ArticlePDF AvailableLiterature Review

Physiological models to understand exercise fatigue and the adaptations that predict or enhance athletic performance


Abstract and Figures

A popular concept in the exercise sciences holds that fatigue develops during exercise of moderate to high intensity, when the capacity of the cardiorespiratory system to provide oxygen to the exercising muscles falls behind their demand inducing "anaerobic" metabolism. But this cardiovascular/anaerobic model is unsatisfactory because (i) a more rigorous analysis indicates that the first organ to be affected by anaerobiosis during maximal exercise would likely be the heart, not the skeletal muscles. This probability was fully appreciated by the pioneering exercise physiologists, A. V Hill, A. Bock and D. B. Dill, but has been systematically ignored by modern exercise physiologists; (ii) no study has yet definitely established the presence of either anaerobiosis, hypoxia or ischaemia in skeletal muscle during maximal exercise; (iii) the model is unable to explain why exercise terminates in a variety of conditions including prolonged exercise, exercise in the heat and at altitude, and in those with chronic diseases of the heart and lungs, without any evidence for skeletal muscle anaerobiosis, hypoxia or ischaemia, and before there is full activation of the total skeletal muscle mass, and (iv) cardiovascular and other measures believed to relate to skeletal muscle anaerobiosis, including the maximum oxygen consumption (VO2 max) and the "anaerobic threshold", are indifferent predictors of exercise capacity in athletes with similar abilities. This review considers four additional models that need to be considered when factors limiting either short duration, maximal or prolonged submaximal exercise are evaluated. These additional models are: (i) the energy supply/energy depletion model; (ii) the muscle power/muscle recruitment model; (iii) the biomechanical model and (iv) the psychological model. By reviewing features of these models, this review provides a broad overview of the physiological, metabolic and biomechanical factors that may limit exercise performance under different exercise conditions. A more complete understanding of fatigue during exercise, and the relevance of the adaptations that develop with training, requires that the potential relevance of each model to fatigue under different conditions of exercise must be considered.
Content may be subject to copyright.
Scand J Med Sci Sports 2000: 10: 123–145
Printed in Denmark ¡All rights reserved
Review Article
Physiological models to understand exercise fatigue and
the adaptations that predict or enhance athletic performance
T. D. Noakes
Bioenergetics of Exercise Research Unit of the Medical Research Council and the University of Cape Town, Sports Science Institute
of South Africa, Newlands, South Africa
Corresponding author: Professor Timothy David Noakes, Bioenergetics of Exercise Research Unit, Department of Physiology,
University of Cape Town, Sports Science Institute of South Africa, Boundary Road, Newlands, 7700, South Africa
Accepted for publication 17 June 1999
A popular concept in the exercise sciences holds that fa-
tigue develops during exercise of moderate to high inten-
sity, when the capacity of the cardiorespiratory system to
provide oxygen to the exercising muscles falls behind their
demand inducing ‘‘anaerobic’’ metabolism. But this car-
diovascular/anaerobic model is unsatisfactory because (i)
a more rigorous analysis indicates that the first organ to
be affected by anaerobiosis during maximal exercise
would likely be the heart, not the skeletal muscles. This
probability was fully appreciated by the pioneering exer-
cise physiologists, A. V. Hill, A. Bock and D. B. Dill, but
has been systematically ignored by modern exercise physi-
ologists; (ii) no study has yet definitely established the
presence of either anaerobiosis, hypoxia or ischaemia in
skeletal muscle during maximal exercise; (iii) the model is
unable to explain why exercise terminates in a variety of
conditions including prolonged exercise, exercise in the
heat and at altitude, and in those with chronic diseases of
the heart and lungs, without any evidence for skeletal
muscle anaerobiosis, hypoxia or ischaemia, and before
The nature of the physiological and biochemical ad-
aptations that occur in response to physical training
has been extensively studied in humans and other
mammals. This information is readily available and
is likely to be well known to most exercise scientists
(Saltin & Gollnick 1983, Holloszy & Coyle 1984).
Similarly there is an extensive literature on the cellu-
lar mechanisms believed to cause the fatigue that de-
velops during exercise (Fitts 1994).
In contrast, fewer studies have evaluated the extent
to which these adaptations explain the improvements
* Based on Keynote addresses presented at the Fourth IOC
World Congress on Sports Sciences, Monte-Carlo, Monaco,
October 22–26, 1997 and the IV Scandinavian Congress on
Medicine and Science in Sports, Lahti, Finland, 5–8 Novem-
ber 1998.
there is full activation of the total skeletal muscle mass,
and (iv) cardiovascular and other measures believed to re-
late to skeletal muscle anaerobiosis, including the maxi-
mum oxygen consumption (VO
max) and the ‘‘anaerobic
threshold’’, are indifferent predictors of exercise capacity
in athletes with similar abilities. This review considers four
additional models that need to be considered when factors
limiting either short duration, maximal or prolonged sub-
maximal exercise are evaluated. These additional models
are: (i) the energy supply/energy depletion model; (ii) the
muscle power/muscle recruitment model; (iii) the bio-
mechanical model and (iv) the psychological model. By
reviewing features of these models, this review provides a
broad overview of the physiological, metabolic and bio-
mechanical factors that may limit exercise performance
under different exercise conditions. A more complete
understanding of fatigue during exercise, and the relevance
of the adaptations that develop with training, requires that
the potential relevance of each model to fatigue under dif-
ferent conditions of exercise must be considered.
in performance that occur with different types of
physical training (Acevedo & Goldfarb 1989, Daniels
et al. 1978, Hawley et al. 1997, Houston et al. 1979,
Moore et al. 1997, Ramsbottom et al. 1989,
Westgarth-Taylor et al. 1997, Weston et al. 1997) and
which presumably result from changes that delay the
onset or development of fatigue. There are at least
three probable reasons for this.
