The Development of a Prediction Equation to Estimate Resting
Metabolic Rate in Healthy, College-aged Students
Brian Vetterkind, Paula Schommer, Amanda Kimmins, Amanda Ekholm
Faculty advisors: Lance Dalleck,
Dalleck PhD and Don Bredle,
Bredle PhD
Department of Kinesiology, University of Wisconsin - Eau Claire
Results
Picture 1: Hydrostatic Weighing Procedure for Body Fat %
Abstract
Our purpose was: 1) to measure resting metabolic rate (RMR) in apparently healthy,
college age (18-26 yr) students, n=30 (18 women, 12 men), 2) compare measured
RMR to estimated values, and 3) to derive a new predictive equation for RMR using
body composition. Each subject completed a 30-min RMR test with a portable
metabolic analyzer and performed hydrodensiometry to obtain an estimate of body
composition. Steady-state VO2 measurements from the last 5-min of each test were
extrapolated over 24 hr to determine kcal/day. Age, weight, body fat%, and RMR
(mean ± SD) for females was 21.7±0.8 yr, 62.0±6.1 kg, 23.5±8.8 %, and 1318±202
kcal/day, and for males was 22.4±1.8 yr, 85.1±16.3 kg, 18.7±7.3%, and 1878±218
kcal/day, respectively). Multiple regression analysis to predict RMR from gender (G)
and lean body
bod mass (LBM) resulted
res lted in the following
follo ing model (R2 = 0.711):
0 711):
RMR (kcal/day) = 781.3 + 313.6(G) + 11.4(LBM)
The standard error of the estimate (SEE) for the prediction of RMR was 193 kcal/day.
The correlation coefficient between the predicted and measured RMR values was r =
0.843. Dependent t-tests resulted in no significant differences (p > 0.05) between
predicted (1545 kcal/day) and measured (1542 kcal/day) RMR values. The equation
developed in the current study can be used by health and fitness professionals to
provide accurate predictions of RMR when oxygen uptake cannot be measured but
body composition can be. This information can be used to design safe and effective
energy balance programs for fat loss and maintenance.
Subjects
18 UWEC students from Biology 354 Physiological Nutrition and 12 student volunteers.
Procedures
Resting Metabolic Rate: All subjects reported to the lab between 6 and 8am having fasted
overnight. Subjects reclined in chair for 30 minutes to return to a resting metabolic
condition. Data (O2 uptake, CO2 production, Respiratory Quotient) were collected for 15
minutes using a Cosmed K4B2 portable metabolic analyzer with an 18mm turbine. (see
picture 2). The first 5 minutes’ data were discarded and the remaining 10 minutes
averaged. Calories expended were calculated from the O2 uptake and the RQ and
expressed for a 24 hour period (kcal/day)
Body Composition: Height and weight were measured using standard scales. BMI was
calculated as weight in kg/ height in m2. Body density was determined via hydrostatic
weighing (see picture 1). Body fat % was calculated from standard equations using body
density, and then lean body mass calculated as total body mass minus fat mass.
St ti ti l analysis
Statistical
l i
All statistical procedures were completed using SPSS statistical software (Version 16.0
SPSS for Windows, SPSS Inc., Chicago, IL). Means and standard deviations as well as
correlations were run on the primary variables. A dependent t-test was performed to
calculate mean differences between estimated and measured RMR values. Data were used
to develop an equation for predicting RMR using multiple regression analyses.
Diagnostic tests were performed to check for outlying observations. Tests were also
performed to detect for any violations of multiple regression model assumptions.
Pearson’s r was calculated to determine the correlation between predicted and measured
RMR values. The standard error of the estimate (SEE) was calculated to determine the
accuracy of the predicted VO2 versus the measured VO2. The probability of making a
Type I error was set at P ≤ 0.05
0 05 for all statistical analyses.
analyses
Descriptive data are shown in Table 1, for all subjects combined as well as averaged
by gender. Measured RMR ranged in our subjects from 850kcal/day to
2172kcal/day. Using the Harris-Benedict equation, RMR was overestimated in 28 of
30 subjects. The combined average overestimation was about 140 kcal/day.
New Predictive Equation
Multiple regression analysis was used to develop a prediction equation for RMR
using the following combination of lean body mass and gender as independent
variables (R2 = 0.711):
RMR (kcal/day) = 781.3 + 313.6(G) + 11.4(LBM) , (Figure 1)
where G = gender (0 for women or 1 for men) and LBM = lean body mass (kg).
