analysis showed a statistically significant and clinically relevant mean bias between the 2 methods. This bias could reflect the underfeeding of our patients. Indeed, we limited enteral intakes before the metabolic measurements were made because continuous enteral feeding increases the thermogenesis from nutrient intake and could affect REE (13). Limiting caloric intake for a long time will reduce REE in malnourished patients and result in false predictions calculated with the Harris-Benedict equations (1). By contrast, Zauner et al (37) showed in healthy subjects that, after short-term fasting (3 d), REE rises as a result of an increase in serum norepinephrine. We considered that a brief suspension of caloric intake did not affect the calculated REE of our patients because we stopped enteral feeding just 6 h before the metabolic measurements began. Moreover, the correction factors proposed by Long et al did not improve the accuracy of the Harris-Benedict equations in our critically ill patients. The use of constant hypermetabolism factors, which are mediated by time (4), could explain the poorer REE prediction. However, time from hospital or ICU admission was not a determinant factor of REE in our patients. Also, our results confirmed that REE estimated with the use of the Harris-Benedict equations and Long et al’s correction factors were not reliable for mechanically ventilated patients.

We established that organ dysfunction, the drugs administered, and therapeutic procedures did not affect the REE in our mechanically ventilated patients. Indeed, there were no significant differences in the REE between patients who received pressuresupport, volume-assisted, or volume-controlled ventilation. This result differs from previously published data (8–11). The combination of factors affecting REE might explain this lack of difference because they cancel each other out. For example, 51% of our patients simultaneously received morphine or its derivatives (which decreased the REE) and inotropic agents (which increased the REE). Such combinations are commonly administered in the ICU. Bruder et al (38) measured REE by indirect calorimetry in 24 patients with head injuries and showed that sedation considerably influenced REE. These authors affirmed that sedation changed REE by changing body temperature, which remains the main determinant of REE. According to these same authors, infection also influences REE independently of body temperature. Moriyama et al (36) showed that the REE of patients with septic systemic inflammatory response syndrome increased in a heterogeneous population, including burn victims and postoperative (heart and digestive tract) patients not receiving respiratory support. In contrast, we did not identify infection as an independent factor associated with REE; however, we used different criteria to define infection, and most of our patients had medical problems. In addition, because the methods used to measure REE in our study and that of Moriyama et al were very different, the REE values are likely not comparable. Moreover, because our sample size was not sufficiently large to reach adequate statistical power in these situations, we were unable to discern significant REE difference between patients with and without curare use or liver failure.

Our multivariate analysis indicated that the independent factors defining REE were those closely linked to metabolism (weight, height, minute ventilation, and body temperature). Roza and Schizgal (1), using the regression equations developed by Moore (34), showed excellent correlation between REE and active cell mass in 337 healthy subjects. To the best of our knowledge, no studies have been conducted in an ICU to evaluate both the REE and the body composition of patients. Anthropometric and bioelectrical impedance measures, both of which can be conducted bedside in the ICU, require constant fat-free mass hydration ({approx}72%) to clinically asess body composition (32, 39). Unfortunately, edema and perturbed cellular hydration are common in critically ill patients and may affect the estimation of body composition with regression equations. Moreover, it is unknown how the regression equations used for estimating fat-free mass and active cell mass are affected by the acute inflammatory process frequently present in ICU patients (40). This is why we merely studied the relation between REE and height2/Z1 and height2/Z2. Unlike Roza and Schizgal, we found no relation between REE and height2/Z1, probably because many of our mechanically ventilated patients had abnormalities in water volume.

