Journal of
PHYSIOLOGICAL
Review
Local Temperature Changes and Human Skeletal Muscle Metabolism
Tiziano Binzoni1) and David Delpy2)
1) Departments of Radiology and Physiology, Faculty of Medicine, University of Geneva, Geneva, Switzerland
2) Department of Medical Physics and Bioengineering, University College London, London, UK.
Abstract The aim of this review is to describe the effects
induced by local temperature changes on human skeletal
muscle metabolism. More specifically, we will consider
t he inf luen ce of tem pe ra tur e o n the mec han ic al
properties of muscle contraction, on aerobic metabolism,
anaerobic metabolism and on the Lohmann reaction. The
text has been voluntarily organized on the basis of a
simple bioenergetic model describing the different
energy fluxes appearing in the muscle system. This
approach should better highlight some of the points that
still need to be investigated. Although it was not always
possible to restrict the discussion to human muscle, the
references report mainly data obtained directly on
humans or on isolated human fibres. A short comment on
skeletal muscle temperature measurement techniques,
on humans, is also included. J Physiol Anthropol 20 (3):
159-174, 2001 http://www.jstage.jst.go.jp/en/
Keywords: skeletal muscle, human, temperature,
metabolism
Introduction
M us c l e f u n c t i o n a n d m e t a b o l i s m a r e s t r o n g l y
influenced by local temperature (T) changes. At first
sight, one might have the impression that a decrease or
an increase in intramuscular T would simply slow down
or speed up the metabolic processes. Variation in T
however may influence physiological mechanisms in a
very surprising manner, showing behaviours that are far
from the intuitive thermodynamical laws. This means
that, for a given T change, different biochemical patterns
may be affected to different extents and that a decrease/
increase in intramuscular T does not necessarily result in
a decreased/increased activity of the reactions (Binzoni
et al., 2001). Moreover, the complexity of the relationship
existing between biochemical reactions, the nervous
system and T also influences muscle function, leading to
situations where for example, the maximum isometric
ANTHROPOLOGY
and Applied Human Science
voluntary contraction is not necessarily found at 37
degrees (Ranatunga et al., 1987).
The present review will focus on the effects of local T
changes on skeletal muscle energy metabolism, where the
variations in the subjects core T may in principle be
neglected (e.g., cooling a hand, a forearm or the legs). This
should enable the highlighting as much as is possible,
pure muscle T responses without the contribution of other
systemic thermoregulatory mechanisms. In this context,
only results coming from experiments performed on
humans have been taken into account.
The material has been organized using the viewpoint of
the bioenergetic approach (di Prampero, 1981). The
reason for this choice was dictated by the wish to organize
in a logical way, the large amount of data now accessible
via electronic data bases. This is of course only one of the
many possible approaches. In undertaking a review, one
must be aware that experimental data is not always of the
same quality because the techniques have changed over
the years. For example, some incorrect results may be
present, some measurements may not be comparable
because they are expressed in different units or they are
acquired under different conditions, etc. In this respect, to
make an analogy, the physiologist is like a meteorologist,
trying to discover knowledge from e.g. incomplete low
quality or inhomogeneous data bases. Thus, to make the
data usable, it is necessary to construct a data “warehouse”
by a careful choice of information extraction, data filtering
cleansing, data preprocessing, etc. (Bramer, 1999). Of
course, the structure of the data “warehouse” depends on
the final use to which the data is put and on the
mathematical model one want to apply to extract new
information. In the same manner, in the present review,
our choice of the bioenergetic model has dictated the
structure of the “physiological” data “warehouse”,
presented in the following sections. In the end however,
the model should help identify and generate specific
results as well as identifying unanswered questions.
Some of the technical difficulties, usually encountered
when making measurements on humans, will also be
159
160 Temperature Dependence of Muscle Metabolism
presented in the text, and we hope that this review will be
a stimulus for the development of new experimental
protocols.
The Basic “Laws” of the Bioenergetic Model
The idea that muscle function and, in fact, the function
of all living tissues, may be described by physical laws
appeared in the second half of the 19th century (for an
extended historical and scientific review on bioenergetics
see di Prampero (1981)). During this period, Helmoltz
(1847) showed that the principle of energy conservation
may be applied to living systems. Based on this principle,
many experimental studies trying to find which were the
muscle “energy yielding molecules”, were performed. By
the beginning of the present century, these studies had
resulted in the theory of muscle contraction proposed by
Hill and Meyerhof (Hill, 1965). This theory and other
work led to the basis of muscle bioenergetics. Today, one
can say that the main mechanisms of muscle function
have been clarified and, in particular, that the main
energy sources have been identified.
If one analyses Kushmerick’s (1983) review paper on
bioenergetics, it is easy to realize the complexity of the
information contained in the experimental data. In fact,
to describe the metabolic processes occurring in the cell,
it is nec essary to take int o a cco unt hundreds o f
biochemical reactions. The practical consequence of this
is that the results obtained in the last twenty years by the
scientists involved in the modelling of the metabolic
control (Fell, 1997) are extremely difficult to apply when
dealing with human subjects or patients. Thus, to solve
this problem, it is necessary to define a “meta-theory” that
highlights only the essential characteristics of the muscle
function. Moreover, it is crucial that the parameters
appearing in the theory are measurable in vivo. The
bioenergetic model, represents one of these possible
“meta-theories”, embodying the behavio ur of the
underlying biochemical reactions in a coherent way.
In this co ntext , t wo pr incip le s may intuitively
summarize the bioenergetic approach. The first can be
expressed as follows: the energy utilised by the muscles
during work (and rest) is stored as ATP (Lundsgaard,
1930). It is noteworthy, that this assertion does not
specify the way ATP is utilized, i.e. any theory is valuable
in th is ca se . T h us, to qua nt if y th e t o ta l e ner g y
expenditure of the muscle (at rest or during exercise), it
would be enough to measure the ATP consumption.
Actually, the ATP concentration is maintained constant
by the muscle in the majority of situations thanks to a
series of biochemical reactions that continuously
resynthesise ATP. As a consequence, in the majority of
the cases, ATP measurements alone do not provide much
information regarding the normal or pathological muscle
state. For this reason it is necessary to refine the first
principle with a second statement: The energy utilized
during muscular activity (and rest) mainly comes from
three sources: aerobic glycolysis, anaerobic glycolysis and
the Lohmann reaction (di Prampero, 1981). The sum of
the energy produced by these three sources yields the
total energy expenditure of the system. Now, the first
principle (and experimental measurements) implies that
the aerobic glycolysis, anaerobic glycolysis and the
Lohmann reaction must continuously replace the ATP
hydrolysed during the muscle contraction. Studies on
bioenergetics (di Prampero, 1981; Kushmerick, 1983;
Kemp and Radda, 1994) have shown that to evaluate the
energy produced by these three sources, and to obtain an
estimation of the number of ATP molecules hydrolysed
for a given muscular activity, it is sufficient to measure
the concentration changes of three key molecules. These
observations can be summarized in formal terms as:
• • • • •
E = [ATP] + α[PCr] + β[La] + γ[O2] [1]
•
where E (the dot is, as usual, the time derivative)
represents the instantaneous energy consumption of the
muscle per unit time (in ATP concentration units, mM s–1)
and [ATP], [PCr], [La] and [O2] are the concentrations of
ATP, phosphocreatine (PCr), lactate (La) and oxygen
(O2), respectively. The first term on the right hand side,
contributes only in extreme conditions when the muscle
cannot succeed in maintaining to maintain a constant
ATP concentration (e.g. fatigue). α, β and γ represent the
number of ATP molecules produced by one molecule of
[PCr], [La] and [O 2 ], respectively. Of course, α[PCr],
β[La] and γ[O2] give the ATP produced by the Lohmann
reaction, anaerobic glycolysis and aerobic glycolysis,
respectively. The essential point here, is that the
molecules appearing in equation [1] are all measurable,
directly or indirectly, in humans.
It is necessary at this point, to define the efficiency (η)
of the ATP utilization to develop mechanical work:
•
w
η= • [2]
E
•
where w represents the mechanical work (J s – 1 ).
Intuitively, η represents the number of ATP molecules
necessary to produce a given amount of mechanical
energy w. Efficiency is also expressed as:
•
w
η∆G = • [3]
E∆GATP
where ∆GATP is the free energy (J mol–1) of ATP. Note
that ∆GATP changes with T. When dealing with isometric
exercise, many authors prefer to use the term “cost”,
defined as:
•
E
ξ= [4]
PO
where PO represents the isometric force (N m–2).
 Binzoni, T and Delpy, D
Equation [1] may be considered as a metabolic
constraint that should enable one to calculate the
“missing” terms, i.e. the terms that in some protocols
cannot be measured experimentally. It must be noticed,
that equation [1] holds for one muscle fibre or for the
whole tissue because, in the latter case, it represents the
total energy of the system (the sum of all the fibres). The
validity and the limits of equation [1] are today well
established in normothermia and for different
pathological conditions (see reviews mentioned above).
However, very little is known about the behaviour of the
different terms in equation [1] and on the efficiency
parameters (equations [2], [3] and [4]) during muscle
hypo- or hyperthermia in humans.
Thus, to summarize one can combine equations [1] and
[2] and explicitly introduce the T dependence as follows:
• • • tension was significantly different only for the lowest
w = η(T){[ATP(T)] + α(T)[PCr(T)]
• •
+ β(T)[La(T)] + γ(T)[O2(T)]} [5]
•
w represents the mechanical power developed by a
subject. In the case of an isometric exercise it is
advantageous to write equation [1] in combination with
equation [4] as:
1 • •
PO = {[ATP(T)] + α(T)[PCr(T)]
ξ(T)
• •
+ β(T)[La(T)] + γ(T)[O2(T)]} [6]
Equations [5] and [6] will dictate the structure of the
present review.
Mechanical Parameters and Temperature
From a practical viewpoint, it is of course essential to
show that a T induced metabolic change has a non
negligible effect on muscle mechanical performance.
•
Thus, in the following sections one will study the term w,
appearing in the left hand side of equation [5], and some
mechanical parameters directly or indirectly related to it
(time-to-peak force, half relaxation time, etc.).
