Am J Physiol Regulatory Integrative Comp Physiol
280: R1887–R1896, 2001.
Prediction of basal metabolism from organ size in the
rat: relationship to strain, feeding, age, and obesity
P. C. EVEN, V. ROLLAND, S. ROSEAU, J.-C. BOUTHEGOURD, AND D. TOMÉ
Unité Mixte de Recherche, Physiologie de la Nutrition et du Comportement alimentaire,
Institut National de la Recherche Agronomique, 75005 Paris, France
Received 15 August 2000; accepted in final form 20 February 2001
body composition; metabolic activity; predictive model; energy expenditure
(BM) for differences in
body weight and body composition is a major challenge
(1, 8, 10, 12, 17, 19, 25, 32, 38, 48). The relationship
between BM and body weight is not linear when different species of grossly different sizes are compared
(8, 25, 32), but also within species, whether the difference was due to age (9, 55) or to under- or overfeeding
or obesity (14, 45, 54). Dimensional analysis of the size
of animals (25, 32, 38, 48) has been demonstrated to
account for variations in BM between subjects of different species and large differences in body size, but
cannot be reliably used within species (25, 39).
A major focus of the efforts to circumvent this problem has been the use of the concept of fat-free mass
(FFM) or lean body mass (LBM) (26), and, until recently, normalization by FFM seemed to be the best
available predictor of basal metabolic rate, because, in
most cases, it enabled correction for variations induced
by age, sex, and weight (37). However, the use of FFM
CORRECTING BASAL METABOLISM
Address for reprint requests and other correspondence: P. C. Even,
Unité INRA-INAPG Physiologie de la Nutrition et du Comportement
alimentaire, Institut national Agronomique Paris-Grignon, 16 rue
Claude Bernard 75005 Paris, France (E-mail: [email protected]).
http://www.ajpregu.org
to normalize BM may introduce a bias in many instances, because FFM is not homogenous but instead
includes tissues with very different specific metabolic
activities (12, 22). Therefore, the use of LBM reaches
limitations when changes in body composition affect
the ratio of LBM to fat mass or, within the LBM
component, change the relative size of the various
tissues.
In theory, whole body BM can be estimated from the
sum of weight of tissues and organs multiplied by their
specific metabolic rate. Elia (12) has been able to determine the specific metabolism of the various tissues
and organs in vivo. Therefore, if one can assess body
composition precisely, it is theoretically possible to
predict BM with a much higher degree of accuracy than
from body weight or any other single component of
body weight. The potential efficacy of such an approach
to correct BM for body size and body composition has
been demonstrated in healthy human adults of normal
body weight (20). The present study aimed to more
precisely evaluate in the rat model the efficacy and
robustness of the use of predictive models, which include various parameters known to affect BM such as
whole body mass, height, LBM, fat mass, body surface,
age, and various combinations of these parameters to
normalize BM of subjects in different experimental
conditions and to account for the wide differences in
body size and body composition between subjects. For
that purpose, BM in adult male Wistar rats used as a
reference group was compared with BM in young
Wistar rats, food restricted/refed adult Wistar rats,
adult lean Zucker rats, adult obese Zucker rats, adult
Sprague-Dawley rats, and aged Sprague-Dawley rats.
METHODS
Animals
Seven groups of male rats were used in these experiments.
They were all fed a standard laboratory maintenance diet
(Extralabo M25, Pietrement, France, 3.2 kcal/g) and were
assigned to seven groups designed to measure the effect of
age, obesity and food restriction on BM. Group 1 comprised
nine Wistar rats weighing 270–520 g. This group was used as
the reference control group. Group 2 comprised seven Wistar
The costs of publication of this article were defrayed in part by the
payment of page charges. The article must therefore be hereby
marked ‘‘advertisement’’ in accordance with 18 U.S.C. Section 1734
solely to indicate this fact.
0363-6119/01 $5.00 Copyright © 2001 the American Physiological Society
R1887
Downloaded from http://ajpregu.physiology.org/ by 10.220.32.247 on June 18, 2017
Even, P. C., V. Rolland, S. Roseau, J.-C. Bouthegourd, and D. Tomé. Prediction of basal metabolism from
organ size in the rat: relationship to strain, feeding, age, and
obesity. Am J Physiol Regulatory Integrative Comp Physiol
280: R1887–R1896, 2001.—Use of the weight of various organs and tissues together with their specific metabolic activity for prediction of basal metabolism (BM) seems to be
promising. In this study we compared the use of this method
with those based on simple or multiple regression analyses.
We observed that 97.4% of differences in BM in a group of
nine adult male Wistar rats weighing 273–517 g could be
accounted for by changes in tissue and organ weights. BM
measured in lean Zucker and Sprague-Dawley rats did not
diverge from the prediction of the model by ⬎1.6%. According
to the organ-based model as well as multiple regression
analyses, but not simple regression analyses, BM was increased 18–21% in young rats, decreased 6–7% in food restricted/refed rats, and decreased 19–21% in aged rats. Only
with obese rats did the predictions of the two methods diverge. The main reason for this discrepancy seems to be the
way adipose tissue size and metabolism are taken into account.
R1888
BASAL METABOLISM AND BODY SIZE
Measurement of BM
BM was measured by indirect calorimetry using a metabolic device described in previous publications (13, 15). Temperature in the metabolic chamber was maintained at 26–
27°C. BM was recorded after the following procedure in all
rats. Food was removed from the home cage of the rat the
previous day at 1800. The rats were housed in the metabolic
chamber at 1200 with only 12 g of food available. With this
restricted amount of food in the cage and the previous overnight food restriction, the food allotment in the metabolic
cage was terminated before 0200. Thereafter basal metabolism decreased progressively and reached a stable value by
0700 in the morning. Recording of respiratory exchange was
terminated at 1100. BM was taken as the average resting
metabolism measured during the last 3 h of recording. Food
was not withdrawn from the food-restricted rats the day
before measurement.
Measurement of Body Composition
Body composition of the rats was measured by means of
dissection and weighing of the fresh weight of the main
organs and tissues of the body. Immediately after measurement of BM, the rat was weighed, anesthetized (pentobarbital sodium, 60 mg/kg, Sanofi Santé), and then exsanguinated
by removal of blood from the abdominal aorta. This led to
collection of 8–15 ml of blood depending on the weight of the
rat. Then the main organs and tissues were removed, blotted
dry, and immediately weighed to the nearest 0.1–0.01 g. The
main organs dissected and weighed were heart, liver, spleen,
kidney, brain, abdominal adipose tissues (mesenteric, retroperitoneal, epididymal), skin, subcutaneous abdominal adipose tissue, tail, and head. After completion of the dissection,
the remainder of the body, i.e., muscle mass plus skeleton
(excluding tail and head), was weighed and classified as
“carcass.” Muscle weight was computed from carcass weight
assuming that in all rats, skeleton accounted for 13.5% of
carcass weight. This value was calculated from direct measurement of skeleton weight in 10 carcasses of adult Wistar
rats.
For the purpose of analysis, three other components were
derived: organ and tissues with a high metabolic activity
(brain, liver, kidney, heart) were grouped under the component “High,” tissues with a low metabolic activity (adipose
depots, skin ⫹ subcutaneous fat, tail) were grouped under
the component “Low,” the residual organs that were not
individually weighed (lungs, testis, etc. ) were grouped under
the component residual (Res). LBM in this study was taken
as body weight minus the component Low.
RESULTS
Body Weight and Body Composition
Table 1 gives the weight of each of the main components of body weight in the different groups, in absolute terms as well as relative to LBM. Because large
differences in body weight in the various groups make
it difficult to compare directly the data in absolute
terms, the comparison in the following sections refers
to the data expressed relative to LBM.
Comparison in the three control groups (adult Wistar
rats, adult lean Zucker rats, and adult Sprague-Dawley
rats). Despite the fact that the three groups of rats
were adults of normal body weight, important differences in body composition were observed between
them. Compared with the Wistar rats, liver weight as
well as weight of the component High were lower in the
lean Zucker rats, whereas in contrast, muscle mass
was greater. In SD rats, the weights of the liver, brain,
kidney, and the component High were lower. Muscle
mass in SD rats also tended to be higher than in Wistar
rats, but the significance remained borderline (P ⫽
0.051). In contrast with the differences observed on the
metabolically more active organs, the weight of the
component Low (adipose tissue ⫹ skin ⫹ tail) was very
similar in the three groups (between 37 and 39% of
LBM).
Comparison between Wistar rats: young vs. adult and
food restricted vs. ad libitum fed. In young rats, body
composition included relatively less tissue with a low
metabolic activity (muscles and the component Low)
and more tissues with a higher metabolic activity (components Res and High) than in adult rats. In foodrestricted rats, the most important change was due to
a large decrease in the weight of the adipose tissue
leading also to a decrease in the weight of the component Low. In contrast, for all the other tissues and
organs, including muscle mass, no significant changes
were observed by comparison with ad libitum-fed rats.
Comparison between lean and obese Zucker rats.
Nearly all of the components of body composition differed between lean and obese Zucker rats. In obese
rats, the adipose stores, but also several of the body
components with a high metabolic activity (Res, High,
liver, heart), were enlarged. In contrast, muscle mass
was much smaller. Comparison of obese Zucker rats
with adult Wistar rats led to the same conclusions.
Comparison between old and adult Sprague-Dawley
rats. In this experiment, skin and subcutaneous adipose tissues were weighed together so that separate
values for adipose tissue and skin are not available in
Sprague-Dawley rats. Old Sprague-Dawley rats were
characterized by an increase in the weight of the component Low and a decrease in brain weight. Weights of
the other organs and tissues were not different. Comparison of the old Sprague-Dawley rats with the adult
Wistar rats led to the same conclusions.
Downloaded from http://ajpregu.physiology.org/ by 10.220.32.247 on June 18, 2017
rats initially weighing 233–304 g and submitted to 3–4 wk of
food restriction to ⬃70% of spontaneous food intake (17
g/days). Four of the rats were used while food restricted and
weighed 230–304 g, and three of the rats were used after 4
days of refeeding and weighed 400–410 g. This group was
designed to measure the effect of long-term mild food restriction and refeeding on BM. Group 3 comprised six young
Wistar rats weighing 180–225 g. This group was used to
compare BM between young growing rats and adult rats.
Group 4 had 10 lean Zucker rats weighing 290–380 g. This
group served as the control for group 5. Group 5 contained 14
obese Zucker rats weighing 500–750 g. This group was used
for measurement of BM in obese subjects. Group 6 had seven
adult Sprague-Dawley rats weighing 350–520 g to serve as
the control group for group 7. Group 7 was made up of six old
Sprague-Dawley rats aged ⬎18 mo and weighing 550–615 g.
This group was used for measurement of the effects of aging
on BM.
R1889
BASAL METABOLISM AND BODY SIZE
Table 1. Weight of the main components of body weight in the various groups
Wistar
(n ⫽ 9)
ZL
(n ⫽ 10)
SD
(n ⫽ 7)
Young
(n ⫽ 6)
FR
(n ⫽ 7)
ZO
(n ⫽ 14)
Old
(n ⫽ 7)
Expressed in g
Weight
AT
Muscle
Liver
Brain
Heart
Low
High
Res
337.34
(7.94)
66.15
(2.14)
165.13
(4.28)
8.04‡
(0.20)
1.93
(0.03)
0.97
(0.02)
2.28*
(0.06)
91.26
(3.27)
13.23†
(0.21)
41.73*
(1.38)
139.34
(1.70)
30.03
(1.27)
64.20
(0.41)
5.08
(0.27)
0.70
(0.05)
0.399
(0.017)
1.007
(0.029)
39.34
(1.70)
7.19
(0.29)
18.51
(0.42)
137.04
(0.83)
26.88*
(0.64)
67.05‡
(0.64)
3.28‡
(0.12)
0.79
(0.01)
0.396
(0.007)
0.929
(0.021)
37.04
(0.83)
5.39‡
(0.13)
17.00
(0.64)
410.47
(21.67)
—
—
194.79
(10.08)
12.44
(0.39)
1.64*
(0.02)
1.18
(0.04)
2.67
(0.10)
115.47
(9.82)
17.92
(0.51)
51.63
(1.21)
208.16‡
(5.92)
42.25‡
(1.69)
91.55‡
(3.54)
8.47†
(0.55)
1.75
(0.03)
0.74†
(0.04)
2.04†
(0.10)
53.90‡
(2.17)
13.00†
(0.64)
35.31†
(2.25)
326.50
(31.92)
62.11*
(6.56)
156.17
(15.67)
11.29
(1.21)
1.89
(0.07)
0.94
(0.13)
2.44
(0.36)
80.39*
(9.01)
16.56
(1.74)
48.81
(3.64)
550.66‡
(18.63)
153.08‡
(7.87)
150.65
(4.20)
17.45
(0.93)
1.92
(0.02)
1.20
(0.04)
2.73
(0.09)
283.58‡
(9.03)
23.31
(1.02)
69.41*
(6.42)
593.77‡
(17.28)
—
—
262.95‡
(9.14)
17.37
(0.98)
1.74
(0.11)
1.49†
(0.05)
3.76‡
(0.13)
195.14‡
(12.16)
24.37*
(1.13)
69.92†
(3.37)
132.37†
(0.95)
25.08†
(0.63)
63.26
(0.65)
4.56
(0.13)
0.79
(0.05)
0.37
(0.02)
0.97
(0.06)
32.37†
(0.95)
6.69
(0.14)
20.09
(0.83)
207.16‡
(3.26)
58.71‡
(4.16)
56.80‡
(1.07)
6.54‡
(0.23)
0.73
(0.03)
0.45†
(0.01)
1.03
(0.03)
107.16‡
(3.26)
8.76‡
(0.25)
25.49‡
(1.33)
149.17†
(3.28)
—
—
65.93*
(0.72)
4.36†
(0.20)
0.44‡
(0.04)
0.38
(0.01)
0.95
(0.03)
49.17†
(3.28)
6.12†
(0.24)
17.58
(0.82)
Expressed in percent of LBM
Weight
AT
Muscle
Liver
Brain
Heart
Kidney
Low
High
Res
138.86
(1.77)
—
—
65.90
(0.72)
4.24*
(0.14)
0.56*
(0.02)
0.40
(0.01)
0.91†
(0.01)
38.86
(1.77)
6.10†
(0.17)
17.63
(0.71)
134.92*
(0.97)
27.37
(0.70)
59.31†
(1.29)
5.49
(0.34)
1.14‡
(0.04)
0.48†
(0.02)
1.32†
(0.07)
34.92*
(0.97)
8.43*
(0.41)
22.92*
(1.46)
Values are means ⫾ SE (in parentheses). Weight of adipose tissue (AT) is not available in Sprague-Dawley (SD) and Old rats because
subcutaneous AT was weighed together with skin. LBM, lean body mass; ZL, Zucker lean; FR, food restricted; ZO, Zucker old; Low,
low-metabolic activity tissues; High, high-metabolic activity tissues; Res, residual organs. See METHODS for more complete descriptions of
groups. *P ⬍ 0.05; †P ⬍ 0.01; ‡P ⬍ 0.001 vs. Wistar (unpaired t-test).
Comparison of Basal Metabolism Between the Group
of Control Wistar Rats and the Other Groups
Normalization of basal metabolism according to body
weight, muscle mass, LBM, High, and Res. The correlation between BM and body weight established in the
group of control Wistar rats was used to predict BM
expected in the other rats according to their own
weight. The differences between the predicted and
measured BM values were then computed for each
group (Fig. 1A and Table 2). This study showed that
measured BM values in the SD control rats were close
to those predicted according to their body weight
(⫺3.56%). In contrast, measured BM values were
larger than predicted in young Wistar rats (⫹30.53%)
and lower than predicted in the lean Zucker rats
(⫺10.77%) as well as in the food-restricted Wistar rats
(⫺5.93%), the obese Zucker rats (⫺27.25%), and the old
Sprague-Dawley rats (⫺26.19%).
Following the same procedure, BM was compared
between the Wistar rats and the other groups after
adjustment of BM with various parameters of body
weight (muscle mass, LBM, High, Res) The results of
these studies are summarized in Table 2 and Fig. 1,
B-E. They showed that BM values derived in the various groups from the above simple regression analyses
differed depending on which body component was used
to adjust BM. As a result, taking BM in the adult
Wistar rats as reference, adjusted BM varied from
⫹7.4 to ⫺15.8% in lean Zucker rats, from ⫹7.5 to
⫺7.1% in control Sprague-Dawley rats, from ⫹2.1 to
Downloaded from http://ajpregu.physiology.org/ by 10.220.32.247 on June 18, 2017
Kidney
390.1
(30.32)
83.17
(5.67)
180.15
(14.77)
14.31
(1.39)
1.89
(0.09)
1.13
(0.12)
2.82
(0.23)
109.79
(8.64)
20.16
(1.73)
51.75
(4.05)
R1890
BASAL METABOLISM AND BODY SIZE
⫹30.5% in young rats, from ⫺16.3 to ⫹9.5% in foodrestricted rats, from ⫺27.2 to ⫹4.0% in obese Zucker
rats, and from ⫺4.8 to ⫺26.2% in aged Sprague-Dawley rats (Fig. 2). This large variability was due to the
differences in the weights of the tissues and organs
relative to each other in the various groups. For example, obese Zucker rats have a much higher weight of
the Res component relative to LBM than adult Wistar
Table 2. Differences between predicted and observed BM in the various groups resulting from the parameters
of the simple linear regression analysis performed in control Wistar rats
ZL
Weight
Muscle
LBM
High
Res
⫺10.769
(2.17)‡
⫺15.827
(2.122)‡
⫺12.625
(2.005)‡
7.447
(2.56)*
⫺5.966
(3.558)
SD
Young
FR
⫺3.556
(2.23)
⫺7.091
(2.315)*
⫺4.669
(1.924)
7.551
(3.451)
⫺2.636
(2.401)
30.528
(2.865)‡
28.032
(3.318)‡
22.669
(2.63)‡
4.814
(1.976)
2.090
(6.058)
⫺5.927
(2.64)
⫺9.956
(2.424)†
⫺10.835
(2.762)†
⫺7.341
(2.471)*
⫺16.318
(5.289)*
ZO
⫺27.253
(1.938)‡
17.315
(3.425)‡
5.620
(2.947)
⫺10.556
(2.005)†
⫺18.488
(4.81)†
Old
⫺26.188
(2.978)‡
⫺22.528
(2.061)‡
⫺19.600
(2.466)‡
⫺4.813
(5.995)
⫺13.439
(6.681)
Values are means ⫾ SE (in parentheses) in %. BM, basal metabolism. * P ⬍ 0.05; † P ⬍ 0.01; ‡ P ⬍ 0.001 vs. Wistar (unpaired t-test).
Downloaded from http://ajpregu.physiology.org/ by 10.220.32.247 on June 18, 2017
Fig. 1. Relationship between basal metabolism (BM) and weight of the main
tissues and organs in the various
groups. The regression line is established with the data of the control
Wistar rats only. It provides an indication of the tightness of the relationship
between organ weight and BM in the
control group. Accordingly, the points
located above the regression line indicate that measured BM was larger
than predicted by the model and vice
versa. Significance of differences between measured and predicted values
of BM in the various groups are reported in Table 2. ZL, Zucker lean rats;
SD, Sprague-Dawley rats; FR, food-restricted Wistar rats; Old, old SpragueDawley rats; Young, young Wistar
rats; Obese, Zucker obese rats.
R1891
BASAL METABOLISM AND BODY SIZE
rats and a smaller muscle mass. Thus when BM was
adjusted relative to Res, BM seemed to be reduced
(⫺18.5%). In contrast, when BM was adjusted relative
to muscle mass, BM seemed to be strongly increased
(⫹17.31%). The same kind of error occurred in the
other groups with the source of error depending on
which component of body weight was increased or
decreased in the group.
These results demonstrate that it is essential to take
into account the relative weights of the main components of total body weight to correct BM for changes in
body composition adequately. The aim of the next
study was to use this methodological approach.
Normalization of BM according to organ-tissue
weights and to their specific metabolic activity. Inasmuch as no full set of data was available to assign a
specific metabolic activity to the various tissues and
organs of the rats, we adjusted for the rat the data
proposed by Elia (12) for humans. This was done by
multiplying the specific metabolic activity of the various tissues and organs in humans by a constant that
allowed the best fit for the value of BM predicted from
the weight of the tissues and organs to the BM values
actually measured in the group of adult Wistar rats.
The multiplicative factor that allowed the best fit of the
organ-based predictive model to the experimental results was adjusted empirically by a “trial and error”
approach and was found to be 4.879 (Table 3).
Figure 3 shows the relative weight and energy production of the four main components of body weight
predicted by the model in the different groups. This
figure shows the large differences in size of the various
tissues between the various groups and the large differences that can occur between tissue size and energy
expenditure.
After multiplication of the specific activity of the
organs as established in humans by 4.879, it appeared
Table 3. Organ and tissue contributions
to metabolic rate
Brain
Liver
Heart
Kidney
Muscle
AT
Res
Human
Rat
1,004.64
837.20
1,841.84
1,841.84
54.42
18.84
50.23
4,901.74
4,084.78
8,986.52
8,986.52
265.51
91.91
490.17
A correction to Rat Res has been made to account for the fact that
in the model of Gallagher et al. (20), skin and skeleton, which both
have a low specific metabolic activity, were included in the component Res of body weight, but in this study they were included in the
components “carcass” and Low, respectively. Because in the model of
Gallagher et al., skeleton and skin amount to 47.9% of total residual
mass, in this study we increased the specific activity of the component Res from 295.08 to 566.37.17 j/g/day [295.08/(1–0.479)]. Data in
humans are extracted from Gallagher et al. according to Elia (12).
Values in rats have been adjusted empirically to data in humans
times 4.879 (see METHODS).
Downloaded from http://ajpregu.physiology.org/ by 10.220.32.247 on June 18, 2017
Fig. 2. Summary of the differences observed in the various groups
between predicted and observed values of BM (dBM) according to the
reference used to normalize BM. The control group is made up adult
Wistar rats for the parameters weight, muscle, residual organs not
individually weighted (Res), tissues with high metabolic activity
(High), and lean body mass (LBM). For Stepwise regression analyses, lean Zucker (ZUC) rats and adult Sprague-Dawley rats have
been included in the control group. REST, restricted.
that the organ-based model accounted for 97.4% of the
differences of BM among adult Wistar rats (Fig. 4). In
the other groups, BM actually measured in these rats
diverged from the prediction of the model by only
⫺0.187% (P ⫽ 0.93) in lean Zucker rats and 1.570%
(P ⫽ 0.48) in SD rats (Table 4). Thus, according to this
procedure of normalization of BM, all three groups of
adult rats, Wistar, Sprague-Dawley, and lean Zucker,
appeared to have the same BM. In contrast, BM appeared to be increased in young Wistar rats only and
appeared decreased in all the other experimental
groups (Table 4).
Normalization of BM by stepwise multiple regression
analysis: adjustment of the model. As the organ-based
model showed that adult Wistar, lean Zucker, and
Sprague-Dawley rats could be considered to have a
similar BM per unit of metabolically active mass, we
pooled the rats of these three groups into a single
larger group of 26 rats from which we tested the possibility of predicting BM from stepwise regression
analyses. We considered it useful to attempt this study
to test 1) whether a statistical model could validate the
incorporation of the Wistar, lean Zucker rats, and
Sprague-Dawley control rats into a single homogenous
group and 2) whether the differences among groups
observed after this statistical procedure agreed with
the differences reported from the organ-based model.
Stepwise regression analysis using BM as the dependent variable and Low, muscle, liver, brain, heart,
kidney, and Res as independent variables led to a
model that finally included only muscle and liver as
significant predictors of BM (Table 4). According to this
model, BM measured in the three control groups clustered tightly along the regression line, confirming that
BM in these three groups was the same after differences in muscle and liver weight were taken into account (Fig. 5A, Table 4). Among the experimental
groups, the statistical model also closely matched the
conclusions drawn from the organ-based model except
in the obese Zucker rat in which the stepwise regression analysis suggested that BM was normal (⫹0.27%),
R1892
BASAL METABOLISM AND BODY SIZE
Fig. 3. Weight and basal metabolism
of the various tissues and organs in the
different groups. Data are presented in
absolute terms (top) or relative to total
weight and basal metabolism (bottom).
Fig. 4. Prediction of BM (W) from the weight (g) of the various
tissues and organ according to the following model: BM ⫽
{(Low 䡠 0.000219907) ⫹ [Carcass 䡠 (0.000636574 䡠 0.865)] ⫹ (liver 䡠
0.00972222) ⫹ (brain䡠0.011574074) ⫹ (heart䡠0.02138889) ⫹ (kidney䡠
0.02138889) ⫹ [Miscellaneous 䡠 (0.0011)]} 䡠 4.8791. The regression line
indicates the correlation between predicted and measured BM values in control Wistar rats. Accordingly, the points located above the
regression line indicate that measured BM was larger than predicted
by the model and vice versa. Significance of differences between
measured and predicted values of BM in the various groups are
reported in Table 4.
derived from the previous stepwise regression analysis.
The results of the stepwise regression analysis thus
confirmed that the three control groups could be considered to have the same BM after differences in body
composition were taken into account, but raised the
question of why the two methods diverged so strongly
only with the obese Zucker rats. We suspected that the
discrepancy was due to the fact that in obese Zucker
rats the large increase in body weight is primarily
related to the accumulation of subcutaneous adipose
tissue (Table 1) (6).
Enlargement of the adipose stores in the obese
Zucker rats is the result of both hyperplasia and hypertrophy of the adipose cells (5, 7, 29, 51). Hypertrophy of adipose cells is mostly due to the accumulation
of inert lipid droplets, increasing adipocyte size by an
average of 4.3 times (7) and therefore decreasing its
specific metabolic activity by the same order of magnitude. Because in the obese Zucker rats average mass of
the adipose tissue was 153.08 ⫾ 7.87 g, whereas the
average mass of the overall Low component was
283.58 ⫾ 9.03 g, we estimated that the average metabolic activity of the component Low in obese rats was
2.3 times smaller than in control rats. By introducing
this correction in the organ-based model, we observed
that the decrease in BM was reduced from 13.3 to
6.84% (Table 4). Despite the fact that the difference
remained significant, this correction reduced the discrepancy observed between the stepwise regression
analysis procedures and the organ-based method.
DISCUSSION
The inability to normalize BM by simply correcting
its value according to body weight soon appeared to be
related to the fact that in many cases, body composition
changes together with either body size or age (2, 16, 23,
Downloaded from http://ajpregu.physiology.org/ by 10.220.32.247 on June 18, 2017
whereas the organ-based model suggested that BM
was decreased (⫺13.31%; Table 4, Fig. 4 vs. 5).
Stepwise regression analyses was also used to test a
simplified model using only the four main components
of body weight, i.e., muscle, High, Low, Res. This led to
a model that finally included only muscle and High as
significant predictors of BM (Fig. 5B, Table 4). This
model led to results that were very close to the ones
R1893
BASAL METABOLISM AND BODY SIZE
Table 4. Differences between predicted and observed BM in the various groups resulting from the prediction of
BM based on the specific metabolic activity of the various tissues and organs
Organ-based model
Stepwise
Muscle-liver
Muscle-high
Organ-based adjusted model
Wistar
ZL
SD
Young
FR
ZO
Old
0.000
(1.393)
⫺0.187
(2.234)
1.570
(2.250)
21.587
(1.611)‡
⫺6.034
(2.259)*
⫺13.308
(1.987)‡
⫺19.233
(2.438)‡
0.561
(0.861)
0.527
(0.902)
—
—
⫺1.580
(2.133)
⫺2.093
(2.139)
—
—
0.265
(2.482)
1.130
(2.483)
—
—
19.844
(0.570)‡
18.448
(0.638)‡
—
—
⫺6.684
(2.228)*
⫺6.791
(2.013)*
—
—
0.267
(2.472)
⫺0.177
(2.452)
⫺6.84
(2.242)†
⫺21.492
(4.281)†
⫺21.194
(4.546)†
—
—
Values are means ⫾ SE (in parentheses) in %. *P ⬍ 0.05; †P ⬍ 0.01; ‡P ⬍ 0.001 vs. Wistar (unpaired t-test).
must be taken into account, but also that FFM should
be subdivided into its main components (21, 37, 43, 50,
59). The main goal of the present study was to verify
whether such a method 1) was usable in rats and 2)
was able to deal with changes in body composition that
occur in relation to various parameters such as strain,
age, feeding status, and obesity. The results showed
that only the methods taking into account the weight of
various organs and tissues of the body seemed to be
able to provide correct adjustment of BM between the
various groups. Indeed the relationship between BM
and body weight in adult Wistar rats was linear and
very tightly correlated within a given interval of body
weight, but the data from young rats, older rats, and
food-restricted rats were far apart in this relationship.
The same was true for obese Zucker rats in which BM
appeared excessively low after correction for body
weight.
Specific Metabolism of the Various Tissues
and Organs
Fig. 5. Prediction of BM (W) from the weight (g) of the various
tissues and organ according to the results of stepwise regression
analyses. A, initial model; B, simplified model. The regression line
indicates the correlation between predicted and measured BM values in the reference group made up the control Wistar rats, the lean
Zucker rats, and the adult Sprague-Dawley rats. Accordingly, the
points located above the regression line indicate that measured BM
was larger than predicted by the model and vice versa. Significance
of the average differences between measured and predicted values of
BM in the various groups are reported in Table 4.
To adjust in the rat the specific metabolic activity of
the various tissues and organs established in humans,
we proposed the hypothesis that the metabolic rate of
the various tissues relative to each other was the same
in rats and in humans. This hypothesis was made
possible because the data available in various animal
species ranging from the mouse to the horse show that
the metabolism of the organs relative to each other
does not exhibit any specific metabolic change with
body size (from H. A. Krebs, 1950, quoted and discussed in Ref. 30). Accordingly, we observed that it was
possible to accurately predict BM in adult Wistar rat
by increasing 4.879 times the specific metabolic activities measured in humans. It is interesting to note that
4.879 is the increase in metabolic rate that, according
to the Brody-Kleiber interspecific correlation (32), accounts for a decrease in body weight from 70 to 0.125
kg. This coefficient is even closer to the theoretical
value expected if, as recommended by Heusner (25),
instead of the 0.75 Brody-Kleiber exponent, surface
area, i.e., the 0.667 exponent, is used. Indeed, in this
case, a 4.879 increase in the specific metabolic activity
of the organs and tissues is explained by a decrease in
body weight from 70 to 0.6 kg. The 4.879 coefficient
empirically established in this study thus appears to
Downloaded from http://ajpregu.physiology.org/ by 10.220.32.247 on June 18, 2017
37, 53, 57), thus altering the proportions of the contribution to energy production of the various tissues.
Major research groups thus consider that, for predicting basal metabolism accurately, fat mass and FFM
R1894
BASAL METABOLISM AND BODY SIZE
Basal Metabolism and Food Restriction, Aging,
or Obesity
Food restriction. Whole body BM declines with food
restriction. The question remains whether BM is still
decreased if changes in body size and body composition
are adequately taken into account, and no definitive
answers have been given (3, 14, 27, 35, 36). In the
present study, BM appeared more reduced after adjustment for LBM (⫺10.8%) than for body weight
(⫺5.9%), certainly because LBM decreased less than
whole body weight. Adjustment according to the organbased model also suggested that BM was decreased
after food restriction. The decrease (6%) was similar to
that suggested after correction for whole body weight
and after correction by stepwise regression analyses
(⫺6.7 to ⫺6.8%). The convergence of the results
brought by these different methods is not surprising
because the main tissue affected by food restriction
was the adipose one whose metabolic activity accounted for only a very small part of BM. Taken together, the present results confirm that BM is indeed
slightly reduced during food restriction. In addition,
because the decline in BM (measured according to the
organ-based method) was similar in the four rats used
while food restricted (⫺5.4 ⫾ 3%) and the three rats
that were previously refed during 4 days (⫺6.87 ⫾
1.85%), the present results confirm recent observations
that the decrease in BM induced by food restriction is
maintained several days during refeeding (11, 34).
Age. The fact that BM can be affected by age is
shown by the many equations derived for use in humans that include the parameter age (1, 58). The
changes in body composition throughout the span of
life are complex, involving at various degrees all the
tissues and organs of the body. This clearly explains
why adjustments of BM founded on one single component of body weight led to such variable results. In
contrast, because adjustment of BM from organ size
was able to account completely for complex differences
in body composition between Wistar, lean Zucker, and
Sprague-Dawley rats, it is clear that it was best fitted
to take into account the complex changes in body composition from youth to aging. Thus the present study
suggests with reasonable confidence that BM is increased in young growing subjects and decreased in
aged but healthy subjects. This confirms an array of
previous results (4, 18, 42, 43) and further suggests two
points. First, it seems that the decrease of BM from
youth to adulthood is rapid because the difference in
body weights and LBMs between the older young rats
and the younger adults was small. Such an observation
was also done in humans in the study of Bitar and
colleagues (4). Second, it seems that the decrease in
BM during aging is more progressive. Indeed, in this
study, five of the aged rats were taken from retired
reproducers (Iffa-Credo, 69592 l’Arbresle Cédex,
France) and were 18–20 mo old. In these five rats, BM
assessed from the organ-based model appeared reduced by 16 ⫾ 1%. The two other rats, remaining from
a previous experiment done in the laboratory, were ⬎2
years old, and one of them was already losing some
weight. In these rats, the decrease in BM was larger,
respectively, 24.6 and 30%. Therefore the decrease in
BM observed during aging is the result of a decrease in
both the weight of the metabolically active tissues and
their specific metabolic activity. Which organ and tissues are more particularly involved remains to be investigated.
Obesity. It is repeatedly suggested that a decrease in
BM may be an important factor leading to obesity (24,
44, 47). The difficulty of correcting BM for body size
and composition in obese subjects is primarily due to
the fact BM must be corrected more specifically in
relation to the hypertrophy of the adipose tissue,
whereas muscle mass and, more generally, lean tissues
are relatively less affected. It is clear thus that any
adjustment of BM according to total body weight or
LBM will introduce large errors. This was clearly demonstrated in this study where, depending on the reference used, BM in obese rats ranged from ⫺27% (correction according to body weight) to ⫹5.6% (correction
according to LBM). The other problem that was revealed in this study is that in obesity adipocytes are
hypertrophied. This phenomenon induces a decrease in
the specific metabolic activity of the adipose tissue. In
this study, we attempted to take this problem into
account by assuming that the specific metabolic activity of the adipose tissue was reduced 4.3 times in the
obese Zucker rats. By doing so we reduced the differences with the stepwise regression analyses, but it is
clear that to be able to establish a really precise measure of BM in obese Zucker rats, specific studies should
be undertaken to get precise information about the
specific metabolic activity of the various tissues, not
only adipose tissue, but also liver, muscles, and skin
that may include more fat than in control animals.
Downloaded from http://ajpregu.physiology.org/ by 10.220.32.247 on June 18, 2017
account well for the size difference between humans
and rats. The other result in favor of the physiological
validity of the procedure is the fact that BM predicted
from organ size did not differ from the BM actually
measured in all three strains of adult rats fed ad
libitum. Such a coincidence observed simultaneously
between predicted and measured BM values in the
three strains could not be observed with any of the
methods that used one single component of body
weight to normalize BM but was confirmed by stepwise
regression analysis. Finally, the capacity of this
method to deal with changes in various components of
body weight is indicated by the fact that lean Zucker
and Sprague-Dawley rats differed in their body composition relative to Wistar rats in different ways, lean
Zucker rats having a relatively larger muscle mass and
a smaller liver mass, Sprague-Dawley rats having no
differences at the level of muscle mass, but smaller
brains, livers, and kidneys. The improvements brought
by the ability to predict BM from the size and weight of
the main organs and tissues constituting the body offer
the possibility to shed some new light on several longstanding questions on whether or not BM changes
during food restriction, aging, and obesity.
BASAL METABOLISM AND BODY SIZE
Nevertheless, the data presently provided by the organ-based model as well as by the stepwise regression
analysis converge to suggest that BM in the obese
Zucker rat is probably not reduced as much as previously thought. Such a result is in line with the most
recent reports (24, 56) and questions the hypothesis
that a reduced basal rate of energy expenditure may
participate in the etiology of obesity.
Conclusion and Perspectives
REFERENCES
1. Arciero PJ, Goran MI, Gardner AM, Ades PA, Tysbir RS,
and Poehlman ET. A practical equation to predict resting
metabolic rate in older females. J Am Geritr Soc 41: 389–395,
1993.
2. Azain MJ, Hausman DB, Kasser TR, and Martin RJ. Effect
of somatotropin and feed restriction on body composition and
adipose metabolism in obese Zucker rats. Am J Physiol Endocrinol Metab 269: E137–E144, 1995.
3. Bessard T, Schutz Y, and Jéquier E. Energy expenditure and
post-prandial thermogenesis in obese women after weight loss.
Am J Clin Nutr 36: 680–694, 1983.
4. Bitar A, Vermorel M, Fellman N, and Coudert J. Twentyfour-hour energy expenditure and its components in prepubertal
children as determined by whole-body indirect calorimetry and
compared with young adults. Am J Clin Nutr 62: 308–315, 1995.
5. Bray GA. Studies on the composition of adipose tissue from the
genetically obese rats. Proc Soc Exp Biol Med 131: 1111–1114,
1969.
6. Bray GA. The Zucker fatty rat: a review. Fed Proc 36: 148–153,
1977.
7. Bray GA, York DA, and Swerloff RS. Genetic obesity in rats.
I. The effects of food restriction on body composition and hypothalamic function. Metabolism 22: 435–442, 1973.
8. Brody S. Bioenergetics and Growth. New York: Reinhold, 1945.
9. Butte NF, Moon JK, Wong WW, Hopkinson JM, and Smith
EO. Energy requirements from infancy to adulthood. Am J Clin
Nutr 62, Suppl 5: 1047S–1052S, 1995.
10. Cunningham JJ. Body composition as a determinant of energy
expenditure: a synthetic review and a proposed general prediction equation. Am J Clin Nutr 54: 963–969, 1991.
11. Dulloo AG and Calokatisa R. Adaptation to low calorie intake
in obese mice: contribution of a metabolic component to diminished energy expenditure during and after weight loss. Int J
Obes 15: 7–16, 1991.
12. Elia M. Organ and tissue contribution to metabolic rate. In:
Energy Metabolism, Tissue Determinants and Cellular Corollaries. New York: Raven, 1992, p. 61–77.
13. Even PC, Mokhtarian A, and Pélé A. Practical aspects of
indirect calorimetry in laboratory animals. Neurosci Biobehav
Rev 18: 435–447, 1994.
14. Even PC and Nicolaı̈dis S. Adaptive changes in energy expenditure during mild and severe feed restriction in the rat. Br J
Nutr 70: 421–431, 1993.
15. Even PC, Perrier E, Aucouturier JL, and Nicolaı̈dis S. The
use of the method of Kalman Filtering for the on-line computation of background metabolism in the rat. Physiol Behav 49:
177–187, 1991.
16. Foster GD, Wadden TA, Mullen JL, Stunkard AJ, Wang J,
Feurer ID, Pierson RN, Yang MU, Presta E, Van Itallie TB,
Lemberg PS, and Gold J. Resting energy expenditure, body
composition, and excess weight in the obese. Metabolism 37:
467–472, 1988.
17. Frield KE, DeLuca JP, Marchitelli LJ, and Vogel JA. Reliability of body-fat estimations from a four compartment model
by using density, body water and bone mineral measurements.
Am J Clin Nutr 55: 764–770, 1992.
18. Fukagawa NK, Bandini LG, and Young JB. Effect of age on
body composition and resting metabolic rate. Am J Physiol
Endocrinol Metab 259: E233–E238, 1990.
19. Fuller NJ, Sawyer MB, and Elia M. Comparative evaluation
of body composition methods and predictions, and calculation of
density and hydration fraction of fat-free mass, in obese women.
Int J Obes 18: 503–512, 1994.
20. Gallagher D, Belmonte D, Deurenberg P, Wang Z, Krasnow N, Pi-Sunyer FX, and Heymsfield SB. Organ-tissue
mass measurement allows modeling of REE and metabolically
active tissue mass. Am J Physiol Endocrinol Metab 275: E249–
E258, 1998.
21. Gallagher D, Visser M, Wang Z, Harris T, Pierson RN Jr,
and Hemsfield SB. Metabolically active component of fat-free
body mass: influence of age, adiposity and gender. Metabolism
45: 992–997, 1996.
22. Garby L and Lammert O. Between-subjects variation in energy expenditure: estimation of the effect of variation in organ
size. Eur J Clin Nutr 48: 376–378, 1994.
23. Going S, Williams D, and Lohman T. Aging and body composition: biological changes and methodological issues. Exerc Sport
Sci Rev 23: 411–458, 1995.
24. Goran MI. Energy metabolism and obesity. Med Clin North Am
84: 347–362, 2000.
25. Heusner AA. Energy metabolism and body size. I. Is the O.75
mass exponent of Kleiber’s equation a statistical artifact? Respir
Physiol 48: 1–12, 1982.
Downloaded from http://ajpregu.physiology.org/ by 10.220.32.247 on June 18, 2017
The present study reveals the potential of the use of
the weights and specific metabolic activities of the
organs and tissues constituting the body for predicting
BM and in this way confirm and extend the observation
of Gallagher et al. (20). This method appears to be able
to deal with large changes in body size and body composition as long as it can be assumed that no significant
changes occur in the composition of the various organs
and tissues making up the body. This limitation was
found for the obese rats of the present study in which
the extremely large variations in the size and composition of the adipose tissue challenged the method. The
same may be true in other experimental conditions or
pathologies in which the level of hydration of the cells
or the composition of specific organs such as lipid
inclusion in muscular dystrophy or steatosis due to
cirrhosis, ketogenic diets, or liver diseases may vary.
One possible solution to this limitation of the method
may be to use additional or alternative parameters
such as the dry weight of the tissues, the weight of the
defatted tissue, or the DNA content per unit of weight
of tissue. With these improvements, and possibly with
further adjustments of the specific metabolic activity
attributed to each organ, there is no doubt that this
method will soon allow us to perform appropriate adjustments of BM whatever the differences in body size
and composition. A precise measurement of organ and
tissue sizes can be done from carcass analysis in animals, as in the present experiment, but can also be
done in humans with noninvasive methods of assessment of body composition such as computed axial tomography and magnetic resonance imaging (20, 26).
Improvements in methods allowing for noninvasive
measurement of body composition (26) and a better
knowledge of the specific metabolic activity in situ of
the various tissues and organs (12) now offer additional parameters permitting us to construct models
able to predict BM in subjects with large differences in
body weight, body composition, age, and sex.
R1895
R1896
BASAL METABOLISM AND BODY SIZE
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
Sutherland PT. Determinants of decline in resting metabolic
rate in aging females. Am J Physiol Endocrinol Metab 264:
E450–E455, 1993.
