Department of Health and Physical Activity, Physical Activity and Weight Management Research Center, University of Pittsburgh, Pittsburgh, PA, USAUnited States
Also available as PDF (figures and tables appear in the PDF).
Background: Energy expenditure is commonly expressed in multiples of the resting metabolic equivalent task (MET), with 1 MET estimated to be equivalent to 3.5 ml/kg/min or 0.250 L/min of oxygen consumption. This investigation examined whether the estimated resting oxygen consumption used to express a MET was significantly different than measured resting oxygen consumption in overweight, obese class I, and obese class II subjects. Methods: Forty-five (age: 37.5±10.5, BMI: 32.4±3.5) overweight (N=11), Class I (N=21), or Class II (N=13) obese subjects participated in this study. Resting energy expenditure (REE) was assessed on two separate days. Following a 30-minute supine resting period in a semidarkened room, REE was assessed using the dilution technique. Data were expressed as relative (ml/kg/min) and absolute (L/min) oxygen consumption. Results: Relative oxygen consumption (3.0±0.6 ml/kg/min, p<0.001) was significantly lower than the reference value for 1 MET (3.5 ml/kg/min), with no difference across BMI categories, but a difference by gender (p=0.002). The reference-MET value over-estimated oxygen consumption for females and under-estimated oxygen consumption for males. There was no difference in the measured absolute resting oxygen consumption compared to the reference of 0.250 L/min; however, there were independent gender (p<0.001) and BMI (p<0.001) main effects. Conclusions: These differences between the measured and estimated oxygen consumption representing 1 MET in overweight and obese adults may have implications with regards to energy expenditure and its assumed impact on body weight regulation. Continued efforts are needed to understand factors that influence metabolism and the variability observed between individuals in energy expenditure.
Keywords: Metabolic equivalent task; Energy expenditure; Obesity; Overweight; Metabolic equivalent task; Energy expenditure; Obesity; Overweight
The metabolic equivalent task (MET) is a measure of resting oxygen consumption that has the benefit of providing a common descriptor of workload or metabolic intensity.1 The MET is considered to be a universal measure of expressing energy expenditure as a multiple of the resting or reference level in relation to body weight.2 Based on work conducted in 1941 that involved heat exchange in a neutral environment under resting conditions, Gagge at el3 are credited with coining the MET terminology, which most closely mirrors the current use of the MET with regard to energy expenditure.
commonly defined as 3.5 ml/kg/min or 0.250 L/min of oxygen consumption. The origins of 3.5 ml/kg/min to represent a resting MET value of 3.5 ml/kg/ min has been agreed to have resulted from the resting VO2 data obtained from one 40 year old Multiples of a resting MET are commonly used to estimate the performed during various activity tasks. Therefore, it is impor-
over-estimation of energy expenditure. Moreover, given the importance of energy expenditure
to the treatment of obesity, an understanding of whether the current estimates of a resting MET are accurate in individuals who are overweight or obese may be of clinical and scientific importance.
Therefore, the purpose of this investigation is to examine whether the measured resting oxygen consumption, which is used to define a resting MET, in individuals who are overweight or obese is consistent with the widely used estimation of a resting MET (3.5 ml/kg/min or 0.250 L/min of oxygen consumption). Moreover, this study examined whether this varied by gender (male or female) or by grade of overweight or obesity.
Data were obtained from 45 overweight or obese, sedentary, but otherwise healthy adults (age: 37.2±10.5 years; body mass index [BMI]: 32.4± 3.5 kg/m2). Subjects included 34 females and 11 males, with 11 overweight (25.0 to <30.0 kg/m2), 21 with Class I obesity (30.0 to <35.0 kg/m2), and 13 with Class II obesity (35.0 to <40.0 kg/m2). Descriptive data are presented in Table 1.
Height was measured to the nearest 0.5 inch via a wallmounted stadiometer and weight was measured to the nearest 0.5 pound on a calibrated scale with subjects wearing a cloth medical gown or light-weight clothing. BMI was calculated by dividing weight in kilograms (kg) by height in meters squared (m2).