Isometric twitch
Ranatunga et al. (1987) demonstrated that the maximal
(electrical stimulation) twitch tension developed by the
first dorsal interosseus muscle decreases by about 50% in
cooling intramuscular T from 35 to 12°C. The tension
decrease was more pronounced below 25°C. The Q 10
values estimated for T values in the range 35–25°C were
1.43 for time-to-peak (time from 0 force to peak force) and
1.7 for half time of relaxation.
Shorter contraction and relaxation times after an
electrical twitch (in warm compared with cold muscles)
were also found by other authors for the triceps surae
(Davies et al., 1982). In this case, the decrease in
maximal twitch tension for T going from 36.5 ± 0.6 to 24.3
± 1.0°C was of the order of 52%. The depression in
161
maximal twitch tension was associated with an increase
in control time-to-peak and half relaxation time from 110
± 12 ms (at 36.5 ± 0.6°C) and 74 ± 12 ms to 179 ± 14 ms
and 145 ± 10 ms when the muscles were “cooled”
(29.5°C), and 188 ± 16 ms and 146 ± 29 ms under “cold”
conditions (24.3°C). Heating had the reverse effect on
the contraction time i.e., time-to-peak and half relaxation
time decreased from 100 ± 19 ms and 78 ± 16 ms (at 36.5
± 0.6°C) to 80 ± 14 ms and 64 ± 13 ms (38.9 ± 0.1°C),
respectively.
In a subsequent experiment on triceps surae (maximal
twitch tension) Davies and Young (1983) found time-to-
peak values of 121 ± 18, 92 ± 15 and 167 ± 25 ms for 36.8
± 1.6, 39.7 ± 0.4 and 28.4 ± 1.2°C, respectively, while
relaxation times were 76 ± 6, 59 ± 6 and 147 ± 42 ms for
the same temperatures respectively. Maximal twitch
temperature i.e., 137 ± 28, 126 ± 18 and 94 ± 26 N for 36.8
± 1.6, 39.7 ± 0.4 and 28.4 ± 1.2°C, respectively.
Hopf and Maurer (1990) were able to calculate a Q10 for
the half relaxation times (maximal electric twitch,
adductor pollicis) of 1.39 (T range 35–25°C) and 1.64 (T
range 30–20°C) while the Q 10 for maximal isometric
tension and time-to-peak were 1.2, 1.63 (T range 35–
25°C) and 1.09, 2.02 (T range 30–20°C), respectively.
Time-to-peak values and half relaxation times increased
with cooling from 35 to 30°C by 2.5 ms/°C (3.87%) and
1.98/°C (3.24%), respectively. Isometric force increased
by 3.1%. Time-to-peak value, half relaxation time and
isometric force at 20°C were 155.6 ± 12.8 ms, 116.3 ± 26.8
ms and 363.4 ± 239.4 g, respectively and at 35°C were
64.5 ± 5.8 ms, 61.2 ± 6.6 ms and 469.3 ± 120.8 g.
Tornbergsen and Stålberg (1986) reported similar
findings (“dorsiflexors of the ankle”) for time-to-peak
(2.46 ms/°C) and isometric force (2.8%/°C) but a value of
7.54 ms/°C for half relaxation time. Values for time-to-
peak were 92 and 108 ms and for half relaxation time
were 92 and 108 ms at 36 and 29.5°C, respectively.
Bigland-Ritchie et al., 1992 (first dorsal interosseus,
maximal twice tension) found a time- to-peak value and a
half relaxation time of 58 and 68 ms at 30 ± 1°C. At ~25°C
the same parameters were ~50% and ~65% lower.
In conclusion, it appears that time-to-peak values and
half relaxation times decrease with decreasing T.
Isotonic twitch
Ruiter and Haan (2000) observed (adductor pollicis) a
decline in maximal shortening velocity, for a maximal
isotonic twitch, with decreasing T. Maximal shortening
velocity Q10 values were 2.2, 1.6 and 1.6 in the T ranges
22.2–25.6, 25.6–31.4 and 31.4–37.1°C, respectively. As a
consequence, maximal power production at 22.2°C was
only 18.7 ± 0.9% of the maximal power produced at
37.1°C. Maximal power production Q 10 values in the
same T ranges were 6.9, 2.9 and 2.0. The velocity at
162 Temperature Dependence of Muscle Metabolism
Fig. 1 Relationship between the frequency of electrical stimulation
and isometric force production of the adductor pollicis muscle at
different temperatures (modified from De Ruiter et al., 1999).
Closed square: 22°C; open square: 25°C; closed circle 31°C; open
circle 37°C. Force is expressed as a percentage of the maximal
values obtained (100%).
which maximal power was obtained significantly
decreased with a decrease of T. At 22.2°C, optimal
velocity was 34.0 ± 3.5% of the initial value obtained at
37.1°C. Velocity for maximal power production Q 10
values were 3.2, 1.9 and 1.7 in the T ranges 22.2–25.6,
25.6–31.4 and 31.4–37.1°C, respectively.
Isometric contractions
Isometric tetanic tension (Davies and Young, 1983;
triceps surae) was shown to depend on T and electrical
stimulation frequency. Thus, for a given T, the effect on
tension may be different depending on the chosen
stimulation frequency. Davies and Young (1983) found
(triceps surae) tetanic tension values of 936 ± 162, 1456
± 164 and 1767 ± 246 N for 10, 20 and 40 Hz at 36.8 ±
1.6°C; 705 ± 162, 1349 ± 81 and 1716 ± 139 N for 10, 20
and 40 Hz at 38.5 ± 0.2°C, and 856 ± 187, 1300 ± 316 and
1594 ± 335 N for 10, 20 and 40 Hz at 28.6 ± 1.4°C. Similar
results were reported in Davies et al. (1982).
The complexity of the behaviour of the tetanic tension
as a function of T and frequency of the electrical
stimulation is shown in Fig. 1 for the human adductor
pollicis muscle (De Ruiter et al., 1999). Cooling the
muscle increases the lowpass filtering effect, i.e. the
amplitude of the force oscillations observed during a train
of subtetanic electrical stimulations is decreased (Rafolt
et al., 1999; triceps surae). These observations were
explained by the T induced changes in the maximal rate
of force development and maximal relaxation rate. This
effect may also be seen in an elegant way in Fig. 2 where
Fig. 2 Force-time relationship of a non tetanic supramaximal
electrical stimulation (10 Hz) of the adductor pollicis at
estimated muscle temperatures of 22 (*) and 37°C (modified
from De Ruiter et al., 1999).
it is shown that at low T the muscle oscillates slower and
thus it has less time to “relax” in between the electrical
twitches. In this context, De Ruiter et al. (1999) obtained
Q 10 values for the maximal rate of force development
during electrical isometric tetanic contraction (adductor
pollicis) of 4.0 (at 22–25°C), 1.8 (at 25–31°C) and 2.0 (at
31–37°C), the corresponding maximal relaxation rates
being 3.7 (22–25°C), 1.9 (25–31°C) and 2.2 (31–37°C).
Moreover, Q10 for rates for 100% to 50% force relaxation
were 3.8, 1.8 and 2.1 and for 50 to 25% force relaxation
were 6.9, 2.2 and 2.1 for T ranges 22–25, 25-31 and 31–
37°C, respectively. At low temperature and high
stimulation frequency (the closed square point at 100 Hz
in F ig . 1) t he mus cle is no t a b le t o “fo l lo w” th e
stimulation, resulting in a lower developed force.
In a subsequent paper work, De Ruiter and Haan (2000)
found (adductor pollicis) the following Q10 values for
maximal rate of force development: 4.4 1.6 and 1.5;
maximal relaxation rate of 3.0, 2.2 and 1.9; rates for 100%
to 50% force relaxation 3.5, 2.0 and 1.7; rates for 50 to 25%
force relaxation of 2.9, 2.9 and 2.0, calculated in the T
ranges 22.2–25.6, 25.6–31.4 and 31.4–37.1°C,
respectively. In their protocol, stimulation frequencies
were adapted at each T to produce maximum rate of force
development.
It is interesting to notice from these data that in spinal
cord injured individuals the natural reduction of T, may
lead to an underestimation of the changes in contractile
properties in term of relaxation rate or the degree of
fusion with low-frequency stimulation during clinical
tests (Gerrits et al., 2000).
The force developed during a maximal voluntary
contraction (MVC, “forearm”) was found to decrease as a
function of T (Clarke et al., 1958). The subjects in this
study were able to exert nearly the same tension after
that the forearm was immersed in water at 18°C for 30
min (muscle temperature ~25–29°C) as at normal T, but
in water cooler than this the maximum tension fell
 Binzoni, T and Delpy, D
eventually to a value only 40% of normal in water at 2°C
(muscle temperature ~15–20°C, from Fig. 5 in Clarke et
al., 1958).
Bergh and Ekblom (1979) found that muscle MVC
developed by the knee extensors is positively related to
intramuscular T (6.5 N m °C–1 or 2.1% °C–1 in the range
30–39°C), whilst Asmussen et al. (1976) observed a
decrease in maximal voluntary isometric tension of 8%
for a T going from 40 to 30°C, for the biceps brachii,
triceps surae and quadriceps femoris considered all
together.
Clarke and Royce (1962) measured an increase in the
rate of development of MVC (“hand” muscles) for cold T.
For normal and hot (bath T 40°C for 10 min) conditions, a
value of 95% of the force (44.53 kg and 45.87 kg,
respectively) was attained in 26.2 10–2 s and 25.5 10–2 s,
respectively. For cold conditions (bath T 10°C for 10
min), 95% of the force (40.11 kg) was attained in 41.4 10–2
s. The rate of relaxation was also slower in cold
conditions.
Binkhorst et al. (1977) conversely found no correlation
at all between MVC and T for the “handgrip muscles” (T
range 22–38°C).
Fig. 3 shows the MVC developed by the first dorsal
interosseus muscle as a function of T (from Ranatunga et
al., 1987). It is noteworthy, that in this case the
maximum tension appears to be around 27 and not 37°C.
A decrease of ~35% in the rate of maximal voluntary
tension development calculated from 5 to 90% MVC for
the biceps brachii, triceps surae and quadriceps femoris
considered all together, was found by Asmussen et al.
(1976), for 10°C decrease (T range ~28–41°C).
Relaxation rates (calculated from 90 to 5% MVC)
decreased by about 50% for the same 10°C decrease.
Wiles and Edwards (1982) have found a Q10 for the
relaxation rate of electrically stimulated isometric
contractions (adductor pollicis) of 2.3 (T range 25–
37°C).