First, many exercise physiologists may consider
this to be the work of the coach, not of the scientist.
Or, accustomed to the tightly controlled conditions
of laboratory research, some scientists may be reluc-
tant to undertake field-based studies of performance
in which all the different variables influencing human
performance are not easily controlled. Human per-
formance is influenced by many variables, not least
those involving the psyche. Many scientists may feel,
perhaps justifiably, that these variables cannot be suf-
ficiently well controlled in field studies for there to be
meaningful findings.
Second, there is a dearth of tools to measure accu-
rately human performance in the laboratory. If sports
performance cannot be measured frequently with a
high degree of precision in the laboratory, then train-
ing-induced changes in exercise performance are not
quantifiable. As a result, most studies use physiologi-
cal surrogates to predict changes in exercise perform-
ance. The most widely used performance surrogate is
the maximum oxygen consumption (VO
max). But,
the very use of this specific measure has helped to
entrench a particular and, perhaps, unquestioning
dogma of the factors that likely determine human ex-
ercise performance (Noakes 1988, 1997, 1998).
As a result, most of the training studies reported
in the literature have measured the physiological and
biochemical responses of the human to training and
have paid less attention (i) to the extent to which hu-
man exercise performance is altered by different
training programmes and (ii) to the specific physio-
logical adaptations which explain training-induced
changes in athletic performance.
Indeed, an important weakness in our current
thinking in exercise physiology is that we lack certain
knowledge of the precise factors that determine fa-
tigue and hence limit performance in different types
of exercise under a range of environmental con-
ditions. In part, this is because some scientists remain
unaware that their research is based on the (subcon-
scious) acceptance usually of one specific model of
human exercise physiology (Noakes 1997, 1998). But
it would be very surprising if one single physiological
model adequately explains human exercise perform-
ance under all conditions.
Accordingly, the aim of this review is not to de-
scribe how the body adapts to physical training. This
information is freely available, largely descriptive and
not particularly contentious, so that its review is un-
likely to challenge how we think about our science.
Rather I will use this opportunity to pose two ques-
tions: What physiological models have exercise scien-
tists developed (and subconsciously accepted) for the
study of the physiological and biochemical determi-
nants of fatigue during exercise? And which specific
physiological, metabolic or biomechanical attributes
Table 1. Current physiological models to understand the physiology of
training for enhanced endurance performance
(a) The cardiovascular/anaerobic model
(b) The energy supply/energy depletion model
(c) The muscle recruitment (central fatigue)/muscle power model
(d) The biomechanical model
(e) The psychological/motivational model
might explain superior athletic performance and en-
hanced resistance to the development of fatigue?
Current physiological models to understand the
physiology of training for enhanced endurance
Table 1 lists five different models that are commonly
used to study and explain the likely physiological and
other training-induced changes that may improve,
especially, endurance performance, probably by de-
laying or preventing the onset of fatigue. Each model
has its own proponents, usually those with a special
expertise in the specific areas embraced by the model.
Thus, the cardiovascular/anaerobic model is pro-
moted usually by cardiovascular and respiratory
physiologists; the energy supply/energy depletion
model is favoured by the exercise biochemists; the
muscle power/muscle recruitment model is advocated
by muscle physiologists, and some biomechanists and
neuro-physiologists; the biomechanical model by bio-
mechanists, and the psychological/motivational
model by sports psychologists.
Yet it is highly improbable that the factors explain-
ing human exercise performance under all conditions
are restricted to one physiological system or to one
scientific discipline. Thus, human performance is un-
likely to be adequately defined by any of these uni-
tary models that are often presented as if they are
mutually exclusive. The complexity of the physiologi-
cal and other factors determining human perform-
ance is emphasized when the limitations of each of
these models are exposed.
The cardiovascular/anaerobic model
Maximal exercise
This model holds that endurance performance is de-
termined by the capacity of the athlete’s large heart
to pump unusually large volumes of blood and oxy-
gen to the muscles. This allows the muscles to achieve
higher work rates before they outstrip the available
oxygen supply, developing skeletal muscle anaerobi-
osis (Fig. 1) (Noakes 1988, 1997, 1998, Bassett &
Howley 1997). This model remains the most popular
for explaining why fatigue develops during exercise;
how the body adapts to training; how these adap-
tations enhance performance and, as a consequence,
how effective exercise training programmes should be
This model predicts that training increases ‘‘cardio-
vascular fitness’’ especially by increasing the body’s
maximum capacity to consume oxygen, measured as
the maximum oxygen consumption (VO
max). This
effect results from an increased maximum capacity of
the heart to pump blood (the cardiac output) and
an enhanced capacity of the muscles to consume that
Physiological models to study exercise
Fig. 1. The Cardiovascular/Anaerobic Model of Exercise Physi-
ology and Athletic Performance holds that the heart has a limit-
ing maximum cardiac output that is reached at the onset of a
‘‘plateau phenomenon’’, ascribed incorrectly to the work of Hill
and colleagues (Hill et al. 1924, Hill 1927). As a result, blood
and hence oxygen flow to the exercising muscles falls behind
demand, inducing anaerobic metabolism. Metabolites of anaer-
obic metabolism, in particular hydrogen ions, then inhibit
muscle contraction, inducing fatigue.
oxygen, the latter by increasing skeletal muscle capil-
larization and mitochondrial mass. It is argued that
these adaptations delay the onset of skeletal muscle
anaerobiosis during vigorous exercise, thereby reduc-
ing blood lactate concentrations in muscle and blood
at all exercise intensities above the so-called ‘‘anaer-
obic threshold’’. The delayed onset of this blood lac-
tate accumulation then allows the exercising muscles
to continue contracting for longer at higher intensit-
ies before the onset of fatigue.
