See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/7488665

Factors influencing variation in basal metabolic rate include fat-free mass,

fat mass, age, and circulating thyroxine but not sex, circulating leptin, or
triiodothyronine

Article  in  American Journal of Clinical Nutrition · December 2005

DOI: 10.1093/ajcn/82.5.941 · Source: PubMed

CITATIONS READS

296 4,162

5 authors, including:

Alexandra M Johnstone Sandra D Murison

University of Aberdeen University of Aberdeen
160 PUBLICATIONS   7,775 CITATIONS    16 PUBLICATIONS   906 CITATIONS   

SEE PROFILE SEE PROFILE

Kellie Anne Watson John Speakman

The Roslin Institute Institute of Genetics and Developmental Biology, CAS
80 PUBLICATIONS   2,115 CITATIONS    751 PUBLICATIONS   32,397 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Impact of plant anatomy on forage quality View project

Whey proteins and tissue growth View project

All content following this page was uploaded by John Speakman on 06 June 2014.

The user has requested enhancement of the downloaded file.

Factors influencing variation in basal metabolic rate include fat-free
mass, fat mass, age, and circulating thyroxine but not sex,
circulating leptin, or triiodothyronine1–3
Alexandra M Johnstone, Sandra D Murison, Jackie S Duncan, Kellie A Rance, and John R Speakman

ABSTRACT cross-sectional studies often report that formerly obese subjects

Background: Basal metabolic rate (BMR) is the largest component have BMRs that are 3–5% lower than expected (11). However,
of daily energy demand in Western societies. Previous studies indi- although some long-term studies have indicated that variation in

Downloaded from www.ajcn.org at University of Aberdeen on November 22, 2005

cated that BMR is highly variable, but the cause of this variation is BMR is associated with subsequent weight gain (12), other lon-
disputed. All studies agree that variation in fat-free mass (FFM) gitudinal studies have failed to replicate this effect (13, 14).
plays a major role, but effects of fat mass (FM), age, sex, and the Second, BMR may show an adaptive response to caloric restric-
hormones leptin, triiodothyrionine (T3), and thyroxine (T4) remain tion, which may predispose individuals to subsequent weight
uncertain. regain. Many studies have called into question whether such
Objective: We partitioned the variance in BMR into within- and changes reflect metabolic adaptation or simply disproportional
between-subject effects and explored the roles of FFM, FM, bone losses of metabolically active FFM (15–18).
mineral content, sex, age, and circulating concentrations of plasma Whatever the role of variations in BMR in the etiology of
leptin, T3, and T4. obesity, or the responses to caloric restriction, it is the largest
Design: This was a cross-sectional study of 150 white adults from component of the daily energy budget in westernized societies
northeast Scotland, United Kingdom. (19, 20). As such, manipulating BMR may provide a route toward
Results: Only 2% of the observed variability in BMR was attribut- treatment or prevention of obesity. A primary conclusion of The
able to within-subject effects, of which 0.5% was analytic error. Of International Dietary Energy Consultancy Group workshop on
the remaining variance, which reflected between-subject effects, energy and protein requirements in 1994 (20) was that “There is
63% was explained by FFM, 6% by FM, and 2% by age. The effects a need to investigate the causes of the relatively large inter-
of sex and bone mineral content were not significant (P 쏜 0.05). individual variations in BMR.” Despite this need being identified
Twenty-six percent of the variance remained unexplained. This vari- a decade ago, however, little progress has been made.
ation was not associated with concentrations of circulating leptin or Leptin is an adipokine produced predominantly by adipose
T3. T4 was not significant in women but explained 25% of the tissue (21). The role of leptin as a signaling molecule indicating
residual variance in men. the level of body fatness or historical levels of energy balance is
Conclusions: Our data confirm that both FFM and FM are signifi- well established (22, 23). Despite the broad correlation between
cant contributors to BMR. When the effect of FM on BMR is re- body fatness and circulating leptin concentrations, this relation
moved, any association with leptin concentrations disappears, which shows tremendous individual variability (24, 25), which may
suggests that previous links between circulating leptin concentra- drive individual differences in residual BMR. Previous studies
tions and BMR occurred only because of inadequate control for the addressing the relation between circulating leptin and BMR
effects of FM. Am J Clin Nutr 2005;82:941– 8. have, however, produced confused results, with some studies
finding positive associations (26 –28), but others finding nega-
KEY WORDS Obesity, metabolism, hormones, health, body
tive (29, 30) or no significant association (31–34). Some of these
composition
differences may reflect problems accounting for the confounding
effects of FM on both BMR and leptin concentrations (35).
INTRODUCTION 1
From the Aberdeen Centre for Energy Regulation and Obesity
The major factor determining basal metabolic rate (BMR) is (ACERO), Division of Energy Balance and Obesity, Rowett Research Insti-
fat-free mass (FFM) (1–3), with some studies finding an addi- tute, Aberdeen, Scotland, United Kingdom (AMJ, SDM, JSD, KAR, and
tional contribution of fat mass (FM) (4, 5), but others failing to JRS), and ACERO, School of Biological Sciences, University of Aberdeen,
find such an effect (6, 7). Even when these factors are known, Aberdeen, Scotland, United Kingdom (JRS).
2
Supported by the Scottish Executive Environment and Rural Affairs
however, substantial residual variation remains. Understanding
Department.
the physiologic nature of this variability is important because it 3
Address reprint requests to A Johnstone, Rowett Research Institute, Aberdeen,
has been implicated in the obesity epidemic in 2 separate ways. Scotland, United Kingdom, AB21 9SB. E-mail: [email protected].
First, low BMR (after mass effects are removed) may be a pre- Received April 5, 2005.
disposing risk factor for the development of obesity (8 –10), and Accepted for publication July 19, 2005.

Am J Clin Nutr 2005;82:941– 8. Printed in USA. © 2005 American Society for Nutrition 941
942 JOHNSTONE ET AL

TABLE 1 For estimates of basal energy requirements, BMR (MJ/d) was

Average subject characteristics, by sex and BMI category1 measured by indirect calorimetry with a ventilated hood system
Characteristic Women (n ҃ 107) Men (n ҃ 43) (Deltatrac II, MBM-200; Datex, Instrumentarium Corporation,
Helsinki, Finland). The subjects were instructed not to move
Age (y) 42 앐 11.4 47 앐 9.7 during the measurement period and were measured while in a
Height (m) 1.63 앐 0.06 1.76 앐 0.08
supine position in a quiet room for a period of 40 min. Basal
Weight (kg) 69.25 앐 16.5 84.79 앐 18.1
energy expenditure was calculated from the gaseous exchange
BMI (kg/m2)
Underweight 18.9 앐 1.0 [11]2 — [0] rates of oxygen and carbon dioxide (39) with the use of the values
Normal weight 22.4 앐 1.4 [42] 23.2 앐 1.2 [16] during the last 20 min of the measurement period. We did not
Overweight 26.6 앐 1.4 [32] 26.8 앐 1.4 [20] measure nitrogenous excretion but rather converted observed
Obese 36.3 앐 16.5 [22] 38.1 앐 0.8 [7] oxygen consumption and carbon dioxide production to energy
expenditure by using the Elia & Livesey (39) equation. The error
1
All values are x៮ 앐 SD.
2
n in brackets.
introduced by this conversion is 앒0.5% when nitrogen excretion
is neglected (assuming 13% of dietary energy intake is from
protein). If the subjects had unphysiologic measurements (respi-
ratory quotient 쏜 1.0 or 쏝 0.73), we rescheduled them for a
Hyper- and hypothyroid conditions lead to changes in BMR, repeat measurement the following week and used only the repeat
but the effects of individual variations in thyroid status within the measure in this analysis. Repeat measurements were made on
euthyroid range of variation remain unclear. Some studies (5, 36)

Downloaded from www.ajcn.org at University of Aberdeen on November 22, 2005

쏝10% of the population and were typically necessary because of
have shown significant links of BMR to circulating triiodothy- hyperventilation or the subject having a cold or cough that re-
ronine (T3) concentrations, yet other studies have failed to find sulted in them being restless or having abnormal breathing pat-
such effects (37, 38). In the present study, we aimed to define the terns. Calibration burns were performed on the machine at 6
roles of FFM, FM, age, and sex and to examine whether residual monthly intervals. These indicated that the Deltatrac machine
variation in BMR once these factors had been accounted for was had a mean carbon dioxide recovery of 99.5% (SD ҃ 0.5%). In
linked to concentrations of circulating leptin and thyroid hor- a separate group of 7 subjects, we measured the within-subject
mones [T3 and thyroxine (T4)]. variation in the BMR measurement by measuring BMR on 3
consecutive days by using the same protocol as in the main
observation series. After the measurement of BMR, the subject’s
SUBJECTS AND METHODS
height was measured to the nearest 0.1 cm with a stadiometer
Subject characteristics (Holtain Ltd, Crymych, Dyfeld, Wales). Body weight was mea-
sured on a portable digital scale (DIGI DS-410; CMS Weighing
One hundred fifty adults (n ҃ 107 women, 43 men) aged Equipment, London, United Kingdom) to the nearest 50 g.
between 21 and 64 y and having a body mass index (in kg/m2) in
the range of 16.7– 49.3 were recruited (Table 1) through news- Body composition
paper advertisements to participate in a study investigating ge-
netic and environmental influences on body weight. The subject Body composition was assessed by dual-energy X-ray absorp-
group was a representative sample of the Scottish population, tiometry (Norland XR-26, Mark II high-speed pencil beam scan-
with 1 in 5 obese (body mass index 쏜 30) and 54% of the ner equipped with dynamic filtration, with version 2.5.2 of the
population collectively overweight and obese (body mass index Norland software; Norland Corporation, Fort Atkinson, WI).
쏜 25). Demographically, Aberdeen is almost completely white, Measurements were conducted in the morning immediately after
with only minor Asian communities. We did not select subjects the BMR and standard anthropometry measurements while the
on the basis of race, but reflecting the local demography, all the subjects were still in a fasted condition. Weekly quality-control
recruited subjects were white. Subjects were included only if checks with a phantom over a period of 7 mo indicated CVs of
they were not consuming any special medical diet; had a stable 0.94% and 1.52% for bone mineral density and bone mineral
weight (weight change of no 쏜2 kg in the previous 3 mo); were content, respectively. The CV (%) in repeatability for this ma-
otherwise healthy on the basis of an extensive medical exami- chine in a total-body scan for mineral determination was 1.4%
nation, screening blood tests (full blood count and renal, liver, (40). The CV of variation for the assessment of FM was 2.6%
and thyroid function), and electrocardiogram (ECG); and took no (41). Obviously, accuracy estimates against the gold standard of
regularly prescribed medication or vitamin or mineral supple- chemical analysis are unavailable.
ments. The study was approved by the Joint Ethical Committee
of Grampian Health Board and The University of Aberdeen. Blood sampling
Written informed consent was obtained. Whole blood was sampled from a large antecubital vein in the
morning before breakfast with an 18G butterfly needle (Sarstedt,
Basal metabolic rate measurement Nuernbrecht, Germany) and adaptor. The blood sample was col-
The subjects attended the Rowett’s Human Nutrition Unit for lected into a 10-mL lithium heparin tube. The sample was spun
measurements of body composition and metabolic rate under (1000 ҂ g at 4 °C for 10 min) in a chilled centrifuge to obtain
standardized conditions. The subjects were instructed to fast plasma, which was stored at Ҁ80 °C for batch analysis.
overnight and to not consume caffeinated products or smoke
before arriving at the unit. Subjects arrived between 0700 and Plasma leptin
0830 and were allowed to relax for 앒30 min before measure- Circulating leptin in plasma was measured by using a radio-
ments were made. immunoassay kit supplied by BioVendor GmbH (Heidelberg,
 VARIABILITY IN BASAL METABOLIC RATE IN HUMANS 943
Germany). One hundred microliters of the diluted standards, TABLE 2
quality controls, and samples (all diluted one-third in dilution Results of repeated measurements of basal metabolic rate (BMR) in a
buffer) was added in duplicate to 96-well microtiter plates coated sample of 7 subjects
with antibody to human leptin, and 100 ␮L of dilution buffer was Mean BMR Mean BMR Mean BMR Mean (앐SD) of
added to the blank wells. The plate was then incubated for 1 h at Subject measurement 1 measurement 2 measurement 3 3 measurements
room temperature with shaking. An automated plate washer was kJ
used to wash the plate 3 times with wash solution, and any 1 4913 4831 5139 4961 앐 160
residual liquid was removed by tapping the plate on tissue paper. 2 5293 5101 5053 5149 앐 127
A total of 100 ␮L of horseradish peroxidase– conjugated anti- 3 5258 4932 5262 5151 앐 189
leptin antibody was added to each well, and the plate was incu- 4 7028 7090 6894 7004 앐 100
bated for 1 h at room temperature with shaking. The plate was 5 5176 4750 4783 4903 앐 237
6 9142 8821 8421 8795 앐 361
washed a second time with wash solution, as described previ-
7 5665 5398 5379 5481 앐 160
ously, before the addition of 100 ␮L of tetramethylbenzidine
substrate to each well. The plate was wrapped in foil and incu-
bated for 10 min at room temperature to allow the blue color to Statistics
develop before stopping the reaction with 100 ␮L of 0.2 mol
sulfuric acid/L. The resulting yellow color was measured at 450 The approach taken was to partition the variation into that
attributable to within-subjects and between-subjects effects by

Downloaded from www.ajcn.org at University of Aberdeen on November 22, 2005

nm within 5 min. The absorbance is directly proportional to the
concentration of leptin. The concentration of each sample (ng/ using the intersubject repeatability data. We also sought to ex-
plain the observed between-subject variation by using general-
mL) was determined from a standard curve produced by plotting
ized linear models with FFM and FM, height, bone mineral
the absorbance values versus the leptin concentrations of the
content, and age as covariates and sex as a factor. Subtracting the
standards. The standards, quality controls, and samples were all
within-subject variance and the between-subject variance ex-
diluted one-third, so multiplying by this dilution factor was not plained by these factors revealed an unexplained residual varia-
necessary. BioVendor provided 2 quality-control standards with tion. We sought to explore how much of this residual variation
expected values of 4.34 and 15.89 ng/mL. We measured each could be explained by individual differences in circulating hor-
standard 5 times, yielding average values of 4.36 and 16.59 mone concentrations, again by using a generalized linear mod-
ng/mL, respectively. Thus, the average accuracy was 2.5%. Pre- eling approach. All statistical analyses were performed with
cision (CV%) across the 5 measurements was 3.5%; hence, ac- MINITAB version 14 (Minitab Inc, State College, PA).
curacy was better than precision. All samples were run in dupli-
cate, and the mean precision (CV %) across all samples averaged
5.2%. RESULTS

Basal metabolic rate

Plasma triiodothyronine and thyroxine concentrations The mean BMR of all subjects was 6279 kJ/d. The minimum
Circulating concentrations of T3 and T4 were measured by was 4301 kJ/d and the maximum was 10 455 kJ/d. The SD was
using a rapid test kit developed by Diagnostic Automation Inc 1199 kJ/d, and the CV was 19.1%. The results of repeated mea-
(Calabasa, CA). Fifty microliters of the standards, samples, and surements in a sample of 7 subjects are shown in Table 2. In this
controls was added to a 96-well microtiter plate coated with sample, the mean (앐SD) BMR across individuals was 5921 앐
antibody to T3 (or T4). After a 10-s mix, 100 ␮L of diluted (1/20) 1459 kJ/d (n ҃ 7), and the within-subject SD (3 repeated mea-
horseradish peroxidase– conjugated T3 (or T4) was added. After sures per subject) averaged 206 kJ/d. The contribution of within-
a further 30-s mix, the plate was incubated at 19 –22 °C degrees subject variance to the total variance was therefore (2062)/
for 1 h. The plate was then washed 5 times with sterile distilled (14592) ҃ 0.0201. About 2% of the total variance in the
water to remove unbound conjugate, and residual water droplets measurements could therefore be attributed to within-subject
were removed by patting firmly onto tissue paper before the variation. Because the analytic precision of the Deltatrac ma-
chine judged from the recovery of carbon dioxide in the alcohol
addition of 100 ␮L tetramethylbenzidine substrate. The plate
burn calibrations was 0.5%, about one-quarter of the within-
was gently mixed for 5 s and the blue color was allowed to
subject variance could be attributed to analytic error.
develop in the dark for 20 min. The reaction was then stopped by
Visual inspection revealed a slight positive skew in all the
the addition of 100 ␮L of stop solution and the plate was mixed
distributions for morphologic traits (height, weight, FFM, FM,
for 15 s to ensure that all the blue color changed to yellow before and bone mineral content). We examined these variables for
reading the optical density at 450 nm. The intensity of the yellow normality by using the Anderson Darling test before analysis,
color formed is directly proportional to the quantity of enzyme and all traits except height were significantly different from
present and inversely proportional to the amount of unlabeled T3 normal (P 쏝 0.01). We therefore log-transformed the variables
(or T4) in the sample. Therefore, by reference to standards as- with skewed distributions before analysis. Most of the between-
sayed under the same conditions, the concentration of T3 in the subject variation in BMR was explained by differences in FFM
samples was quantified in ng/mL and the concentration of T4 in (Figure 1), which alone explained 62.3% of the variation. Once
ng/0.5 mL. No quality-control standards were available, so we the effects of loge FFM had been removed, the remaining vari-
could not assess the accuracy of the measurements. All samples ation was significantly related to loge FM (Figure 1; r2 ҃ 0.18).
were run in duplicate, and the mean precision (CV%) across all Once the combined effects of FM and FFM had been removed,
individuals was 2.46% for T3 and 3.6% for T4. there was a significant negative effect of age (y), with older
944 JOHNSTONE ET AL

Downloaded from www.ajcn.org at University of Aberdeen on November 22, 2005

FIGURE 2. Partitioning of the variance in basal metabolic rate measured
in 150 adults. Most of the variation was between subjects. Of this variation,
most was explained by fat-free mass (P 쏝 0.001). There were smaller effects
of fat mass (P 쏝 0.01) and age (P 쏝 0.03), and a large portion of the variance
remained unexplained. Effects of potential predictor variables were evalu-
ated by using generalized linear modeling.

analytic error effects. Of the between-subject variance, most was

explained by FFM, with relatively minor contributions of FM
and age. There was a large unexplained residual variation be-
tween individuals that accounted for 26% of the total variance.

Circulating leptin, triiodothyronine, and thyroxine

concentrations
We obtained assay results for circulating leptin for 145 of the
150 subjects. Concentrations of circulating leptin (ng/mL) were
positively related to FM (Figure 3). In addition, there was a
significant effect of sex, with the women having higher leptin
concentrations than did the men at any given FM (Figure 3).
FIGURE 1. Basal metabolic rate (BMR) in a sample of 150 Scottish
adults as a function of fat-free body mass (FFM) and the residual variation in Together, these 2 variables explained 78% of the total variation
loge BMR once the effects of FFM were removed in relation to loge fat mass in circulating leptin concentrations. Height, FFM, age, and bone
(FM). In both plots, men (n ҃ 43) are shown as open symbols and women mineral content were all not significant (P 쏜 0.05). T3 concen-
(n ҃ 107) as closed symbols. The effects on BMR of both FFM (P 쏝 0.001) trations (ng/mL) were not significantly related to any of the
and FM (P 쏝 0.01) were significant (least-squares linear regression). Ana-
lytic error for BMR was 앒0.5%, and within-subject variation was 앒2%.
morphologic traits measured (height, FFM, FM, or bone mineral
Precision error (CV%) for body-composition traits was 2.6%. content) or to sex or age. T4 concentrations were also unrelated
to variability in morphology, but an effect of FFM approached
significance (P ҃ 0.078).

Relations between residual BMR and hormone

people having lower resting metabolic rates independent of any
concentrations
age-related changes in their body compositions, although age
explained only 6.3% of the residual variation. In a multiple re- We removed the effects of FFM, FM, and age on BMR to
gression analysis, height, sex, and bone mineral content were all derive a residual variance. We then sought associations between
nonsignificant (P 쏜 0.05). The best-fit least-squares regression this residual variation and the residual variation in circulating
including the 3 significant factors was as follows: leptin concentrations (after removing the effects of FM and sex)
and concentrations of T3 and T4. There was no significant asso-
loge (BMR) kJ/d ⫽ 0.908 ⫹ 0.619 loge (FFM) ciation between residual variation in BMR and residual variation
in circulating leptin concentrations (Figure 4; P ҃ 0.455) and
⫹ 0.132 loge (FM) ⫺ 0.00262 age (y) (1)
also no relation to circulating concentrations of T3 (Figure 5;
which explained 71.4% of the variation in loge BMR (F ҃ P ҃ 0.532). For T4 concentrations, however, the relation to
121.34, P 쏝 0.001). residual BMR was more complex (Figure 5; P ҃ 0.199). Using
The partitioning of the variance in BMR in this population is a general linear model, we found a significant overall effect of T4
illustrated in Figure 2. Most of the variation resided between concentrations (F ҃ 8.07, P 쏝 0.005) but also an effect of sex
subjects, with relatively little variance due to within-subject and (F ҃ 5.68, P ҃ 0.019) and an interaction between these effects
 VARIABILITY IN BASAL METABOLIC RATE IN HUMANS 945

Downloaded from www.ajcn.org at University of Aberdeen on November 22, 2005

FIGURE 3. Circulating leptin concentrations in plasma plotted against
loge fat mass (FM) in a sample of 145 Scottish adults. Women (n ҃ 106) are
shown as closed symbols and men (n ҃ 39) as open symbols. There were
significant effects of both fat mass (FM; P 쏝 0.001) and sex (P 쏝 0.001) on
concentrations of circulating leptin as evaluated by using general linear
modeling. Accuracy of the leptin assay against quality controls was 2.5%.
Average precision (CV%) in repeated measurements of the leptin assay
across all subjects was 5.2%. Precision error (CV%) for fat mass was 2.6%.

(F ҃ 8.78, P ҃ 0.004). At any given T4 concentration, men on

average had higher residual BMRs. The relation between circu-
lating T4 and residual BMR was highly significant in the men
(r2 ҃ 0.25, F ҃ 13.02, P 쏝 0.001) but was not significant in the
women (r2 ҃ 0.0, F ҃ 0.01, P ҃ 0.904). Removing the single
point in Figure 5 that had a large influence on the regression
for men diminished but did not abolish the significance of the
relation. FIGURE 5. Residual variation in basal metabolic rate (BMR) once the
effects of fat-free mass, fat mass, and age were removed plotted against
circulating concentrations of triiodothyronine (T3) and thyroxine (T4). Men
(n ҃ 39) are shown as open symbols and women (n ҃ 106) as closed symbols.
The relation with T3 was not significant (P 쏜 0.05), but there was a significant
association with T4, which differed by sex. The relation was significant (P 쏝
0.001) in men, but not women (P 쏜 0.05) by linear regression. Analytic
precision error (CV) averaged across duplicate analyses of all subjects av-
eraged 2.4% for T3 and 3.6% for T4. Removing the point marked with an
arrow in the plot for T4 diminished but did not abolish the significance of the
relation in men.

DISCUSSION
Many previous studies have quantified the repeatability of
BMR measurements when using hood-based measurement sys-
tems. Previous estimates of within-subject repeatability fell in
the range of 2– 4% of the mean estimate (reviewed in reference
20), with the extent of repeatability depending on the interval
between measurements. Our estimate of repeatability of 2.8%
falls within the expected range given the short time scale between
FIGURE 4. Relation between residual variation in basal metabolic rate
(BMR) once the effects of fat-free mass, fat mass, and age were removed and repeated measurements in our study. Because BMR is a highly
the residual variation in circulating leptin concentrations once the effects of repeatable trait, the contribution of within-subject variation to
fat mass and sex were removed. Men (n ҃ 39) are shown as open symbols and the total variance was trivially small (2%), which indicates that
women (n ҃ 106) as closed symbols. The relation was not significant (P 쏜 most of the variance in BMR resided between the individuals
0.05) by least-squares linear regression. Accuracy of the leptin assay against
quality controls was 2.5%. Average precision (CV) in repeated measure-
under study.
ments of the leptin assay across all subjects was 5.2%. Analytic error for Consistent with previous studies (1–3, 42) the dominant factor
BMR was 앒0.5%, and within-subject variation was 앒2%. influencing this variation between individuals was the extent to
946 JOHNSTONE ET AL

which the individuals varied in their FFM. The additional role in brain size by the energy metabolism of brain tissue gives a
played by FM as an independent factor influencing BMR has contribution to variation in BMR of 70.6 kJ/d. This compares
been a matter of previous debate, with some studies finding a with the residual variation (SD) in BMR that amounted to 529.8
significant independent effect (4, 5, 43) and others failing to kJ/d. The percentage contribution of the undetected variance
replicate this result (6, 7). Our studies confirmed that, in this (SD2) in brain size to the residual variance in BMR (SD2) was
sample of adults, FM was an independent factor—although the therefore (70.62/529.82) ҂ 100 ҃ 1.3%. It therefore seems un-
contribution to the total variance that was explained by differ- likely that undetected variation in tissue sizes of highly energetic
ences in FM was relatively small compared with that explained organs was a significant contribution to the observed residual
by FFM. In the derived multiple regression, the coefficient re- variation in BMR.
lating BMR to FFM was 0.619, compared with 0.132 for the The independent effects of FM and sex on circulating leptin
coefficient relative to FM. This suggests that each kilogram of concentrations were consistent with previous reports (35). The
lean tissue exerted about 5 times more effect on BMR than did large degree of individual variation about the average trend in
each kilogram of fat tissue. This ratio of 5:1 is at odds with in vitro this relation also agrees with previous observations. Studies in
estimates of the energy metabolism rates of fat and lean tissue [in rodents have implicated circulating leptin as a key signal related
both humans (44), and rodents (45)], for which the differences in to resting metabolism (23, 58). For example, the mutant ob/ob
energy metabolism between FM and FFM ranged from 10 to 100 mouse that lacks functional leptin has a lowered resting meta-
fold. Nevertheless, this ratio is consistent with similar multiple bolic rate and a lowered body temperature (21). Several previous
regression models linking energy metabolism to body composi- studies indicated that the between-subject variation in BMR is

Downloaded from www.ajcn.org at University of Aberdeen on November 22, 2005

tion in rodents (46). linked to between-subject variation in leptin concentrations (26 –
These data emphasize that the derived regression equation is a 28), but others found a negative relation (29) or no significant
statistical rather than a physiologic model of the BMR. The association (31–34). Consistent with the findings of Neuhauser-
regression coefficients do not necessarily imply that the energy Berthold et al (35) in much older subjects, we found no associ-
expenditure of FFM is 5 times greater than the energy expendi- ation between this residual variation in circulating leptin con-
ture of FM. The coefficient with respect to FM may be greater centrations and the residual variance in BMR once the effects of
than anticipated physiologically because the presence of FM FFM, FM, and age had been removed. In other words, if an
stimulates the metabolic rate of other tissues as a result of adi- individual had a particularly high concentration of circulating
pokine secretions that are themselves positively associated with leptin for their FM and sex, this had no effect on whether they had
FM. The negative effect of age on BMR, independent of any a high or low BMR for their FFM, FM, and age. This indicates
age-related changes in body composition reported here, has been that the unexplained variation between subjects in BMR could
previously reported (3, 47– 49). not be attributed to the variability between subjects in their leptin
Like all previous studies (50 –55), we found enormous varia- production by the FM. This observation in our data is consistent
tion in between-subject BMR (26% of total variance) that was not with the observation that humans lacking functional leptin pro-
explained by differences in body composition or age (or sex). duction do not have a significantly altered BMR (59). The dif-
This variation could not be attributed to either within-subject ferences in the effects of leptin between humans and rodents may
error or analytic error, which together only summed to 2% of the in part be because laboratory rodents are routinely kept under
total variance (less than one-tenth of the unexplained variation). thermoregulatory stress at around 20 °C, which is 10 °C below
In part, this unexplained variance may reflect the inadequacies of the lower critical limit of the thermoneutral zone (60). Hence, the
our characterization of body composition into only 3 compart- effects of leptin absence on BMR of the ob/ob mouse, and its
ments (FFM, FM, and bone mineral content). Clearly, FFM is not lowered body temperature, may reflect the absence of stimula-
a homogeneous tissue (55), and, consequently, variations in the tion of brown adipose tissue thermogenesis (58). In humans who
size of the organs contributing to FFM might result in variations have their BMR measured at thermoneutral temperatures, such
in BMR that our analysis would not detect. However, studies effects of leptin would not be anticipated, particularly because in
using magnetic resonance imaging to model the contributions of adults, levels of brown adipose tissue are probably not of func-
different organs that make up the FFM (56) also leave a relatively tional significance for thermoregulation.
large residual variation unexplained, although the mean pre- The effects of pathologic variations in concentrations of the
dicted and observed values across a sample of individuals are thyroid hormones and metabolic rate have long been established.
close. To further explore the potential effects of unquantified Hyper- and hypothyroid conditions both lead to major changes in
variability in organ tissue sizes to the unexplained variation, we basal metabolism. However, despite these major effects, the role
modeled the effect of variation in brain size on the BMR. In of individual variations in thyroid status within the euthyroid
making this calculation, we assume that the dual-energy X-ray range of variation remains unclear. Some studies (5, 36) have
absorptiometry machine is blind to any variation in brain size and shown significant links of BMR to circulating T3 concentrations,
thus that brain size variations can contribute to variation in BMR, yet other studies have failed to find such links (37, 38). We also
but not FFM or FM. Elia (57) suggested that the metabolism of failed to establish any link between circulating concentrations of
the brain averages 1008 kJ · kgҀ1 · dҀ1. Gallagher et al (56) T3 and the unexplained variation in BMR. An effect of T4 con-
estimated the variation (SD) in the brain size of adults, measured centrations on BMR in men but not women was previously de-
by magnetic resonance imaging, to be 0.14 kg in women and 0.15 scribed (61), although in that previous study the effect was neg-
kg in men. However, there was a positive correlation between ative: higher concentrations of T4 were associated with lower
brain size and body size (r ҃ 0.54), so some of this variability will BMR. The variability in previous studies with respect to the
be linked to the variation in body size. At a fixed body size, we effects of T4 cannot be accounted for by the different sex effects,
assumed the variation in brain size might amount to one-half of because previous studies that showed relations included studies
the total variation (ie, SD ҃ 0.07 kg). Multiplying this variation focused only on women (36). The difference in effects of T4
 VARIABILITY IN BASAL METABOLIC RATE IN HUMANS 947
between the sexes and between studies points to the complexity 5. Svendsen OL, Hassager C, Christiansen C. Impact of regional and total
of the interactions on the residual variation in BMR. body composition and hormones on resting energy expenditure in over-
weight postmenopausal women. Metabolism 1993;42:1588 –91.
The absence of significant hormone effects on residual BMR 6. Bogardus C, Lillioja S, Ravussin E, et al. Familial dependence of the
was not due to analytic precision error in the hormone measure- resting metabolic rate. N Engl J Med 1986;315:96 –100.
ments. To evaluate the effect of these errors on the significance 7. Segal KR, Gutin B, Albu J, Pi-Sunyer FX. Thermal effects of food and
of the relations between the given traits and BMR, it is necessary exercise in lean and obese men of similar lean body mass. Am J Physiol
to consider the gradient of the estimated relation between the 1987;252:E110 –7.
8. Roberts SB, Savage J, Coward WA, Chew B, Lucas A. Energy expen-
traits in question. When this gradient is unity, an x% precision
diture and intake in infants born to lean and overweight mothers. N Engl
error in the x variable will translate to an equivalent error in the J Med 1988;318:461– 6.
y variable. Thus, if the precision error (CV) in the x variable were 9. Ravussin E, Lillioja S, Knowler WC, et al. Reduced rate of energy
30%, the maximum variation in y explained by the x variable expenditure as a risk factor for body-weight gain. N Engl J Med 1988;
would be 앒70% (100% Ҁ 30%). When the gradient declines 318:467–72.
10. Weyer C, Snitker S, Bogardus C, Ravussin E. Energy metabolism in
below unity, the effect of error in the x variable becomes less and
African Americans: potential risk factors for obesity. Am J Clin Nutr
less important, until at a gradient of 0, the error in x imparts no 1999;70:13–20.
effect on the relation. Conversely, when the gradient exceeds 11. Astrup A, Gotzsche PC, van de Werken K, et al. Meta-analysis of resting
unity, the effect of errors in x become more significant. Hence, if metabolic rate in formerly obese subjects. Am J Clin Nutr 1999;69:
the gradient were 2.0, rather than unity, the maximum variation 1117–22.
in y explained by x would only be 앒40% (ie, 100% Ҁ 2 ҂ 30%). 12. Seidell JC, Muller DC, Sorkin JD, Andres R. Fasting respiratory ex-
change ratio and resting metabolic rate as predictors of weight gain: the

Downloaded from www.ajcn.org at University of Aberdeen on November 22, 2005

On the other hand, if the gradient were 0.5, the maximum ex- Baltimore Longitudinal Study on Aging. Int J Obes Relat Metab Disord
pected explained variation would be 앒85% (100% Ҁ 0.5 ҂ 1992;16:667–74.
30%). In the case of the hormones (and body-composition traits) 13. Goran MI, Shewchuk R, Gower BA, Nagy TR, Carpenter WH, Johnson
that we tested, the precision errors in the x variables were be- RK. Longitudinal changes in fatness in white children: no effect of
childhood energy expenditure. Am J Clin Nutr 1998;67:309 –16.
tween 2% and 5% (see methods), and the fitted gradients on
14. Katzmarzyk PT, Perusse L, Tremblay A, Bouchard C. No association
logged traits were all substantially less than unity. Thus, the between resting metabolic rate or respiratory exchange ratio and subse-
amount of unexplained variation in the y trait (BMR) that could quent changes in body mass and fatness: 5 1/2 year follow-up of the
be attributed to errors in the x traits would be on the order of Quebec Family Study. Eur J Clin Nutr 2000;54:610 – 4.
1–3%. This compares with the actual unexplained variation by 15. McNeill G, Rivers JPW, Payne PR, Debritto JJ, Abel R. Basal metabolic
body composition that was 26%, and the unexplained residual rate of Indian men: no evidence of metabolic adaptation to a low plane
of nutrition. Hum Nutr Clin Nutr 1987;41C:473– 83.
variation by hormones that was 쏜95% for leptin, T3, and T4 in 16. Soares MJ, Shetty PS. Basal metabolic rates and metabolic economy in
women and 80% for T4 in men. Overall, the effect of the precision chronic undernutrition. Eur J Clin Nutr 1991;45:363–73.
errors in the assays could not account for the absence of signif- 17. Shetty PS. Chronic undernutrition and metabolic adaptation. Proc Nutr
icant relations or the high residual variation in BMR. Soc 1993;52:267– 84.
Like many other studies, our work has indicated a large 18. Weinsier RL, Nagy TR, Hunter GR, Darnell BE, Hensrud DD, Weiss
HL. Do adaptive changes in metabolic rate favor weight regain in
between-individual variation in BMR that cannot be explained weight-reduced individuals? An examination of the set-point theory.
by morphologic characteristics, including sex and age. We show Am J Clin Nutr 2000;72:1088 –94.
here that 2 important physiologic candidates that might explain this 19. Black AE, Coward WA, Cole TJ, Prentice AM. Human energy expen-
variability, namely, variations in circulating concentrations of leptin diture in affluent societies: an analysis of 574 doubly-labelled water
and T3, do not account for this unexplained variation in a population measurements. Eur J Clin Nutr 1996;50:72–92.
20. Scrimshaw NS, Waterlow JC, Schurch B. Energy and protein require-
of 150 white adults. In men, however, 25% of the residual variance ments. Proceedings of an IDECG Workshop 31 October to 4 November
was associated with variation in circulating concentrations of T4. No 1994. Eur J Clin Nutr 1996;50:S1–197.
equivalent effect was found in women. 21. Zhang YY, Proenca R, Maffei M, Barone M, Leopold L, Friedman JM.
Positional cloning of the mouse obese gene and its human homolog.
We thank Sylvia Hay, Marion Scott, and Jean Bryce for assistance in the Nature 1994;372:425–32.
Human Nutrition Unit and Zydrune Visockiene for assistance with the RMR 22. Flier JS. What’s in a name? In search of leptin’s physiologic role. J Clin
repeatability measurements. Endocrinol Metab 1998;83:1407–13.
JRS, KAR, and AMJ were responsible for the study concept and design. 23. Schmidt I, Doring H, Stehling O, Nuesslein-Hildesheim B, Steinlechner
AMJ, JSD, and SDM were responsible for the data collection and laboratory S, Schwarzer K. Leptin disinhibits rather than stimulates sympatheti-
analysis. JRS was responsible for the data analysis and drafting of the manu- cally mediated energy expenditure. In: Blum WF, Kiess W, Rascher W,
script. JRS, KAR, and AMJ were responsible for critical revision of the eds. Leptin: the voice of adipose tissue. Heidelberg-Leipzig, Germany:
manuscript for important intellectual content. None of the authors had a J A Barth Verlag, 1997:133– 47.
24. Maffei M, Halaas J, Ravussin E, et al. Leptin levels in human and rodent:
conflict of interest.
measurement of plasma leptin and ob RNA in obese and weight-reduced
subjects. Nat Med 1995;1:1155– 61.
REFERENCES 25. Considine RV, Sinha MK, Heiman ML, et al. Serum immunoreactive
1. Weinsier RL, Schutz Y, Bracco D. Reexamination of the relationship of leptin concentrations in normal-weight and obese humans. N Engl J Med
resting metabolic rate to fat-free mass and to the metabolically active 1996;334:292–5.
components of fat-free mass in humans. Am J Clin Nutr 1992;55:790 – 4. 26. Jorgensen JOL, Vahl N, Dall R, Christiansen JS. Resting metabolic rate
2. Cunningham JJ. Body composition as a determinant of energy expen- in healthy adults: relation to growth hormone status and leptin levels.
diture: a synthetic review and a proposed general prediction equation. Metabolism 1998;47:1134 –9.
Am J Clin Nutr 1991;54:963–9. 27. Nicklas BJ, Toth MJ, Goldberg AP, Poehlman ET. Racial differences in
3. Fukagawa NK, Bandini LG, Young JB. Effect of age on body compo- plasma leptin concentrations in obese postmenopausal women. J Clin
sition and resting metabolic rate. Am J Physiol 1990;259:E233– 8. Endocrinol Metab 1997;82:315–7.
4. Nelson KM, Weinsier RL, Long CL, Schutz Y. Prediction of resting 28. Toth MJ, Gottlieb SS, Fisher ML, Ryan AS, Nicklas BJ, Poehlman ET.
energy expenditure from fat-free mass and fat mass. Am J Clin Nutr Plasma leptin concentrations and energy expenditure in heart failure
1992;56:848 –56. patients. Metabolism 1997;46:450 –3.
948 JOHNSTONE ET AL

