Study Design:
- Click here for explanation of classification scheme.
Quality Rating:
Research Purpose:

For many hospitalized patients, a growing trend is to calculate a patient’s energy requirement based on the resting metabolic rate (RMR) measured by indirect calorimetry (IC).

However, little attention has been directed toward the errors involved in deriving the daily RMR by IC.

  1. To examine the validity of resting metabolic rate (RMR) obtained by indirect calorimetry (IC)-a single 3-5 minute measurement obtained after an overnight fast; this value is then extrapolated to 24 hr to provide a value for daily resting energy losses
  2. By measuring the within- and between-day variability in RMR.

To do so:  the RMR was measured hourly in 14 healthy adults after an overnight fast on 2 separate days.

Inclusion Criteria:
  1. adults
  2. volunteers
  3. healthy (definition not provided)
  4. able to consent
Exclusion Criteria:
  1. Refusal to consent
  2. Not meeting inclusion criteria
Description of Study Protocol:


  • Ht measured? yes (method not described)
  • Wt measured? yes (method not described)
  • Fat-free mass measured? Yes, using equations of Durin and Womersley
  • Skinfold thicknesses measured? Yes—biceps, triceps, iliac crest, and subscapular skinfold thicknesses using standardized techniques


  • Monitored heart rate? yes
  • Body temperature? yes
  • Medications administered? Not mentioned

Resting energy expenditure:

  • Protocol the same for all volunteers.
  • Patients were admitted to research facility in morning on 2 separate days after overnight fast.
  • IC type: 2 gas analyzers and a digital pneumotachograph with a face mask and meteorological balloon
  • Equipment of Calibration: yes each morning and between each measurement
  • Coefficient of variation using std gases: all gas volumes were corrected to standard temperature, pressure and density
  • Rest before measure (state length of time rested if available): Rested in bed 1 hr prior to initial measurement; after at least 30 minutes for each measurement thereafter. All patents were confined to bed for 8 hour measurement period, except for bathroom privileges
  • Measurement length: specific length not specified but as measurement were made after at least 30 min of bed rest and they were measured hourly they could not be longer than 30 minutes.
  • Steady state: RMR measurements were recorded when the subject stabilized at a consistent VE for 3 consecutive cycles of the pneumotachograph.
  • Fasting length: Overnight fast.
  • Exercise restrictions prior to test?: non mentioned
  • Room temp:  not mentioned
  • No. of measures within the measurement period: 8 measurements were taken on each day (8 am:measurement 1—4 pm: measurement 8); measurements taken hourly each day
  • Were some measures eliminated? no
  • Were a set of measurements averaged?
  • The hourly RMR were averaged as well as a set of hourly RMR to assess RMR reliability
  • Training of measurer? Single trained measurer to limit interobserver measurement variabililty
  • Subject training of measuring process? Study and measurement techniques described to each subject prior to IC

DIETARY:  Not measured

Statistical tests:  Reproducibility or reliability of the RMR was estimated by calculation of the intraclass correlation coefficient (ICC).

  • ICC established by one-way analysis of variance (ANOVA) for within and between subject variance
  • ICC of 1.00 indicated perfect reliability between measures (i.e., no within-subject variance)
  • ICC of 0 or less indicated no reliability between RMR measurements
  • ICC also determined for average of several measurements
  • Statistical significance set at p<0.05.


  • Ht (cm): mean 167.4 + 2.0
  • Wt (kg): mean 60.4 + 2.8
  • Fat-free mass (kg): mean 41.5 + 1.3
  • Fat mass (kg): mean 18.9 +2.3

*not broken down by gender


  • Circulatory hormones: not mentioned
  • Breathing ability: Spontaneously breathing; “healthy” adults
  • Medical tests/procedures: none other than IC and body composition measurements
  • Chemicals: none mentioned


Number of Measurements: 8 measurements were taken on each day (8 am:measurement 1—4 pm: measurement 8); measurements taken hourly each day

Length of Measurements

specific length not specified but as measurement were made after at least 30 min of bed rest and they were measured hourly they could not be longer than 30 minutes.

Steady State

RMR measurements were recorded when the subject stabilized at a consistent VE  for 3 consecutive cycles of the pneumotachograph.



Data Collection Summary:

Outcome(s) and other measures

  1. Measured RMR with IC
  2. Minute ventilation rate (VE) was measured using a digital pneumotachograph.
  3. RMR measurements were recorded when the subject stabilized at a consistent VE for 3 consecutive cycles of the pneumotachograph.
  4. Caloric expenditure was determined using the nonprotein caloric equivalent for oxygen.
  5. Fat and fat-free body mass calculated according to the equations of Durin and Womersley
  6. Independent variables of weight, height, age, fat-free mass, fat mass
  7. Hourly respiratory quotient (RQ, VCO2/VO2) were compared for days 1 and 2.
  8. Hourly measured oral body temperature was compared for days 1 and 2.
  9. RMR measurements were compared within day and between days 1 and 2.

Blinding used: No.
Description of Actual Data Sample:
  • N = 14 healthy adults volunteered for the study
  • N = 2 males; 12 females
  • (*mean age 26.8 + 1.7 y; range 22- 47)
  • (*not broken down by gender)
Summary of Results:

RQ: In regard to RQ, the trend was that the RQ decreased throughout the day on both day 1 and day 2.