The MVC values (triceps surae) given by Davies and
Young (1983) are 2109 ± 480, 2098 ± 523 and 1707 ± 545
N at 36.8 ± 1.6, 39.7 ± 0.4 and 28.4 ± 1.2°C, respectively.
Only the MVC value at the lowest T is significantly
different. In another study on triceps surae, Davies et al.
(1982) found for the MVC: a) 1852 ± 352 (at 36.5 ± 0.6°C,
control), 1892 ± 270 (at 38.9 ± 0.1°C); b) 1793 ± 214 N (at
36.5 ± 0.6°C, control), 1562 ± 222 N (at 29.5 ± 0.5°C) and
c) 1816 ± 222 (at 36.5 ± 0.6°C, control), 1493 ± 254 N (at
24.3 ± 1.0°C). Only c) was significantly different from
control.
Q 10 values, found by De Ruiter and Haan (2000), for
electrical stimulated adductor pollicis maximal isometric
force were 1.3, 1.2 and 1.2 in the T range 22.2–25.6, 25.6–
31.4 and 31.4–37.1°C, respectively. This means that, at
22.2°C maximal isometric force was reduced to 79.3 ±
2.9% of the force obtained at 37.1°C.
163
Fig. 3 Mean for maximum voluntary contraction (first dorsal
interosseus) for 5 subjects as a function of intramuscular
temperature (modified from Ranatunga et al., 1987). Vertical
bars represent standard deviation.
A study performed on elderly men (70.2 ± 5.4 years,
triceps surae) has shown a decrease in the time-to-peak
value induced by an electrical twitch and in the half
relaxation time (Davies and Young, 1985). Time-to-peak
force values were 150 ± 21 and 114 ± 17 ms at 36.2 ± 0.2
and 39.6 ± 0.3°C respectively, the corresponding
relaxation times being 112 ± 21 and 81 ± 17 ms at 36.2 ±
0.2 and 39.6 ± 0.3°C, respectively. In this instance,
heating significa ntly reduced the time cour se o f
contraction parameters towards that of young males
under control conditions (121 ms ± 18 for time-to-peak
force and 76 ± 6 for relaxation time). The force-
generating capacity of the elderly remained unaffected by
the increase in T except for a significant reduction in low-
frequency 10-Hz tetanic tension.
Maximal isometric tension in skinned human fibres
increases almost 2-fold between 12 and 17°C, but
increases very little between 17 and 30°C (Bottinelli et
al., 1996; Stienen et al. 1996).
To summarise, it appears that dynamic parameters
such as time-to-peak value, half relaxation time, etc.
decrease with decreasing T. Electrically stimulated
maximum isometric contraction values seem also to
decrease with decreasing T. However, the evidence is
more equivocal over whether the MVC displays in all
cases an optimum value of 37°C or at some lower value.
Isotonic contractions
Bergh and Ekblom (1979) have found that muscle
maximal force developed by the knee extensors, at fixed
angular velocities, is positively related to intramuscular T
(range 35.8–37.9°C). The changes in maximal force were
164 Temperature Dependence of Muscle Metabolism
9.8 (4.7% °C–1) and 8.3 (4.9% °C–1) N m °C–1 for angular
velocities of 90° and 180° s –1 , respectively. The peak
power output was also positively related to T, the rates of
change being 28.5 (5.6% °C–1) and 15.8 (4.9% °C–1) W °C–1
at the same angular velocities of 90° and 180° s–1.
In a study by Binkhorst et al. (1977) on force-velocity
curves as a function of T (“handgrip muscle”, voluntary
contractions) they found that the Q10 for the maximum
velocity of contraction (at zero load) was 1.2 (T range 22–
38°C), whilst the same T range, the Q10 for the maximal
power was 1.3.
Sargeant (1987) has studied the effects of changing T
on muscle force and power output during a maximal 20 s
sprint effort performed at constant pedalling rates (95
crank rev min–1) on an isokinetic cycle ergometer. The
maximal peak force reached was 817 ± 107, 737 ± 111,
660 ± 107 and 594 ± 123 N for an intramuscular thigh T of
39.3 ± 0.4, 36.6 ± 0.5, 31.9 ± 0.7 and 29.0 ± 1.7°C,
respectively. Maximal peak power was 1456 ± 196, 1325
± 201, 1172 ± 208 and 1047 ± 213 W for the same T. It is
interesting to notice, that maximum peak force and
power also depended on the pedalling rate (Sargeant,
1987). In fact, the optimum velocity for maximum power
increased (88, 95, 109 and 125 rev min–1) as
intramuscular temperature increased (29.0 ± 1.7, 31.9 ±
0 .7 , 36.6 ± 0.5 a nd 39 .3 ± 0 .4°C). The pr ac tic al
consequence being that the effect of warming increases
the power output by ~2% per °C at 54 rev min–1 and by
~10% per °C at 104 rev min–1.
F er re tti e t a l. (199 5) fo und t hat the m in im um
mechanical power necessary to elicit maximal oxygen
consumption on a cycle ergometer is 21 W lower at 31.0 ±
2.0°C than at 36.0 ± 1.4°C, whilst Davies and Young
(1983) fo und that sho rt 10 s exercise on a cycle
ergometer at a preselected optimal pedalling speed, gave
a peak power of 3059 ± 430, 3176 ± 458 and 2096 ± 243
for 36.8 ± 1.6, 39.5 ± 0.1 and 28.5 ± 1.0°C, respectively.
Only the peak power value at the lowest T was
significantly different. Bergh and Ekblom (1979) have
also demonstrated that the shortest time in which 20
revolutions could be performed on a mechanically braked
cycle ergometer (~8.5 kJ) is inversely related to
intramuscular T (range 35.7–37.8°C). The work time
increased from 10.72 s at 38.3°C to 15.27 s at 31.4°C.
The height of a vertical jump decreases with decreasing
T (Bergh and Ekblom, 1979; T range 35.8–37.8°C) at a
rate of 2.1 cm °C –1 (4.2% °C–1), while peak velocity was
also positively related to T and changed by 0.51 m s–1 °C
(4.4% °C–1). Asmussen and Boje (1945) and Asmussen et
al. (1976) found a variation of 4.4% °C–1 for jumping from
a semisquatting position (in the T range 27–40°C) and
5.3% °C–1 for sprinting velocity (in the T range 36–40°C).
Davies and Young (1983) measured a change in the
height of jump of 2.38 cm °C–1 and a decrease in the peak
power (maximum instantaneous power) as a function of
T of 112.4 W °C –1 (range 28.5–39.5°C). Ferretti et al.
(1992) found a change in the maximum instantaneous
power, for a typical “mean” subject of 70 kg, of 111.9 W
°C–1. Davies and Young (1985) observed that in elderly
(70.2 ± 5.4 years, triceps surae) the maximal
instantaneous power during a vertical is some 52% of that
produced by 22 years old males. Following heating (from
36.2 ± 0.2 to 39.6 ± 0.3°C) there was a 20% rise in power.
At the microscopic level, the Q10 value for maximum
unloaded shortening velocity of isolated human skinned
fibres is 5.0, in the T range 12–22°C (Bottinelli et al.,
1996). On the other side, He et al. (2000) found Q 10
values (T values of 12 and 20°C) of 3.85 and 3.46 for slow
and fast fibres, respectively. For the same T values, Q10
values for maximal power output were found to be 5.35
and 5.41 for slow and fast skinned fibres, and it was
estimated that, the average force per attached myosin
head (at any speed) is higher in fast than in slow fibres,
and ~1.5 times higher at 20°C than at 12°C in both fibres.
Muscle fatigue (isometric contractions)
Muscle fatigue is modulated by T changes. In the 1940’,
application of hot “fomentations” (hot moist dressing) on
a limb have been shown to reduce strength and duration
of isometric contractions (Hall et al., 1944). Nukada
(1955) showed that when a limb was immersed for a short
period in water ranging from 20 to 40°C, the duration of
voluntary isometric contraction to fatigue steadily
increased from the higher to the lower water
temperature.
Edwards et al. (1970) have demonstrated that the time
taken to reach fatigue (quadriceps femoris) during one
voluntary isometric contraction (70% of MVC) decreases
from 65.6 ± 6.8 (SEM) to 41.4 ± 4.4 (SEM) seconds for an
intramuscular T of 31.6 and 38.6°C, respectively. If the
same fatiguing protocol was repeated 7 times at 20 s
intervals, then the holding times for the 1st, 2nd and 7th
contraction were 65.6 ± 6.8 (SEM), 27.5 ± 1.9 (SEM), 16.5
± 3.4 (SEM) s at 31.6°C and 41.5 ± 4.4 (SEM), 15.7 ± 1.9
(SEM), 11.7 ± 2.6 (SEM) s at 38.6°C. No significant
differences were found between 22.5 and 31.6°C.
This is in agreement with previously findings by Lind
and Samueloff (1957) where the fatiguing time of a single
voluntary isometric contraction (arm muscles) was
shown to decrease from ~264 to ~186 s when going from
a water T of 18 to 34°C. In the case of repeated
contractions, the fatiguing time was dependent on the
recovery time between contractions. At 18°C water
temperature, the fatiguing time of the 5th contraction
was ~193 and ~122 s for 20 and 7 min recovery,
respectively, and at 34°C, ~119 and ~107 for 20 and 7 min
recovery. Thus, in contradiction to what was observed at
18°C, when the water was at 34°C the 5th contraction
with 7 min recovery intervals was not significantly
different from that with 20 min recovery interval.
 Binzoni, T and Delpy, D
In a subsequent work, Clarke et al. (1958) showed that
for a voluntary isometric contraction (forearm, 1/3 of the
MVC) the fatiguing time has a maximum for a muscle
temperature near ~ 27°C (234 s). It is of note, that this T
corresponds to the value were Ranatunga et al. (1987)
measured a maximum value for MVC. For 17–20°C and
35–38°C the durations were ~47.5 s and ~87.05 s,
respectively (these data are estimated from Fig. 6; Clarke
et al., 1958). As in Lind and Samueloff (1957), 5 repeated
contractions, spaced by 20 min recovery, resulted in a
decreased force with different force values depending on
muscle temperature. Similar preliminary data work was
proposed by the same authors in Clark et al. (1957).
