Proposed shorter duration protocols for measuring exercise energetics utilizing whole room indirect calorimetry

Abstract

Previously, 30 min is the shortest testing duration that is validated for exercise energetics (EXEE; kcal) in human subjects utilizing a whole room indirect calorimeter (WRIC). The objective of this analysis was to show the validity of extrapolating 30-min EXEE from the first 15 min of moderate intensity steady state cycling in human subjects utilizing a WRIC. Metabolic data from human subjects that had 30-min EXEE measured within a WRIC (10,000 L), along with simulated EXEE of the same duration determined by propane combustion, were utilized for this analysis. Metabolic data for 30 min of oxygen consumption (VO₂; liters), carbon dioxide production (VCO₂; liters), the respiratory quotient (RQ; VCO₂/VO₂) and EXEE (kcal), were calculated. Thereafter, the first 15 min of data from both actual and simulated metabolic measurements from human subjects and propane combustion, respectively, were utilized to extrapolate 30-min VO₂, VCO₂, RQ and EXEE. Statistical (p < 0.05) comparisons between actual and extrapolated metabolic parameters were determined by SPSS (version 30). No significant differences existed between actual and extrapolated metabolic parameters for both human subjects and that from propane combustion. In conclusion, utilizing the first 15 min of steady state cycling exercise data to reflect a 30-min measurement duration is valid.

Introduction

Until recently metabolic carts were the only means to obtain measurement of exercise energetics (EXEE) with typical durations from 20–45 min^1,2. This does not include time allotted for warmup/stretching and cool-down. Furthermore, older model metabolic carts only included EXEE expressed in calories, calculated from oxygen and carbon dioxide concentrations during exercise. However, with newer metabolic carts³, EXEE results also include carbohydrate and fat oxidation, derived either from embedded formulas within their software, or that applied after initial data analysis. Finally, for the context of this report, EXEE will refer only to caloric expenditure during steady state cycling exercise without reference to carbohydrate and fat oxidation.

The main disadvantage of metabolic carts for EXEE measurements is the necessity for subject connection to the instrumentation using a head apparatus, which sometimes includes a nose clip. Furthermore, wearing of the head apparatus/nose clip could cause discomfort⁴ that may affect EXEE measurements⁵. Finally, this limits the type of physical activities that can be performed for EXEE measurements⁶. Many individuals who receive exercise prescriptions may not enjoy walking/running or cycling. Therefore, the ability to obtain short duration EXEE measurements of just about any physical activity with whole room indirect calorimetry, without the possible effects of subject anxiety, may go a long way to increasing success rates of obesity treatment programs.

Many older indirect calorimetry methodologies have an inherent disadvantage. This is the inability to accurately correct for the presence of water vapor in the baseline and sample gases, as well as the air flow rates, to allow expression of the metabolic results as Standard Temperature Pressure Dry^7,8. These errors may be greatly enhanced due to the greater amount of water vapor produced by exercising subjects⁹. Finally, this may partly explain why metabolic carts tend to underestimate EXEE by up to 30%⁶. The utilization of newer technologies, such as the Sable Systems Promethion Integrated metabolic instrumentation, incorporated direct continuous measurement of water vapor in the sample and baseline gases, as well as the air flow rates through the whole room indirect calorimeter (WRIC). This eliminates the need for water vapor removal¹⁰ and possible related underestimates of EXEE.

Currently, the gold standard for validation of WRIC’s for EXEE is propane combustion¹¹. However, the torch utilized emulates the energy expenditure output for a person under sedentary conditions. This is not an appropriate method for validating WRIC’s for EXEE since the energetic output is equivalent to that for a sedentary person and not one performing moderate intensity exercise.

This report proposes a new methodology for obtaining shorter duration measurements of EXEE utilizing a WRIC. The total proposed measurement duration within the WRIC including warmup, steady state exercise and cool down would be 30 min. Furthermore, this report also presents a new propane combustion methodology for validating WRIC’s for EXEE utilizing a new larger burner. Therefore, the aim of this analysis is to show that 30-min steady state EXEE, extrapolated from 15-min of actual EXEE data, are valid utilizing a WRIC. This places the total duration of EXEE measurements more on par with that of metabolic carts^1,2, including time for warm up and cool-down. Finally, shorter duration EXEE measurements utilizing a WRIC may make it more cost effective for various research and outpatient clinics, as well as more convenient for individuals being tested.