Indirect calorimetry requires 3–4 h of metabolic and hemodynamic stability to obtain accurate REE measurements—a long time for critically ill patients. Moreover, estimating REE with the use of Fick’s method is not reliable in the ICU (41, 42). The predictive equation derived from our multivariate analysis was strongly correlated with measured REE, and the mean bias between the 2 methods was close to zero. The limits of agreement between the 2 methods were clinically acceptable. However, according to our selection criteria, metabolic stability is required to calculate REE with this predictive equation. In our clinical experience, most ventilated patients briefly satisfy these criteria, thereby allowing their REE to be estimated with the predictive equation but not with indirect calorimetry. Furthermore, overestimation of the quality of the prediction is possible for 2 reasons: 1) because the equation was developed in a specific population, which limits its generalization to other populations, and 2) because any equation developed with a given data set must be validated with another set. Finally, although our results suggest that, for stable mechanically ventilated patients, REE estimated with the use of weight, height, minute ventilation, and body temperature is clinically more reliable than is REE estimated with predictive models, further prospective validation is needed to confirm these results in a larger group of ICU patients.


   ACKNOWLEDGMENTS
 
We acknowledge Gilles Chatellier (Department of Medical Informatics and Biostatistics, Hôpital Européen Georges Pompidou) for statistical advice.