In addition, these changes increase the capacity of
the muscles to use fat as a fuel during exercise, there-
by enhancing endurance performance (according to
the Energy Depletion model, described subsequently)
(Saltin & Gollnick 1983, Holloszy & Coyle 1984). An
important but unrecognized prediction of this model
is that increases in coronary blood flow must be an
essential adaptation to training (Noakes 1998). The
higher coronary blood flow allows a greater pumping
capacity of the heart producing a greater cardiac out-
put to perfuse the exercising muscles, which can then
achieve a higher exercise capacity.
This model finds strong support from the confir-
mation that these changes, with the exception of a
greater coronary flow which is inferred, not proven,
do indeed result from training, as fully documented
in the literature. The key question is whether these
changes are causally linked; that is, do these changes
cause the change in exercise performance or do they
occur pari-passu with other adaptation(s) that are the
real cause of changes in exercise performance. For
there are important deficiencies in this model which
are fully argued (Noakes 1988, 1997) and counter-
argued (Noakes 1998, Bassett & Howley 1997) else-
where and will not be repeated here.
Perhaps the major but overlooked limitation of this
model is that, if the pumping capacity of the heart
does indeed limit oxygen utilization by the exercising
skeletal muscle, then the heart itself will be the first
organ affected by any postulated oxygen deficiency
(Noakes 1998). This was first recognized by Hill and
his colleagues as early as 1925 (Hill et al. 1924). Para-
doxically it was the incorrect interpretation, by
others, of the work of Hill and his colleagues, in par-
ticular their supposed description of a plateau phe-
nomenon (Noakes 1998), that forms the (mythical)
foundation for the cardiovascular/anaerobic model of
exercise physiology. Yet those who popularized this
mythical interpretation of the work of Hill and his
colleagues, failed also to record what Hill considered
to be the physiological cause (and equally the conse-
quence) of the fatigue that develops during maximal
For the interpretation of Hill and his colleagues
was unequivocal: ‘‘Certain it is that the capacity of
the body for muscular exercise depends largely, if not
mainly, on the capacity and output of the heart. It
would obviously be very dangerous for the organ to
be able, as the skeletal muscle is able, to exhaust itself
very completely and rapidly, to take exercise far in
excess of its capacity for recovery ... When the oxygen
supply becomes inadequate, it is probable that the
Fig. 2. The weakness of the Cardiovascular/Anaerobic Model
of Exercise Physiology and Athletic Performance is that the at-
tainment of a maximum cardiac output has more serious conse-
quences for the heart than it does for the skeletal muscles as
the first organ to be affected by a maximum cardiac output
would be the heart itself. Continuing to exercise with a fixed
(maximum) cardiac output would cause myocardial ischaemia.
The heart, unable to increase coronary flow (dependant on an
increase in cardiac output), would be unable to balance the in-
creased myocardial oxygen demand caused by the very increase
in workrate, theoretically necessary to define the mythical ‘‘pla-
teau phenomenon’’ of Hill and his colleagues. Many early re-
searchers believed that the heart did indeed develop ischaemia
during maximal exercise, leading to a fall in cardiac output (Hill
et al. 1924, Bainbridge 1931, Hill 1927, Dill 1938).
heart rapidly begins to diminish its output, so avoid-
ing exhaustion ...’.
The point identified by Hill and his colleagues, and
since ignored by all subsequent generations of exer-
cise physiologists, is that the heart is also a muscle,
dependent for its function on an adequate blood and
oxygen supply. But, unlike skeletal muscle, the heart
is dependent for its blood supply on its own pumping
capacity. Hence any intervention that reduces the
pumping capacity of the heart, or demands the heart
somehow to sustain an increased work output by the
exercising muscles without any increase in cardiac
output and coronary flow (as theoretically occurs
when the ‘‘plateau phenomenon’’ develops), imperils
the heart’s own blood supply. Any reduction in coro-
nary blood flow will consequently reduce the heart’s
pumping capacity, thereby inducing a vicious cycle of
progressive and irreversible myocardial ischaemia
(Fig. 2). It would seem logical that human design
should include controls to protect the heart from ever
entering this vicious circle.
Hence if (skeletal) muscle function fails when its
oxygen demand exceeds supply then, for logical con-
sistency, the inability of the pumping capacity of the
heart to ‘‘raise the cardiac output’’ at the VO
(Rowell 1993), must also result from an inadequate
(myocardial) oxygen supply caused by a plateau in
Fig. 3. According to the logic of the Cardiovascular/Anaerobic
Model of Exercise Physiology and Athletic Performance, coro-
nary flow must be the first physiological variable to ‘‘plateau’
during progressive exercise to exhaustion. The peak in coronary
flow would then induce a plateau in cardiac output as a result
of a progressive myocardial ischaemia. Continuing to exercise
with a fixed cardiac output and coronary flow would rapidly
cause an ischaemia-induced fall in cardiac output and in coro-
nary flow, and hence in whole body oxygen consumption. This
logic was accepted by the early proponents of this model (Hill
et al. 1924, Bainbridge 1931, Hill 1927, Dill 1938) but has since
been overlooked by exercise physiologists for the past 75 years.
coronary flow. This limiting coronary blood flow in-
duces myocardial ‘‘fatigue’’, causing the plateau in
cardiac output and hence in the VO
max leading,
finally, to skeletal muscle anaerobiosis. Thus, by this
logic, the coronary blood flow must be the first
physiological function to show a ‘‘plateau phenom-
enon’’ during progressive exercise to exhaustion (Fig.
3). All subsequent physiological ‘‘plateaus’’ must re-
sult from this limiting coronary flow (Noakes 1998).
