Resting Metabolic Rate in Moderate Obesity

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 6

Resting Energy Expenditure in Moderate Obesity

Predicting Velocity of Weight Loss

KONSTANTIN N. PAVLOU, SC.D. MARTHA A. HOEFER, B.A. GEORGE L. BLACKBURN, M.D., PH.D.

The predicted resting metabolic rate (pRMR), as is estimated From the Nutrition/Metabolism Laboratory, Cancer Research
by the Harris-Benedict equation (HBE), was compared with the Institute, New England Deaconess Hospital, Harvard Medical
actual resting metabolic rate (mRMR), as assessed by indirect School, and Nutritional Management, Inc.,
calorimetry, in 31 moderately obese (X above ideal body weight Boston, Massachusetts
= 44 ± 2.8%) male subjects (X age = 48 ± 4.5 years; X weight
= 107.3 ± 17.1 kg; X% fat = 34 ± 3.9). Measured resting met-
abolic rate (mRMR) (1,942 ± 293 kcal/day) was found to be
significantly (p < 0.001) lower than pRMR (2,108 ± 270 kcal/ rate of weight loss. Some of this weight loss is fat due to
day), but significantly higher (p < 0.001) than pRMR kilojoule deficit; the remainder is due to losses in protein
(1,636 ± 133 kcal/day), if ideal body weight was used in the
HBE formula. Individual variation of the mRMR and pRMR and water.7 The inability of many obese patients to sustain
ranged from 65-105% and 95-155% of the expected normal a predetermined rate of weight loss, especially as they near
population values, respectively. The findings suggest that if the their ideal body weight goal, is well known to clinicians.
reduced daily caloric needs observed are added to the metabolic A recently published study attempted to explain this
suppression occurring during dietary restriction, it might explain discrepancy.8 These researchers report that morbidly obese
why many obese individuals experience difficulties in maintaining
predicted rates of weight loss. An equation was derived to predict subjects preparing for gastric bypass surgery demonstrate
RMR in moderately obese male patients. a lower resting metabolic rate (RMR) than is estimated
using the Harris-Benedict Equation (HBE). However, they
O BESITY is considered one of the most common were unable to explain a significant amount of the variance
medical disorders.' Current estimates from the in RMR attributable to the independent variables and
National Center for Health Statistics cite the in- therefore could not propose a clinically useful equation
cidence of obesity (defined as > 130% of ideal body weight) to predict more accurately the daily kilojoule needs for
at 14% for males and 27% for females.2 Its treatment is the obese.
credited with a concomitant reduction in the incidence In an effort to develop a new standard predictive equa-
of such life-threatening conditions as coronary heart dis- tion of the RMR of the obese patients, we compared the
ease, hypertension, and diabetes.3'4 predicted resting metabolic rate (pRMR), as is estimated
The various treatment techniques available5 attempt by the HBE, with measured resting metabolic rate
to create an energy imbalance in which daily caloric ex- (mRMR), as assessed by indirect calorimetry in moder-
penditure exceeds the prescribed kilojoule intake. To ately obese male subjects.
achieve this, standard predictor equations are used to es-
timate resting energy needs6 with an additional 20% for Methods
daily physical activity.
During the initial weeks of dietary treatment, almost Thirty-one moderately obese male subjects participated
every patient demonstrates a psychologically gratifying in a study of weight loss resulting from a 4180 kj (1000
kcal) balanced deficit diet and an aerobic exercise pro-
gram. Analysis of the baseline RMR measurements of
Reprint requests: Dr. Pavlou, Cancer Research Institute, 194 Pilgrim these men prior to dietary intervention is presented here.
Road, Boston, MA 02215. They ranged from 30 to 60 years ofage (X age = 48 ± 8.5
Submitted for publication: September 16, 1985. years), and 21 to 70 (X = 44 ± 15.6)% above ideal body

136
Vol. 203 * No. 2 RESTING ENERGY EXPENDITURE IN MODERATE OBESITY 137
TABLE 1. Anthropometric Characteristics*
Age Height Weight LBM
N (years) (cm) (kg) AIBW % (kg) FM %
31 48.5 ± 4.5 178.1 ± 5.1 107.3 ± 17.1 44.5 ± 15.8 68.4 ± 10.7 36.4 ± 11.2
* Mean ± S.D. LBM = lean body mass. AIBW = above ideal body weight. FM = fat mass. Patients ranged from 21 to 70% above ideal
body weight.