A maximum on the duration of one voluntary isometric
contraction to reach fatigue (2/3 MVC, quadriceps
femoris, same protocol as in Edwards et al. (1970), see
above) has also been found by Edwards et al. (1972). The
duration times were ~65.6, ~51.8 and ~40.9 s in the
in ter va ls 3 0.3– 32 .6, 19 .4– 25 .8 an d 38 .2 –3 9.1 °C,
respectively. At the end of the 7th contraction the
duration times were ~15.9, ~15.9 and ~11.8 s in the
in ter va ls 3 0.3– 32 .6, 19 .4– 25 .8 an d 38 .2 –3 9.1 °C,
respectively (all the data from Fig. 4, Edwards et al.,
1972).
On the other hand, De Ruiter et al. (1999) have
demonstrated that cooling of the adductor pollicis
significantly reduces the fatiguing effect on developed
muscle tension of 24 repetitive isometric tetanic
electrical stimulations (90% MVC, 1.5 s interval between
contractions, 1 s duration each). The measurements were
performed in this case during arterial cuff occlusion. The
decline in isometric force (to 73.0 ± 2.9%) following
repetitive stimulations at 37°C, was not significantly
different from that at 31°C (to 79.9 ± 1.3%) but it was
significantly greater than the decrease at 25°C (to 81.4 ±
0.8%) and 22°C (to 82.6 ± 1.4%). Thus, most of the
temperature-dependent differences found in non fatigued
muscle become smaller or even disappeared following
fatigue. The same authors found again (De Ruiter and
Haan, 2000; same protocol as in De Ruiter et al., 1999) for
the isometric force: 83.5 ± 1.1% at 37.1°C and 87.8 ± 0.9%
at 22.2°C (first contraction 100%).
Similar results were presented by Davies et al. (1982)
for the triceps surae. The fatigue test consisted of a
series of supramaximal 20 Hz stimuli given for 300 ms
each second for two minutes. A fatigue index was
calculated as the tension ratio of the 120th to the 1st
response. The respective fatigue index for 36.5 ± 0.6,
38.9 ± 0.1, 29.5 ± 0.5 and 24.3 ± 1.0°C were 0.63 ± 0.09,
0.64 ± 0.05, 0.81 ± 0.05 and 0.79 ± 0.06. No significant
differences were observed between heating (38.9°C) and
control (36.5°C). The resistance to fatigue (trains of 50
µ s stimuli at 20 Hz, lasting 330 ms, repeated every
second, for 2 min) also remains unaltered when heating
(from 36.2 ± 0.2 to 39.6 ± 0.3°C) in the muscles of elderly
165
men (70.2 ± 5.4 years, triceps surae; Davies and Young,
1985).
Muscle fatigue (isotonic contractions)
Wade et al. (2000) have found that during voluntary
isotonic contractions (250 g weight-lifting exercise at a
“as fast as possible” rate; first dorsal interosseus), the
time taken to reach fatigue (loss of the ability to lift the
weight) was 102.1 ± 10.7 (SD), 88 (n/a) and 130.4 ± 42.1
(SD) s for a skin T of 21.5 ± 0.6 (SD), 25.2 ± 0.1 (SD) and
30.5 ± 0.3 (SD) °C, respectively. At local low T, this
phenomenon might be explained in part by the fact that,
the nervous system is unable to adapt the motoneuron
firing rate and adjust to the slower properties of the
colder nerve/muscle, which may also increase the risk for
neuromuscular transmission failure (Bigland-Ritchie et
al., 1992). In fact, Bigland-Ritchie et al. (1992) found that
after a 60 s MVC fatiguing protocol, (first dorsal
interosseus) the motoneuron firing rate was dimished by
30% of the control. The same fatiguing protocol after
cooling the muscle by 5°C produced no measurable
change in the firing rate. For investigative purposes, this
problem may of course be bypassed by using electrical
stimulation and adapting the stimulation rate to the
experimental T (De Ruiter and Haan, 2000).
Sargeant (1987) showed the effects induced by
changing T upon muscle fatigue, during a maximal 20 s
sprint effort performed at constant pedalling rates (95
crank rev min – 1 ) on an isokinetic cycle ergometer.
Associated with the concomitant increase in maximal
power with muscle temperature (see above) was a faster
rate of fatigue over 20 s exercise. Hence, the maximal
power declined by ~33, ~27, ~22 and ~19 Ws –1 for an
intramuscular T of 39.3 ± 0.4 (SD), 36.6 ± 0.5 (SD), 31.9 ±
0.7 (SD) and 29.0 ± 1.7 (SD) °C (data from Fig. 3;
Sargeant, 1987).
Conclusions
It seems to be clear at this point that the effect of a T
change on muscle mechanical properties strongly
depends on the chosen type of exercise or electrical
stimulation protocol. Electrically stimulated or voluntary
contractions may lead to a different T-dependent muscle
behaviour. Moreover, in the case of isotonic exercise, the
velocity of contraction seems also to define the amplitude
of the T induced changes on the force developed.
Probably, other passive mechanical properties, e.g. T
changes in viscoelasticity (Asmussen et al., 1976; Price
and Lehman, 1990), may also influence muscle activity.
Thus, it is reasonable to expect that the metabolic
energy yielding reactions and the efficiency parameters,
represented by the right hand side of equation [5], may
also depend on the exercise protocol. This might be the
•
case even if the developed power, w, remains the same.
The link existing between the mechanical properties (in
166 Temperature Dependence of Muscle Metabolism
• •
this case w) and the energy metabolism is given by the
efficiency parameter η appearing in equation [5]. This is
the subject of the next section.
ATP Consumption Rate and Efficiency
As explained above, equation [1] describes the
contribution of the main energy yielding reaction during
rest or exercise. It must be noticed that, for a fixed
•
mechanical work load, E may have different values,
depending on the type of exercise performed by the
subject. For example, concentric and eccentric (negative
work) exercises have different η at the same absolute
mechanical power (Ryschon et al., 1997). In other words,
the ATP consumption rate is higher during concentric
than during eccentric exercise. Thus, care must be taken
when trying to generalize the results concerning the
effects of T on different kind of exercises. To our
knowledge, the only studies trying to determine by
“direct” measurements the effects of T on the efficiency of
ATP utilization in humans were performed on isolated
muscle fibres. He et al. (2000) measured a peak efficiency,
η∆G, increase for isotonic contraction from 0.21–0.27 to
~0.4 when the T was raised from 12 to 20°C. When we
consider that the ATP free energy was taken as constant
for any T, η ∆ G is in this case proportional η. No
differences were observed between fast or slow twitch
fibres. Surprisingly, as observed by Bottinelli and
Reggiani (2000), in isometric conditions, an increase of T
(from 12 to 20°C) increases the cost of force production
(ξ, see equation [4]) from 0.56 to 0.65 pmol mN–1 mm–1 s–1
for type I fibres and from ~1.51 to ~3.04 pmol mN–1 mm–1
s–1 for type II fibres (calculated from: Stienen et al., 1996).
It is clear from the reduced amount of data on humans,
that this domain need to be further investigated.
A na lysin g t he abo ve r esult s a t t he light o f t he
bioenergetic model (equation [6]) has a direct
consequence on the efficiency value parameter (γ) of the
aerobic glycolysis. In fact, if one hypothesize that the
results obtained on isolated human fibres (Bottinelli and
Reggiani, 2000) may be compared with those obtained in
vivo (Binzoni et al., 2001 and Fig. 4) it follows that, for a
constant pure aerobic isometric exercise, equation [6]
can be simplified to:
1 •
PO = {γ(T)[O2(T)]}. [7]
ξ(T)
Thus, at constant PO (i.e. the force developed during the
•
i so m e tr ic co n t r a c t io n) , t he de c r e a se in O2 w it h
decreasing T (Binzoni et al., 2001 and Fig. 4) seems to
balance in a coherent way the decrease in the cost ξ
(Bottinelli and Reggiani, 2000). In this case, the aerobic
glycolysis efficiency parameter γ might also remain
constant. This has never been proved experimentally. A
similar argument might also be considered for the
remaining energy sources represented by α(T)[PCr(T)]
•
and β(T)[La(T)].
ATP and PCr Concentrations
Equation [5] describes the constraint existing between
the fluxes of ATP, PCr, La, O 2 and the developed
•
mechanical work, w. In this context, even if equation [5]
describes molecular fluxes, a knowledge of the ATP and
PCr concentrations may be important for at least three
reasons: 1) It was hypothesised that, at the maximal
instantaneous power, the ATP consumption imposed by
the work load is proportional to [ATP] (Ferretti et al.,
1992), where the proportionality factor represents the
velocity constant of ATP hydrolysis; 2) PCr is present in a
•
fixed amount in the cell concentration at rest thus, [PCr]
may remain non zero only until exhaustion of the PCr
stocks (PCr is resynthetised only at the end of exercise);
3) The chemical equilibrium between [ATP] and [PCr]
may determine [ADP], a powerful stimulator of the
•
mitochondrial activity and thus, of O2.
Again in this area, very little data on humans can be
found in the literature. To summarize, with increasing T
(interval 28–37°C), [ATP], [PCr] and pH decrease by –0.17
± 0.05, –0.54 ± 0.05 mM °C–1 and –0.016 ± 0.001 pH °C–1,
respectively (Binzoni et al., 2000). The data are in
agreement with previous findings in animals
(Kushmerick, 1983). Febbraio et al. (1996) have found,
performing biochemical analysis on bioptic samples, that
[ATP] and [PCr] are unaffected by T changes. This result
might be explained by the fact that their measurements
were performed at 37.3 ± 1.4 and 35.4 ± 0.1°C. For this
small change in T, [ATP] and [PCr] would in principle
only increase from ~8.2 to 8.52 and from ~32 to 33.02
mM, respectively. These variations were probably below
the sensitivity of the method. Wade et al. (2000) also
found no changes in [ATP] and [PCr] for the first dorsal
interosseus as a function of T (T measured at the skin
level, cooling the hand during 10 min). Unfortunately,
the authors did not clamp or measure T during the 31P-
NMR measurements thus, it is difficult in this case to
obtain an estimation of the concentrations.
To evaluate if there is any influence of the above
concentration changes on the mechanical and metabolic
activity one can consider the relationship existing
between the maximal instantaneous power and T. In
humans, the maximal instantaneous power may be
estimated for example during a maximal vertical jump on
a force platform (Davies and Rennie, 1968). In this case,
the power developed by the subject last a few
milliseconds and all the energy comes directly from the
ATP stores (Ferretti et al., 1992). Thus, equation [5] can
be written as:
• •
w = η(T)[ATP(T)] [8]
 Binzoni, T and Delpy, D 167

all the remaining terms being nil during the jump.