Whereas the most influential modern exercise
physiologists have enthusiastically embraced this
mythical basis for a ‘‘plateau phenomenon’’ for the
past 75 years, none seems to have grasped this logical
prediction of the ‘‘plateau phenomenon’’, which is
that the ‘‘plateau phenomenon’’ requires the heart to
fatigue first before skeletal muscle fatigue can de-
velop. But this was clearly a concept with which the
pioneering exercise physiologists were entirely
comfortable. Thus, in addition to the conclusion of
Hill and his colleagues, already quoted, both Bock
and Dill (Bainbridge 1931) also believed that myocar-
dial hypoxia causes a fall in the cardiac output at the
point of fatigue during high intensity exercise:
‘‘The blood supply to the heart, in many men, may
be the weak link in the chain of circulatory adjust-
ments during muscular exercise, and as the intensity
of muscular exertion increases, a point is probably
reached in most individuals at which the supply of
oxygen to the heart falls short of its demands, and
the continued performance of work becomes difficult
or impossible’’ (p. 15). Hence they proposed that:
‘‘Another factor, which may contribute to the pro-
duction of this type of fatigue, is fatigue of the heart
itself ’’ (p. 229).
‘‘Although the occurrence of fatigue of the heart
in health is not very clearly established, a temporary
lowering of the functional capacity of the heart, in-
duced by fatigue of its muscular fibres, might gradu-
ally bring about during exercise an insufficient blood
supply to the skeletal muscles and brain. The lassi-
tude and disinclination for exertion, often experi-
enced on the day after a strenuous bout of exercise,
has been ascribed to fatigue of the heart as its pri-
mary cause’’ (p. 229). Hence they concluded: ‘‘The
heart, as a rule, reaches the limit of its powers earlier
than the skeletal muscles, and determines a man’s
capability for exertion’’.
In summary, the early physiologists who believed
that skeletal muscle anaerobiosis limits maximal exer-
cise clearly understood that any plateau in cardiac
output, necessary for there to be a limiting skeletal
muscle blood flow, must result from a plateau in coro-
nary blood flow which would expose the heart to a
progressive myocardial ischaemia that would worsen
as exercise was prolonged.
Perhaps the reluctance of modern physiologists to
acknowledge these concepts stems from the current
Physiological models to study exercise
appreciation that progressive myocardial ischaemia
does not occur during maximal exercise in healthy
athletes (Raskoff et al. 1976), even though there is
good evidence that it is a limiting cardiac output that
probably determines the VO
max (Rowell 1993).
Thus, one postulate might be that even if cardiac out-
put limits maximal exercise as seems likely (Noakes
1997), termination of exercise must occur before the
heart actually reaches that maximum and hence well
before skeletal muscle anaerobiosis can develop
(Noakes 1998). Hence for 75 years, exercise physiol-
ogists may have focused on the incorrect organ as the
site of any potential anaerobiosis that may develop
during maximal exercise (Hill et al. 1924, Bainbridge
1931, Hill 1927, Hill et al. 1924).
How might a maximal cardiac output be reached
without the development of myocardial ischaemia?
The argument that the rate of cardiac filling, due
either to a limiting venous return or the effects of a
restrictive pericardium (Stray-Gundersen et al. 1986)
may limit the maximal cardiac output, whilst super-
ficially attractive, is still unable satisfactorily to ex-
plain which physiological events terminates exercise.
Such an argument fails for the reason that the con-
tinuation of exercise beyond that (however limited)
maximal cardiac output must still cause a progressive
myocardial ischaemia to develop (Fig. 2). Hence, even
if the cardiac output is limited by factors unrelated
to the development of myocardial ischaemia (for ex-
ample, a limiting venous return), the continuation of
exercise beyond that point of limitation must induce
myocardial ischaemia and the development of chest
pain (angina pectoris) that would terminate exercise.
Perhaps it is more logical to speculate that maxi-
mal exercise terminates as part of a regulated process
before the absolute maximum cardiac output and
coronary blood flow are achieved. Interestingly Hill
and his colleagues seem to have been the first to sug-
gest a solution to this dilemma as early as 1924:
‘‘From the point of view of a well co-ordinated mech-
anism, ... it would clearly be useless for the heart to
make an excessive effort if by doing so it merely pro-
duced a far lower degree of saturation of the arterial
blood; and we suggest that, in the body (either in the
heart muscle itself or in the nervous system), there is
some mechanism which causes a slowing of the circu-
lation as soon as a serious degree of unsaturation oc-
curs, and vice versa. This mechanism would tend to
act as a governor maintaining a high degree of satu-
ration of the blood’’ (Hill et al. 1924, p. 161–162).
Clearly no such governor has yet been discovered,
perhaps because no physiologists have yet searched
for it. But there is clear physiological evidence for the
existence of such a governor. The evidence comes
from studies of skeletal and cardiac muscle function
at altitude. For if oxygen deficiency really does de-
velop in either heart or skeletal muscle during maxi-
mum exercise, its appearance will likely be more eas-
ily identifiable during exercise at altitude under con-
ditions of hypobaric hypoxia. Furthermore, such
experiments should identify in which organ – heart
or skeletal muscle – anaerobiosis first becomes ap-
parent; the heart, according to the ideas of the pion-
eering British and North American exercise physiol-
ogists, or the skeletal muscles, according to the influ-
ential group of modern exercise physiologists
(Noakes 1998).
The original studies of exercise at altitude were
undertaken by a research group co-ordinated by Dill
and his colleagues from the Harvard fatigue labora-
tory. This research established two crucial findings.
First, that peak blood lactate concentrations during
maximum exercise fell with increasing altitude (Ed-
wards 1936), a phenomenon since labelled the ‘‘lac-
tate paradox’’ (Hochachka 1989). Second, that maxi-
mum heart rate and cardiac output likewise fell dur-
ing exercise at increasing altitude (Christensen 1938,
Dill 1938).
Edwards (1936) interpreted the ‘‘lactate paradox’’
at altitude accordingly: ‘‘The inability to accumulate
large amounts of lactate at high altitudes suggests a
protective mechanism preventing an already low ar-
terial saturation from becoming markedly lower ... It
may be that the protective mechanism lies in an inad-
equate oxygen supply to essential muscles, e.g. the
diaphragm or the muscles’’.