Ravussin E. Energy metabolism in obesity. Studies in the Pima
Indians. Diabetes Care 16: 232–238, 1993.
Refinetti R. Computation of the effective body mass for metabolic studies of lean and obese rats. Metabolism 38: 763–766,
1989.
Roberts SB, Fuss P, Heyman MB, and Young VR. Influence
of age on energy requirements. Am J Clin Nutr 62, Suppl 5:
1053S-1058S, 1995.
Rothwell NJ and Stock MJ. Luxuskonsumption, diet-induced
thermogenesis and brown fat: the case in favour. Clin Sci (Colch)
64: 29–23, 1983.
Schultz AR. Energy metabolism in the whole animal revisited.
Respir Physiol 73: 11–20, 1988.
Seale JL, Rumpler WV, and Moe PW. Description of a directindirect room-sized calorimeter. Am J Physiol Endocrinol Metab
260: E306–E320, 1991.
Sparti A, DeLany JP, de la Bretonne JA, Sander GE, and
Bray GA. Relationship between resting metabolic rate and the
composition of the fat-free mass. Metabolism 46: 1225–1230,
1997.
Stern JS and Greenwood MR. A review of development of
adipose cellularity in man and animals. Fed Proc 33: 1952–1955,
1974.
Sun M, Reed GW, and Hill JO. Modification of a whole room
indirect calorimeter for measurement of rapid changes in energy
expenditure. J Appl Physiol 76: 2686–2691, 1994.
Vaughan L, Zurlo F, and Ravussin E. Aging and energy
expenditure. Am J Clin Nutr 53: 821–825, 1991.
Verga S, Buscemi S, and Caimi G. Resting energy expenditure and body composition in morbidly obese, obese and control
subjects. Acta Diabetol 31: 47–51, 1994.
Weinsier RL, Schutz Y, and Bracco D. Reexamination of the
relationship of resting metabolic rate to fat-free mass and to the
metabolically active components. Am J Clin Nutr 55: 790–794,
1992.
Weyer C, Snitker S, Bogardus C, and Ravussin E. Energy
metabolism in African Americans: potential risk factors for obesity. Am J Clin Nutr 70: 13–20, 1999.
Wolden-Hanson T, Marck BT, Smith L, and Matsumoto
AM. Cross-sectional and longitudinal analysis of age-associated
changes in body composition of male brown Norway rats: association of serum leptin levels with peripheral adiposity. J Geront
54A: B99–B107, 1999.
Wong WW, Butte NF, Hergenroeder AC, Hill RB, Stuff JE,
and Smith EO. Are basal metabolic rate equations appropriate
for female children and adolescents? J Appl Physiol 81: 2407–
2414, 1996.
Zurlo F, Larson K, Bogardus C, and Ravussin E. Skeletal
muscle metabolism is a major determinant of resting energy
expenditure. J Clin Invest 86: 1423–1427, 1990.
Downloaded from http://ajpregu.physiology.org/ by 10.220.32.247 on June 18, 2017
26. Heymsfield SB, Wan Z, Baumgartner RN, and Ross R.
Human body composition: advances in models and methods.
Annu Rev Nutr 17: 527–558, 1997.
27. Hill JO, Latiff A, and DiGirolamo M. Effects of variable
caloric restriction on utilization of ingested energy in rats. Am J
Physiol Regulatory Integrative Comp Physiol 248: R549–R559,
1985.
28. Holliday MA. Metabolic rate and organ size during growth from
infancy to maturity and during late gestation and early infancy.
Pediatrics 47: 167–177, 1971.
29. Johnson PR, Zucker LM, Cruce JA, and Hirsch J. Cellularity of adipose depots in the genetically obese Zucker rat. J Lipid
Res 12: 706–714, 1971.
30. Kayser C and Heusner A. Etude comparative du métbolisme
énergétique dans la série animale. J Physiol (Paris) 56: 489–524,
1964.
31. Klausen B, Toubro S, and Astrup A. Age and sex effects on
energy expenditure. Am J Clin Nutr 65: 895–907, 1997.
32. Kleiber M. Body size and metabolic rate. Physiol Rev 7: 511–
541, 1947.
33. Kraemer WJ, Volek JS, Clark KL, Gordon SE, Incledon T,
Puhl SM, Triplett-McBrid NT, McBride JM, Putukian M,
and Sebastianelli WJ. Physiological adaptations to a weight
loss dietary regimen and exercise programs in women. J Appl
Physiol 83: 270–279, 1997.
34. Luz J and Griggio MA. Food restriction and refeeding in
growing rats. Comp Biochem Physiol 99A: 477–481, 1991.
35. McCarter RJ and McGee JR. Transient reduction of metabolic rate by food restriction. Am J Physiol Endocrinol Metab
257: E175–E179, 1989.
36. Nelson KM, Weisner RL, James LD, Darnell B, Hunter G,
and Long CL. Effect of weight reduction on resting energy
expenditure, substrate utilization, and the thermic effect of food
in moderately obese women. Am J Clin Nutr 55: 924–933, 1992.
37. Nelson KM, Weinsier RL, Long CL, and Schutz Y. Prediction of resting energy expenditure from fat-free mass and fat
mass. Am J Clin Nutr 56: 848–856, 1992.
38. Nevill AM. The need to scale for differences in body size and
mass: an explanation of Kleiber’s 0.75 mass exponent. J Appl
Physiol 77: 2870–2873, 1994.
39. Nevill AM, Ramsbottom R, and Williams C. Scaling physiological measurements for individuals of different body size. Eur
J Appl Physiol 65: 110–117, 1992.
40. Noblet J, Dourmad JY, Etienne M, and Le Dividich J.
Energy metabolism in pregnant sows and newborn pigs. J Anim
Sci 75: 2708–2714, 1997.
41. Ortigues I, Martin C, Vermorel M, and Anglaret Y. Energy
cost of standing and circadian changes in energy expenditure in
the preruminant calf. J Anim Sci 72: 2131–2140, 1994.
42. Pace N, Rahlmann DF, and Smith AH. Scaling of metabolic
rate on body mass in small laboratory mammals. Physiologist 23:
115–116, 1980.
43. Poehlman ET, Goran MI, Gardner AW, Ades PA, Arciero
PJ, Katzman-Rooks SM, Montgomery ST, Toth MJ, and