Integrating muscle cell biochemistry and whole-body physiology in humans:31P-MRS data from the InSight trial

Article (PDF Available)inScientific Reports 3:1182 · January 2013with48 Reads
DOI: 10.1038/srep01182 · Source: PubMed
We acquired (31)P-MRS data from skeletal muscle of subjects of mixed gender and ethnicity, combined with a panel of physiological characteristics, and tested several long-standing hypotheses regarding relationships between muscle cell biochemistry and whole-body physiology with unusually high statistical power. We hypothesized that i) whole-body VO(2)max would correlate with muscle respiratory capacity, ii) resting muscle phosphocreatine concentration ([PCr]) would negatively correlate with delta efficiency and iii) muscle mitochondrial function would positively correlate with both resting VO(2) and total daily energy expenditure (TDEE). Muscle respiratory capacity explained a quarter of the variation in VO(2)max (r(2) = 26, p < .001, n = 87). There was an inverse correlation between muscle [PCr] and delta efficiency (r = -23, p = 046, n = 87). There was also a correlation between [PCr] recovery halftime and TDEE (r = -23, p = 035, n = 87). Our data not only provide insights into muscle cell chemistry and whole-body physiology but our mixed cohort means that our findings are broadly generalizable.


Integrating muscle cell biochemistry and
whole-body physiology in
P-MRS data from the InSight
Lindsay M. Edwards
*, Graham J. Kemp
, Renee M. Dwyer
, Justin T. Walls
, Huddy Fuller
Steven R. Smith
& Conrad P. Earnest
School of Biological Sciences, University of Essex, Colchester, UK,
Department of Musculoskeletal Biology, Faculty of Health and
Life Sciences, University of Liverpool, Liverpool, UK,
School of Medicine, University of Tasmania, Australia,
Sanford Burnham
Institute, Orlando, Florida, US,
Translational Research Institute for Metabolism and Diabetes, Florida Hospital, Orlando, Florida,
Pennington Biomedical Research Institute, Baton Rouge, Louisiana, US.
We acquired
P-MRS data from skeletal muscle of subjects of mixed gender and ethnicity, combined with a
panel of physiological characteristics, and tested several long-standing hypotheses regarding relationships
between muscle cell biochemistry and whole-body physiology with unusually high statistical power. We
hypothesized that i) whole-body VO
max would correlate with muscle respiratory capacity, ii) resting
muscle phosphocreatine concentration ([PCr]) would negatively correlate with delta efficiency
and iii) muscle mitochondrial function would positively correlate with both resting VO
and total daily
energy expenditure (TDEE). Muscle respiratory capacity explained a quarter of the variation in VO
5 26,
, .001,
5 87). There was an inverse correlation between muscle [PCr] and delta efficiency
5223, p 5 046,
5 87). There was also a correlation between [PCr] recovery halftime and TDEE
(r 5223, p 5 035,
5 87). Our data not only provide insights into muscle cell chemistry and whole-body
physiology but our mixed cohort means that our findings are broadly generalizable.
uclear magnetic resonance spectroscopy (MRS) offers the opportunity to study the relationships between
cell metabolism and whole-body physiology by providing noninvasive, quantitative, temporally- and
spatially-resolved measurements of cell biochemistry in vivo. This has been especially true of skeletal
muscle, where 31-phosphorus (
P)-MRS has been much used, in living humans and animals, to measure
phosphorus biochemistry, energy metabolism, mitochondrial function, proton handling and contractile effi-
ciency (among others)
. As biology enters an era in which both reductionist and systems approaches are con-
nected by studies of in vivo function, MRS continues to play an important role in understanding the relationships
between cell/tissue biology and whole-body physiology i.e. integrating cell, organ and organism.
As part of a large prospective longitudinal study (InSight), we acquired
P-MRS data from skeletal muscle in a
large number of subjects (,90), in parallel with an extensive panel of important physiological and morphological
characteristics including maximal whole-body respiratory capacity (VO
max), body composition and muscle
performance. This provided an opportunity to study the relationships between in vivo muscle chemistry, muscle
mitochondrial function and whole-body physiology with high statistical power. Furthermore, the cohort involved
in this trial, while young, was mixed in both gender and ethnicity. Therefore any relationships that might be
uncovered should be more broadly generalizable than those from existing studies that have been conducted in
more homogeneous cohorts. Nevertheless, our results are still restricted to subjects who are young and relatively
fit and should not be extended beyond these.
We made a number of hypotheses before analyzing the data, based on current evidence regarding how muscle
cell biology translates upwards into physi ological function. These are summarized in Table 1, with associated
references. The meaning of the symbols, and the methods used, are described in Methods. We also used untar-
geted correlations and an unsupervised multivariate statistical method (principal components analysis) for data
21 June 2012
24 September 2012
31 January 2013
Correspondence and
requests for materials
should be addressed to
L.M.E. (lindsay.
* Current address:
Centre of Human &
Sciences, King’s
College London, UK.
SCIENTIFIC REPORTS | 3 : 1182 | DOI: 10.1038/srep01182 1
In addition to the study of relationships between muscle and
whole-body physiology, the InSight data allowed us to report ref-
erence values and ranges for several important phosphorus metabo-
lites and metabolite ratios in a young, disease-free, racially and
gender-mixed cohort, which will be of general use for assessing indi-
vidual patients’ data. Finally, we tested for potentially important
differences between groups that would need to be accounted for in
future studies.