The existence of the ‘‘lactate paradox’’ was con-
firmed during the epic laboratory experiment of exer-
cise and acclimatization at simulated high altitude,
Operation Everest II (Green et al. 1989). That study
found that muscle lactate concentrations achieved
during maximal exercise at the highest equivalent alti-
tude achieved during that experiment (8848 m –
equivalent to the summit of Mount Everest) were no
higher than when at rest at sea level.
Hence, in as much as high muscle lactate concen-
trations would have to be present if the exercising
muscles were contracting ‘‘anaerobically’’, this study
proves that exercise at extreme altitude terminates
when the exercising muscles are contracting in fully
aerobic conditions.
Similarly Operation Everest II (Sutton et al. 1988)
confirmed these original and subsequent studies
(Pugh 1964, Vogel et al. 1974) showing that heart rate
and cardiac output are substantially reduced during
exercise at extreme altitude. The key observation is
that the peak cardiac output falls with increasing alti-
tude. This response is equally paradoxical for those
who believe that the delivery of an adequate oxygen
supply to the exercising muscles is the cardinal prior-
ity during exercise (Noakes 1998). For logic demands
that if the principal responsibility of the cardiovascu-
lar system during exercise is the achievement of an
(ultimately inadequate) oxygen supply to skeletal
muscle, then the maximum cardiac output during ex-
ercise at increasing altitude must either stay the same
or even increase at increasing altitude in order to limit
the effects of the progressive reduction in the arterial
oxygen content.
Yet the evidence is absolutely clear. The heart
makes the exactly opposite adjustment – maximum
cardiac output falls with increasing altitude (Sutton
et al. 1988). The reduction is due to the reduction in
heart rate; stroke volume and myocardial contrac-
tility are, if anything, enhanced during peak exercise
at altitude (Reeves et al. 1987, Suarez et al. 1987).
Hence the conclusion must be that some currently
unrecognized mechanism must exist to insure that the
heart does not become ‘‘anaerobic’’ during maximal
exercise at any altitude – from sea level to the summit
of Mount Everest – in healthy humans.
Interestingly Christensen, but not Dill, interpreted
this phenomenon correctly: ‘‘Christensen and I dif-
fered in our interpretation of his measurements of
respiratory and circulatory function in exercise (at
altitude). In his opinion, the chief limiting factor is
the ventilation of the lungs. In the hardest grade of
work at any station, the pulmonary ventilation
reached about as high a value as at sea level, while the
maximal cardiac output became less as the altitude
increased. He thinks this means that the heart has an
untapped reserve; it is circulating blood fast enough
to carry to the tissues all the oxygen supplied by the
lungs’’ (Dill 1938, p. 170–171).
These studies invite two precise conclusions. First,
that the oxygen demands of the skeletal muscles are
not the cardinal priority and hence are not ‘‘pro-
tected’’ during maximum exercise, at least at extreme
altitude. Second, neither the skeletal muscles nor the
heart becomes ‘‘anaerobic’’ during maximal exercise
Fig. 4. The ‘‘governor’’ postulated by Hill could be activated by
a limiting myocardial oxygen delivery. As a result, there would
be a reduced activation of the exercising skeletal muscle by the
cerebral motor cortex. There is compelling evidence for the
presence of this reflex during exercise at altitude (Kayser et al.
1994) with the result that neither the heart (Sutton et al. 1988,
Reeves et al. 1987, Suzrez et al. 1987) nor the skeletal muscles
(Green et al. 1989) develop ‘‘anaerobiosis’’ during maximal ex-
ercise at extreme altitude.
under conditions of hypobaric hypoxia. The sole con-
clusion must be that some type of ‘‘governor’’, as
originally proposed by A.V. Hill, must limit maxi-
mum exercise at altitude. Furthermore, it would be
difficult to explain why the same control mechanism
should not act similarly during maximum exercise at
sea level.
In summary, a number of famous studies have
shown that under the precise conditions likely to in-
duce anaerobiosis in either the heart or skeletal
muscles – maximal exercise at altitude – neither the
heart nor the skeletal muscle show any evidence
whatsoever for ‘‘anaerobic’’ metabolism. This unex-
pected finding can be explained only if there is a
‘‘governor’’, probably in the central nervous system,
whose function is likely to prevent the development
of myocardial ischaemia. The same governor could
also serve the identical function also at sea level,
thereby preventing the development of myocardial
ischaemia during maximum exercise at sea level, ac-
cording to Fig. 2. As Dill (1938) concluded, probably
correctly: ‘‘The capacity of the heart, as has already
been suggested, is restricted at high altitude because
of the deficiency in supply of oxygen to it’’ (p. 15).
But the important point is that the heart never actu-
ally develops an oxygen deficiency at altitude or at
sea level; the governor acts to terminate exercise be-
fore that deficiency becomes apparent.
The final confirmation for the presence of this the-
oretical governor comes from the study of Kayser et
al. (1994). They showed that skeletal muscle recruit-
ment, measured as skeletal muscle EMG activity at
peak exercise, falls with increasing altitude, but in-
creases acutely with oxygen administration. They
conclude: ‘‘during chronic hypobaric hypoxia, the
central nervous system may play a primary role in
limiting exhaustive exercise and maximum accumu-
lation of lactate in blood’’. This study therefore
proves the existence of the neural effector limb of
Hill’s postulated governor (Fig. 4) and its activity
during exercise at altitude.