Ferretti et al. (1992), found the following relationship
• •
between w and T (experimental range 24–37°C):
• rest (“forearm”) as a function of local T were performed
w = (0.031T–0.12) × 51.6 [9]
where 51.6 W Kg–1 is the maximal power at 35.8°C (mean
of 6 subjects, see also section: “isotonic contractions”). To
• •
explain the results, the same authors proposed that [ATP
] should be expressed as a pseudo first order reaction:
• obtained from blood samples extracted from a forearm
[ATP(T)] = k(T)[ATP(T)] [10]
where k repr esents the velocity constant o f ATP
hydrolysis, and thus equation [8] becomes:
• (“forearm”) are slightly different if the blood samples are
w = k(T)η(T)[ATP(T)] [11]
This means that, for a change from 37 to 27°C, [ATP]
changes from 8.2 (Harris et al., 1974) to 9.9 mM (see
results above; Binzoni et al., 2000). It follows, that the
effect of the concentration alone should increases the
• •
maximum ATP flux and, as a consequence w, from 100 to
108.3%, a non-negligible quantity. On the other hand,
from equation [9], which is based on the real
experimental measurements, one would expect a
• •
decrease of w from 100 to 69.8% for 37 and 27°C,
respectively. This means that, if equation [11] is
considered a good model, then the product k(T)η(T) must
by definition decrease. Unfortunately, insufficient
measurements of k and η exist in the literature to allow
relevant comparison of the experimental data with the
theoretical prediction. The only observation in this
instance, was the η decrease measured in isolated muscle
fibres (isotonic contraction) for decreasing T (He et al.,
2000; see “ATP consumption rate and efficiency”).
The last interesting observation is that it has been
calculated (Binzoni et al., 2000) that the T-dependence of
[ATP], [PCr] and pH produces a decrease of ~0.83 µM °C–1
in [ADP] (at rest) with decreasing T. Considering that
[ADP] modulates the mitochondrial activity, this
•
phenomenon might contribute to the reduced O2 observed
at low T (at rest). This will be the subject of the next
section.
Oxygen Consumption
Oxygen consumption at rest
It is well known that at normothermia the efficiency, γ,
of the aerobic glycolysis (moles of ATP produced per O2
mole) depends on the substrate utilized by the muscle. In
vivo, γ can range between ~5.6 and ~6.2 for free fatty acid
and glycogen, respectively (di Prampero, 1981). Hot
(Febbraio, 2000) or cold (Doubt, 1991) environments
may change in some cases the fractional contribution of
the different substrates, modulating for example the level
of circulating catecholamines. In addition to the effects
of catecholamines, T per se seems to play a role in
mediating the substrate response (Febbraio, 2000; Doubt,
1991).
One of the first measurements of human muscle O2 at
•
by Abramson et al. (1957). The authors found O2 values
of 0.332, 0.173 and 0.071 ml min–1 (100 ml)–1 limb volume
at a mean T of 39.9, 34.5 and 24.0°C, respectively. The
•
Q 10 for O2 was found to have a mean value of 2.5. O2
values were calculated using the Fick principle on data
vein. Blood flow was measured by plethysmography.
In another series of measurements Abramson et al.
•
(1958a) demonstrated that the calculated rest O2 values
taken from a deep vein or a superficial vein. The authors
explained that this difference is probably due to the
different tissue compartments drained by the veins when
changing T. At high T, this results in a underestimation of
•
O2 when based on blood from superficial vein and an
overestimation of O2 reading when based on blood from
•
deep vein. For “deep vein” measurements O2 values were
0.369 ± 0.051, 0.199 ± 0.023 and 0.071 ± 0.006 ml min–1
(100 ml vol)–1 at 38.6, 35.8 and 28.3°C, respectively. For
“superficial vein” measurements O2 values were 0.261,
0.165 and 0.070 ml min–1 (100 ml vol)–1 at 38.6, 35.8 and
28.3°C, respectively.
•
Abramson et al. (1958b) measured an increase of O2
(“forearm”) immediately after 5 minutes arterial
occlusion, compared to the control condition with no
•
arterial occlusion. The increase of O2 values during
reactive hyperhemia were 1.340 and 0.769 ml min–1 (100
•
ml limb vol)–1 at 38.9 and 28.3°C, respectively. The O2
curves at both T decreased exponentially as a function of
time towards the resting value levels. It is of note that, at
•
high T, O2 debt calculated multiplying O2 at rest times 5
min ischemia, was different from the measured value (the
integral under the O2-time curves). In fact O2 debt values
at 38.9°C were 2.021 and 0.804 ml (100 ml limb vol)–1 for
predicted and measured values, respectively. O2 debt
values at 28.3°C were 0.348 and 0.410 ml (100 ml limb
vol) – 1 which are not significantly different. This
observation was explained with the hypothesis that at
high T, when the arterial occlusion is applied, a large
complement of blood is trapped in the vascular tree due
to the resulting vasodilation. Thus, some of this blood is
available as a O2 source during the occlusion, decreasing
in this case the O2 debt. This observation seems to be in
a g r e e m e n t w it h ne a r i n f r a r e d m e a s ur e m e nt s o f
haemoglobin saturation, were it was demonstrated that,
for 4–5 min arterial occlusion at normothermia, the
metabolism of the forearm (Hamaoka et al., 1996) or calf
(Binzoni et al., 1998) may remain reasonably aerobic. At
low temperature, due to vasoconstriction, blood stores
should be reduced and thus the measured O2 debt values
should be near the theoretical values (Abramson et al.,
168 Temperature Dependence of Muscle Metabolism
1958b).
Binzoni et al. (1999) and Binzoni et al. (2001) found, by
near infrared spectroscopy, a Q 10 value of ~2 for the
• •
human “forearm” muscle O2 at rest. Typical O2 values
were ~2 and ~4 µ mol 100–1 g –1 min–1 at ~26 and ~36°C
respectively.
• •
Koga et al. (1997) measured rest VO2 values (O2
•
measured at the lung level, VO2) 355 ± 71, 343 ± 134, 285
± 38 and 381 ± 102 ml min–1 at 35.3 ± 0.4, 35.4 ± 0.4, 38.9
± 0.1 and 38.6 ± 0.3°C (vastus lateralis), respectively. No
significant differences were observed between the data.
This is in contrast with measurements obtained on the
•
forearm at rest were a Q 10 of 2–2.5 was found for O2
(Abramson et al., 1957; Abramson et al., 1958; Binzoni et
al., 1999; Binzoni et al., 2001). A possible explanation of
• •
this discrepancy, is that in Koga et al. (1997) VO2 take
•
into account the whole body O2. One can estimate that,
•
the lower body utilize ~20% of the VO2 (Janssen et al.,
2000). In this case, only (355/100) * 20=71 ml min–1 at
35.3°C are needed for the legs. If only the legs are heated,
with a T increase of 38.6– 35.3=3.3°C and a Q10 of ~2.5,
one obtains ~96 ml min–1 at 38.6°C. Thus, from 35.3 to
•
38.6°C VO2 will increase by 96– 71=25 ml min –1 . This
variation is within the experimental error estimated and
thus not measurable. Shiojiri at al. (1997) found also not
• •
significant differences for resting VO2 (vastus lateralis)
with values of 357 ± 51 and 378 ± 22 ml min–1 at 36.8 ± 0.4
and 30.2 ± 1.0°C, respectively. The same conclusions
•
were drown by Beelen and Sargeant (1991) were VO2 at
rest was 370 ± 60 and 350 ± 40 ml min–1 after cooling the
legs from 37.1 ± 0.3 to 33.0 ± 1.3°C (T measured at 4 cm
depth in the vastus lateralis). Finally, Blomstrand et al.
• •
(1984) measured a VO2 of 380 ± 0.06, 350 ± 0.08 and 380
± 0.06 ml min–1 at 29.2 ± 1.1, 34.6 ± 1.2 and 34.3 ± 0.6°C
(vastus lateralis), respectively. Esophageal T remained
•
constant and no VO2 differences were observed. In a
subsequent study, Bloomstrand et al. (1986) found rest
VO2 values (vastus lateralis) of 460 ± 120 and 400 ± 50
ml min–1 at 29 ± 2.8 and 35 ± 0.9°C, respectively.
On the other hand, Ishii et al. (1992) found significantly
• •
different VO2 of 440 ± 90 and 340 ± 60 ml min–1 at 35.5 ±
0.5 and 37.1 ± 0.3°C, respectively (T measured at 3 cm
depth in the vastus lateralis). It is interesting to notice
• •
that in this case, o ne o bta ins a VO2 increase for
decreasing T. This is probably due to the fact that the
authors measured a significant decrease in rectal (~core)
T (from 37.1 ± 0.3 to 35.5 ± 0.5°C). The decrease in rectal
T may have induced a thermoregulatory response mixing
in this case with “pure muscle” response. Significant
•
differences in VO2 as a function of T were also reported in
Ferretti et al. (1995) with the same cooling protocol as in
•
Ishii et al. (1992). In this case, VO2 values were 360 ± 50
and 520 ± 60 ml min –1 at 37.1 ± 0.3 and 35.5 ± 0.5°C,
respectively (T measured at 3 cm depth in the vastus
lateralis).
It appears from the above examples that to study
•
human muscle O2 over a large T range, it is essential to
•
develop techniques to enable local O2 measurements. In
•
fact, VO2 measured at the lung level implies the heating/
cooling of large muscle mass to improve signal to noise
ratio. This may change core T, introducing in this
manner confounding factors induced by
thermoregulation (Beelen and Sargeant, 1991).