Materials and methods

The premise of this analysis was to show that previously published⁶ actual (human subject) and simulated (propane combustion) steady state EXEE from 30-min metabolic measurements could be recalculated to reflect the same results in just 15 min. This analysis proposed to take just the first 15 min of actual⁶ and simulated EXEE data, respectively and extrapolate it up to 30-min. Finally, an improved propane combustion methodology for validating exercise WRIC’s is also presented. All the data obtained for this analysis was from the exercise WRIC (10,000 L interior volume) at Mount Sinai Morningside Hospital, Manhattan, NY⁶.

Instrumentation

The Sable Systems Promethion (Model GA3m2/FG250) integrated system (Sable Systems International, North Las Vegas NV, USA) was connected to the WRIC and utilized to obtain metabolic data for this analysis⁶. This instrument contains two channels, each comprising of separate fuel cell oxygen, near inferred carbon dioxide and thin film capacitive water vapor pressure sensors. A novel switching system is utilized to ensure continuous measurements of the sample gases from within the WRIC¹⁰. Another unique feature of the Promethion integrated instrumentation is the continuous correction of the baseline air, sample gases and air flow rates through the WRIC, for the presence of water vapor without the necessity for its removal^7,10. Upon completion of metabolic tests, continuous differences in oxygen and carbon dioxide concentrations between the baseline air and sample gases were used for final calculation of all the metabolic parameters utilizing Expedata software (version 1.9.27, Sable Systems International, North Las Vegas, NV).

Human subject data analysis for extrapolation of 30-min EXEE

Exercise metabolic data involving two female and 13 male healthy adult non-smoking subjects (mean ± standard deviation; age: 28.3 ± 10.8 years and BMI: 25.1 ± 3.9 kg/m²) from a prior study⁶ were utilized for this analysis¹². The sample size for this analysis was similar to two prior studies of EXEE^1,2. Medical histories were obtained by physician’s familiar with the study. Exclusion criteria were age < 18 years old, recent or prescribed medications affecting metabolism, any current or past injury that would not allow exercise and pregnancy in the women as verified by an on-site HCG urine test.

After the informed consent was obtained, the subjects were instructed to fast for 12 h and refrain from caffeine, strenuous physical activity and alcohol for one day prior to exercise metabolic testing. Anthropometrics were obtained upon arrival at the laboratory. Thereafter, the subjects were instructed on how to perform the cycling exercise protocol. Prior to cycling exercise subjects were allowed a 15-min stretching and warm-up period while sitting on the cycle ergometer (Monark Cardio Care 927E, Monark Exercise AB, Vansbro Sweden) within the WRIC, set to their comfort, until instructed to begin the protocol. During the 30-min exercise portion, subjects cycled at 65% of age predicted heart rate max (beats/minute), calculated by the Karvonen formula¹³, at a cadence of 60 revolutions/minute. Heart rate was monitored continuously during exercise using a Polar monitor (Model# POLAR FT1, Polar Electro Inc. Lake Success, NY). The exercise heart rates were maintained within 2 beats/minute of that calculated for subjects. This was achieved by subjects adjusting the resistance on the cycle ergometer themselves while maintaining cadence⁶.

The temperature of the WRIC was maintained at 24 ± 1 °C at ~ 30% relative humidity.

The protocol for this prior study⁶ was approved by the Institutional Review Boards of both St. Luke’s-Roosevelt Hospital and Columbia University, in accordance with the Declaration of Helsinki.

All 30-min extrapolated metabolic data¹² were calculated based on a theoretical 15-min of actual exercise. This involved taking the average across the first 15 min of each subject’s steady state cycling exercise metabolic data from the prior study⁶ for oxygen consumption (VO₂, liters/minute), carbon dioxide production (VCO₂, liters/minute) and EXEE (kcal/minute) and multiplying by 30. The respiratory quotient (RQ; VCO₂/VO₂) was extrapolated by just taking the mean across the respective first 15 min of the exercise data. The actual and extrapolated subject data for all exercise metabolic parameters were then compared ¹².

Propane combustion data analysis for extrapolation of 30 min simulated

EXEE

Ten 60-min propane (99.2% purity, Air Liquide, Houston, TX) combustion tests were performed as described in a prior study of exercise energetics⁶. This duration was chosen to match that from the prior study of EXEE⁶ and to enable a similar recalculation scheme as described above for the human subjects. However, a unique larger burner (Coleman Model 5431B, The Colman Company, Wichita, Kansas USA) was utilized to better simulate moderate intensity steady state exercise (Fig. 1). Propane is completely combusted according to the following balanced equation:

Fig. 1.

Photo of the large propane burner setup utilized for simulated EXEE combustion tests. Analytical balance shown with permission from Mettler Toledo LLC, Columbus, OH, USA.