CF was responsible for conceiving and designing the study, collecting and analyzing the data, and drafting the report. EG and J-LD were responsible for interpreting the data and drafting the report. JL and J-YF were responsible for revising the report. No author had any financial or personal interest in any company or organization relevant to the field of research in connection with this work.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Roza AM, Schizgal HM. The Harris Benedict equation re-evaluated: resting energy requirements and the body cell mass. Am J Clin Nutr 1984;40:168–82.[Abstract/Free Full Text]
  2. Rigaud D, Hassid J, Meulemans A, Poupard AT, Boulier A. A paradoxical increase in resting energy expenditure in malnourished patients near death: the king penguin syndrome. Am J Clin Nutr 2000;72:355–63.[Abstract/Free Full Text]
  3. Giner M, Laviano A, Meguid MM, Gleason JR. In 1995 a correlation between malnutrition and poor outcome in critically ill patients still exists. Nutrition 1996;12:23–9.[Medline]
  4. Long CL, Schaffel N, Geiger JW, Schiller WR, Blackmore WS. Metabolic response to injury and illness: estimation of energy and protein needs from indirect calorimetry and nitrogen balance. JPEN J Parenter Enteral Nutr 1979;3:452–6.[Abstract]
  5. Weissman C, Kemper M, Askanazi J, Hyman A, Kinney JM. Resting metabolic rate of the critically ill patients: measured versus predicted. Anesthesiology 1986;64:673–9.[Medline]
  6. Weissman C, Kemper M, Damask MC, Askanazi J, Hyman A, Kinney JM. Effect of routine intensive care interactions on metabolic rate. Chest 1984;86:815–8.[Abstract/Free Full Text]
  7. Swinamer DL, Phang PT, Jones RL, Grace M, King EG. Effect of routine administration of analgesia on energy expenditure in critically ill patients. Chest 1988;92:4–10.
  8. Manthous CA, Hall JB, Kushner R, Schmidt GA, Russo G, Wood LD. The effect of mechanical ventilation on oxygen consumption in critically ill patients. Am J Respir Crit Care Med 1995;151:210–4.[Abstract]
  9. Viale JP, Annat G, Bouffard Y, Delafosse B, Bertrand O, Motin JP. Pressure support ventilation versus continuous positive airway pressure. Chest 1988;93:506–9.[Abstract/Free Full Text]
  10. Bursztein S. Oxygen cost of breathing during weaning. Intensive Care Med 1991;17:193–4.[Medline]
  11. Staudinger T, Kordova H, Roggla M, et al. Comparison of oxygen cost of breathing with pressure-support ventilation and biphasic intermittent positive airway pressure ventilation. Crit Care Med 1998;26:1518–22.[Medline]
  12. Giovannini I, Chiarla C, Boldrini G, Castagneto M. Impact of fat and glucose administration on metabolic and respiratory interactions in sepsis. JPEN J Parenter Enteral Nutr 1989;13:141–6.[Abstract]
  13. Schutz Y. The basis of direct and indirect calorimetry and their potentials. Diabetes Metab Rev 1995;11:383–408.[Medline]
  14. Wang Z, Heshka S, Gallagher D, Boozer CN, Kotler DP, Heymsfield SB. Resting energy expenditure—fat-free mass relationship: new insights provided by body composition modeling. Am J Physiol Endocrinol Metab 2000;279:E539–45.[Abstract/Free Full Text]
  15. Harris JA, Benedict FG. A biometric study of basal metabolism in man. Washington, DC: Carnegie Institute of Washington, 1919. (Publication no. 279.)
  16. Carlsson M, Nordenstrom J, Hedenstielna G. Clinical implications of continuous measurement of energy expenditure in mechanically ventilated patients. Clin Nutr 1984;3:103–10.
  17. Mann S, Westenskow DR, Houtchens BA. Measured and predicted caloric expenditure in the acutely ill. Crit Care Med 1985;13:173–7.[Medline]
  18. Wynn M, Hill DM, Miller DR, Waxman K, Eisner ME, Gazzinaga AB. Management of pancreatic and duodenal trauma. Am J Surg 1985;150:327–32.[Medline]
  19. Van Lanschot JJ, Feenstra BW, Vermeij CG, Bruining HA. Calculation versus measurement of total energy expenditure. Crit Care Med 1986;14:981–5.[Medline]
  20. Bouffard Y, Viale JP, Annat G, Delafosse B, Guillaume C, Motin J. Energy expenditure in the acute renal failure patient mechanically ventilated. Intensive Care Med 1987;13:401–4.[Medline]
  21. Bouffard Y, Delafosse B, Annat G, Viale JP, Bertrand O, Motin J. Energy expenditure during severe acute pancreatitis. JPEN J Parenter Enteral Nutr 1989;13:26–9.[Abstract]
  22. Bursztein S, Elwyn DH, Askanazi J, Kinney JM. Energy metabolism, indirect calorimetry and nutrition. Baltimore: Williams and Wilkins, 1989.
  23. Branson RD. The measurement of energy expenditure: instrumentation, practical considerations and clinical application. Respir Care 1990;35:640–56.
  24. Jéquier E, Felber JP. Indirect calorimetry. Clin Endocrinol Metab 1987;4:911–35.
  25. Baker JP, Detsky AS, Stewart S, Whitwell J, Marliss EBL. Randomized trial of total parenteral nutrition in critically ill patients: metabolic effects of varying glucose-lipid ratios as the energy source. Gastroenterology 1984;87:53–9.[Medline]
  26. Askanazi J, Rosenbaum SH, Hyman A, Silverberg PA, Milic-Emilie J, Kinney JM. Respiratory changes induced by the large glucose loads of total parenteral nutrition. JAMA 1980;243:1444–7.[Abstract]
  27. Ultman JS, Bursztein S. Analysis of error in the determination of respiratory gas exchange at varying FiO2. J Appl Physiol 1981;50:210–6.[Abstract/Free Full Text]
  28. Weissman C, Sardar A, Kemper M. An in vitro evaluation of an instrument designed to measure oxygen consumption and carbon dioxide production during mechanical ventilation. Crit Care Med 1994;22:1995–2000.[Medline]
  29. Le Gall JR, Lemeshow S, Saulnier F. A new simplified acute physiology score (SAPS II) based on a European/North American multicenter study. JAMA 1993;270:2957–63.[Abstract]
  30. De Weir JB. New methods for calculating metabolic rate with special reference to protein metabolism. J Physiol (Lond) 1949;109:1–9.
  31. Boulier A, Fricker J, Thomasset AL, Apfelbaum M. Fat free mass estimation by the two-electrode impedance method. Am J Clin Nutr 1990;52:581–5.[Abstract/Free Full Text]
  32. Tatara T, Tsuzaki K. Segmental bioelectrical impedance analysis improves the prediction for extracellular water volume changes during abdominal surgery. Crit Care Med 1998;26:470–6.[Medline]
  33. Vallentinuzzi ME, Morucci JP, Felice CJ. Bioelectrical impedance techniques in medicine. Part II: Monitoring of physiological events by impedance. Crit Rev Biomed Eng 1996;24:353–466.[Medline]
  34. Moore FD. Energy and the maintenance of the body cell mass. JPEN J Parenter Enteral Nutr 1980;4:228–60.[Medline]
  35. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307–10.[Medline]
  36. Moriyama S, Okamoto K, Tabira Y, et al. Evaluation of oxygen consumption and resting energy expenditure in critically ill patients with systemic inflammatory response syndrome. Crit Care Med 1999;27:2133–6.[Medline]
  37. Zauner C, Schneeweiss B, Kranz A, et al. Resting energy expenditure in short-term starvation is increased as a result of an increase in serum norepinephrine. Am J Clin Nutr 2000;71:1511–5.[Abstract/Free Full Text]
  38. Bruder N, Raynal M, Pellissier D, François G. Influence of body temperature, with or without sedation, on energy expenditure in severe head-injured patients. Crit Care Med 1998;26:568–72.[Medline]
  39. Durnin JVJA, Womersley J. Body fat assessed from total body density and its estimation from skinfold thickness: measurements on 481 men and women aged from 16 to 72 years. Br J Nutr 1974;32:77–97.[Medline]
  40. McMillan DC, Scott HR, Watson WS, Preston T, Milroy R, McArdle CS. Longitudinal study of body cell mass depletion and the inflammatory response in cancer patients. Nutr Cancer 1998;31:101–5.[Medline]
  41. Ogawa AM, Shikora SA, Burke LM, Heetderks-Cox JE, Bergren CT, Muskat PC. The thermodilution technique for measuring resting energy expenditure does not agree with indirect calorimetry for the critically ill patient. JPEN J Parenter Enteral Nutr 1998;22:347–51.[Abstract]
  42. Epstein CD, Peerless JR, Martin JE, Malangoni MA. Comparison of methods of measurements of oxygen consumption in mechanically ventilated patients with multiple trauma: the Fick method versus indirect calorimetry. Crit Care Med 2000;28:1363–9.[Medline]
Received for publication June 14, 2002. Accepted for publication January 14, 2003.