Oxygen consumption during exercise •
Koga et al. (1997) measured VO2 as a function of T
during rest to work transitions on a cycle ergometer. The
measurements were performed during two different
protocols: “moderate work” (50 W) and “heavy work”
defined as 50% of the difference between lactate
•
threshold and peak VO2. VO2 time constants for the on-
transient response during “moderate work” were 28.1 ±
10.4 and 28.8 ± 5.3 s at 35.3 ± 0.4 and 38.6 ± 0.3°C (vastus
lateralis), respectively. For “heavy work”, the time
constants were 31.2 ± 5.1 and 29.3 ± 3.2 s at 35.4 ± 0.4
and 38.9 ± 0.1°C, respectively. Off-transient time
constant at the same T values for “moderate work” and
“heavy work” were 34.9 ± 9.6, 35.4 ± 7.3, 30.8 ± 3.4 and
31.7 ± 5.7 s respectively, so no significant effect of T on
exercise was observed on either the on and off transients
time constant for VO2. No significant differences were
•
also found between hot and cold for the fast VO2
•
component (first phase) at the onset of exercise. The VO2
amplitudes were in this case: 703 ± 170 and 728 ± 102 ml
min –1 at 35.3 ± 0.4, 38.6 ± 0.3°C (“moderate work”),
respectively; 1960 ± 343 and 2025 ± 335 ml min–1 at 35.4
± 0.4 and 38.9 ± 0.1°C (“heavy work”), respectively. The
increment in VO2 between the 3rd and 6th min of “heavy
exercise” was significantly smaller for “hot” than for
“control” conditions 138 ± 66 and 205 ± 70 ml min –1 ,
respectively. It is worth to note that the change in T (due
to the natural heating during exercise) observed during
“control” conditions was larger than during “hot”
•
conditions. This might explain in part the greater VO2
increase in “control” conditions.
Shiojiri et al. (1997) found similar VO2 values at the
end of a rest-to-work transition on a cycle ergometer (50
W, vastus lateralis) i.e., 981 ± 110 and 865 ± 71 ml min–1
at 36.8 ± 0.4 and 30.2 ± 1.0°C, respectively. VO2 time
constants for the on-transient response were significantly
greater at 30.2 ± 1.0°C (36.0 ± 7.7 s) than at 36.8 ± 0.4°C
•
(27.5 ± 4.4 s). Phase 1 amplitude for VO2 at 30.2 ± 1.0°C
(156 ± 34 ml min–1) was significantly lower than at 36.8 ±
0.4°C (240 ± 53 ml min–1).
On the other hand, Ishii et al. (1992) observed no
•
significant differences in VO2 time constants for the on-
transient response on the cycle ergometer. The values at
75 W were 36.2 ± 6.7 and 41.4 ± 10.0 s at 35.4 ± 0.9 and
27.5 ± 1.7°C (vastus lateralis), respectively. At 125 W
the values were 41.6 ± 8.6 and 43.8 ± 14.0 s at 35.6 ± 1.1
 Binzoni, T and Delpy, D 169

and 28.4 ± 1.5°C, respectively. Even if not significant, it

must be noticed that the changes are on the order of that
observed in Shiojiri et al. (1997). In fact, in Shiojiri et al.
(1997) the measurements were repeated 4 times on each
subject to decrease the error on the measurements. This
might explain the difference between the two studies.
Ferretti et al. (1995) found the same values for the
•
maximal VO2 on cycle ergometer at 31.0 ± 2.0°C (3.44 ±
0.82 l min –1 ) and 36.0 ± 1.4°C (3.55 ± 0.75 l min –1 ).
•
H o w e v e r , s i n c e V O2 w a s d i f f e r e n t a t r e s t , t h e
•
corresponding net VO2 turned out to be significantly
lower at 31°C than at 36°C i.e., 2.92 ± 0.85 and 3.19 ± 0.78
•
l min –1 , respectively. VO2 time constants for the on-
•
transient response at minimum power for maximal VO2
(289 ± 51 W) was 30.2 ± 6.2 s at 36.0 ± 1.4°C. No
significant differences were observed for the same
temperature at 260 ± 55 W (33.3 ± 5.0 s). By contrast, at
31.0 ± 2.0°C and 260 ± 55 W the time constant value was Fig. 4 Oxygen consumption curve as a function
significantly different (43.4 ± 8.6 s). This result is of intramuscular (“forearm”) temperature
explained by a greater oxygen deficit (more lactate (from Binzoni et al., 2001) of a subject
during rest (dashed-line) and 4% of the
production during the transient, see below). Thus, in this
maximum isometric contraction (closed-
case the change in the time constant are not completely circles).
explained by the modulation of the speed of the oxidative
reactions but also by the fact that another (anaerobic)
energy source participate to the energy production
during the transient phase (see equation [5]). Donald, 1998) might mask “pure muscle” responses,
Blomstrand et al. (1984) measured no differences in maintaining an unchanged core T. This is the reason
•
VO2 values after 1 min exercise (vastus lateralis) at 350 why, in the present review, experiments using hole body
•
W. VO2 were 2.39 ± 0.50, 2.39 ± 0.54 and 2.46 ± 0.49 l cooling have been discarded.
•
m in – 1 a t 29.2 ± 1.1 , 34.6 ± 1.2 a nd 34.3 ± 0.6°C, It is also the case that, an increase in O2 may also mean
respectively. In a later study at 370 ± 34 W, Bloomstrand a change in the efficiency of the biochemical reactions. A
•
et al. (1986) found VO2 at the end of the first 60 s of 2.3 ± decrease in efficiency during cooling might help to
0.44 and 2.3 ± 0.43 l min–1 at ~29 and ~35°C, respectively. produce more heat and maintain tissue temperature.
At the end of ~1.5 min exercise, the same authors found a However, it seems not to be always the case. In fact, it
•
VO2 value of 3.5 ± 0.39 and 3.5 ± 0.40 l min–1 at 33 ± 1.9 appears that during an isometric contraction (on small
and 36 ± 0.9°C, respectively. muscle mass or human fibres) the efficiency of the
•
Binzoni et al. (2001) found a decrease in O2 with aerobic metabolism and/or ATP utilization seems to
decreasing T, during 4% MVC (“forearm”). Fig. 4 shows a increase (see section: “ATP consumption rate and
typical subject. It is interesting to notice that the efficiency”).
•
difference of the two curves, for a given T, gives the O2 for The choice of the type of exercise, e.g. isometric versus
the isometric work (without rest). Another curious isotonic, may also be a possible explanation of the
• •
observation is the O2 peak observed at low T (Fig. 4). In different behaviours observed on the O2 -T curves.
• •
fact, a methodical O2 increase was detected for all the Unfortunately, to our knowledge, no O2-T measurements
subjects (although sometimes slightly shifted toward have been performed on small muscle mass during
lower T). isotonic exercise allowing to test if, even in this case, the
It is clear at this point that the size of the muscle chosen efficiency of the aerobic metabolism and/or ATP
to be cooled or heated seems to play a major role in the utilization is improved.
•
determination of the muscle O2 -T curves. In fact, the
choice of a large or a small muscle mass may result in an La Production
increase, decrease or leave unchanged the evaluated
•
muscle O2 . The explanation for this probably, is the In the previous section we have considered situations
difficulty in maintaining the experimental criteria of a were the workload was not purely aerobic. This means
constant core T during the cooling of a muscle and hence that other energy sources, other than those represented
•
obtain “pure muscle” responses to T changes. by the O2 term in equation [5], were involved in the ATP
Nonshivering thermoregulation (Florez-Duquet and Mc resynthesis (e.g. during exercise on a cycle ergometer). A
170 Temperature Dependence of Muscle Metabolism
knowledge of these energy sources is of course essential if
one want to estimate the efficiency of the aerobic
reactions. Neglecting the anaerobic sources would, for
example, lead to an overestimation of γ.
Edwards et al. (1972) have measured by biopsy a rest
[La] (quadriceps femoris) of 4.3 ± 1.6 (SEM), 7.0 ± 3.1
(SEM) and 8.6 ± 1.9 (SEM) µmoles g–1 dry muscle in the
intervals 19.4–25.8, 30.3–32.6, and 38.2–39.1°C,
respectively. The same authors also demonstrated that at
the end of the 1st, 2nd and 7th contraction, of a series of
voluntary isometric contractions maintained each to
fatigue (70% of MVC, quadriceps femoris) and spaced by
20 s, [La] was: 45.4 ± 7.7 (SEM), 60.3 ± 8.9 (SEM) and
92.5 ± 12.5 µmoles g–1 (SEM) for the first contraction in
the intervals 19.4–25.8, 30.3–32.6 and 38.2–39.1°C,
respectively; and for the 2nd and 7th contraction, at the
same temperatures, 71.9 ± 11.3 (SEM), 91.2 ± 11.5 (SEM),
104.1 ± 11.6 (SEM) and 94.0 ± 11.8 (SEM), 92.8 ± 12.1
(SEM), 111.1 ± 7.3 µmoles g–1 (SEM). Thus, [La] seems to
increase with increasing temperature.
Kaijser (1970) did not find any significant differences
in the deep vein lactate of the forearm after exhaustive
exercise for two different T (difference in forearm vein T,
~2.4°C).
Alternatively however, Bloomstrand et al. (1986) found
[La], at the end of the first 60 s exercise (370 ± 34 W) on
a cycle ergometer, of 37 ± 21 and 21 ± 11 mmol kg–1 dry
weight for a starting temperature of 29 ± 2.8 and 35 ±
0.9°C, respectively. Value at rest were 9.3 ± 4.7 (29 ±
2.8°C) and 10 ± 2.5 (35 ± 0.9°C) mmol kg–1 dry weight
(vastus lateralis).
In a similar protocol (370 ± 34 W, exercise lasting 1.5 ±
0.2 min, vastus lateralis), Blomstrand and Essen-
Gustavsson (1987) found an higher levels of [La] (2–4
times) at the end of the exercise performed at 28 ± 0.6°C
compared to 35 ± 1.0°C.
In a further study again, Bloomstrand et al. (1984)
found at the end of an exhaustive exercise on cycle
ergometer (350 W) [La] values (vastus lateralis) of 60 ±
36 and 46 ± 33 mmol kg–1 dry weight at 29.2 ± 1.1 and 34.3
± 0.6°C, respectively.
More recent results tend to confirm the above findings
on cycle ergometer. Fore example, Ishii et al. (1992)
measured early La values (in the venous blood) of 0.28 ±
0.44 (35.4 ± 0.9°C) and 0.95 ± 0.54 (27.5 ± 1.7°C) mM at
75 W. At 125 W, the early [La] values were 1.07 ± 0.92
and 1.38 ± 0.99 mM for 35.6 ± 1.1 and 28.4 ± 1.5°C,
respectively. Ferretti et al. (1995) obtained early [La]
values on cycle ergometer (260 ± 55 W, corresponding to
•
VO2 max in cold conditions) of 7.7 ± 2.5 and 9.6 ± 2.6 mM
at 36.0 ± 1.4 and 31.0 ± 2.0°C, respectively. Similarly,
Beelen and Sargeant (1991) measured a peak blood [La]
during exercise on cycle ergometer (at 70% maximun
•
VO2 ) that were 60% higher for low T than for control
conditions. The values were 8.7 ± 2.4 mM and 5.5 ± 1.3
mM at 33.0 ± 1.3 and 37.1 ± 0.3°C.