Descriptive data. Our subjects (n 5 89) comprised 39 males and 50
females. They had a median age of 26 years (interquartile range
(IQR) 5 8 years) and their median mass-corrected maximal aero-
bic capacity was 36 mL kg
(IQR 5 13 mL kg
Their BMI (separated by gender) was 23.4 6 0.4 kg m
and 22.3 6 0.3 kg m
(females). Fasting glucose (GLU) was 86 6
1 mg/dl (4.78 6 0.06 mmol/l) and fasting insulin (INS) was (median
6 IQR) 3.9 6 3.3 mU/mL.
Normal values and ranges for resting quadriceps muscle phos-
phorus metabolites. Table 2 gives the means and standard errors
for each phosphorus metabolite that was directly measured. We did
not attempt absolute quantification, so the values are given in two
forms –a ratio relative to the c-phosphorus of ATP, and as a milli-
molar concentration based on an assumed concentration of ATP in
(see Methods).
Resting muscle pH was 7.07 6 0.003. We also calculated several
indices of muscle respiratory capacity. Although these would some-
what depend upon the exact protocol used, the means for each were:
5 21.9 6 0.6 s; k 5 0.034 6 0.001; ATPmax 5 1.13 6
0.03 mM s
; Qmax 5 0.70 6 0.03 mM s
We tested for differences in all these values by both gender and
ethnicity, to highlight important physiological differences that may
need to be accounted for in future studies. The only significant dif-
ference by group was resting muscle pH between genders: Male (n 5
39): 7.09 6.004, vs. Female (n 5 50): 7.06 6.003; p , 10
Predicted relationships. As expected, there were significant corre-
lations between all MR measures of in vivo muscle mitochondrial
function and both relative (mass-corrected) and absolute VO
(Figure 1).
Interestingly, the respiratory exchange ratio (RER) at VO
max (an
indicator of the contribution of glycolytic metabolism at maximal
exercise) and the initial rate of PCr resynthesis after cessation of
plantar flexion exercise (an indicator of the contribution of oxidative
metabolism at the end of plantar flexion exercise) were inversely
correlated (r 52.25, p 5 .023, n 5 86), showing a link between
the contributions of various metabolic pathways in muscle and at the
whole-body level.
Table 2
Mean values of phosphorus metabolite ratios and con-
centrations in a young, ethnically-diverse and gender-mixed
cohort. (n 5 89)
Metabolite Relative to c-ATP mM
PCr 4.1 6 0.03 33 6 0.3
Pi 0.47 6 0.007 3.9 6 0.06
PME 0.58 6 0.05 4.8 6 0.4
PDE 0.76 6 0.05 6.2 6 0.4
NADP 0.12 6 0.01 1.0 6 0.08
r = -.45, p < .0001
10 20 30 40
1/2t (s)
10 20 30 40
1/2t (s)
0.01 0.03 0.05 0.07
0.01 0.03 0.05 0.07
0.01 0.51 1.01 1.51
Qmax (mM/s)
0.01 0.51 1.01 1.51
Qmax (mM/s)
0.5 1 1.5 2
ATPmax (mM/s)
0.5 1 1.5 2
ATPmax (mM/s)
r = -.52, p < .0001
r = .44, p < .0001 r = .53, p < .0001
r = .22, p = .039 r = .26, p = .015
r = .41, p = .0001 r = .53, p < .0001
Figure 1
Correlations between
P-MRS estimates of mitochondrial
function and either absolute (left column) or mass-adjusted (right
column) VO
max (
5 86). Absolute VO
max is in L min
, mass-
adjusted VO2max is in mL min
Table 1
Predicted correlations between
P-MRS measures of skeletal muscle biology and whole-body physiology
xyDirection Hypothesis Refs Hypothesis supported?
max Mitochondrial function will positively correlate with both
absolute and mass-corrected VO
Yes (p , .0001)
Qmax, ATPmax VO
max 1
Resting [PCr] Delta efficiency Resting muscle [PCr] will negatively correlate with delta
Yes (p 5 .046)
Resting VO
, TDEE Mitochondrial function will positively correlate with both
resting VO
and total daily energy expenditure
TDEE: Yes (p 5 .032)
:NoQmax, ATPmax Resting VO
, TDEE 1
SCIENTIFIC REPORTS | 3 : 1182 | DOI: 10.1038/srep01182 2
We had predicted that, due to the relationship between resting
muscle phosphocreatine (PCr) content and muscle fibre-type distri-
bution, we would observe an inverse correlation between [PCr] and
delta efficiency (DE). This was found, with resting [PCr] significantly
inversely correlated with DE (r 5 .22, p 5 .046, n 5 87). Although
DE did not correlate with Qmax or ATPmax, it was significantly
correlated with k (r 5 .23, p 5 .037, n 5 86) and negatively correlated
with PCr
(r 52.29, p 5 .006, n 5 86), suggesting that better
mitochondrial function was associated with higher delta efficiency.
Total daily energy expenditure (TDEE), but not resting oxygen
uptake, was correlated with measures of muscle mitochondrial func-
tion. Both k (r 52.22, p 5 .038, n 5 86) and PCr
(r 52.23, p 5
.032, n 5 86) were significantly related to TDEE, signifying a positive
relationship between muscle mitochondrial function and daily
energy expenditure (Figure 2).
Data mining untargeted correlations. There were a number of
unpredicted and highly significant (p , 10
) correlations between
Biodex measures of muscle strength and intramuscular pH. On
closer examination, these were the result of the previously obser-
ved difference in muscle pH between genders, highlighting the
importance of accounting for this difference in future studies. In
addition, there were two significant correlations between MR mea-
sures of muscle chemistry and muscle performance. First, there was a
significant correlation between muscle PCr/Pi and average power
during the five 60 deg s
leg extensions (r 52.29, p 5 .008).