weight9 (X% fat = 36.4 ± 11.2) (Table 1). Medical screen- HBE calculations were performed using a Student's two
ing found them free of any physical, psychological, or tailed t-test for paired data, with a level of significance set
metabolic impairment. These men had previously at- at alpha = 0.05. All analyses were done using the Statistical
tempted to lose weight. Many had failed to lose significant Analysis System (SAS, SAS Institute Inc., Cary, NC).
amounts of weight, citing a discouraging inability to Linear regression analysis was performed to assess the
maintain the rate of weight loss demonstrated in the early relationship between mRMR and pRMR, using both ideal
phase of their weight loss regimen; others had been unable weight and current weight in the HBE. Multiple regression
to maintain the weight loss they had achieved. analysis was used to describe the relationship between
To prevent any confounding of the measurement of mRMR and the predictive variables height, age, weight,
RMR, patients were required to report to the pulmonary and per cent above ideal body weight (% AIBW) and to
laboratory in the postabsorptive state (8-12 hours after develop a clinically useful predictor equation for appli-
the last meal). In addition, caffeine-containing beverages cation in obese male populations.
were prohibited during the premeasurement period, and
subjects did not engage in any physical activity or smoke Results
any cigarettes during that period. Statistical analysis showed that the predicted values of
Patients rested for 30 to 45 minutes, reclined in a dark- RMR were indeed significantly different from the mea-
ened, quiet, and comfortable room. Thereafter, while sured values in our obese, male population. Measured
subjects rested supine in bed, one author (KP, with the resting metabolic rate (mRMR = 8118 ± 1246 kj/day)
help of an assistant, a trained, hospital-based pulmonary was found to be significantly lower (p < 0.001) than pre-
technician) measured resting metabolic rate (mRMR) ev- dicted resting metabolic rate using current weight in HBE
ery 30 seconds by indirect calorimetry, using the Beckman (pRMR (current) = 8811 ± 1129 kj/day). Measured RMR
Horizon metabolic measurement cart (Beckman Instru- (mRMR) was significantly higher (p < 0.001) than pRMR
ments, Inc., Schiller Park, IL). The system measures ox- (X = 6838 ± 556 kj/day) if ideal body weight was used
ygen consumption (V02 ml/min) and carbon dioxide in the calculation (Table 2).
production (CO2 ml/min). Resting metabolic rate was au- Measured RMR (mRMR) was 92 ± 10% of the pRMR
tomatically calculated by the abbreviated Weir formula'0 (current weight) and 119 ± 12% of the pRMR (ideal) (Ta-
and expressed in kcal/min and kcal/day by the equation: ble 2). Individual variation of the mRMR ranged from
RMR in kcal/day 65 to 109% of the expected normal values, with only 64%
of them having RMR within ± 10% of the expected (Fig.
= 3.94 X V02(L/min) + 1.1 X VCO2(L/min). 2). When mRMR was expressed as per cent of expected,
Kcalories were converted to kj by multiplying them with with ideal body weight in the HBE, only 26% of the pa-
the factor 4.18. To assure the validity of our RMR mea-
surements, mean values achieved during the last 10 min- w
utes of steady state (Fig. 1) were used in the calculations.
Predicted resting metabolic rate (pRMR) was calculated I-
from the Harris-Benedict Equation6 for males: Averoge value for
6.0 '(calculating
0 _ mRMR
RMR in kj/day 4
E
5.8
5.6
= [5(H) + 13.7(W) + 66 - 6.8(A)] X 4.18, _ 5.4
5.2
z
where H = height in cm, W = weight in kg (current and CO 1 2 3 4 5 6 7 8 9 40
ideal), A = age in years. Comparisons between measured w
MINUTES
RMR (mRMR) and predicted RMR using current weight
[pRMR (current)] or ideal weight [pRMR (ideal)] in the FIG. 1. Pattern of caloric needs during steady state measurement.
PAVLOU, HOE-ER, AND BLACKBURN Ann. Surg. * February 1986
138
TABLE 2. Measured and Predicted Resting Metabolic Rate (RMR)*
Predicted (kj/day) RMR (Expressed as % Expected)
Measured
Current Weight Ideal Weight (kj/day) Ideal Weight Current Weight
8811±1129t 6838 ±556t 8118±1246
(2108 ± 270 kcal) (1636 ± 133 kcal) (1942 ± 298 kcal) 119 ± 12 92 ± 10
Predicted = using Harris-Benedict equation. Measured = using Indirect Calorimetry. * = mean + S.D.; t = p < 0.001 vs. measured.