29. Niskanen L, Haffner S, Karhunen LJ, Turpeinen AK, Miettinen H, Uus- Tucker HN, eds. Energy metabolism: tissue determinants and cellular
itupa MIJ. Serum leptin in relation to resting energy expenditure and fuel corollaries. New York: Raven, 1992:61– 80.
metabolism in obese subjects. Int J Obes 1997;21:309 –13. 45. Krebs HA. Body size and tissue respiration. Biochim Biophys Acta
30. Bobbioni-Harsch E, Asimacopoulos-Jeannet F, Lehmann T, Munger R, 1950;4:249 – 69.
Allaz AF, Golay A. Leptin plasma levels as a marker of sparing-energy 46. Selman C, Lumsden S, Bunger L, Hill WG, Speakman JR. Resting
mechanisms in obese women. Int J Obes Relat Metab Disord 1999;23: metabolic rate and morphology in mice (Mus musculus) selected for high
470 –5. and low food intake. J Exp Biol 2001;204:777– 84.
31. Roberts SB, Nicholson M, Staten M, et al. Relationship between circu- 47. Poehlman ET. Energy expenditure and requirements in aging humans. J
lating leptin and energy expenditure in adult men and women aged 18 Nutr 1992;122:2057– 65.
years to 81 years. Obes Res 1997;5:459 – 63. 48. Roberts SB. Effects of aging on energy requirements and the control of
32. Kennedy A, Gettys TW, Watson P, et al. The metabolic significance of food intake in men. J Gerontol A Biol Sci Med Sci 1995;50:101– 6.
leptin in humans: gender-based differences in relationship to adiposity, 49. Piers LS, Soares MJ, McCormack LM, O’Dea K. Is there evidence for an
insulin sensitivity, and energy expenditure. J Clin Endocrinol Metab age-related reduction in metabolic rate? J Appl Physiol 1998;85:2196 –
1997;82:1293–300. 204.
33. Nagy TR, Gower BA, Shewchuk RM, Goran MI. Serum leptin and 50. Booyens J, Mccance RA. Individual variations in expenditure of energy.
energy expenditure in children. J Clin Endocrinol Metab 1997;82:4149 – Lancet 1957;1:225–9.
53. 51. Schofield WN. Predicting basal metabolic rate, new standards and re-
34. Campostano A, Grillo G, Bessarione D, De Grandi R, Adami GF. Re- view of previous work. Hum Nutr Clin Nutr 1985;39(suppl):5– 41.
lationships of serum leptin to body composition and resting energy 52. Welle S, Nair KS. Relationship of resting metabolic rate to body com-
expenditure. Horm Metab Res 1998;30:646 –7. position and protein turnover. Am J Physiol 1990;258:E990 – 8.
35. Neuhauser-Berthold M, Herbert BM, Luhrmann PM, et al. Resting met- 53. Ismail MN, Ng KK, Chee SS, Roslee R, Zhang Y. Predictive equations
abolic rate, body composition, and serum leptin concentrations in a for the estimation of basal metabolic rate in Malaysian adults. Mal J Nutr

Downloaded from www.ajcn.org at University of Aberdeen on November 22, 2005

free-living elderly population. Eur J Endocrinol 2000;142:486 –92. 1998;4:81–90.
36. Astrup A, Buemann B, Christensen NJ, et al. The contribution of body 54. Weyer C, Walford RL, Harper IT, et al. Energy metabolism after 2 y of
composition, substrates, and hormones to the variability in energy ex- energy restriction: the Biosphere 2 experiment. Am J Clin Nutr 2000;
penditure and substrate utilization in premenopausal women. J Clin 72:946 –53.
Endocrinol Metab 1992;74:279 – 86. 55. Heymsfield SB, Gallagher D, Kotler DP, Wang ZM, Allison DB, Heshka
37. Bernstein RS, Thornton JC, Yang MU, et al. Prediction of the resting S. Body-size dependence of resting energy expenditure can be attributed
metabolic rate in obese patients. Am J Clin Nutr 1983;37:595– 602. to nonenergetic homogeneity of fat-free mass. Am J Physiol Endocrinol
38. Welle S, Jozefowicz R, Statt M. Failure of dehydroepiandrosterone to Metab 2002;282:E132– 8.
influence energy and protein metabolism in humans. J Clin Endocrinol 56. Gallagher D, Belmonte D, Deurenberg P, et al. Organ-tissue mass mea-
Metab 1990;71:1259 – 64. surement allows modeling of REE and metabolically active tissue mass.
39. Elia M, Livesey G. Energy expenditure and fuel selection in biological Am J Physiol Endocrinol Metab 1998;38:E249 –58.
systems: the theory and practice of calculations based on indirect calo- 57. Elia M. Energy expenditure in the whole body. In: Kinney JM, ed.
rimetry and tracer methods. World Rev Nutr Diet 1992;70:68 –131. Energy metabolism: tissue determinants and cellular collaries. New
40. Tothill P, Avenell A, Love J, Reid DM. Comparisons between Hologic, York: Raven Press Ltd, 1992:19 – 49.
Lunar and Norland dual-energy X-ray absorptiometers and other tech- 58. Scarpace PJ, Matheny M, Pollock BH, Tumer N. Leptin increases un-
niques used for whole-body soft tissue measurements. Eur J Clin Nutr coupling protein expression and energy expenditure. Am J Physiol En-
1994;48:781–94. docrinol Metab 1997;36:E226 –30.
41. Tothill P, Avenell A, Reid DM. Precision and accuracy of measurements 59. Montague CT, Farooqi IS, Whitehead JP, et al. Congenital leptin defi-
of whole-body bone mineral: comparisons between Hologic, Lunar and ciency is associated with severe early-onset obesity in humans. Nature
Norland dual-energy X-ray absorptiometers. Br J Radiol 1994;67:1210 –7. 1997;387:903– 8.
42. Karhunen L, Franssila Kallunki A, Rissanen A, Kervinen K, Kesaniemi 60. Speakman JR, Rossi FP. No support for socio-physiological suppression
YA, Uusitupa M. Determinants of resting energy expenditure in obese effect on metabolism of paired white mice (Mus sp. ). Funct Ecol 1999;
non- diabetic Caucasian women. Int J Obes Relat Metab Disord 1997; 13:373– 82.
21:197–202. 61. Stenlof K, Sjostrom L, Fagerberg B, Nystrom E, Lindstedt G. Thyroid
43. Ferraro R, Ravussin E. Fat mass in predicting resting metabolic rate. hormones, procollagen III peptide, body composition and basal meta-
Am J Clin Nutr 1992;56:460(letter). bolic rate in euthyroid individuals. Scand J Clin Lab Invest 1993;53:
44. Elia M. Organ and tissue contribution to metabolic rate. In: Kinney JM, 793– 803.

View publication stats