Resting oxygen consumption (VO2rest) was measured with a metabolic cart using the dilution technique. Measurements were obtained between 7:30 AM and 10:30 AM. Pre-test instructions included: fasting for at least 12 hours the night before testing, avoiding consumption of any over-the-counter medications, abstaining from all vigorous physical activity the day before testing, and vehicle transportation to the research center the morning of testing. Study participants were questioned to confirm adherence to these pre-testing instructions upon arrival at the research center. Subjects were placed in a supine position in a semi-darkened room for a period of 30 minutes prior to data collection. Data collection occurred for at least 15 minutes with 5 consecutive minutes representing a steady state condition, de-
(BMI=25.0-29.9 kg/m )
(BMI=30.0-34.9 kg/m ) Obese Class 2 (N=13) (BMI=35.0-39.9 kg/m )
Table 1. Demographic characteristics of subjects.
fined as the range of energy expenditure across this 5 minutes differing by <150 kcal/d.8 This is consistent with a technique that defined steady state of resting energy expenditure at a coefficient of variation of no more than 5% for both oxygen consumption and carbon dioxide production.9 The initial 5 minutes were discarded to allow for the subject to acclimate to the dilution canopy, with the average of a subsequent five consecutive data points meeting the <150 kcal/d difference criteria used to represent VO2rest.
Statistical analyses were conducted using IBM SPSS Statistics (release version 21.0.0.0). One-sample t-tests were used to compare relative and absolute measured VO2rest to the reference-MET (relative=3.5ml/kg/min, absolute=0.250 L/min) value in all subjects. A multivariate analysis of variance was used to examine the difference measured versus estimated VO2rest between genders (males versus females) and BMI categories (overweight, Class I obesity, Class II obesity). Main effects were further examined using post-hoc analysis with Bonferroni adjustment.
Measured relative VO2rest was significantly less than the reference-MET value (3.0±0.6 ml/kg/min versus 3.5 ml/kg/min; p<0.001). There was no significant Gender X BMI interaction. There was also no significant main effect by BMI category for difference between measured and the reference-MET value for relative VO2rest (Table 2). However, there was a significant difference (p=0.002) between males and females for the difference between measured and the reference-MET value for relative VO2rest (Table 2). The reference-MET value over-estimated VO2rest for females by 0.7±0.5 ml/kg/min and under-estimated VO2rest for males by 0.2±0.4 ml/kg/min.
Overall, measured absolute VO2rest (0.275±0.083 L/min) did not differ from reference-MET value (0.250 L/min) (Table 2). While there was no significant Gender X BMI interaction, there was a significant main effect for both Gender (p<0.001) and BMI category (p<0.001). The reference-MET over-estimated by 0.008±0.59 for females while it under-estimated VO2rest by 0.125±0.066 L/min for males. The difference between the
All Subjects (n=45) 91.0±13.6 32.4±3.5 37.2±10.5
Females (n=34) 87.3±12.0 32.7±3.8 35.6±11.1
Males (n=11) 102.5±11.7 31.5±2.1 42.2±6.5 Overweight (N=11) 2 77.0±11.7 28.2±1.5 36.4±11.8
Obese Class I (N=21) 2 93.2±10.4 32.0±1.3 40.6±9.1
2 99.3±10.8 36.7±1.6 32.5±10.1
Relative VO2rest Female N 8 14 Measured 2.73±0.59 2.73±0.56 Difference* 0.77±0.59 0.77±0.56 Male N 3 7 Measured 3.43±0.23 3.77±0.47 Difference* 0.07±0.23 -0.27±0.47 Absolute VO2rest
Female N 8 14 Measured 0.197±0.053 0.239±0.049 Difference** 0.054±0.053 0.011±0.049 Male N 3 7 Measured 0.315±0.034 0.393±0.064 Difference** 0.065±0.034 0.143±0.064 *relative reference-MET=3.5 ml/kg/min; positive value indicates reference-MET over-estimated, negative value indicated reference-MET under-estimates **absolute reference-MET=0.250 L/min; positive value indicates reference-MET over-estimated, negative value indicated reference-MET under-estimates
Table 2. Comparison of relative and absolute VO2rest to reference-MET values (mean±standard deviation).
measured and reference-MET value for absolute VO2rest was significantly less in the overweight category compared to both the Class I (p=0.010) and Class II (p=0.026) obesity categories.