  • Sleep or rest: complete bed rest; rested 1 hour prior to initial measurement and further measurements were done after at least 30 minutes of rest
  • Physical activity: not mentioned
  • Food Intake: Subjects fasted overnight for 2 separate day studies
  • Various times in the day: Metabolic rate measured hourly by IC from 8 am (measurement 1) to 4 pm (measurement 8) on each day; repeated for 2 days

ICC for Day 1*

Measurement number:

  • 1 to 8: 8 am to 4 pm sr = 0.74; ar = 0.96
  • 1 to 3: 8 am to 10 am sr = 0.7; ar = 0.87
  • 4 to 6: 11 am to 1 pm sr =0.9; ar = 0.96
  • 7 to 9: 2 pm to 4 pm sr = 0.7; ar = 0.87
  • 2 to 4: 9 am to 11 am sr = 0.88; ar = 0.96

ICC for Day 2

Measurement number:

  • 1 to 8: 8 am to 4 pm; sr = 0.85; ar = 0.98
  • 1 to 3: 8 am to 10 am; sr = 0.87; ar =0.95
  • 4 to 6: 11 am to 1 pm; sr = 0.91; ar = 0.97
  • 7 to 9: 2 pm to 4 pm; sr = 0.87; ar = 0.95
  • 2 to 4: 9 am to 11 am; sr = 0.88; ar = 0.96

*sr= single measure reliability (0.85 and above = good reliability, 0.76-0.84 = questionable reliability, and 0.75 and below = poor reliability); ar= average reliability of 3 to 8 measures (0.90 and above = good reliability, 0.81-0.89 = questionable reliability, and 0.80 and below = poor reliability)

Analysis of variance of the individual data indicated that there was no significant between-day difference in measurements in RMR; overall the RMR measurements were the same on a daily basis.

The greatest difference in average RMR between days 1 and 2 was 5 kcal/hr and was much less for most measurements. Very similar RMR measurements were obtained in the middle part of both days.

During day 1, RMR remained virtually unchanged from early morning to late afternoon, and there was no significant difference between RMR measurements.

In regard to average ICC results, for day 1, a good estimate of RMR was obtained by averaging measurements 2 to 4 (middle measurements)

RMR reliability was improved by averaging all measurements taken during the day, and maximum reliability was obtained by averaging the middle 3 RMR measurements.

A similar trend was observed during day 2, but the overall reliability was improved compared to day 1.

For day 2, the best (most reliable) measure of RMR was found by averaging measurements 4 to 6; however, the other groupings 1 to 3 and 7 to 9 provided very good estimates of RMR. These averages were more reliable than the comparable measurements on day 1.

Between-measurement variability using 95% confidence intervals were:

Day 1:

  • Measurements 1-2=+11 kcal/hr (264 kcal/d);
  • Measurements 1-3=+9.3 kcal/hr (223 kcal/d);
  • Measurements 2-3=+6.0 kcal/h (144 kcal/d);
  • Measurements 2-4=+4.0 kcal/h (96 kcal/d);

Day 2:

  • Measurements 2-4=+4.0 kcal/hr (96 kcal/d)
Author Conclusion:
  • The results suggest that RMR can be reproducibly measured in the same subject on different days. This was indicated by the very close agreement of the average of eight hourly measurements on the 2 days.
  • However, from examination of the individual RMR curves on both days and the ICC coefficients, it can be seen that the greatest chance of obtaining a large between-day and within-day difference (unreliability) in RMR is in the first few and last few measurements.
  • Thus, the results suggested that although the average RMRs on days 1 and 2 were not statistically significant, the current practice of establishing a patients’ RMR based upon a single measurement potentially can lead to large errors in determining energy needs.
  • Measurement reliability can be improved by serially measuring RMR, eliminating the initial measurement, and averaging the remaining two or three values.
  • The results suggest that new guidelines are recommended that minimize the errors due to measurement variability.
  • Some workers suggest that collecting respiratory gases via a ventilated canopy system rather than a face mask system used in this study may be more reproducible. The expense and importance of these RMR values to the patient dictates that future studies address this issue.
Funding Source:
Government: NIH Grant RR0039
University/Hospital: Emory University, including the Carlyle Heart Fund
Reviewer Comments:


  • Small sample size; especially males
  • generalizability; convenience sample and relatively young age group
  • (Self-selection bias)
  • How the subjects were determined to be “healthy” was not discussed
  • Not discussed whether heights and weights were self-reported or measured by study personnel
  • Declining RQ during study indicated progressive lipolysis with duration of fasting; concern with ketonuria with the 12-to 20 hour fasting (medical complications)
  • Length of study; participants confined to bed for 8 hours on 2 days
  • Use of face mask rather than ventilated canopy system may not have been as reliable; need further testing

Working Group Members:

Had questions regarding the sample size since it was small and if it had the ability to detect difference between measurements (i.e., low power to detect differences and ability of an outlier effect). It was not clear whether variance reported in the article was due to inter-subject variability or intra-subject variability. The method may not be generalizable to the outpatient setting; potential generalizability in a research setting but not a treatment setting; Relevance to clinical setting questioned.