The SEE for the prediction of RMR was 193 kcal/day. The correlation coefficient between
predicted RMR and measured RMR values was significant (P < 0.05), r = 0.843. Mean differences
between predicted RMR and measured RMR values averaged only 3 kcal/day (P > 0.05)
25 00
SEE = 192.6 kcal/day
r = .843
20 00
Prredicted R MR (kcal/day)
Methods
15 00
10 00
50 0
5 00
1 00 0
1 50 0
2 00 0
250 0
M easured RM R (kcal/day)
Figure 1. (Predicted vs. measured resting metabolic rate.
I
Introduction
d i
C
Conclusions
l i
and
dR
Recommendations
d i
Creeping obesity is defined as slow weight gain associated with aging. Statistics
show that obesity is a prevalent disease in today’s society. Beginning in childhood,
habits in nutrition and physical activity are implemented throughout the lifespan.
Reaching adulthood, specifically college-age, is where those habits become part of a
lifestyle. Being able to better balance caloric output with caloric intake would
greatly assist individuals with preventing, slowing, or reversing obesity. The
greatest percentage of daily caloric output comes from Resting Metabolic Rate
(RMR), which refers to the number of calories burned in a 24-hour period before
any increase in body metabolism because of daily activities or planned exercise.
The most accurate determination of RMR is through measurement of oxygen
consumption in a completely rested state. However, because the lab equipment to
do this RMR measurement is expensive, and the requires much expertise and time,
this procedure is rarely done.
Alternatively, there are numerous equations to help individuals estimate their RMR.
However, these equations, which are derived from gender, age, height, weight
and/or lean body mass, have been derived from studies done mainly on sedentary
adults. These equations may differ in their validity when applied to more
specialized populations.
Due to this limitation, there is a need to further study RMR in more specialized
groups, which may include people such as athletes, diseased, age-specific, etc.
Through
g our research, we found little information based solelyy on the college-age
g g
population (18-26 yrs). The purposes of our study were:
1. To accurately measure RMR in the college-age population
2.Compare measured RMR with the available estimated equations
3.Use our data of RMR and body composition to derive a new, preliminary
predictive RMR equation specific to college-age individuals
•The old, but still widely used Harris-Benedict equation overestimated our subjects
daily caloric expenditure by about 140 kcals. This could help explain why even
people who carefully monitor their calories in and out are still gaining weight.
We hypothesize that there will be a significant difference between measured RMR
and the currently available estimations, and also that our new predictive equation
would be closer to the measured RMR value than the available estimative equations.
•Measurement of RMR via oxygen uptake is preferred to estimates when caloric
balance needs to be closely calculated.
• When measurement of RMR is not possible, but body composition can be
assessed, our proposed equation for predicting RMR should be more accurate for
college-age individuals.
Pi t
Picture
2:
2 Rest
R t andd RMR data
d t collection
ll ti
•This equation
eq ation can be used
sed by
b health and fitness professionals to provide
pro ide acceptably
acceptabl
accurate RMR in non-laboratory based settings.
Table 1. Physical and metabolic characteristics of participants. (mean ± SD).
•Our multiple regression analysis suggests that the strongest factors affecting resting
metabolic rate are gender and lean body mass.
Women
(n=18)
21.7± 0.8
Men
(n=12)
22.4± 1.8
Combined
(n=30)
22.0± 1.3
168.1± 5.1
179.8± 7.6
172.7± 8.4
Body mass (kg)
62.0± 6.1
85.1± 16.3
71.2± 16.0
BMI (kg/m2)
22.0± 2.1
26.2± 3.7
23.6± 3.5
References
Lean Body Mass (kg)
47.2± 5.5
68.9± 10.1
55.9± 13.2
Goldman, Ronald. in The science of anti-aging medicine. American
Academy of Anti-Ageing Medicine,Chicago, IL. 2003, p 167-173.
Age (yr)
Height (cm)
Estimated RMR (kcal/day)*
Measured RMR (kcal/day)
Body Fat Percentage
1456± 65
2027± 232
1685± 322
1318± 202
1878± 218
1542± 346
23.5± 8.8
18.2± 7.3
21.4± 8.6
*Harris-Benedict Equation, circa 1919
•Only when neither oxygen uptake nor body composition measures are possible,
should the Harris-Benedict type estimates be used.
•Overall, this information can be used to design safe and effective nutrition/exercise
strategies for fat loss and maintenance programs.
Burke, Louise. Practical Sports Nutrition. Human Kinetics, Champaign,
IL. 2007, p 35-36
Acknowledgments
The authors would like to thank Ms. Susan Krueger and the students in her Biology 354 class.