This study examined if measured VO2rest differed from widely accepted reference-MET values (3.5ml/kg/min, 0.250 L/min) in adults who are overweight or obese. This study found that measured VO2rest (3.0±0.6 ml/kg/min) in a sample of adults who are overweight or obese is less than the typically used reference-MET value of 3.5 ml/kg/min (Table 2). Results from this study are similar to results from others in which measured VO2rest was less than the reference-MET value of 3.5 ml/kg/ min. For example, in a sample of 36 males (age=40.0±3.3 years; BMI=25.9±3.8 kg/m2) measured relative VO2rest (3.0±0.3 ml/kg/ min) was significantly lower than the reference-MET value (3.5 10 11 ml/kg/min).10 Gunn et al11 have also reported a similar pattern in a sample of 50 males (age=60.6±3.2 years; BMI=26.7±3.2 kg/ 2 12 m ). In a mixed sample of males and females, Byrne et al also reported that VO2rest (2.6±0.4 ml/kg/min) was significantly lower than the reference-MET value (3.5 ml/kg/min). However, Byrne et al12 did not examine whether there was a gender influence on the difference between the measured and reference-MET value. By comparison, the current study did report a gender difference between the measured and reference-MET value, with reference-MET value over-estimating VO2rest for females and underestimating VO2rest for males.
Another unique contribution of the current study is the examination of whether the difference between measured and the reference-MET value was influence by BMI. Results showed
(25 to <30 kg/m ) (30 to <35 kg/m ) (35 to <40 kg/m ) BMI Category
2.86±0.54 0.64±0.54 0.002 0.742 0.668
3.51±n/a -0.01± n/a
0.277±0.053 0.027±0.053 <0.001 0.006 0.697
0.435±n/a 0.185±n/a
no difference across BMI categories when examining relative VO2rest; however, the difference between measured and the reference-MET value for absolute VO2rest (L/min) was less in the overweight category compared to the Class I and Class II obesity categories. This may suggest that at higher levels of BMI the reference-MET for absolute VO2rest (0.250 L/min) has less utility as an accurate measure. Others studies that have directly compared the measured and reference-MET value for VO2rest have not examined the potential influence of BMI. This finding also has implications when using the reference-MET value of 0.250 L/min to estimate the energy costs of physical activity, as this may over-estimate the energy cost more as BMI increases.
Currently, VO2rest is used to represent the resting MET, and multiples of this resting value are universally used to ex- 1,2 press energy expenditure of various forms of physical activity.1,2 The origins of the MET to express energy expenditure during different forms of physical activity relative to resting energy expenditure appears to date back to approximately 1890,13 which 3,14 was followed by similar observations made decades later. Thus, the findings of the current study, which are similar to the 10,12 finding of others, may suggest that the energy cost of a variety of physical activities may be over-estimated when using common reference-MET values. Moreover, the data from this study may suggest that this over-estimation may be of particular concern for women and for individuals at a higher BMI. However, given the relatively small sample and the limited inclusion criteria, these findings should be interpreted with caution and warrant further replication.
In summary, this study demonstrated that the conventional estimates of VO , represented as the MET, may differ
from measured values in adults who are overweight or obese. These findings may have implications on estimates of resting and physical activity energy expenditure. Thus, while additional research to confirm these findings is warranted, this may suggest the need to establish new estimates of the resting MET that can be used broadly in clinical applications aimed at prevention or treatment of obesity.
The authors would like to acknowledge the staff and students at the Physical Activity and Weight Management Research Center at the University of Pittsburgh for their contributions to the project.
The authors declared conflicts of interest.
RJR: Manuscript development; Statistical analyses.
JMJ: Data collection; Statistical analyses; Manuscript revision.
As per University of Pittsburgh Instructional Review Board Guidelines, all original signed consent forms have been retained by the principal investigator.