Second, there were significant correlations between all measures of
mitochondrial function and power output during the last five
extensions of the fatigue index protocol (for example, PCr
: r 5
2.38, p 5 .0004).
We performed 63 pairwise correlations of anthropometry vs. MRS
data (Anthropometry: waist, waist/hip, weight, BMI, percent body
fat, fat mass and lean mass (seven variables); vs. MRS: PCr, Pi, ADP,
, ATPmax and Qmax (nine variables)).
These variables are clearly not independent, making assessment of
significance problematic, although both Bonferroni and Sidak cor-
rections suggest an adjusted significance threshold of approximately
p , .001. There were correlations between all measures of mitochon-
drial function and both percent body fat and fat mass (Figure 3).
Although no single correlation had p , .001, several were borderline
(p 5 .0032.004) and, taken together, these correlations provided
persuasive evidence of a link between fatness and mitochondrial
function. There were no other significant correlations between any
measures of anthropometry or body composition and muscle bio-
chemistry or mitochondrial function. Likewise, there were no sig-
nificant correlations between any measured plasma metabolite
(HDL, LDL, TAG, CHOL, GLU, INS) and muscle phosphorus chem-
istry or mitochondrial function.
Data mining principal components analysis. In order to look for
unpredicted patterns in the
P-MRS data we performed principal
components analysis (PCA) on the known components of the resting
muscle spectra. Two principal components (PC1 and PC2) were
sufficient to explain 84% of the variance in the data, suggesting
substantial collinearity between phosphorus resonances. This colli-
nearity was not due to resonances from the same metabolite (for
example, the expected correlation between the resonances from the
a, b and c phosphorus nuclei of ATP). In particular, the loading
coefficients for PC1 showed significant collinearity between phos-
phocreatine and phospho- di- and mono-esters. However, as can be
seen from Figure 4, there was no obvious clustering or patterning in
the dimensionally-reduced data.
Since a period of rapid technical discovery and development in the
1980s, nuclear MRS has provided an unrivalled tool to study the
biochemistry (and biophysics) of living cells in real time. Now that
the reconciliation of cell and tissue data with whole organism
10 20 30 40
TDEE (kcal)
PCr1/2t (s)
Figure 2
Correlation between PCr
(see main text for details) and
total daily energy expenditure (TDEE); relationship is significant at
, .05.
10 20 30 40
1/2t (s)
10 20 30 40
1/2t (s)
0.01 0.03 0.05 0.07
0.01 0.03 0.05 0.07
0.01 0.51 1.01 1.51
Qmax (mM/s)
0.01 0.51 1.01 1.51
Qmax (mM/s)
0.5 1 1.5 2
ATPmax (mM/s)
0.5 1 1.5 2
ATPmax (mM/s)
r = .31, p = .004 r = .26, p = .017
r = -.32, p = .003 r = -.28, p = .009
r = -.27, p = .013 r = -.21, p = .050
r = -.31, p = .004 r = -.29, p = .007
Figure 3
Correlations between
P-MRS estimates of mitochondrial
function and percent body fat (left column) or fat mass (right column, in
kg) (
5 86).
SCIENTIFIC REPORTS | 3 : 1182 | DOI: 10.1038/srep01182 3
physiology is becoming a priority, MR will continue to offer unique
advantages over more destructive methods (mass spectrometry) or
those that can only be used under very specific circumstances (live
cell imaging). In this paper, we report our results from an unusual
data set combined
P MRS and applied physiology measurements
made on a particularly large cohort of healthy humans of mixed
ethnicity and gender. As a result, we were able to test a number of
hypotheses relating muscle biochemistry to whole body physiology
in normal humans with unusually high statistical power.
In general we found that muscle biochemistry was reflected in
whole body physiology in a manner that was consistent with existing
theory and data, and with our a priori hypotheses. For example,
based on earlier findings in smaller, more specialized cohorts
we predicted that, in a large cohort of mixed gender and ethnicity,
whole-body VO
max would correlate with muscle respiratory ca-
pacity, measured using
P-MRS. Our hypothesis was supported:
(negatively) and Qmax and ATPmax (positively) correlated
with both absolute and mass-corrected VO
max (Figure 1). Al-
though the correlation coefficients appear low this supports the
notion that VO
max is a complex measure of integrative physiology
that is only partly determined by mitochondrial function. Indeed,
the magnitude of these relationships suggested that approximately
20–25% of the variation in mass-corrected VO
max could be ex-
plained by variations in muscle respiratory capacity, a finding that
was also consistent with previous work. The remainder of the vari-
ation in VO
max was presumably due to extra-muscular physio-
logical factors such as diffusive and/or convective oxygen delivery
(and possibly the neuropsychological factor of motivation). The
halftime of PCr recovery appeared to be a more robust indicator
of VO
max than either of the extrapolated/calculated values (Qmax
or ATPmax), while ATPmax was a better predictor of both absolute
and mass-adjusted VO
max than was Qmax. This may however
simply reflect the greater mathematical/theoretical simplicity of
ATPmax, resulting in reduced noise (but not necessarily better
reflecting underling physiology).
Delta efficiency during cycling is related to the percentage of oxid-
ative fibres in the locomotor muscles
. Given that PCr concentration
in oxidative fibres is lower than in glycolytic fibres
and therefore
mean PCr concentration in mixed muscle reflects fibre type distri-
, we tested the hypothesis that muscle PCr content, mea-
sured using
P-MRS, would be positively correlated with DE (the
slope of VO
against work rate at work rates below VT
) during
incremental exercise testing. In this case our hypothesis was sup-
ported by the data (r 5 .22, p 5 .046), strengthening not only the
earlier (but disputed
) findings of Coyle et al.
, but also the utility
of [PCr] as an index of muscle fibre type distribution between
There is an ongoing debate as to whet her muscle mitochondrial
function and delta efficiency are related (cf the introduction in
). For
example, Lucia et al. reported an inverse relationship between
max and efficiency (during a steady-state exercise task)
gesting an inverse relationship between muscle oxidative capacity
and delta efficiency. However, this study was carried out in a
highly-selective group (world-class endurance cyclists), so the results
may have reflected not ‘normal’ physiology but the result of a select-
ive process within tight constraints - those with marginal aerobic
capacity may have been nevertheless able to succeed as professional
road cyclists precisely because they were highly efficient. Thus we
decided to test for a relationship between delta efficiency (as defined
above) and muscle respiratory capacity. We found a significant
inverse correlation between PCr
and DE (r 52.29, p 5 .006),
such that better mitochondrial function was associated with higher
delta efficiency. Our findings are at odds with those of Hunter et al.
), who reported an inverse correlation between exercise economy
(defined as the oxygen cost of a given walking velocity during steady-
state walking) and muscle respiratory capacity (measured, as here,
P-MRS). However, these disparate results might be recon-
ciled by differences in both methodology (DE compared with VO
during steady state walking) and cohort (pre-menopausal women
only, compared with a mixed gender and ethnicity cohort). Given
that we also found a positive correlation between PCr concentration
and delta efficiency, it is tempting to suggest that both correlations
([PCr] vs. DE and PCr
vs. DE) are the result of underlying differ-
ences in fibre type distribution. Unfortunately it is a limitation of
the present study that we did not measure fibre type distribution
Given an earlier finding that resting metabolic rate was positively
related to muscle ATP-synthase content
, we decided to test our
hypothesis that resting metabolic rate was related to muscle respir-
atory capacity. The data supported our hypothesis: PCr
was sig-
nificantly related to total daily energy expenditure (TDEE) (r 5
2.23, p 5 .030) (Figure 2). Therefore those with higher muscle
respiratory capacity had increased daily energy expenditure. The
nature of causality between these variables is, as yet, unclear. One
might reasonably hypothesize that increased daily activity led to
improved muscle respiratory capacity and this view is somewhat
supported by the absence of a positive relationship between resting
and muscle oxidative capacity.
Intriguingly, we found good evidence linking body composition to
muscle mitochondrial function. Again, the direction of causality here
is unclear. One would expect those who are more active to use more
energy and therefore to have higher muscle (and whole-body) res-
piratory capacities and less body fat. By Occam’s razor, this explana-
tion would seem to be more likely than the concept that increased fat
mass was somehow inhibiting muscle mitochondrial function
However, there was a very significant relationship between resting
oxygen uptake and TDEE (r 5 .43, n 5 86, p 5 0.00003), such that
one might also reasonably argue that some subjects were genetically
predisposed to having a higher resting energy cost, better muscle
mitochondrial function and therefore had a higher TDEE and lower
fat mass.
There were no consistent differences in muscle biochemistry
between genders or by ethnicity, with one exception. Muscle pH
was significantly lower in females compared with males (7.06
6.003 vs. 7.09 6.004, p , 10
). Although this appears at first to
be only a minor difference it amounts to a 7% higher proton con-
centration in males. The determinants of intracellular pH setpoint in
themselves remain far less well understood in skeletal muscle than in
cardiac muscle
, although the principle is that basal pH is deter-
mined by the steady state balance between a number of sarcolemmal
1000 500 0 500 1000
Scores on PC1 (71%)
Scores on PC2 (13%)
Figure 4
Principle components analysis scores biplot for
P-MRS data
from 87 subjects of mixed gender and ethnicity.
SCIENTIFIC REPORTS | 3 : 1182 | DOI: 10.1038/srep01182 4
proton exchangers and uniporters. Furthermore the significance of
this pH difference by gender is difficult to determine, although it
must be considered if comparing any pH-dependent variable
between groups that are unbalanced for gender. Beyond this, our
unsupervised multivariate analysis of the resting spectra revealed
no unexpected grouping.
This dataset represented a unique opportunity to study relation-
ships between non-invasive measures of muscle biochemistry with
whole-body exercise-physiology, with particular reference to oxid-
ative metabolism, in a group of normal subjects large enough to
establish these with high statistical power. Unlike earlier studies in
this area we did not study the effects of formal exercise training, and
especially not adaptions in professional athletes where, as noted in
the Discussion, selection effects can give apparently paradoxical
results. The results broadly confirmed our pre-specified hypotheses
that a substantial fraction of variability in whole body aerobic fitness
could be explained by variations in muscle mitochondrial function,
and that [PCr] correlated with delta efficiency in a way consistent
with its being a surrogate for average muscle fibre type composition.
We also found that muscle mitochondrial function was a positive
predictor of both daily energy expenditure and leanness. Finally, we
found that there was a previously unreported gender difference in
resting muscle pH. These findings are important for those with an
interest in the genetic and physiological determinants of whole-body
aerobic capacity, as well as those with broader interests in human
physiology and health (given that VO
max is a strong predictor of
all-cause mortality
Ethical approval. This study is a registered clinical trial (
identifier:NCT00945633) and was approved by the ethical review board of the
Pennington Biomedical Research Institute. Written informed consent was obtained
from all participants prior to the study, which conforms to the guidelines in the
Declaration of Helsinki.
Participants. In brief, the InSight study is a prospective, longitudinal, clinical study
using an epidemiological approach to identify dietary, physiological, genetic and
behavioral determinants of unhealthy weight gain over a ten-year period. Inclusion
criteria for the InSight study were as follows: men and women aged 20–35 y, BMI ,
27.5 kg m
, fasting blood glucose , 126 mg/dl (7.0 mmol L
); clinical and physical
examination to confirm health prior to acceptance (exclusion criteria detailed below).
Participants were 50% black and 50% Caucasian with an equal gender mix within
each ethnic group. The design was intended to allow the identification of subtle
differences in physiology and behavior that tracked along the level of familial risk.
Exclusion criteria included the following: history of diabetes, history of obesity
(BMI $ 30 kg/m
), history of known inherited medical conditions that might
influence future health status, current or planned medication usage that might
influence future health status, prior serious injuries/surgeries that might influence
future health status, women who were pregnant or breastfeeding at recruitment
(pregnancy did not cause already-enrolled subjects to be removed from the study, nor
were pregnancy or childbirt h considered adverse events), women who were
, 6 months post-partum, or women who had discontinued breastfeeding ,
3 months prior to screening, history of cancer (including skin cancer) within 5 years,
history of organ transplant, previous diagnosis with HIV, Hepatitis B or C, or
tuberculosis, abuse of alcohol or illegal drugs, abnormal ECG, presence of a pace-
maker, defibrillator, or implanted metal, history of eating disorders and abnormal
psychological scores at screening.
Participants entering the InSight study were initially screened over 3 visits to the
Pennington Biomedical Research Center (PBRC) in order to determine eligibility and
obtain height, weight, waist and hip measurements, vital signs, a resting ECG, a
medical history and physical examination, lab work (full blood count and standard
blood chemistry (high-density lipoproteins (HDL), low-density lipoproteins (LDL),
triacylglycerols (TAG), cholesterol (CHOL), fasting glucose (GLU) and fasting
insulin (INS)), urinalysis, urine drug screen) and the completion of various ques-
tionnaires, including a PBRC Screening Health Questionnaire and demographic
questionnaire. During these visits the participants were assessed on a range of skeletal
muscle and whole-body characteristics including cardiorespiratory fitness (VO
muscle strength and performance, anthropometry and
P-MRS measures of muscle
chemistry and metabolism (in 89 participants), the methods for which are described
Anthropometry. We measured body mass in a gown after voiding and waist
circumference using a standardized protocol. Height was measured on a calibrated
stadiometer. Percent body fat, fat mass and lean mass were measured on a Hologic
Dual Energy X-ray Absorptiometer (DXA, QDR 4500A, Hologic, Inc., Waltham,
MA.) During the DXA, scan, participants were asked to lie on a table wearing a
hospital gown for 4–6 minutes during the scanning process. Two distinct energies
were used to determine body mineral and soft tissue content. An attenuation ratio was
determined from a known tissue content. Variations of the attenuation ratio
determined the fat content of the tissue at each pixel thereby calculating the
percentage body fat. The scans were analyzed with QDR software for Windows V11.1.
The coefficient of variation (CV) for the body composition measurement of lean
mass, fat mass and percentage body fat was 0.8%, 1.6%, and 1.7%, respectively.
Exercise testing. Maximal cardiorespiratory testing. We performed all maximal
cardiorespiratory exercise testing (VO2max) in the PBRC Exercise Biology Exercise
Testing Core. During the VO
test, we monitored each participant using a 12-lead
electrocardiogram (Q-Stress, Quinton Instruments, Seattle, WA). All exercise tests
were performed using a standardized graded exercise testing protocol administered
on a treadmill (Trackmaster 425, Newton, KS). The participants performed a brief,
self-selected walking speed as a warm-up, then the test was initiated a speed of 1.7
mph and 1% grade. Each testing stage lasted two minutes and speed and/or grade
were increased every two minutes until the participant reached a state of temporary
exhaustion and could no longer continue the test. The determination of VO
required the achievement of two of the three following conditions:
(1) a plateau or rise in VO
(L min
) , 100 mL,
(2) RER . 1.1, and
(3) a maximal heart rate within 10 beats/min of the participants age predicted
During the testing all ventilatory measures were collected breath-by-breath and
then ave raged into one minute epochs for later analysis to determine VO
max and
underlying measures of exercise capacity and function. These included maximal heart
rate, maximal oxygen uptake (VO
), pulmonary ventilation (VE), ventilatory
equivalents for oxygen (VE/VO
), carbon dioxide (VE/CO
), end-tidal partial pres-
sure of oxygen (PETO
) and carbon dioxide (PETCO
), and the calculation of delta
efficiency (DE). DE was calculated as the reciprocal of the slope of the relationship
between work accomplished per minute and energy expended per minute during the
incremental exercise test
Muscular strength and endurance testing. After 5-min of low- to moderate-ef fort
warm-up exercise on a treadmill, concentric isokinetic strength of the knee flexors
and knee extensors (right leg) were assessed on a Biodex System 3 dynamometer
(Biode x Medical Systems, Shirley, New York). Two tests were performed in or der
to determine leg strength and endurance: five 60u/second isokinetic knee exten-
sions and flexions, followed by thirty 180u/second isokin etic knee extensio ns and
knee flexion. Each te st was preceded by two rehearsal trial s using 3–5 repetitions
and all tests were interleaved with one-min ute recovery peri ods. Average power,
average work and peak torque were then calculated for each test and expressed in
absolute terms. We further calculated a muscle fatigue index from the 180u/second
isokinetic test that was defined as the pe rcent decrease in peak torque during the
30 repetition set as follows:
Percent decrease 5 1002[(mean of the last 5 repetitions)/mean of the highest
consecutive 5 repetitions) 3 100].
Total daily energy expenditure. Doubly-labelled water (DLW) was used to obtain an
accurate measure of total daily energy expenditure (TDEE) and, during a period of
energy balance, TDEE is equal to energy intake
. TDEE was adjusted for changes in
body weight following the methods outlined by Schulz et al
. Briefly, we assumed that
2/3 of any change in body weight was metabolic and 1/3 was water. In addition, we
assumed that 3/4 of this change in metabolic weight was fat mass (FM) and 1/4 was
fat-free mass (FFM)
. For weight loss, we assumed 9 kcal/g of fat mass (FM) and
1 kcal/g of fat free mass (FFM). For weight gain, we assumed 13.2 kcal/g of FM and
2.2 kcal/g of FFM
. These methods are consistent with those used in Tataranni
et al
Although these methods are commonly used to measure energy intake with DLW
when energy balance is not present, the PBRC’s Human Physiology Laboratory
recently found that correcting TDEE data for change in energy stores results in less
accurate measures of energy intake during periods of significant calorie restriction
. Therefore, we used DLW when participants were essentially weight stable.
We also closely evaluated changes in energy stores and corrected for minor deviations
in energy balance using the methods outlined above. Based on previous research, we
are confident that the timing of the DLW measures ensured that we obtained accurate
measures of energy intake using DLW.
Participants were given a mixture of 1 part deuterium and 19 parts
O, followed by
a 100-ml of tap water used to rinse the dose container. All subjects received a dose of
1.5 g/kg total body weight. The dose was given early in the morning after an overnight
fast and collection of two baseline urine specimens. Participants’ body weight was
measured and recorded. Two urine specimens were collected and discarded 1.5 hours
and 3 hours post dose. A urine specimen was collected 4.5 hours and 6 hours post
dose. Body weight also was measured and recorded in the morning on day 7, and a
morning urine sample was collected in the clinic (another sample was collected prior
to the participant leaving the clinic on day 7). This procedure was repeated on Day 14,
along with a DXA scan (see above).
SCIENTIFIC REPORTS | 3 : 1182 | DOI: 10.1038/srep01182 5
Magnetic resonance spectroscopy protocol. Muscle phosphorus metabolite
concentrations, muscle pH and mitochondrial function were determined using a 3T
GE Signa MNS magnet (GE, Milwaukee, WI) and
P-tuned surface coil positioned
over the distal vastus lateralis. Following the acquisition of a fully relaxed spectrum,
P spectra were acquired every 6 seconds at rest and continuously during a 24-, 30- or
36-sec ballistic exercise protocol that comprised ‘kicking’ against Velcro straps
positioned tight across the leg and thigh. To correct for the effects of partial saturation
due to the short TR (6 sec) we calculated correction factors from the fully relaxed
spectrum and 6-sec spectra acquired from the same volume of resting muscle.
Exercise time and intensity was targeted to reduce muscle phosphocreatine (PCr)
content by 1/3 to 1/2 from resting values whilst ensuring pH did not fall below 6.8
(to avoid well-known technical complications in interpreting estimates of
mitochondrial function in the presence of large pH change
Magnetic resonance data processing. Spectra were processed using jMRUI version
and quantified using a non-linear least squares algorithm (AMARES
). Resting
ATP and total creatine concentrations were assumed to be 8.2 mM
42 mM
respectively. These commonly-used concentrations are based on
extensive published values, and are reliable in healthy humans
. The chemical shift of
the inorganic phosphate (Pi) peak relative to phosphocreatine (PCr) (s, in parts per
million) was used to determine intracellular pH. The myocellular free adenosine
diphosphate (ADP) concentration was calculated making the standard assumption
that that the creatine kinase reaction is at equilibrium, and allowing for changes in
Indices of mitochondrial function. We calculated four commonly-used indices of
muscle mitochondrial function. The first, the halftime of PCr recovery after moderate
exercise (PCr
), is an inverse index of in vivo mitochondrial function, and was
determined by fitting a monexponential equation expressing PCr concentra tion
(normalized to resting values) as a function of time (t), PCr (t) 5 1 2 exp (2kt).
Microsoft Solver was used to estimate the second the rate-constant (k) - by least-
squares fit, from which PCr
5 ln (2)/k
. The third, ATPmax, was calculated using
the PCr recovery time constant (t 5 1/k) (derived as above) and the concentration of
PCr in the same muscle at rest: ATPmax 5 resting [PCr]/t
. The final index of
mitochondrial function, Qmax, is based on the sigmoidal relationship between [ADP]
and respiratory rate (V)
and is calculated from values for both [ADP] and V at
exercise cessation: Qmax 5 {V 1 (K
}, where K
is the concentration of
[ADP] at which respiration is half-maximal (25 mM) and n is a Hill coefficient 5
. One subject had values for these estimates of mitochondrial function that were
more than five standard deviations from the group mean, and was removed from this
part of the analysis.
Statistics. As much of the analysis presented herein relies on measures of correlation,
and because Pearson’s r is a distribution-free statistic (in other words, it does not rest
on the assumption of a normal distribution
), we did not routinely check our data for
normality. However, in cases where this assumption was made (for example, testing
for differences between males and females) the assumption of normality was checked
using Shapiro-Wilk and violations were accounted for if necessary.
Where an effect was predicted in ad vance (for example, the positive relationship
between muscle mitochondrial function and whole-body VO
max), statistical
significance was accepted at p , .05. However, to lessen the risk of Type I errors
when relationships were not predic ted in advance, the significance threshold was
adjusted as described in the main text, although statistically marginal results were
given extra credence if accompanied by a ro bust biological explanation. Values are
reported as mean 6 standard error. Data that we re not normally-distributed are
reported as median 6 interquartile range (IQR). All statistical analyses except for
principal components analysis (PCA) were conduc ted using PASW 18.0 (SPSS
Inc., Chicago, Illinois, USA). For PCA, the data were processed in AMARES as
described above. However, rather than normalize the spectra to the intensity of
the c-ATP resonance we normalized each spectrum by the sum of all peaks
(integral-sum normalization). The data were mean-centered and Pareto-scaled
before being analyzed using PCA, all in PLS Toolbox 6.2 (Eigenvector Research
Inc., Wenatchee, US).
1. Kemp, G. J. & Radda, G. K. Quantitative interpretation of bioenergetic data from
P and
H magnetic resonance spectroscopic studies of skeletal muscle: an
analytical review. Magn Reson Q 10, 43–63 (1994).
2. Coyle, E. F., Sidossis, L. S., Horowitz, J. F. & Beltz, J. D. Cycling efficiency is related
to the percentage of type I muscle fibers. Med Sci Sports Exerc 24, 782–788 (1992).
3. Livingstone, M. B. & Black, A. E. Markers of the validity of reported energy intake.
J Nutr 133 Suppl 3, 895S–920S (2003).
4. Schoeller, D. A. How accurate is self-reported dietary energy intake? Nutr Rev 48,
373–379 (1990).
5. Schulz, L. O., Alger, S., Harper, I., Wilmore, J. H. & Ravussin, E. Energy
expenditure of elite female runners measured by respiratory chamber and doubly
labeled water. J Appl Physiol 72, 23–28 (1992).
6. Forbes, G. B. in Human body composition. Growth, aging, nutrition, and activity.
209–247 (Springer-Verlag, 1987).
7. Forbes, G. B., Brown, M. R., Welle, S. L. & Lipinski, B. A. Deliberate overfeeding in
women and men: energy cost and composition of the weight gain. Br J Nutr 56,
1–9 (1986).
8. Pullar, J. D. & Webster, A. J. The energy cost of fat and protein deposition in the
rat. Br J Nutr 37, 355–363 (1977).
9. Spady, D. W., Payne, P. R., Picou, D. & Waterlow, J. C. Energy balance during
recovery from malnutrition. Am J Clin Nutr 29, 1073–1088 (1976).
10. Tataranni, P. A. et al. Body weight gain in free-living Pima Indians: effect of energy
intake vs expenditure. Int J Obes Relat Metab Disord 27, 1578–1583 (2003).
11. de Jonge, L. et al. Validation study of energy expenditure and intake during calorie
restriction using doubly labeled water and changes in body composition. Am J
Clin Nutr 85, 73–79 (2007).
12. Jubrias, S. A., Crowther, G. J., Shankland, E. G., Gronka, R. K. & Conley, K. E. Vol.
553, 589–599 (2003).
13. Naressi, A., Couturier, C., Castang, I., de Beer, R. & Graveron-Demilly, D. Java-
based graphical user interface for MRUI, a software package for quantitation of in
vivo/medical magnetic resonance spectroscopy signals. Comput Biol Med 31,
269–286 (2001).
14. Vanhamme, L., van den Boogaart, A. & Van Huffel, S. Improved method for
accurate and efficient quantification of MRS data with use of prior knowledge.
J Magn Reson 129, 35–43 (1997).
15. Kemp, G. J., Meyerspeer, M. & Moser, E. Absolute quantification of phosphorus
metabolite concentrations in human muscle in vivo by
P MRS: a quantitative
review. NMR Biomed 20, 555–565 (2007).
16. Golding, E. M., Teague, W. E. & Dobson, G. P. Adjustment of K’ to varying pH and
pMg for the creatine kinase, adenylate kinase and ATP hydrolysis equilibria
permitting quantitative bioenergetic assessment. J Exp Biol 198, 1775–1782
17. Braun, M. Differential Equati ons and their Applications. 4 edn, (Springer,
18. Amara, C. E. et al. Mitochondrial function in vivo: Spectroscopy provides window
on cellular energetics. Methods 46, 312–318 (2008).
19. Jeneson, J. A. L., Wiseman, R. W., Westerhoff, H. V. & Kushmerick, M. J. The
signal transduction function for oxidative phosphorylationiIs at least second
order in ADP. J. Biol. Chem. 271, 27995–27998 (1996).
20. Rodg ers, J. L. & Nicewander, W. A. Thirteen ways to look at the correlation
coefficient. The American Statistician 42, 59–66 (1988).
21. Larson-Meyer, D. E., Newcomer, B. R., Hunter, G. R., Hetherington, H. P.
& Weinsier, R. L. 31P MRS measurement of mitochondrial function in skeletal
muscle: reliability, force-level sensitivity and relation to whole body maximal
oxygen uptake. NMR Biomed 13, 14–27 (2000).
22. McCully, K. K., Fielding, R. A., Evans, W. J., Leigh, J. S. & Posner, J. D.
Relationships between in vivo and in vitro measurements of metabolism in
young and old human calf muscles. Journal of Applied Physiology 75, 813–819
23. Conley, K. E. et al. Ageing, muscle properties and maximal O(2) uptake rate in
humans. J Physiol 526 Pt 1, 211–217 (2000).
24. Wyss, M. & Wallimann, T. Creatine metabolism and the consequences of creatine
depletion in muscle. Mol Cell Biochem 133–134, 51–66 (1994).