This equates to 1 g of propane requiring 2.54 L of oxygen for complete combustion. The larger burner utilized was set to consume approximately 0.5 g of propane per minute to emulate the caloric expenditure of moderate intensity exercise of approximately 200 kcal for 30 min. The WRIC was aired out for 30 min in between propane combustion tests⁶. The burn rate (g/minute) for propane was determined by obtaining the weight (g) prior to and after completion of each combustion test using a calibrated analytical balance (Mettler Toledo Model MS1602S/03, Mettler Toledo LLC, Columbus, OH) and then dividing by the total testing duration (minutes). The burn rate for propane was assumed to be linear as determined in a prior study¹². Prior to the commencement of the combustion tests, the Promethion metabolic instrumentation (Sable Systems International, North Las Vegas, NV) was calibrated according to the manufacturer’s instructions. Furthermore, Expedata software (version 1.9.27; Sable Systems International, North Las Vegas, NV) was utilized to calculate energy expenditure (kcal/minute) according to the Weir equation¹⁴, along with oxygen consumption (VO₂, liters/minute), carbon dioxide production (VCO₂, liters/minute) and the RQ (VCO₂/VO₂) from the propane burn rates. The results from each combustion test were then compared to that calculated from the respective propane stoichiometries.

Extrapolation to 30 min for all metabolic parameters was similar to that as described for the human subjects^6,12.

Statistics

Statistical analyses utilized SPSS (version 30, Chicago, IL). Paired t-tests were utilized to determine differences between actual and extrapolated 30-min VO₂, VCO₂, RQ and EXEE in human subjects⁶. Furthermore, regression analysis was utilized to determine the relationship between actual and extrapolated 30-min EXEE. For propane combustion independent t-tests were utilized to determine differences between stoichiometry and combustion for both the actual and simulated extrapolated 30-min VO₂, VCO_2, RQ and EXEE.

The Bland–Altman¹⁵ limit analysis was applied to the 30-min human subject EXEE data to determine agreement between actual and that extrapolated from 15-min for VO₂, VCO₂, RQ and EXEE. This same analysis was applied to the propane combustion data determining the magnitude of agreement between actual and extrapolated 30-min simulated metabolic parameters but not including the RQ. Furthermore, proportional bias (p < 0.05) was first determined for human subjects by regressing mean differences between actual and extrapolated 30-min VO₂, VCO₂, RQ and EXEE on mean values obtained by both methods (actual and extrapolated). A similar proportional bias was performed for the propane combustion data but did not include RQ. All results are presented as mean ± SD (p < 0.05), unless otherwise noted.

Results

Human subject data analysis for extrapolation of 30-min EXEE

All descriptive statistics for both actual and extrapolated 30-min VO₂, VCO₂, RQ and EXEE for the human subjects are shown in Table 1. According to paired t-tests, no differences were found between actual and extrapolated 30-min VO₂, VCO₂, or EXEE during steady state cycling exercise. Furthermore, the overall percentage differences between actual and extrapolated values for any of these metabolic parameters was less than 1.6 ± 6.5%, when the RQ is not included (Table 1). However, there was a difference in the RQ (3.4 ± 1.8%; p < 0.05) between the two methods of calculating the EXEE data (Table 1). The significant relationship (r = 0.97, p < 0.01) between the actual and extrapolated 30-min derived EXEE is shown in Fig. 2A. Furthermore, the regression residuals for this relationship ranged from -25.4 to 22.8, with a mean of 0.0 ± 14.8 kcal.

Table 1.

Descriptive statistics for both actual and extrapolated (EXP) metabolic parameters from subjects and propane combustion.

Parameter	Actual 30-min data			EXP from 15-min			Statistics
Subjects4	Mean	± SD1	SE2	Mean	± SD	± SD	p < 0.05	Delta (%)3
EXEE (kcal)5	253.6	52.6	13.6	252.1	58.5	15.1	0.71	-1.0 ± 6.1
VO2 (liters)6	52.7	11.0	2.8	52.1	12.2	3.2	0.48	-1.6 ± 6.6
VCO2 (liters)7	43.5	9.3	2.4	44.4	10.7	2.8	0.21	1.6 ± 6.5
RQ (VCO2/VO2)8	0.83	0.05	0.01	0.85	0.05	0.01	0.01	3.4 ± 1.8

Propane (larger burner)	30-min stoichiometry			EXP from 15-min combustion			Statistics
Burn rate (g/minute)	0.5661	0.1180	0.0373	–––-	–––	–––	–––	––––––
EXEE (kcal)	202.4	42.2	13.3	196.8	43.2	13.7	0.87	-2.9 ± 2.5
VO2 (liters)	43.2	9.0	2.8	42.8	9.5	3.0	0.87	-1.2 ± 2.8
VCO2 (liters)	25.9	5.4	1.7	25.5	5.4	1.7	0.84	-1.8 ± 1.9
RQ (VCO2/VO2)	0.60	0.00	0.00	0.60	0.01	0.00	1.00	0.5 ± 1.9

¹SD = Standard deviation.