This article has been cited by other articles:


Home page
 JPEN J Parenter Enteral NutrHome page
J. A. Hoher, P. J. Z. Teixeira, F. Hertz, and J. da S. Moreira
A Comparison Between Ventilation Modes: How Does Activity Level Affect Energy Expenditure Estimates?
JPEN J Parenter Enteral Nutr, March 1, 2008; 32(2): 176 - 183.
[Abstract] [Full Text] [PDF]

Home page
 Nutr Clin PractHome page
N. Zijlstra, S. M. t. Dam, P. J. M. Hulshof, C. Ram, G. Hiemstra, and N. M. de Roos
24-Hour Indirect Calorimetry in Mechanically Ventilated Critically Ill Patients
Nutr Clin Pract, April 1, 2007; 22(2): 250 - 255.
[Abstract] [Full Text] [PDF]

Home page
 Nutr Clin PractHome page
D. Frankenfield
Energy Expenditure and Protein Requirements After Traumatic Injury
Nutr Clin Pract, October 1, 2006; 21(5): 430 - 437.
[Abstract] [Full Text] [PDF]

Home page
 Nephrol Dial TransplantHome page
E. Fiaccadori, U. Maggiore, C. Rotelli, R. Giacosa, E. Picetti, E. Parenti, T. Meschi, L. Borghi, D. Tagliavini, and A. Cabassi
Effects of different energy intakes on nitrogen balance in patients with acute renal failure: a pilot study
Nephrol. Dial. Transplant., September 1, 2005; 20(9): 1976 - 1980.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Clin. Nutr.Home page
N. Barak
Assessment of resting energy expenditure in mechanically ventilated patients
Am. J. Clinical Nutrition, February 1, 2004; 79(2): 341 - 342.
[Full Text]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Faisy, C.
Right arrow Articles by Fagon, J.-Y.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Faisy, C.
Right arrow Articles by Fagon, J.-Y.
Agricola
Right arrow Articles by Faisy, C.
Right arrow Articles by Fagon, J.-Y.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS