Twenty-four-hour energy expenditure and resting metabolic rate in obese, moderately obese, and control subjects

Abstract

Twenty-four-hour energy expenditure (24-EE), resting metabolic rate (RMR) and body composition were determined in 30 subjects from three groups; control (103 +/- 2% ideal body weight, n = 10), moderately obese (129 +/- 1% ideal body weight, n = 6), and obese (170 +/- 5% ideal body weight, n = 14) individuals. Twenty-four EE was measured in a comfortable airtight respiration chamber. When expressed as absolute values, both RMR and 24-EE were significantly increased in obese subjects when compared to normal weight subjects. The RMR was 7592 +/- 351 kJ/day in the obese, 6652 +/- 242 kJ/day in the moderately obese, and 6118 +/- 405 kJ/day in the controls. Mean 24-EE values were 10043 +/- 363, 9599 +/- 277, and 8439 +/- 432 kJ/day in the obese, moderately obese, and controls, respectively. The larger energy expenditure in the obese over 24 h was mainly due to a greater VO2 during the daylight hours. However, 92% of the larger 24-EE in the obese, compared to the control group, was accounted for by the higher RMR and only 8% by other factors such as the increased cost of moving the extra weight of the obese. The higher RMR and 24-EE in the obese was best related to the increased fat free mass.

Full text

The American Journal of Clinical Nutrition 35: MARCH 1982, pp. 566-573. Printed in U.S.A.
© 1982 American Society for Clinical Nutrition
Twenty-four-hour energy expenditure and
resting metabolic rate in obese, moderately
obese, and control subjects1’ 2
Eric Ravussin, Ph. D., Bernard Burnand, M. D., Yves Schutz, Ph. D., and Eric J#{233}quier,M.D.
ABSTRACT Twenty-four-hour energy expenditure (24-EE), resting metabolic rate (RMR)
and body composition were determined in 30 subjects from three groups: control ( 103 ±2% ideal
body weight, n =10), moderately obese (129 ±1% ideal body weight. n =6), and obese (170 ±5%
ideal body weight, n =14) individuals. Twenty-four EE was measured in a comfortable airtight
respiration chamber. When expressed as absolute values, both RMR and 24-EE were significantly
increased in obese subjects when compared to normal weight subjects. The RMR was 7592 ±351
kJ/day in the obese, 6652 ± 242 kJ/day in the moderately obese, and 61 18 ± 405 kJ/day in the
controls. Mean 24-EE values were 10043 ±363, 9599 ±277, and 8439 ±432 kJ/day in the obese,
moderately obese, and controls, respectively. The larger energy expenditure in the obese over 24 h
was mainly due to a greater VO2 during the daylight hours. However, 92% of the larger 24-EE in
the obese, compared to the control group, was accounted for by the higher RMR and only 8% by
other factors such as the increased cost of moving the extra weight of the obese. The higher RMR
and 24-EE in the obese was best related to the increased fat free mass. Am J Clin Nutr 1982;35:
566-573.
KEY WORDS Obesity, daily energy expenditure, resting metabolic rate, activity, body com-
position
Introduction
Many studies have been reported on basal
(BMR) and resting metabolic rates (RMR) of
obese subjects in order to investigate whether
a low energy expenditure could play a role in
the energy imbalance that characterizes the
dynamic phase of obesity. Although it is dif-
ficult to define what is a “normal” metabolic
rate in an obese subject, the reported values
of BMR and RMR, expressed in relationship
with the lean body mass (1) or the body
surface area (1,2, 3), are usually within the
normal range. However, as pointed out by
James et al. (4), “measuring only the RMR
in the obese is unlikely to help in understand-
ing the pathogenesis of obesity”. We need to
know more about the energy expended for
longer periods of time. There have been very
few studies of total energy expenditure during
24 h (24-EE) in control or overweight subjects
(5, 6). By using a respiratory chamber, con-
tinuous measurements of respiratory ex-
changes can be obtained. These measure-
ments permit us to have an accurate estimate
566
of the total daily energy expenditure and to
follow the energy profile during the course of
the day and night. Having developed such a
chamber, the present study was designed in
an attempt to: 1) compare both RMR and 24-
EE in obese and control subjects under stan-
dardized conditions; 2) measure the energy
expenditure profile throughout the day and
night in both control and obese subjects; 3)
study the relationship between anthropomet-
ric data and RMR or 24-EE in control and
obese individuals.
Materials and methods
Subjects
Thirty individuals, 16 women and 14 men, aged 20 to
46 yr. took part in the study. Their physical characteris-
1From the Institute of Physiology, University of Lau-
sanne, 101 1 Lausanne, Switzerland.
2Address reprint requests to: Dr. E. Ravussin, Insti-
tute of Physiology, Faculty of Medicine, Rue du Bugnon
7, 1011 Lausanne, Switzerland.
Received June 10, 1981.
Accepted for publication September 18, 1981.
by guest on July 13, 2011www.ajcn.orgDownloaded from

tics are presented in Table 1. They were classified into
three groups on the basis of their ideal body weight
(IBW) according to the Metropolitan Life Insurance (7),
10 controls (22.4 ±0.6 yr, 61. 1 ±2.5 kg, 103 ±2% IBW),
six moderately obese (24.2 ±1 .2 yr, 78.8 ± 3.2 kg, 129
± 1% IBW) and 14 obese (29.8 ± 2.2 yr, 97.9 ± 4.0 kg,
170 ± 5% IBW).
Experimental design (Fig. 1)
The subjects lived in a respiration chamber for 24
consecutive h from 8 AM on day I until 8 AM on day 2.
They were free to move within the room, to read, write,
telephone, listen to the radio, and watch television, but
they were not allowed to perform strenuous exercise.
Three standardized meals and one snack providing 10418
kJ/day (2490/Kcal/day) (15% protein, 39% fat, and 46%
carbohydrate) were fed at 0800, 1200, 1800, and 2200 h,
respectively.
At 8 AM on day 2, 10 hours after the evening snack,
RMR was measured with a hood device for l’A, h and
integrated values were recorded every 5 mm. The results
during the last hour of measurement were used to cal-
culate the subject’s RMR.
Apostabsorptive blood sample was drawn for glucose
immunoreactive insulin (IRI) (8), free fatty acid (FFA)
(9), glucagon and thyroid hormones [T3 RIA (10), TSH
(1 1), total T4 and free T4, rT3, TBGJ.
RMR
RMR was measured using an open circuit ventilated
hood system. A transparent ventilated hood was placed
over the subject’s head and secured around the neck. Air
was drawn though the hood at a constant rate in order
to maintain the CO2 concentration in the outfiowing air
between 0.8 and 1.0% (range 25 to 35 1/mm according
to the subject). The flow rate was measured by a mass
14) subjects (mean ±
TABLE 1
Physical characteristics of control (n, =10), moderately obese (n2 =6), and obese (n3 =
SEM)
Subjects Sex Age Ht Body wt Ideal wi Body fat Fat mass FFM
Controls
1
2
3
4
5
6
7
8
9
10
M
M
M
M
M
F
F
F
F
F
yr
22
22
21
25
24
25
21
20
20
24
22.4 ± 0.6
cm
175
179.5
174
176
174
161
163
163.5
171
171
170.8 ± 2.0
kg
69.1
67.4
67
69.3
65.6
47
50.5
58
61.4
56
61.1 ± 2.5
%
108
100
106
107
103
93
98
112
108
98
103.3 ± 1.9
%bodywi
19.9
18.5
15.8
16.8
16.4
22.9
24.1
23.5
24.8
19.9
20.3 ± 1.1
kg
13.8
12.5
10.6
11.6
10.8
10.7
12.2
13.6
15.2
11.1
12.2 ± 0.5
kg
55.3
54.9
56.4
57.7
54.8
36.3
38.3
44.4
46.2
44.9
48.9 ± 2.5
Moderately
obese
11
12
13
14
15
16
M
M
M
F
F
F
22
24
28
27
20
24
24.2 ± 1.2
172
185
174
172
175
164.5
173.8 ± 2.7
80
89
85.5
74.5
76.6
67
78.8 ± 3.2
128
125
135
130
128
128
129 ± 1.4
25.2
22.5
29.9
33.6
34.3
33.4
29.8 ± 2.0
20.2
20.0
25.6
25.0
26.3
22.4
23.3 ± 1.1
59.8
69.0
59.9
49.5
50.3
44.6
55.5 ± 3.7
Obese 17
18
19
20
21
22
23
24
25
26
27
28
29
30
M
M
M
M
M
M
F
F
F
F
F
F
F
F
22
25
44
25
30
23
22
46
24
23
40
30
31
32
29.8 ± 2.2
179
178
169
179
175
188
164.5
152
174
166
165
163
157.5
157
161.1 ± 2.7
105
119
106
114
118
110
78
80.9
89
78
94
92
106
81
97.9 ± 4.0
156
179
177
170
184
150
149
177
151
146
179
179
219
168
170.3 ± 5.2
29.4
31.7
29.1
30.9
32.2
27.0
39.6
41.7
41.2
37.8
43.2
40.5
41.8
38.6
36.1 ± 1.5
35.0
37.7
30.8
35.2
38.0
29.7
30.9
33.7
36.7
29.6
40.6
37.3
44.3
31.3
35.1 ± 1.2
70.0
81.3
75.2
78.8
80.0
80.3
47.1
47.2
52.3
48.4
53.4
54.7
61.7
49.7
62.9 ± 3.7
24-H ENERGY EXPENDITURE IN MAN 567
by guest on July 13, 2011www.ajcn.orgDownloaded from

DAY 1
8AM
T
t t
M2
MEASUREMENTS
DAY 2
8AM 11AM
11iF 1
11
M3 S
VO2, VCO2.
ACTIVITY BY RADAR
BLOOD
VO2, VCO2 I
+
r=0.89
#{163}
#{163}
#{163} #{163}
#{149}Obese
#{163}Controls
0 5 10 15 20 25 30 35 40 45
568 RAVUSSIN ET AL.
FIG. 2. Relationship between percentage fat estimated by skinfolds (15) and percentage fat estimated by whole
body densitometry (16).
flow meter (Setaram, Lyon). A fraction ofthe outfiowing
air was continuously sampled, dried and analyzed for
oxygen (Magnos 2T, Hartmann & Braun; full scale range
19 to 21%) and carbon dioxide (Uras 2T, Hartmann &
Braun; full scale range 0 to 2%). The two analyzers were
calibrated using a gas mixture from a proportional mix-
ing pump (H. Wosthoff, Bochum). The respiratory quo-
tient (RQ), the oxygen consumption (VO2) and the met-
abolic rate (MR) was calculated as previously described
(12).
24-EE
24-EE was measured in an airtight respiration cham-
ber (5 m long, 2.5 m wide, and 2.5 m high) which is part
of an open circuit ventilated indirect calorimeter. The
concentrations of oxygen and carbon dioxide were mea-
sured continuously using a thermomagnetic oxygen an-
alyser (Magnos 2T, full scale range 19 to 21%) and an
infrared carbon dioxide analyser (Uras 2T, full scale
range 0 to 1%). The flow rate of the air extracted from
the chamber was measured continuously using a pneu-
motachograph with differential manometer (HP 47303
A). The two analysers were calibrated before and after
each test using a mixture of gases obtained directly from
a proportional mixing pump (H. Wosthoff, Bochum).
The air flow rate and 02, CO2 concentrations of the
outflowing air were instantly computed by a data acqui-
sition system (HP 3052 A) interfaced to an HP 9825 A
calculator to obtain VCO2, V02, RQ, and consequently
the EE. The results were calculated every minute and
mean values were printed every half hour. During the
2’4 PM
214 HENERGY EXPENDITURE
BODY
COMPOSITION (B C)
14
DENS hOME TRY
GLUCOSE -IRI -FFA
GLUCAGON -THYROID -HORMONES
FIG. 1. Experimental design. M,, M2, and M3 =meals; S =snack.
45 #{176}/Fat Idensitometry)
40
35
30
25
20
15
10
5
#{149}Moderately obese
% Fat (skinfolds)
by guest on July 13, 2011www.ajcn.orgDownloaded from

TABLE 2
Mean energy expenditure and RQ measured during 24 h and during the resting metabolic rate
24-EE RMR
Subjects % Activ.
kJ/24 h W/m2 RQ .kJ/24 h W/m2 RQ
‘ty
1
2
3
4
5
6
7
8
9
10
9205
7979
9979
9368
10397
6385
6477
7602
8217
8778
8439 ± 432
58
58
64
59
67
50
49
54
55
61
57.5 ± 1.8
0.778
0.810
0.83 1
0.813
0.832
0.807
0.837
0.814
0.868
0.842
0.823 ± 0.008
6.9
7.9
10.5
8.4
7.6
8.6
7.4
4.8
5.0
5.3
7.2 ± 0.6
6962
6607
7029
7540
7849
4531
4360
4749
5443
6109
61 18 ± 405
44
41
45
47
51
36
33
34
37
42
41.0 ± 1.9
0.809
0.845
0.825
0.821
0.804
0.805
0.829
0.793
0.856
0.802
0.819 ± 0.006
1 1
12
13
14
15
16
9904
10334
9590
8393
9376
9996
9599 ±277
59
56
55
51
57
67
57.5 ± 2.2
0.833
0.818
0.829
0.857
0.871
0.826
0.839 ± 0.008
6.3
4.7
7.5
7.1
5.9
8.9
6.7 ± 0.5
6502
7012
5971
6163
6661
7602
6652 ± 242
39
38
35
39
40
51
40.3 ± 2.2
0.8 16
0.807
0.802
0.828
0.822
0.795
0.812 ± 0.005
17
18
19
20
21
22
23
24
25
26
27
28
29
30
11188
11749
11669
11134
10749
12205
8184
9092
9150
8682
8970
8832
10033
8970
10043 ± 363
58
58
63
56
54
60
51
59
52
54
52
52
57
57
55.9 ± 1.0
0.828
0.790
0.828
0.800
0.778
0.809
0.843
0.808
0.831
0.814
0.783
0.807
0.761
0.822
0.807 ± 0.006
9.7
5.6
5.7
8.8
5.7
8.0
5.3
8.6
6.5
8.2
6.1
7.1 ± 0.5
9037
9037
9858
8841
8167
8514
6640
5552
6418
7368
5945
7100
7017
6795
7592 ± 351
47
45
53
44
41
42
41
36
36
46
34
42
40
43
42.1 ± 1.3
0.738
0.787
0.811
0.778
0.805
0.817
0.772
0.767
0.809
0.880
0.786
0.770
0.764
0.808
0.792 ± 0.009
24-H ENERGY EXPENDITURE IN MAN 569
first 8 to 10 h of the test, the extraction flow rate was
kept at a low level (5 to 10 1/mm) in order to quickly
reach a CO2 concentration of about 0.5% in the chamber.
Then this flow was increased (40 to 60 1/mm) in order to
maintain the CO2 concentration between 0.4 and 0.6%.
During one period of measurement both V02 and VCO2
were calculated taking into account the decreased (02)
or increased (C02) volume within the chamber, and the
volume introduced (02) or extracted (CO2) by the air
flow through the chamber. By doing this, we were able
to obtain a fast response for the respiratory exchange
measurements from the slow response in gas concentra-
tions as discussed by McLean and Watts (13). The
response time of the chamber measured by gas dilution
techniques or burning alcohol was estimated to be 3 mm
in usual conditions of measurements.
The chamber was built in an attempt to mimic the
comfort and facilities of a one bedroom apartment.
Visual contact with the subject is possible through a
large window. Food, containers, and sampling bottles
are introduced in or withdrawn from the chamber
through a small airtight compartment. An air condition-
ing system maintains air temperature and humidity
within a predetermined range in the chamber. Sponta-
neous physical activity was monitored using a radar
system based on the Doppler effect as previously de-
scribed (14). This activity was expressed as a percentage
of the time during which the subject was moving.
Body composition
Body composition was determined using two meth-
ods. Skinfold thickness was measured at four sites (tn-
ceps, biceps, subscapular, and suprailiac) using Harpen-
den calipers and body density was calculated using the
equations of Durnin and Rahaman (15). Body density
was also estimated in 19 of the subjects using a modifi-
cation of the whole body densitometer described by
Garrow et al. (16). A good correlation was found (r =
0.893, p <0.001) between these two techniques (Fig. 2).
Since not all subjects could be measured by the whole
body densitometer the results reported here are those
using the skinfold method.
by guest on July 13, 2011www.ajcn.orgDownloaded from

#{163}10 Cutrsls
U6N.dsrately obese
#{149}14 Obese
200
100
0
I:SLEEP
S
Thee (hours)
812  11 #{149}  i0  4S
570 RAVUSSIN ET AL.
FIG. 3. Pattern of VO2 over 24 h in obese, moderately obese and control subjects (mean ± SEM).
Analysis of results
All the results given in the tables are expressed as
mean ± SEM. Statistical differences were assessed using
the ttest and single correlations were calculated between
the different variables. Single and multiple linear regres-
sion analyses were performed using the statistical pack-
age for the Social Sciences Programs.
Results
Body composition
The three groups differed markedly with
respect to fat mass which was less than 16 kg
for controls, between 20 and 26 kg for mod-
erately obese, and more than 29 kg in the
obese group (Table 1).
RMR
When expressed in absolute values the
RMR varied considerably from subject to
subject ranging from 4360 to 9858 kJ/day
(Table 2). The obese subjects had signifi-
c#{226}ntlyhigher mean RMR than the control
subjects (7592 ±351 >6118 ±405 U/day p
<0.01). However, when RMR was expressed
on the basis of surface area, or lean body
mass, there was no significant difference be-
P00 V (#{149}#{149}lu)
500
400
300
tween the control (41.0 ± 1.9 W/m2 and 124
±3 kJ/day X k FFM), moderately obese
(40.3 ±2.2 W/m and 123 ±1 1 kJ/day x kg
FFM) and obese (42. 1 ±1.3 W/m2 and 123
±4 kJ/day Xkg FFM) subjects (1 W =1 J/
5=86.4 kJ/day =20.7 kcal/day). The resting
respiratory quotients were higher in the con-
trol subjects (0.819 ±0.006) than in the obese
subjects (0.792 ± 0.009; p <0.05) (Table 2).
24-EE
The 24-EE also varied considerably, rang-
ing from 6385 to 12205 kJ/day. Mean values
were 8439 ±432, 9599 ±277, and 10043 ±
363 kJ/day for the control, moderately obese,
and obese subjects respectively (Table 2). The
mean values of 24-EE for each group were
significantly differentwhereas the mean per-
centage activities were not different among
the three groups; 7.2 ±0.6% for control, 6.7
±0.5% for moderately obese, and 7. 1 ±0.5%
for obese subjects. When expressed as a func-
tion of surface area or lean body mass, the
mean 24-EE was similar in the three groups:
57.5 ±1.8, 57.5 ±2.2, and 55.9 ±1.0 W/m2
and 173 ± 4, 176 ± 1 1, and 163 ±4 kJ/day
by guest on July 13, 2011www.ajcn.orgDownloaded from

24-H ENERGY EXPENDITURE IN MAN 571
TABLE 3
Mean oxygen consumption (ml STP/min) during the day (8 to 22 h), during sleep and during RMR (mean ±
SEM)
Day (8-22 h) Sleep’ RMR Day/sleep RMR/sleep
Controls 325 ± 17 194 ± 8 210 ± 14 1.67 1.08
Moderately obese 369 ± l3t 210 ± 7 230 ± 8 1.76 1.10
Obese 393 ±lStt 216 ±7j 264 ±13ff l.82t l.22f
aLowest 1-h value measured during the night; the percentage activity measured by radar was zero.
tSignificantly different from the controls; p <0.05.
tt Significantly different from the controls; p <0.01.
TABLE 4
Simple correlations for 24-EE or RMR (Power part) as
dependent variables (n =30)
.
Independent vanables Correlation
.
coefficients (r)
Statistical
.
signtficance (p<)
1) RMR 0.888 0.001
2) Fat free mass 0.886 0.001
3) Surface area 0.847 0.001
4) Body wt 0.797 0.001
5) Ht 0.532 0.01
6) Plasma insulin 0.473 0.01
7) Fat mass 0.449 0.05
1) Fat free mass 0.822 0.001
2) Surface area 0.787 0.001
3) Body wt 0.741 0.001
4) Ht 0.489 0.01
5) Plasma insulin 0.480 0.01
6) Fat mass 0.4 19 0.05
xkg FFM for control, moderately obese, and
obese subjects, respectively.
The oxygen consumption (V02) profiles
over 24-h test are shown in Figure 3. Lowest
1-h values of V02 during sleep were similar
in the three groups in absolute terms. Only
the oxygen consumption of the obese was
slightly higher than that ofthe controls. How-
ever, the most noticeable differences occurred
during the day, where VO2 increased more in
the obese subjects (82%) than in the controls
(67%) (p <0.05) using sleeping V02 as base-
line. During sleep, there was also a greater
depression of V02 below the RMR in the
obese (22%) than in the moderately obese
(10%) or control (8%) subjects (Table 3).
The mean 24-h RQ was found to be 0.823
±0.008 in the controls, 0.839 ±0.008 in
moderately obese subjects and 0.808 ±0.006
in the obese (Table 2). The difference was
significant only between the moderately
obese and the obese group (p <0.05).
TABLE 5
Prediction of 24-EE and RMR in men and women of
various fat free mass (FFM) (N =30)
Regression equations Variance
explained
a) Simple
RMR (U/day) =1971 +87.1 FFM (kg) 68
24 EE (kJ/day) =3782 +99.3 FFM 78
(kg)
24 EE (kJ/day) =2920 +0.94 RMR 79
b) Multiple
24 EE (ki/day) =2754 +0.52 RMR 86
(kJ/day) +53.8 FFM (kg)
Correlations
Single correlations were performed be-
tween 24-EE or RMR and the different var-
iables characterizing the subjects. The van-
ables which significantly correlated with en-
ergy expenditure are presented in Table 4.
With both 24-EE and RMR, the anthropo-
metric variable which showed the best cor-
relation coefficients was the fat free mass.
Surface area and body weight had slightly
lower correlation coefficients than fat free
mass. RMR was also highly correlated with
24-EE.
The results of the multiple regression anal-
ysis are presented in Table 5. A good predic-
tion of the RMR can be obtained from the
measurement of FFM. Three different pre-
dictions of the 24-EE in a respiratory cham-
ber or perhaps by extension in a metabolic
ward are proposed knowing the FFM, RMR,
or both (Table 5).
Blood parameters
The blood parameters are presented in
Table 6. Immunoreactive insulin and free
by guest on July 13, 2011www.ajcn.orgDownloaded from

572 RAVUSSIN ET AL.
TABLE 6
Meanplasma levels (mean ±SEM) in control, moderately obese, and obese subjects
Controls Moderately obese Obese
Plasma glucose (mmol/l) 4.86 ±0.12 4.78 ±0.22 5.08 ±0.14
Immunoreactive insulin (U/ml) 13.0 ±1.2 19.1 ±2.0* 26.8 ±2.6*
Free fatty acids (j.tmol/l) 368 ± 26 454 ± 57 541 ± 32*
Glucagon (pg/mi) 98 ± 8 103 ± 20 1 1 1 ± 9
T3 radioimmunoassay (nmol/l) 1.89 ±0.12 1.85 ±0.05 1.79 ±0.09
rT3 (nmol/l) 0.247 ±0.031 0.248 ± 0.033 0.219 ± 0.028
FreeT4(pmol/l) 16.0± 1.8 14.8± 1.8 11.8± 1.7
TotalT4(nmol/l) 91 ±9 89± 11 93±9
Thyroid-stimulating hormone (jU/ml) 3.50 ± 0.64 4.33 ±1.91 2.22 ± 0.45
*Significantly different from the controls; p <0.01.
fatty acids plasma levels were significantly
higher in obese than in control subjects, but
there was no difference in thyroid hormones
and glucagon levels.