Interestingly, had the human body been designed
to function according to the modern physiologists’
cardiovascular/anaerobic model, which requires that
anaerobiosis first develops in skeletal muscle before
maximal exercise is terminated, no climber would
ever have reached the summit of Mount Everest or
other high mountains, even with the use of sup-
plemental oxygen. Rather, all would have succumbed
to a combination of myocardial ischaemia and cer-
ebral hypoxia whilst their skeletal muscles were exer-
cising vigorously and unrestrainedly, in pursuit of an-
aerobiosis and fatigue, according to the model de-
picted in Fig. 1.
Figure 4 therefore summarizes the hypothetical
existence and action of the ‘‘governor’’, first proposed
by A.V. Hill. It is postulated that receptor(s) exist in
Physiological models to study exercise
the heart, to assess the adequacy of any of all of the
following: coronary blood flow, coronary oxygen de-
livery or myocardial or coronary venous oxygen ten-
sion. Before any of these reach some predetermined
limit, the motor cortex in the brain reduces skeletal
muscle activation. As a consequence, skeletal muscle
recruitment either fails to rise further or it falls, limit-
ing the work output of the body, indicating the onset
of ‘‘fatigue’’. The fall in work output by the body
reduces myocardial oxygen demand and, as a conse-
quence, the threat of myocardial ischaemia is averted.
Alternatively it may be that myocardial adenosine
triphosphate (ATP) concentrations are sensed and
‘‘defended’’ in much the same way as appears to be
the case for skeletal muscle, as discussed subsequently
(Fitts 1994, Spriet et al. 1987). Reduction of myocar-
dial ATP concentrations could lead directly to a re-
duction in myocardial contractile force as occurs in
‘‘myocardial stunning’’ (Braunwald & Kloner 1982).
This could explain the onset of cardiac failure during
maximal exercise in persons with coronary artery dis-
ease but could not explain why, at altitude, left ven-
tricular function is enhanced during maximal exercise
and shows no evidence for ‘‘fatigue’’.
Accordingly, it is proposed that maximal exercise is
limited by a regulated process that terminates exercise
before the development of a progressive myocardial
ischaemia, that would precede the development of
skeletal muscle anaerobiosis. This model further pre-
dicts that peak coronary blood flow is an important
determinant of maximum exercise performance, and
that interventions, including exercise training, that in-
crease the maximum cardiac output probably also in-
crease the maximum coronary blood flow, as their
more important effect.
But this model does not exclude the possibility that
interventions could also improve exercise perform-
ance by altering either skeletal muscle or myocardial
contractile function or the efficiency of oxygen util-
ization, or both (Fig. 5).
Interestingly, the presence of a ‘‘governor’’ prevent-
ing the development of anaerobiosis in either heart
or skeletal muscle during exercise at altitude has in-
teresting implications for theories of the value of ex-
ercise training at altitude. For its presence means that
any beneficial effect of altitude training cannot result
from repeated exposure of either the heart or exercis-
ing skeletal muscle to greater levels of ‘‘anaerobiosis’
than can be achieved during maximal exercise at sea
level. This might explain why there remains consider-
able controversy about the proven value of high in-
tensity training at altitude (Boning 1997).
The additional paradoxes that (i) adaptation to ex-
treme altitude is associated with reduced skeletal
muscle mitochondrial volume and enzyme content
(Green et al. 1989, Oelz et al. 1986, Hoppeler et al.
1990); (ii) the skeletal muscle morphology of altitude-
Fig. 5. According to the Hill/Noakes Cardiovascular/Neural
Model of Exercise Physiology and Athletic Performance, per-
formance during maximal exercise is ultimately limited by the
peak coronary blood flow. However, the actual workrate
achieved at that peak coronary blood flow would be determined
by the efficiency and contractility of both the heart and the
active skeletal muscles.
adapted Nepalese Sherpas is not different from that
of acclimatized Caucasian climbers (Kayser et al.
1991) except that (iii) the volume density of skeletal
muscle mitochondria is significantly smaller in Sherp-
as than in untrained sedentary subjects (Kayser et al.
1991), can best be explained if exercise performance
at altitude is more likely determined by factors pro-
ducing superior oxygenation of the heart, than of the
skeletal muscles. This model therefore predicts that
coronary blood flow and perhaps cardiac mitochon-
drial mass would be higher in high altitude natives
and superior performers at high altitude. This would
explain why high altitude natives achieve almost simi-
lar cardiac outputs at sea level and at altitude (Vogel
et al. 1974), indicating a lesser activation of the ‘‘gov-
ernor’’ during exercise at altitude.
Indeed the ability of high altitude natives to
achieve high heart rates and cardiac outputs during
exercise at altitude is associated with ‘‘relatively high
oxygen tension and saturation’’ (Vogel et al. 1974)
compatible with this postulate that superior myocar-
dial oxygenation might be an important factor deter-
mining exercise capacity at altitude.
Prolonged submaximal (endurance) exercise
Many physiologists, notably in the past (Bainbridge
1931, Hill 1927, Dill 1938) but even today (Bassett &
Howley 1997), have used this cardiovascular/anaer-
obic model also to explain the fatigue that develops
during prolonged submaximal exercise and conse-
quently have evoked changes in cardiovascular func-
tion to explain the mechanisms by which exercise
training improves (endurance) performance during
prolonged submaximal exercise.
Yet it is not entirely apparent why changes in the
Table 2. World rankings of male Kenyan runners in 1996
Distance Rankings in world top 10
800 m 1*, 4, 6, 7, 10
1500 m 4, 5, 7, 8
5000 m 1, 7, 8, 9, 10
10000m 3,4,5,6
3000 m steeplechase 1, 2, 3, 4, 5, 6, 8, 9, 10
3000 m 1, 3, 4, 5
Marathon 4, 6
* The naturalized Dane, Wilson Kipketer, is considered a Kenyan for the
purposes of this analysis.
Table 3. Performances of Kenyans in the I.A.A.F. world cross-country
championships (1986–1997)
Kenyan Senior Men, 1st for the last 12 years (1986–1997)
Kenyan Junior Men, 1st for the last 10 years (1988–1997)
Kenyan Senior Women, 1st 5 times in the last 7 years
Kenyan Junior Women, 1st 8 times in the last 9 years
Total: 35 Championships in 49 competitions including 24 individual
maximum capacity to transport and utilize oxygen
must also explain alterations in performance during
submaximal exercise when oxygen transport cannot
be limiting. An early proponent of this (il)logic was
Sir Roger Bannister who wrote in 1956 that: ‘‘The
muscular effort in long-distance running appears to
be limited by cardio-respiratory failure as a whole
and not by premature failure of any part of the inte-
gration’’ (Bannister 1956).
The obvious point is that, whereas the cardiovascu-
lar system could indeed set the limit for maximal ex-
ercise performance because of a limiting capacity to
increase blood flow first to the heart, and then to the
active muscles, it is not clear why cardiovascular func-
tion should limit prolonged submaximal exercise
when blood flow and oxygen supply to muscle must
be adequate. An Olympic analogy from my (African)
continent highlights the issues that require debate.
In the years since Wilson Kiprugut won Kenya’s
first Olympic medal by finishing third in the 800 m at
the 1964 Olympic Games, the dominance by
Africans, especially Kenyans, in distance running has
become a phenomenon unequalled in any other sport
in the world (Bale & Sang 1996, Tanser 1997). Two
measures of that dominance are provided by the
world rankings of male Kenyan track runners in 1996
(Tanser 1997) (Table 2) and of the performances of
the men and women’s team, both senior and junior,
in the World Cross-Country championships over the
past 12 years (Tanser 1997) (Table 3). Of particular
interest is the almost total dominance of the 3000 m
steeplechase by Kenyans (Table 2). Indeed in excess
of 90% of the 100 fastest-ever 3000 m steeplechase
times in the world have been set by Kenyans. Any
physiological explanation for the Kenyan’s success
must be able also to explain why these physiological
attributes, uniquely common in Kenyan runners, of-
fer an even greater advantage in cross-country events
and in the steeplechase, rather than at other distances
in which repeated jumping and changes in speed do
not occur.
Two studies of Kenyan runners failed to provide a
definitive physiological answer for their manifest su-
periority as distance runners although two of the best
Kenyan runners were the most efficient runners yet
studied (Saltin 1996, Saltin et al. 1995a,b) (Fig. 6).
The overriding conclusion was that the Kenyans’ VO
max values were not inordinately high; hence, a su-
perior capacity for oxygen consumption during maxi-
mum exercise did not explain the Kenyans’ manifest
superiority during more prolonged submaximal exer-
cise. In the words of Bengt Saltin, the senior author:
‘‘A comparison of some data on some of the very
best runners in Kenya during the last decades and
world class runners in Scandinavia does not reveal
much that was not already known or could be antici-
pated’’ (Saltin 1996).
The only other study of elite (South) African dis-
tance runners is that of Coetzer et al. (1993). That
study, which reported physiological data in the best
group of distance runners yet evaluated anywhere in
the world, has been largely ignored, for reasons that
are not immediately clear. The sole weakness of the
study was that the physiological characteristics of el-
Fig. 6. In comparison to a group of elite Scandinavian runners
(the highest running curve), the running economy (VO
at any
running speed) of former world marathon record holder, Derek
Clayton, and two of the greatest Kenyan runners, Julius Korir
and John Ngugi, is substantially better. The Biomechanical
Model of Exercise Physiology and Athletic Performance pre-
dicts that superior running economy would aid elite athletic
performance by reducing both the rate at which heat is pro-
duced during exercise as well as the extent to which Stretch
Shortening Cycle Fatigue develops during both training and
racing. This theory predicts that the superior running ability of
the Kenyans may relate, at least in part, to characteristics of
their skeletal muscle and tendon elasticity (see also Fig. 12).
Physiological models to study exercise
Table 4. Characteristics of elite South African distance runners
Middle Long
distance distance
Height (cm) 181 169*
Weight (kg) 70 56*
Aerobic capacity (ml/kg/min) 72 71
Fatigue resistance (% VO
max for 21 km) 82 90*
Muscle fibre composition (% Type I) 63 53
Data from Coetzer et al. 1993.
ite black South African distance runners were, per-
force, compared to those of white South African
middle distance runners. This comparison group was
not ideal but was necessary because no other readily
available group could match the performances of the
black distance runners at any distance over 5 km. But
the performances of the black and white runners in
that study were at least equal in races of 1–3 km.
Table 4 lists the important findings of that study.
The black distance runners were lighter and
smaller, as also reported by Saltin (1996), with a
slightly lower proportion of type I muscle fibres. But
the key finding was that the black runners were able
to run substantially faster at all distances beyond 5
km despite VO
max values that were the same as
those of the middle distance runners. Hence the car-
diovascular/anaerobic model failed to explain the su-
perior endurance capacity of the black distance run-
ners in that study and perhaps also in the studies of
Saltin and colleagues (Saltin 1996, Saltin et al.
Rather, the important difference was that the black
runners were able to sustain a substantially higher
proportion of their VO
max when racing. This is
shown graphically in Fig. 7, which compares the %
max sustained by the black and white runners
at different racing distances. At distances beyond 5
km, black runners sustained a significantly higher %
max than did the white runners and the differ-
ence increased with increasing racing distance. Thus,
the crucial finding was that the black distance run-
ners have superior fatigue resistance, not a higher
aerobic capacity (VO
max). Hence factors distal to
the heart, perhaps in the brain or in the muscles, ap-
pear to distinguish the very best runners in the world
from those who are almost as good.