²SE = Standard error.

³Delta % = calculated as the percentage difference from either actual subject data or propane stoichiometry.

⁴Data from a previous study ⁶.

⁵EXEE = Exercise energetics.

⁶VO₂ = Oxygen consumption.

⁷VCO₂ = Carbon dioxide production.

⁸RQ = Respiratory quotient.

Fig. 2.

Relationships between actual and extrapolated (EXP) 30-min (min) derived EXEE (kcal) in 15 healthy adult non-smoking human subjects. This plot also includes the line of equality (plot A). The Bland–Altman limits of agreement analysis between actual and extrapolated (EXP) 30-min derived EXEE (kcal) in 15 healthy adult non-smoking human subjects (plot B).

The related comparisons in terms of the Bland–Altman limit analysis are presented in Table 2. There were good agreements between actual 30-min VO₂, VCO₂, RQ and EXEE and that extrapolated from 15 min for subjects. For example, the Bland–Altman plot (Fig. 2B) represents the close agreement between 30-min actual EXEE and that extrapolated from 15-min. Furthermore, delta extrapolated 30-min EXEE was within 1.4 ± 15.3 kcal of zero (Fig. 2B). These close agreements are further reflected by the narrow confidence intervals and much tighter limits of agreement for each of these metabolic parameters (Table 2). Finally, no significant proportional bias existed for any of these metabolic parameters (Table 2) suggesting that larger average values do not lead to greater disagreements between actual and that extrapolated from 15 min.

Table 2.

Bland–Altman limit analysis for comparison between actual and extrapolated 30-min EXEE for subjects and propane combustion.

Comparison	Parameter	Bias	95% CI1		Limits of Agreement2		Proportional bias (p < 0.05)
15-min subject metabolic data extrapolated to 30-min versus 30-min subject metabolic data	EXEE (kcal)3	1.4 ± 15.3	-7.0	10.0	-29.2	32.0	0.15
	VO2 (liters)4	0.6 ± 3.5	-3.4	10.2	-6.2	7.4	0.19
	VCO2 (liters)5	-0.9 ± 2.7	-2.4	0.6	-6.3	4.5	0.06
	RQ (VCO2/VO2)6	-0.03 ± 0.01	-0.03	-0.02	-0.05	0.01	0.57
15-min propane combustion data extrapolated to 30-min versus 30-min propane stoichiometry	EXEE (kcal)	5.6 ± 5.3	1.8	9.4	-5.0	16.2	0.59
	VO2 (liters)	0.4 ± 1.4	-0.5	1.4	-1.3	3.0	0.36
	VCO2 (liters)	0.5 ± 0.5	0.1	0.8	-0.5	1.5	0.97
	RQ (VCO2/VO2)	0.00 ± 0.01	-0.01	0.01	-0.02	0.02	––

¹Confidence interval.

²Left and right-hand columns equal lower and upper values, respectively.

³EXEE = Exercise energy expenditure.

⁴VO₂ = Oxygen consumption.

⁵VCO₂ = Carbon dioxide production.

⁶RQ = Respiratory quotient.

Propane combustion data analysis for extrapolation of 30 min simulated

EXEE

All descriptive statistics for extrapolated simulated 30-min VO₂, VCO₂, RQ and EXEE, when derived from 15-min recalculated propane combustion data, are shown in Table 1. No differences existed between actual 30-min propane stoichiometry and that extrapolated from 15-min combustion data for any of the metabolic parameters (Table 1). Furthermore, the overall percentage differences between stoichiometry and combustion for any of these metabolic parameters was less than 3.0% (Table 1).

The Bland–Altman limit analysis for the related comparisons between stoichiometry and combustion for the extrapolated 30-min duration are presented in Table 2. There was good agreement between 30-min propane stoichiometry and that extrapolated from 15 min of combustion data for VO₂, VCO₂, and simulated EXEE, as reflected by the narrow confidence intervals and much tighter limits of agreement (Table 2). Moreover, no significant proportional bias existed for any of these parameters (Table 2), suggesting that larger average values do not lead to greater disagreements between propane stoichiometry and combustion.