Discussion
The mean RMR expressed as an absolute
value was found to be increased in obese
patients compared to normal weight subjects
(Table 2), confirming other reports (2, 4). The
increased RMR may be partly explained by
a greater fat free mass in obese than in control
subjects, as already suggested (1, 4), since
both variables were found to be highly cor-
related (Table 4). For this reason, a predictive
equation of the RMR is proposed knowing
the FFM. The RMR value of about 125 kJ/
day x kg FFM observed in each group stud-
ied is similar to that found by James et al.
(17) in their obese male and female subjects.
The postabsorptive resting RQ (Table 3) was
found to be significantly lower in the obese
compared to the control subjects, confirming
previous results (18, 19). This indicates a
greater contribution of lipid oxidation to rest-
ing energy expenditure certainly related to
the increased availability of free fatty acids
in the obese subjects (Table 6).
As previously discussed (14), 24-EE can be
partitioned into several components and mea-
suring RMR by itself gives an incomplete
view of the overall energy metabolism of the
obese patient. The observation that 24-EE
was significantly correlated with RMR may
be explained by the large contribution of
RMR to total energy expenditure in subjects
confined to a respiration chamber. The con-
tinuous measurement of energy expenditure
in subjects allowed relative unrestrained cx-
istence over 24 h may characterize more ac-
curately obese individuals’ metabolism. Just
as for RMR, 24-EE expressed in absolute
terms was found to be higher in obese mdi-
viduals than in normal weight subjects (Table
2). This increased 24-EE was not due to a
greater level of physical activity in the obese
subjects, since spontaneous activity measured
by radar was similar in both groups. It is
interesting to note that the larger 24-EE in
the obese than in the control group was
mainly due to a greater RMR: 92% of the
difference was due to RMR, and only 8% to
other factors such as the increase cost of
moving the extra-weight of the obese (Table
2). Whether or not the postprandial thermo-
genesis component of both groups is similar
remains to be established in these conditions
of three large equilibrated meals. For exam-
plc, Pittet et al. (20) have shown a decrease
thermic effect of carbohydrate in obese
women but further work on the thermic effect
of usual meals is required. Thus, in subjects
restricted to a comfortable thermo-controlled
respiration chamber, ingesting approximately
10.42 mJ/day, it is possible to obtain three
different equations to predict 24-EE from
either FFM, RMR, or both FFM and RMR.
These equations can be used to predict total
daily energy expenditure in subjects living in
a metabolic ward on a maintenance diet.
The changes in energy expenditure during
the 24-h period are also important to assess,
in order to elucidate precise differences be-
tween groups in diurnal and nocturnal met-
abolic rates. We found a greater fall in VO2
during sleep in obese individuals than in
controls when the RMR was used as a base-
line (Table 3). The reasons for this greater
fall in metabolism are unknown. However,
by guest on July 13, 2011www.ajcn.orgDownloaded from

24-H ENERGY EXPENDITURE IN MAN 573
major quantitative differences between
groups occurred primarily during the waking
state rather than the sleeping state (Fig. 2).
In order to allow comparison between mdi-
viduals of different body weight, surface area
is often used as a standard (1,2). However,
we found the fat free mass to be a more
suitable reference both for RMR and 24-EE
(Table 4). This may appear convenient when
predictions of energy expenditure are made
both in resting and moving individuals. We
conclude that the most important factor de-
termining both the increased RMR and 24-
EE in obese subjects when compared to con-
trol subjects is the increased FFM.
In conclusion, this study revealed that the
total daily energy expenditure is higher in
obese compared to normal weight subjects.
The larger energy expenditure in the obese is
mainly due to a higher RMR than that of
normal weight subjects. In addition, the
RMR is best correlated with the size of the
lean body mass. These data emphasize the
necessity to maintain the lean body mass of
obese individuals during weight reduction
(e.g., by physical exercise) since both RMR
and the total energy expenditure over 24 h
are closely dependent on the size of the lean
body mass.
The authors thank Professor Therese Lemarchand for
the thyroid hormones determinations and Evelyne Mae-
der for the other biochemical measurements. They also
express their gratitude to Drs. Elliot Danforth, Jr. and
John S. Garrow for critically reading the manuscript.
This study was supported by a research grant of Nestle
Company, Vevey.
References
I. Halliday D, Hesp R, Stalley SF, Warwick P, Altman
DG, Garrow JS. Resting metabolic rate, weight,
surface area and body composition in obese women.
Int J Obesity l979;3: 1-6.
2. Hoffmans M, Pfeifer NA, Gundlach BL, et al. Rest-
ing metabolic rate in obese and normal weight
women. Int J Obesity l979;3:l 11-18.
3. Bray G, Schwartz M, Rozin R, Listen J. Relation-
ships between oxygen consumption and body com-
position of obese patients. Metabolism l970;l9:4l8-
29.
4. James WPT, Davies HL, Bailes J, Dauncey MJ.
Elevated metabolic rates in obesity. Lancet 1978; 1:
1122-5.
5. Dauncey MJ. Metabolic effects of altering the 24 h
energy intake in man, using direct and indirect cal-
onimetry. Br J Nutr l980;43:257-9.
6. Irsigler K, Veitl V. Sigmund A, Tschegg E, Kunz K.
Calorimetric results in man: energy output in normal
and overweight subjects. Metabolism l979;28: I 127-
32.
7. Metropolitan Life Insurance Company. Stat Bull
l959;40:40.
8. Hales CM, Randle PJ. Immunoassay of insulin with
insulin antibody precipitate. Biochem J 1963;88: 137-
46.
9. Heindel ii, Cushman SW, Jeanrenaud B. Cell-as-
sociated fatty acid levels and energy-requiring proc-
esses in mouse adipocyte. Am J Physiol l974;226:
16-24.
10. Burger A, Sakoloff C, Staeheli V, Vallston MB,
Ingher SH. Radioimmunoassay of 3,5,3’-tniiodo-l-
thyronine with and without a prior extraction step.
Acta Endocrinol l975;80:58-69.
I 1. Lemarchand Th. Radioimmunoassay of thyroid-
stimulating hormone (TSH) in plasma. In: Breuer H,
Hamel D, Kuskemper HL, eds. Methods of hormone
analysis. Stuttgard: Georg Thieme Verlag. 1976:22-
35.
12. J#{233}quier E. Long term measurement of energy cx-
penditure in man: direct or indirect calorimetry. In:
Bj#{246}nntop P, Caivella M, Howard A, eds. Recent
Advances in Obesity Research III. London: John
Libbey, 1981:130-5.
13. McLean JA, Watts PR. Analytical refinements in
animal calorimetry. J Appl Physiol 1976;40:827-3 1.
14. Schutz Y, Ravussin E, Diethelm R, J#{233}quierE. Spon-
taneous physical activity measured by radar in obese
and control subjects studied in a respiration cham-
ber. Int J Obesity (in press).
15. Durnin JVGA, Rahaman MM. The assessment of
the amount of fat in the human body from measure-
ments ofskinfold thickness. B J Nutr 1967;2l:681-9.
16. Garrow JS, Stalley 5, Diethelm R, Pittet Ph, Hesp
R, Halliday D. A new method for measuring the
body density ofobese adults. Br J Nutr l979;42: 173-
83.
17. James WPT, Dauncey MJ, Jung RT, Shetty PS,
Trayhurn P. Comparison of genetic models of obe-
sity in animals with obesity in man. In: Animal
models of obesity. Festing MF, ed. New York:
McMillan, 1979.
18. Ravussin E, Pahud P. Thelin-Doerner A, Arnaud
MJ, J#{233}quier E. Substrate utilization during pro-
longed exercise after ingestion of 3C-glucose in
obese and control subjects. Int J Obesity l980;4:235-
42.
19. Ravussin E, Doerner A, Acheson KJ, Pahud P.
Arnaud Mi, J#{233}quier E. Carbohydrate utilization in
obese subjects after an oral load of 100 g of naturally-
labelled ‘3C-glucose. Br J Nutr 1980;43:28 1-8.
20. Pittet Ph, Chappuis Ph, Acheson K, de Techtermann
F, J#{233}quier E. Thermic effect of glucose in obese
subjects, studied by direct and indirect calorimetry.
Br J Nutr l976;35:28l-92.
by guest on July 13, 2011www.ajcn.orgDownloaded from