tients were found to be within ± 10% of the expected nor- lojoule intake has been attributed to the metabolic re-
mal, with individual variation ranging from 76 to 150% sponse to kilojoule restriction imposed. This restriction
of expected values (Fig. 3). decreases resting metabolic rate (RMR), as reported in
Linear regression analysis showed a statistically signif- previous studies.' 2 The present investigation reports that
icant (p < 0.001) correlation (r = +0.79) between mRMR even without kilojoule restriction, obese individuals are
and pRMR predicted by the HBE (Table 3). Multiple characterized by a suppressed RMR when compared to
regression analysis with height, weight, age, and per cent individuals of normal body weight and body composition.
of above ideal body weight as independent variables Our data are in general agreement with the findings of
showed a statistically significant (p < 0.001) correlation Feurer's recently published study8 reporting that obese
(r = +0.81) between these indices and mRMR (Table 3). individuals preparing to undergo gastric bypass surgery
The regression equation resulting from our analysis: demonstrate a lower resting metabolic rate (mRMR) than
mRMR (kj/day) = [2089.7 - 8. 1(Ht) + 16.8 l(Wt) is estimated by the Harris-Benedict formula (HBE). This
should be of no surprise, since the HBE was derived from
- 8.9(Age) - 1.03(% AIBW)] X 4.18 136 male subjects with normal body weight and com-
accounts for 66% of the variance in measured RMR. position. No obese subjects were included in the popu-
lation studied to develop the HBE.6
Discussion Although Feurer's study clearly demonstrates that "the
resting energy expenditure of morbidly obese persons
The main reason for the observed reduction in the rate cannot be estimated accurately by the Harris-Benedict
of weight loss among obese individuals with restricted ki- formula," it failed to derive a new equation that will more
precisely estimate daily kilojoule needs. Feurer's corre-
lation coefficients of +0.46 (r2 = 0.18) for the male pop-
40 ulation accounted for only a small amount of the variance
35-
30_ Norma
z 25-

0.

*10
5-
0 60 70.. 80 10090 o 120
MesrdRMR:as % of Epce 80 90 {00 41 0 120 t30 140
(alon -CuretWeigt) Mea d RM as % of Expced
(Balon
FIG. 2. When current body weight was used in the calculations, only
64% of the obese patients, as compared to 92% of Boothby's13 normal, FIG. 3. When ideal body weight was used in the calculations, only 26%
healthy volunteers, had measured resting metabolic rates that were within of the patients had measured resting metabolic rates within ± 10% of the
+ 10% of expected values. expected values.
-
Vol. 203 No. 2 RESTING ENERGY EXPENDITURE IN MODERATE OBESITY 139
TABLE 3. Regression Equation for Resting Metabolic Rate (RMR) in Men
r r2 p
RMR kj/day = [-169.1 + 1.02 pRMR (current)] X 4.18 0.79 0.63 <0.001
RMR kj/day = [2089.7 - 8.17 H + 16.81 W - 8.9 A - 1.03 %AIBW] X 4.18 0.81 0.66 <0.001
H = height (cm); W = weight (kg); A = age (years). pRMR = predicted rent weight. %AIBW = per cent above ideal body weight.
resting metabolic rate with Harris-Benedict equation. WCW = with cur-

and thus are not of much clinical use. Their inability to the relationship and indicates that the relative degree of
develop a new equation may be attributed to the heter- obesity provides a significant factor in more precisely es-
ogeneity of the study population, which demonstrated a timating RMR in moderately obese male subjects. The
degree of obesity ranging from 57-226% above ideal body implications of this finding might be important, since,
weight (% AIBW) and a mean (% AIBW) of 111 ± 35%. during the first 2 to 3 weeks on the dietary treatment,
In comparison, we chose to study a more homogeneous almost every patient demonstrates a psychologically grat-
group of obese male subjects, with a degree of obesity ifying rate of weight loss. As mentioned earlier, some of
ranging from 21-70% AIBW (X = 44 ± 16% AIBW). this weight loss is due to kilojoule deficit and some is due
The significant correlation coefficient of +0.81 to protein and water losses.7 But a rather large number
(r2 = 0.66) found between mRMR and age, height, weight, of patients drop out of treatment at 4 to 5 weeks, especially
and % AIBW indicates that these indices adequately pre- those patients on balanced deficit diets. Most often, these
dict the caloric needs of the metabolically active tissues patients are generally classified as "noncompliant" despite
in moderately obese male subjects (Table 3). The addition the patients' insistence of the opposite.
of the index of % AIBW as an independent variable in To demonstrate the extent of the possibility to miscal-
the multiple regression analysis increases the strength of culate daily kilojoule needs in obesity when the HBE is

TABLE 4. Weight Loss in Two Patients


Patient A Patient B
(Subject #6) (Subject #30)
pRMR* Measured pRMR Measured
Weight (kg) 91.5 92.2
Lean Body Mass (kg) 42.5 (46.5%) 64.5 (70%)
Kcal/kg. LBM-' 27.8 28.7
Height (cm) 180.0 183.0
Age (yrs) 59.0 53.0
Resting energy expenditure (kj/day) 7603 4932 7511 7729
(1819 kcal) (1180 kcal) (1797 kcal) (1849 kcal)
20% added for daily activity (kj/day) 1521 986 1651 1534
Total kilojoule need (kj/day) 9125 5910 9012 9263
Minus 10% metabolic suppression due
to diet (kj/day) 911 594 903 928
Actual kilojoule need (kj/day) 8176 5325 8109 8335
Minus kilojoule intake (kj/day) 4180 4180 4180 4180
Daily kilojoule deficit -4034 -1145 -3929 -3946
Kilojoule deficit in 7 weeks of -197,639 -56,121 -192,531 -203,591
treatment (-47,282 kcal) (-13,426 kcal) (-46,060 kcal) (-48,706 kcal)
Weight loss (kg) (kj + 33,440) 5.9 1.7 5.7 6.1
Comments: Dropped out during week 4 of the Loss of 7.2 kg after 7 weeks of treatment.
treatment without weight loss.
* pRMR = using current weight in the HBE.
140 PAVLOU, HOEFER, AND BLACKBURN Ann. Surg. February 1986
(X = 76.5 ± 4.6) and kj per minute (X = 5.6 ± 0.8) was
*
used in the calculations, it is worth looking at two extreme
cases from this study. Subjects #6 and #30 (Table 4) are similar to values reported in other studies for obese in-
characterized by similar age, weight, and height. If we dividuals, and lower than those of normal weight and
utilize the Harris-Benedict Equation using current weight, body composition.'5 However, when mRMR was ex-
we see that both patients are predicted to demonstrate pressed in kj per kg of LBM, obese individuals exhibit
similar daily kilojoule needs and would expect to dem- normal metabolic rate, as is demonstrated by this (X
onstrate similar rates of weight loss. However, the RMR = 119.7 ± 64.8 kj X kg-' of LBM) and other studies'4"18
of Subject #6 was overestimated by 54%, or 2888 kj per even in extreme cases, as is seen in Subjects A and B of
day. Both subjects participated in a 7-week weight loss this study (Table 4).
program with 4180 kj daily intake. Subject #6 dropped The statistically significant differences between the
out of the study at week 4, unable to sustain a satisfactory mRMR and pRMR noticed in this study are due to
rate of weight loss beyond the first 2 weeks of the study. changes in body composition resulting from the obese
As is illustrated in Figure 2, only 64% of the patients state, unaccounted for by the HBE. The increases in body
had metabolic rates within ± 10% of the expected values weight in obesity are mainly due to increases in fat deposits
using current weight, whereas 92% of the normal body and an accompanied moderate increase in LBM, which,
weight, healthy volunteers studied by Boothby13 had rest- although it creates a higher metabolic need, is not high
ing metabolic rates within ± 10% of the predicted. The enough to compensate for the parallel increase in body
mRMR of Patient B (Subject #30) of this study occurred weight (fat). Thus, an explanation is provided for the lower
at 101% of the expected. As is shown in Table 4, the weight metabolic needs noticed in this study.
loss predicted by using current weight in the HBE (5.7 The HBE was developed to predict daily kilojoule re-
kg) is similar in his case to predicted weight loss using quirements for normal weight and body composition in-
mRMR in the calculations. In addition to the 4180 kj per dividuals. The metabolically altered state of obesity re-
day diet, the patient participated in a three-times-per-week quires a new equation that will more accurately account
exercise program and achieved 7.2 kg of weight loss. for these differences. The equation developed from this
Like 36% of the subjects in this study, Patient A (Subject study accurately predicts daily energy expenditure for
#6) demonstrated a metabolic rate of <90% (66%) of ex- moderately obese male subjects with a degree of obesity
pected (Fig. 1); his mRMR was only 66% of expected. ranging from 21-70% above ideal body weight, and its
Patient A's metabolic rate was overestimated by 54% concluding possible applications and ways to test their
(4932 vs. 7603 kj/day). It should be clear from this dis- usefulness should be considered.
cussion that for Patient A to register a similar rate of weight
loss to Patient B, he should be placed on a very different Acknowledgments
weight loss regimen, such as a very-low-calorie-diet (2090
kj/day) requiring strict medical monitoring. The authors wish to thank the subjects who volunteered for the study
It is of interest to note the similarities of the distribution and the pulmonary technician, Patricia Carroll, for her assistance in
between Feurer's data and ours when mRMR is expressed indirect calorimetry.
as per cent of expected, based either on current weight
(Fig. 2) or ideal body weight (Fig. 3). When compared to References
normal weight and body composition population, the 1. Bray GA. The Obese Patient. Philadelphia: WB Saunders Co, 1976.
distribution clearly illustrates that RMR of obese individ- 2. National Center for Health Statistics. Abraham S, Obese and over-
uals, as a group, is clearly suppressed (Fig. 2) but higher weight adults. In U.S. Vital and Health Statistics, Series 11, No.
than if the individual's ideal body weight is used in the 230, DHHS Pub. No. 83-1680 Public Health Service. Washing-
ton, DC: U.S. Government Printing Office, January 1983.
calculation (Fig. 3). 3. Rabkin SW, Mathewson FA, Hsu PH. Relation of body weight to
Although our data show that moderately obese subjects development ofischemic heart disease in a cohort of young North
demonstrate lower RMR than is estimated by the HBE, American men after a 26-year observation period: the Manitoba
Study. Am J Cardiol 1977; 39:452-471.
they are in general agreement with previous reports in- 4. Hubert HB, Feinleib M, McNamara PM, et al. Obesity as an in-
dicating that obese subjects have higher metabolic rates dependent risk factor for cardiovascular disease: a 26-year follow-
when compared to individuals with normal weight and up of participants in the Framingham Heart Study. Circulation
1983; 67:968-977.
body composition.'4"15 This increased metabolic rate is 5. Stunkard A. Obesity. In Kaplan HI, Freedman AM, Sardock BG,
probably due to greater lean body mass (LBM) present eds. The Comprehensive Textbook of Psychiatry, vol II, 3rd ed.
in obese when compared to lean subjects,'6"'7 since LBM Baltimore: Williams and Wilkins, 1980.
6. Harris JA, Benedict FG. Biometric Studies of Basal Metabolism in
has been found to correlate highly with RMR.' "8"9 Rest- Man. Publication Number 297. Washington, DC: Carnegie In-
ing metabolic rate expressed in kj per kg of body weight stitute of Washington, 1919.
Vol. 203 * No. 2 RESTING ENERGY EXPENDITURE IN MODERATE OBESITY 141
7. Yang MU, VanItallie TB. Composition of weight lost during short- 14. James WPT, Davies HL, Bailes J, Douncex MJ. Elevated metabolic
term weight reduction: metabolic responses of obese subjects to rates in obesity. Lancet 1978; 1: 1122-1125.
starvation and low-calorie ketogenic and nonketogenic diets. J 15. Ravussin E, Burnand B, Schutz Y, Jequiez E. Twenty-four-hour
Clin Invest 1976; 58:722-730. energy expenditure and resting metabolic rate in obese, moder-
8. Feurer ID, Crosby LO, Buzby GP, et al. Resting energy expenditure ately obese, and control subjects. Am J Clin Nutr 1982; 35:566-
in morbid obesity. Ann Surg 1983; 197:17-21. 573.
9. Metropolitan Life Insurance Co. data in: Bray GA, ed. Obesity in 16. Pavlou KN, Steffe WP, Lerman RH, Burrows BA. Effects of dieting
Perspective, vol. I. DHEW Publication (NIH) 75.708. Washing- and exercise on lean body mass, oxygen uptake, and strength.
ton, DC: U.S. Government Printing Office, 1975; 72. Med Sci Sports Exerc 1985; 17:466-471.
10. Der Weir JB. New methods for calculating metabolic rate with special
reference to protein metabolism. J Physiol 1949; 109:1-9. 17. Benoit FL, Martin RL, Watten RH. Changes in body composition
11. Apfelbaum M, Bostsarron J, Lacatis D. Effect of caloric restriction during weight reduction in obesity. Ann Intern Med 1965; 63:
and excessive caloric intake on energy expenditure. Am J Clin 604-612.
Nutr 1971; 24: 1405-1409. 18. Webb P. Energy expenditure and fat-free mass in men and women.
12. Bray GA. Effect ofcaloric restriction on energy expenditure in obese Am J Clin Nutr 1981; 34:1816-1826.
patients. Lancet 1969; 2:397-399. 19. Halliday D, Hesp R, Stalley SF, et al. Resting metabolic rate, weight,
13. Boothby WM. In DuBois EF, ed. Basal Metabolism in Health and surface area and body composition in obese women. Int J Obes
Disease, 3rd ed. Philadelphia: Lea and Febiger, 1936; 163-164. 1979; 3:1-6.

You might also like