Discussion

Human subject data analysis for extrapolation of 30-min EXEE

The results from this analysis suggest that only 15 min of actual steady state cycling exercise is necessary for interpretation of the amount of energy expended for 30 min utilizing a WRIC. Furthermore, including the suggested 10 and 5 min of warmup and cool-down, respectively, means the entire exercise metabolic testing session can be completed in 30 min. The warmup/stretching period also allows the subject to get up to the appropriate moderate intensity for any physical activity before data acquisition for EXEE begins. This is a major advantage for room calorimeter laboratories that may perform measurements of EXEE in similar sized WRIC’s for research purposes.

There have been previous attempts to circumnavigate the issues of subject discomfort with the use of metabolic carts with an enhanced head gear design². This head gear relies on exhaled breath passing through a sensor set that hangs a few centimeters in front of the mouth. However, the resultant air space between the sensor set and mouth is affected by external air currents, and the entire apparatus (head gear and back worn electronics) weighs approximately 1.4 kg. Other wearable calorimeters have been studied to provide more versatility for EXEE measurements^16–18. However, they all still suffer from a similar issue of additional weight being carried by the exercising subject, as well as needed head gear thus possibly affecting the EXEE results. Utilizing whole room indirect calorimetry may eliminate some of the subjects’ discomfort and anxieties associated with the use of metabolic carts or wearable electronics thus possibly reducing or even eliminating the errors in EXEE.

Propane combustion data analysis for extrapolation of 30 min simulated

EXEE

This is the first report that utilized a larger propane burner to simulate EXEE. A previous study of EXEE utilized a standard torch⁶ which produces the amount of energy equivalent to an average subject undergoing measurement of resting metabolic rate¹⁹. This equates to approximately 2500 kcal/day or 52.1 kcal in 30 min. However, the larger propane burner allowed energy expenditure of approximately 200 kcal in 30 min, more in line with moderate intensity exercise. Furthermore, this also increases the amount of water vapor as well⁹, as found with exercising subjects. Finally, this larger burner can be adjusted to increase the amount of energy expended thus allowing for measurement of more intense EXEE.

This analysis also demonstrated that extrapolated 30-min simulated EXEE can be achieved from just a 15-min proposed measurement period.

Summary

This analysis showed that whole room indirect calorimetry can produce accurate EXEE results in as little as 15 min. This is advantages in that whole room indirect calorimetry could become an adjunct for many weight loss programs where a convenient and comfortable short duration accurate measurement of EXEE of various physical activities is required. This is important since many overweight individuals undergoing weight loss treatment that requires an exercise component, will be more apt to perform an activity they enjoy. For example, 30-min EXEE during yoga (243 kcal), shadow boxing (267 kcal), playing musical instruments such as the guitar (81 kcal) or the flute (61 kcal) has been determined in a WRIC⁶. Finally, a visit to any wellness or weight loss clinic that has a WRIC could be completed in an hour. This would include 30 min for all client intake requirements while the remaining time would be for the actual EXEE measurement. Currently, the Physiolab at Mount Sinai Morningside Hospital (https://www.mountsinai.org/locations/physiolab) utilizes the only commercially licensed WRIC in the United States for one-hour EXEE measurements. They have been performing these EXEE metabolic measurements since validation of this original protocol in 2016⁶. Now it may be more convenient for clients, and more profitable to the institution, to obtain assessments EXEE in only a 30-min. This is similar to what is currently available in many institutions, including Physiolab, utilizing metabolic carts.

One of the main disadvantages of this methodology is that a WRIC is required. However, this trend is changing as more WRIC’s are being either planned or constructed worldwide in the coming years⁸. This will eventually provide greater access to more accurate EXEE testing, in a comfortable environment, for many individuals suffering from obesity or other metabolic disorders.

In the future additional research, including collaboration with other whole room indirect calorimetry laboratories, needs to be performed regarding the energetics of many other types of physical activities. This will allow updating many of the caloric expenditure estimates for these physical activities²⁰ that are provided to practitioners who operate wellness or obesity treatment clinics.

Author contributions

Dr. Russell Rising: Designed the study, performed the statistical analysis, and prepared the manuscript. Ms. Hannah Kittrell: Assisted in the design and performance of the metabolic measurements. Dr. Jeanine Albu: Aided in the design of the study and review of this manuscript.

Funding

This project was supported in part by an Award from the National Institutes of Health (#P30DK020541).

Data availability

The datasets generated during and/or analyzed for the current study are available from the corresponding author upon reasonable request.

Declarations

Competing of interests

The authors declare no conflict of interest or financial gain from the data analyzed for this manuscript or the association with Physiolab at Mount Siani Morningside Hospital.