25. Madhu, B., Lagerwall, K. & Soussi, B. Phosphorus metabolites in different muscles
of the rat leg by 31P image-selected in vivo spectroscopy. NMR Biomed 9, 327–332
26. Takahashi, H. et al. Relationships between fiber composition and NMR
measurements in human skeletal muscle. NMR Biomed 9, 8–12 (1996).
27. Wasserman, K., Hansen, J. E., Sue, D. Y., Stringer, W. W. & Whipp, B. J. in
Principles of Exercise Testing and Interpretation (Lipincott Williams and Wilkins,
28. Mallory, L. A. et al. Influence of peak VO
and muscle fiber type on the efficiency
of moderate exercise. Med Sci Sports Exerc 34, 1279–1287 (2002).
29. Hunter, G. R. et al. Inverse relationship between exercise economy and oxidative
capacity in muscle. Eur J Appl Physiol 94, 558–568 (2005).
30. Lucia, A., Hoyos, J., Perez, M., Santalla, A. & Chicharro, J. L. Inverse relationship
between VO2max and economy/efficiency in world-class cyclists. Med Sci Sports
Exerc 34, 2079–2084 (2002).
31. Edwards, L. M. et al. Uncoupling protein 3 inversely correlates with resting
metabolic rate in fit, young men. Proc Physiol Soc 11, C49 (2008).
32. Vaughan-Jones, R. D., Spitzer, K. W. & Swietach, P. Intracellular pH regulation in
heart. J Mol Cell Cardiol 46, 318–331 (2009).
33. Kampert, J. B., Blair, S. N., Barlow, C. E.& Kohl, H. W., 3rd. Physical activity,
physical fitness, and all-cause and cancer mo rtality: a prospective study of men
and women. Ann Epidemiol 6, 452–457 (1996).
34. Mogensen, M., Bagger, M., Pedersen, P. K., Fernstrom, M. & Sahlin, K. Cycling
efficiency in humans is related to low UCP3 content and to type I fibres but not to
mitochondrial efficiency. J Physiol 571, 669–681 (2006).
Author contributions
These experiments were conducted at PBRC. Data analysis was conducted at the University
of Tasmania Medical School and the School of Biological Sciences, University of Essex. SRS
and CPE conceived and designed of the experiments. LME, GJK, RMD, HF, JTW and CPE
collected, analysed and interpreted the data. LME, GJK, RMD, JTW, SRS and CPE drafted
the article or revised it critically for important intellectual content.
SCIENTIFIC REPORTS | 3 : 1182 | DOI: 10.1038/srep01182 6
Additional information
Competing financial interests: The authors declare no competing financial interests.
License: This work is licensed under a Creative Commons
Attribution-NonCommercial-NoDerivs 3.0 Unported License. To view a copy of this
license, visit
How to cite this article: Edwards, L.M. et al. Integratin g muscle cell biochemistry and
whole-body physiology in humans:
P-MRS data from the InSight trial. Sci. Rep. 3, 1182;
DOI:10.1038/srep01182 (2013).
SCIENTIFIC REPORTS | 3 : 1182 | DOI: 10.1038/srep01182 7
  • [Show abstract] [Hide abstract] ABSTRACT: Magnetic resonance spectroscopy (MRS) can give information about cellular metabolism in vivo which is difficult to obtain in other ways. In skeletal muscle, noninvasive (31) P MRS measurements of the post-exercise recovery kinetics of pH, [PCr], [Pi] and [ADP] contain valuable information about muscle mitochondrial function and cellular pH homeostasis in vivo, but quantitative interpretation depends on understanding the underlying physiology. Here, by giving examples of the analysis of (31) P MRS recovery data, by some simple computational simulation, and by extensively comparing data from published studies using both (31) P MRS and invasive direct measurements of muscle O2 consumption in a common analytical framework, we consider what can be learnt quantitatively about mitochondrial metabolism in skeletal muscle using MRS-based methodology. We explore some technical and conceptual limitations of current methods, and point out some aspects of the physiology which are still incompletely understood. This article is protected by copyright. All rights reserved.
    Full-text · Article · Apr 2014
  • [Show abstract] [Hide abstract] ABSTRACT: Objective: To investigate the relationship among moderate-to-vigorous physical activity (MVPA), mental health symptoms, and weight gain in young adults. Methods: Fifty-nine healthy men and women (age 26.8 ± 4.7 years, body mass index [BMI] 22.4 ± 2.3 kg/m(2) ) were assessed at baseline and year 2. Weight, BMI, fat mass by dual-energy X-ray absorptiometry, anxiety and depression by the SCL-90, and MVPA by accelerometry were assessed. Sex-stratified linear regression models were used to examine associations between baseline anxiety/depression and body composition change, then repeated controlling for baseline MVPA. Results: Overall, weight, BMI, fat mass, and subcutaneous adiposity significantly increased at year 2 (P < 0.05). For women, a higher depression and anxiety score predicted increased weight, BMI, fat mass, and subcutaneous adiposity (P < 0.05). Controlling for MVPA attenuated these associations to nonsignificance. For men, MVPA did not alter the associations between anxiety and increased weight, BMI, fat mass, and subcutaneous adiposity (P < 0.05) or the associations between depression and decreased weight, BMI, fat mass, and subcutaneous adiposity (P < 0.05). Conclusions: Anxiety and depression were related to body composition change for young adults. In women, MVPA attenuated these associations to nonsignificance. MVPA may buffer the adverse effects of depression and anxiety symptoms on young women's weight gain.
    Article · Jul 2016