•
Thus, as for O2, there seems to exist a difference in La
production depending on the chosen muscle size and type
of exercise. Again, this discrepancy has been not yet
explicitly explained through suitably designed
experiments.
One more technical point should be taken into account
concerning [La] measurements. It is often the case in
experiments that [La] is measured in the blood and not
directly in the muscle tissue. In these cases, one must
take into account the fact that La may be washed out by
other organs (e.g. the liver) and that T may influence
blood fluxes and as a consequence La wash out (Brooks et
al., 1996). This might lead to incorrect estimations of the
energy produced by the anaerobic metabolism.
Temperature Measurements: A Real Problem
At this point, one cannot conclude this review on
effects of T on muscle metabolism without considering
the problem of T measurement. The measurement of
intramuscular T in humans, whether by invasive or non
invasive means is not a simple task, especially when it is
matter of determining its spatial distribution. The first
measurements in humans were obtained by Becquerel
and Breschet (1835), utilizing a thermocouple mounted
inside a needle. One hundred and sixty years later, the
most widely used method is still the same and even with
the introduction of very fine and flexible thermocouples
a n d t h e u s e o f m ul t i p l e s e n so r s , t h e n um b e r o f
measurement points remains limited. Thus, to obtain a
temperature distribution map it has been necessary to
find an alternative method. Only since the middle 1980’s,
with the introduction of new nuclear magnetic resonance
imaging techniques has it been possible to estimate
noninvasively, the full three dimensional temperature
distribution in a tissue. Many different MRI approaches
to the problem of temperature mapping have been
proposed. For example, it is possible to exploit the
temperature dependence of the longitudinal relaxation
time (Dickinson et al., 1986), the chemical shift (Hall and
Talagala, 1985) or, the diffusion coefficient (Le Bihan et
al., 1989) of the water resonance. Actually, none of the
above methods is perfect, the major problem being the
absolute calibration of the measured T values. It is
remarkable, that this represents the only real progress
since 1835! Alternative techniques, based on
ultrasonography or CT (Fallone et al., 1982) have been
proposed, however, these methods have never really been
successful. This is probably due to the poor spatial
r eso lut ion o r , in t he c ase o f ult r aso unds, to th e
insufficient anatomical information.
Of course, not all experimental protocols require a
detailed knowledge of the spatial T distribution. In these
cases, the zero heat flow (ZHF) thermometers may
 Binzoni, T and Delpy, D
become extremely useful (Fox and Solman, 1971). This
non invasive method is based on the hypothesis that if the
heat flux going through the skin is reduced to zero, then
the T at the surface is equilibrated with the deep tissue T,
i.e. this is equivalent to creating a zero T gradient from
the underlying tissues to the surface. The zero heat flux
condition may, in principle, be obtained with an ideal
thermal insulation material applied to the skin surface.
Unfortunately, all insulating materials are far from ideal.
For this reason, Fox and Solman (1971) proposed an
electronic servo-control system, capable of reproducing
the effects of a perfect thermal insulation. This technique
has allowed one to measure, non invasively on humans,
the deep T of a muscle from the surface. The ZHF probe
has the disadvantage of a slow response time of 15–20
min. Thus, care must be taken when the temperature
change are rapid, as is the case in extreme conditions
such as in the cooling/heating procedures during
cardiopulmonary bypass (Muravchick, 1983). However, a
significant advantage of the ZHF probe is that it
eliminates the T gradients in the time. In a forearm
immersed in cold water at 20°C, radial T gradient values
reaching 2°C cm–1 have been reported (Ducharme et al.,
1991) which means that the measurement of total tissue
metabolism often includes contributions from several
subsets of tissue operating at different temperatures. The
ZHF probe, allows one to obtain a uniform T region (of
some centimetres cubes) from which precise metabolic
measurements can be obtained. These probes may also be
modified to allow measurements directly in water
(Binzoni et al., 1999). ZHF technology has recently
helped in the development and calibration of a promising
new near infrared T- measuring technique (Hollis et al.,
2001). This technique should permit a higher time
resolution of T measurements compared to ZHF and
avoids the stabilization time needed by the technique
(usually 15–20 minutes).
There is no doubt that a large amount of technical
development is critical for further studies on the effects
induced by local T changes.
Conclusions
In the present review we have shown that a change in
local muscle T change does not induce the same effect
under different physiological situations. In fact, it
appears that the type of exercise performed by the subject
(e.g. isotonic vs isometric) might elicit a different
physiological response to T. Moreover, when studying
“pure” T phenomena, it is essential to work on a small
muscle mass to ensure the exclusion of non detectable
central thermoregulatory responses. In each of the
different sections, of this review, we also highlighted the
significant shortage of experimental data concerning the
effects of T upon the efficiency constants of the energy
171
yielding reactions. The exact interplay existing between
aerobic metabolism, anaerobic metabolism and the
Lohmann reaction when changing T also needs to be
better defined. This includes, for example, a better
knowledge of the dynamic parameters (e.g. half times
during rest-to-work transitions) for all the proposed
energetic reactions and their specific T modulation.
The present review has not been able to fully cover all
of this subject but, we hope, that the particular approach
chosen for the organization of the paper will generate
constructive comments and promote new research
activity in the area.
Aknowledgments The authors would like to thank and
acknowledge the support of the Swiss National Science
Foundation (#31-58759.99).
References
Abramson DI, Kahn A, Jr Tuck S, Turman GA, Rejal H,
Fleischer CJ (1957) Relationship between a range of
tissue temperature and local oxygen consumption in
the resting forearm. J Lab Clin Med 50: 789
Abramson DI, Kahn A, Jr Tuck S, Turman GA, Rejal H,
Fleischer CJ (1958a) Relationship between a range of
tissue temperature and local oxygen uptake in the
human forearm. I. Changes observed under resting
conditions. J Clin Invest 37: 1031-1038
Abramson DI, Kahn A, Jr Tuck S, Turman GA, Rejal H,
Fleischer CJ (1958b) Relationship between a range of
tissue temperature and local oxygen uptake in the
human forearm. II. Changes observed after arterial
occlusion, in the period of reactive hyperemia. J Clin
Invest 37: 1039-1048
Asmussen E, Boje O (1945) Body temperature and
capacity for work. Acta Physiol Scand 10: 1-22
Asmussen E, Bonde-Petersen F, Jorgensen K (1976)
Mechano-elastic properties of human muscles at
different temperatures. Acta Physiol Scand 96 (1): 83-
93
Beelen A, Sargeant AJ (1991) Effect of lowered muscle
temperature on the physiological response to exercise
in men. Eur J Appl Physiol Occup Physiol 63 (5): 387-
392
Becquerel AC, Breschet G (1835) Annales de Sciences
Naturelles 3 (2nd sér): 257
Bergh U, Ekblom B (1979) Influence of muscle
temperature on maximal muscle strength and power
output in human skeletal muscles. Acta Physiol Scand
107 (1): 33-37
Bigland-Ritchie B, Thomas CK, Rice CL, Howarth JV,
Woods JJ (1992) Muscle temperature, contractile
speed, and motoneuron firing rates during human
voluntary contractions. J Appl Physiol 73 (6): 2457-
2461
172 Temperature Dependence of Muscle Metabolism
Binkhorst RA, Hoofd L, Vissers AC (1977) Temperature
and force-velocity relationship of human muscles. J
Appl Physiol 42 (4): 471-475
Binzoni T, Hiltbrand E, Terrier F, Cerretelli P, Delpy D
(2000) Temperature dependence of human
g a st r o c ne m iu s p H a n d h i g h - e ne r g y p h o s p h a t e
concentration by noninvasive techniques. Magn Reson
Med 43 (4): 611-614
Binzoni T, Quaresima V, Barattelli G, Hiltbrand E, Gurke
L, Terrier F, Cerretelli P, Ferrari M (1998) Energy
metabolism and interstitial fluid displacement in
human gastrocnemius during short ischemic cycles. J
Appl Physiol 85 (4): 1244-1251
Binzoni T, Springett R, Dalton JC, Delpy D (1999) A new
combined deep-body-temperature/ NIRs probe for
noninvasive metabolic measurements on human
skeletal muscle. Adv Exp Med Biol 471: 623-629
Binzoni T, Ngo L, Hiltbrand E, Springett R, Delpy D
(2001) Non-standard O2 consumption- temperature
curves during rest and isometric exercise in human
skeletal muscle. (submitted)
Blomstrand E, Bergh U, Essen-Gustavsson B, Ekblom B
(1984) Influence of low muscle temperature on muscle
metabolism during intense dynamic exercise. Acta
Physiol Scand 120 (2): 229-236
Blomstrand E, Essen-Gustavsson B (1987) Influence of
reduced muscle temperature on metabolism in type I
and type II human muscle fibres during intensive
exercise. Acta Physiol Scand 131 (4): 569-574
Blomstrand E, Kaijser L, Martinsson A, Bergh U, Ekblom
B (1986) Temperature-induced changes in metabolic
and hormonal responses to intensive dynamic exercise.
Acta Physiol Scand 127 (4): 477-484
Bottinelli R, Canepari M, Pellegrino MA, Reggiani C
(1996) Force-velocity properties of human skeletal
muscle fibres: myosin heavy chain isoform and
temperature dependence. J Physiol 495 (Pt 2): 573-586
Bottinelli R, Reggiani C (2000) Human skeletal muscle
fibres: molecular and functional diversity. Prog Biophys
Mol Biol 73: 195-262
Bramer MA (1999) Knowledge discovery and data
mining. IEE, London, United Kingdom
Brooks GA, Fahey TD, White TP (1996) Metabolic
response to exercise: lactate metabolism during
e x er c ise a nd r e co v er y , ex c es s p o ste x er c ise O 2
consumption (EPOC), O 2 deficit, O 2 debt, and the
“anaerobic threshold”. In Exercise Physiology. Human
bioenergetics and its applications. Mayfield Publishing
Company, Mountain View, California, 173-197
Clarke RSJ, Hellon RF, Lind AR (1957) The influence of
muscle temperature on sustained contractions to
fatigue. J physiol 136: 41P
Clarke RSJ, Hellon RF, Lind AR (1958) The duration of
sustained contractions of the human forearm at
different muscle temperatures. J Physiol 143: 454-473
Clarke DH, Royce J (1962) Rate of muscle tension
development and release under extreme temperatures.
Int Z Physiol einschl Arbeitsphysiol 19: 330-336
Davies CT, Young K (1983) Effect of temperature on the
contractile properties and muscle power of triceps
surae in humans. J Appl Physiol 55 (1 Pt 1): 191-195.
Davies CT, Young K (1985) Effect of heating on the
contractile properties of triceps surae and maximal
p o w e r o ut p ut d ur i n g j um p i n g i n e l d e r ly m e n .
Gerontology 31 (1): 1-5
Davies CT, Mecrow IK, White MJ (1982) Contractile
properties of the human triceps surae with some
observations on the effects of temperature and
exercise. Eur J Appl Physiol Occup Physiol 49 (2): 255-
269
Davies CT, Rennie R (1968) Human power output. Nature
217 (130): 770-771
Dickinso n R J, H all AS, Hind AJ , Y oug IR (198 6 )
Measurement of changes in temperature using MR
imaging. J Comp Assist Tom 10: 468-472
Doubt TJ (1991) Physiology of exercise in the cold. Sports
Med 11 (6): 367-381
Ducharme MB, Van Helder WP, Radomski MW (1991)
Tissue temperature profile in the human forearm
during thermal stress at thermal stability. J Appl
Physiol 71 (5): 1973-1978
Edwards RH, Harris RC, Hultman E, Kaijser L, Koh E,
Nordesjo LO (1970) Ener gy metabolism during
isometric exercise at different temperature of M.
quadriceps femoris in man. Acta Physiol Scand 80 (4):
17A-18A
Edwards RH, Harris RC, Hultman E, Kaijser L, Koh D,
Nordesjo LO (1972) Effect of temperature on muscle
energy metabolism and endurance during successive
isometric contractions, sustained to fatigue, of the
quadriceps muscle in man. J Physiol 220 (2): 335-352
Fallone BG, Moran PR, Podgorsak EB (1982) Noninvasive
thermometry with a clinical x-ray CT scanner. Med
Phys 9 (5): 715-721
Febbraio MA (2000) Does muscle function and
metabolism affect exercise performance in the heat?
Exerc Sport Sci Rev 28 (4): 171-6
Fell D (1997) Understanding the control of metabolism.
Portland Press Ldt, London
Ferretti G (1992) Cold and muscle performance. Int J
Sports Med 13 Suppl 1: S185-S187
Ferretti G, Binzoni T, Hulo N, Kayser B, Thomet JM,
Cerretelli P (1995) Kinetics of oxygen consumption
during maximal exercise at different muscle
temperatures. Respir Physiol 102 (2-3): 261-268
Ferretti G, Ishii M, Moia C, Cerretelli P (1992) Effects of
temperature on the maximal instantaneous muscle
power of humans. Eur J Appl Physiol Occup Physiol 64
(2): 112-116
Florez-Duquet M, McDonald RB (1998) Cold-induced
 Binzoni, T and Delpy, D
thermoregulation and biological aging. Physiol Rev 78
(2): 339-358
F o x RH , So lm a n A J (1 97 1) A ne w t ec hn ique fo r
monitoring the deep body temperature in man from the
intact skin surface. J Physiol 212: 8P-10P
Gerrits HL, De Haan A, Hopman MTE, Van Der Woude
LHV, Sargeant AJ (2000) Influence of muscle
temperature on the contractile properties of the
quadriceps muscle in humans with spinal cord injury.
Clin Sci 98 (1): 31-38
Hall VE, Schamp HM, Brown GE, Davis MN (1944) Effect
of Kenny fomentation on the strength of voluntary
muscular contraction in man. Arch Phys Med 25: 96-99
H a ll LD, T a la g al a SL (1 9 85 ) M a p p ing o f p H a nd
temperature distribution using chemical-shift- resolved
tomography. J Magn Reson 65: 501-505
Hamaoka T, Iwane H, Shimomitsu T, Katsumura T,
Murase N, Nishio S, Osada T, Kurosawa Y, Chance B
(1996) Noninvasive measures of oxidative metabolism
on working human muscles by near-infrared
spectroscopy. J Appl Physiol 81 (3): 1410-1417
Harris RC, Hultman E, Nordesjo LO (1974) Glycogen,
glycolytic intermediates and high-energy phosphates
determined in biopsy samples of musculus quadriceps
femoris of man at rest. Methods and variance of values.
Scand J Clin Lab Invest 33 (2): 109-120
He ZH, Bottinelli R, Pellegrino MA, Ferenczi MA, Reggiani
C (2000) ATP consumption and efficiency of human
single muscle fibers with different myosin isoform
composition. Biophys J 79 (2): 945-961
Helmholtz H (1847) Über die Erhaltung der Kraft. G
Reimer
Hill AV (1965) Trails and trials in physiology. Arnold,
London
Hollis VS, Binzoni T, Delpy D (2001) Non-invasive
monitoring of brain tissue temperature by near-
infrared spectroscopy. Proc SPIE (in press)
Hopf HC, Maurer K (1990) Temperature dependence of
the electrical and mechanical responses of the adductor
pollicis muscle in humans. Muscle Nerve 13 (3): 259-
262
Ishii M, Ferretti G, Cerretelli P (1992) Effects of muscle
temperature on the VO2 kinetics at the onset of
exercise in man. Respir Physiol 88 (3): 343-353
Janssen I, Heymsfield SB, Wang ZM, Ross R (2000)
Skeletal muscle mass and distribution in 468 men and
women aged 18-88 yr. J Appl Physiol 89 (1): 81-88
Kaijser L (1970) Limiting factors for aerobic muscle
performance. The influence of varying oxygen pressure
and temperature. Acta Physiol Scand Suppl 346: 1-96
Kemp GJ, Radda GK (1994) Quantitative interpretation of
bioenergetic data from 31P and 1H magnetic resonance
spectroscopic studies of skeletal muscle: an analytical
review. Magn Reson Q 10 (1): 43-63
Koga S, Shiojiri T, Kondo N, Barstow TJ (1997) Effect of
173
increased muscle temperature on oxygen uptake
kinetics during exercise. J Appl Physiol 83 (4): 1333-
1338
Kushmerick M (1983) Energetics of muscle contraction.
In Peachey LD, Adrian RH, Geiger RS eds. Skeletal
muscle. The American Physiological Society, Bethesda,
MD, USA, 189-236
Le Bihan D, Delannoy J, Levin RL (1989) Temperature
mapping with MR imaging of molecular diffusion:
application to hyperthermia. Radiology 171: 853-847
Lind AR, Samueloff M (1957) The influences of local
temperature on successive sustained contractions. J
Physiol 136: 12P
Lundsgaard E (1930) Untersuchungen über
Muskelko ntr a ktio nen oh ne Milchsä ure bildung.
Biochemische Zeitschrift 217: 162-177, 227: 51-82
Muravchick S (1983) Deep body thermometry during
general anesthesia. Anethesiol 58: 271-275
Nukada A (1955) Hauttemperatur und Leistungsfähigeit
in Extremitäten bei stätischer Halterarbeit.
Arbeitsphysiologie 16: 74-80
di Prampero PE (1981) Energetics of muscular exercise.
Reviews in Physiology. Biochem Pharmacol 89: 143-222
Price R, Lehmann JF (1990) Influence of muscle cooling
on the viscoelastic response of the human ankle to
sinusoidal displacements. Arch Phys Med Rehabil 71
(10): 745-748
Rafolt D, Gallasch E, Mayr W, Lanmüller H (1999)
Dynamic force responses in electrically stimulated
triceps surae muscles: effects of fatigue and
temperature. Artif Organs 23 (5): 436-439
Ranatunga KW, Sharpe B, Turnbull B (1987)
Contractions of a human skeletal muscle at different
temperatures. J Physiol 390: 383-95
Ryschon TW, Fowler MD, Wysong RE, Anthony A,
Balaban RS (1997) Efficiency of human skeletal muscle
in vivo: comparison of isometric, concentric, and
eccentric muscle action. J Appl Physiol 83 (3): 867-74
De Ruiter CJ, Jones DA, Sargeant AJ, de Haan A (1999)
Temperature effect on the rates of isometric force
development and relaxation in the fresh and fatigued
human adductor pollicis muscle. Exp Physiol 84 (6):
1137-1150
De Ruiter CJ, De Haan A (2000) Temperature effect on
the force/velocity relationship of the fresh and fatigued
human adductor pollicis muscle. Pflugers Arch 440 (1):
163-170
Sargeant AJ (1987) Effect of muscle temperature on leg
extension force and short-term power output in
humans. Eur J Appl Physiol Occup Physiol 56 (6): 693-
698
Shiojiri T, Shibasaki M, Aoki K, Kondo N, Koga S (1997)
Effects of reduced muscle temperature on the oxygen
uptake kinetics at the start of exercise. Acta Physiol
Scand 159 (4): 327- 333
174 Temperature Dependence of Muscle Metabolism
Stienen GJ, Kiers JL, Bottinelli R, Reggiani C (1996)
Myofibrillar ATPase activity in skinned human skeletal
muscle fibres: fibre type and temperature dependence.
J Physiol 493 (2): 299- 307
Torbergsen T, Stålberg E (1986) Effects of temperature
on muscle twitch. Muscle Nerve 9: 574
Wade AJ, Broadhead MW, Cady EB, Llewelyn ME, Tong
HN, Newham DJ (2000) Influence of muscle
temperature during fatiguing work with the first dorsal
interosseous muscle in man: a 31P-NMR spectroscopy
study. Eur J Appl Physiol 81(3): 203-209
Wiles CM, Edwards RH (1982) The effect of temperature,
ischaemia and contractile activity on the relaxation
rate of human muscle. Clin Physiol 2 (6): 485-497
Received: March 23, 2001
Accepted: April 9, 2001
Correspondence to: Tiziano Binzoni, Centre Médical
Universitaire, Département de Physiologie 1211 Genéve
4, Switzerland.