Interestingly, physiologists have known for at least
two decades that the % VO
max that athletes can
sustain during exercise is an important predictor of
performance (Costill et al. 1973, Davies & Thompson
1979). Yet we have perhaps failed to emphasize, prob-
ably because of a devotion to the cardiovascular/an-
aerobic model, that this is likely to be a more import-
ant determinant of performance in prolonged exer-
cise than is the VO
max alone. Furthermore, it has
not been appreciated that the % VO
max sustained
during exercise is a measure of the athlete’s resistance
to fatigue.
Hence, the important finding of that study was to
show that the cardiovascular/anaerobic model may be
unable to discriminate between very good and su-
perior performance in events lasting more than a few
minutes and which constitute the bulk of sporting
events. It is consistent with the finding that the VO
max is a relatively poor predictor of endurance per-
formance in athletes whose abilities are relatively
homogenous (Noakes 1988, 1997, 1998, Davies &
Thompson 1979, Noakes et al. 1990). The failure
stems from the inability of this model to measure or
predict fatigue resistance during prolonged submaxi-
mal exercise on the basis of physiological variables
and performance measured during a single bout of
progressive, maximal exercise to exhaustion. This
confirms that the VO
max test does not measure all
the physiological variables determining success dur-
ing more prolonged exercise.
Further support for this explanation can be sur-
mised from other information in Fig. 7, which shows
that these athletes run at 100% or greater of their
max in race distances of 1–2 km. Yet it is not at
those distances that the Kenyans’ dominance is most
apparent. If the Kenyans’ success was due to their
unusually high VO
max values, one would expect
Kenyans also to be dominant at race distances of 800
m to the mile, which is not the case (Table 2).
Indeed, comparison of the performances of the
Fig. 7. The%VO
max sustained by elite South African black
and white distance runners falls with increasing racing distance.
However, black runners sustain a significantly higher % VO
max at race distances of 10 km and 21 km, indicating superior
fatigue resistance. Such superiority cannot be explained by the
Cardiovascular/Anaerobic Model of Exercise Physiology and
Athletic Performance as VO
max values of black and white
athletes in this study (Coetzer et al. 1993) were the same.
great British runner, Sebastian Coe, with those of a
current Kenyan champion, Daniel Komen, provides
further evidence for this interpretation. Remarkably,
Komen’s best time for the mile is 1 s faster than Coe’s
best. Yet the real performance difference occurs at 5
km: Komen’s best time is 83 s faster than Coe’s best,
a performance difference of 10%.
In summary, there are serious theoretical flaws in
the proposed cardiovascular/anaerobic model of exer-
cise physiology and athletic performance (Noakes
1998), not least because the model predicts that a
‘‘plateau’’ in cardiac output must develop before skel-
etal muscle anaerobiosis can begin to occur. But any
‘‘plateau’’ in cardiac output requires that myocardial
ischaemia be present either to cause that plateau (ac-
cording to the theory that anaerobiosis limits muscle
function) or as a result of it, as the cardiac output
determines both coronary and skeletal muscle blood
flow. As myocardial ischaemia has never been shown
to develop during maximal exercise in healthy
humans, so it would seem unlikely that skeletal
muscle anaerobiosis can develop during progressive
exercise to exhaustion (Noakes 1998). Rather, it
would seem that ‘‘fatigue’’ during maximal exercise
of short duration is part of a regulated neural process
that prevents the development of myocardial isch-
aemia during maximal exercise.
Whilst this mechanism is designed to protect the
heart from myocardial ischaemia, only indirectly does
it determine the actual peak work rate achieved dur-
ing maximal exercise (Fig. 5). The actual peak work-
rate achieved will depend on the ‘‘quality’’ of the skel-
etal and cardiac muscle. Superior myocardial contrac-
tility and efficiency of oxygen use would increase the
maximum cardiac output achieved at any maximum
(limiting) coronary flow. Similarly at any maximum
skeletal muscle blood flow, superior contractility and
efficiency of skeletal muscle contraction would in-
crease the peak workrate achieved at that maximum
cardiac output. This hypothesis forms what might be
called the Cardiovascular/Neural Recruitment Model
of Exercise Physiology and Athletic Performance.
Thus, this analysis of the traditional cardiovascu-
lar/anaerobic model of exercise performance leads to
the alternate hypothesis that superior fatigue resis-
tance, determined perhaps by the central nervous sys-
tem or skeletal muscle contractile function, might ex-
plain superior performance in events lasting more
than a few minutes. This superior fatigue resistance
cannot be predicted by the cardiovascular/anaerobic
model which uses exercise tests of short duration and
in which the fatigue resistance component of endur-
ance performance is not measured. By extension, it
would seem that fatigue resistance is not causally de-
termined by the magnitude of the athlete’s cardio-
vascular capacity. There is also no logical reason to
believe that fatigue resistance during submaximal ex-
ercise is determined by either the presence or absence
of skeletal muscle anaerobiosis.
Accordingly, changes in exercise performance that
result from endurance training are unlikely to be de-
termined solely by changes in cardiovascular func-
tion, with the exception that increases in maximum
coronary blood flow would likely be crucial for any
increases in maximal cardiac output and hence in
max. It is of interest that the vasodilator ca-
pacity of the major epicardial coronary vessel is
greatly increased in veteran long distance runners
(Haskell et al. 1993). Perhaps this indicates that an
important effect of endurance training, possibly at
some critical growth periods, may be to increase
maximum coronary blood flow as shown in animal
models (Scheuer & Tipton 1997).
The energy supply/energy depletion model
The energy supply model
The central premise in the cardiovascular/anaerobic
model is that it is the provision of a substrate (oxy-
gen) to muscle that limits exercise performance so
that fatigue is a direct consequence of a failure of
oxygen delivery to the exercising muscles. A subtle
extension of this idea produces a second model which
proposes that fatigue during high intensity exercise
may, alternatively, result from the inability to supply
another substrate (ATP) at rates sufficiently fast to
sustain exercise. Nobel Laureate A. V. Hill, whose re-
search in the 1920s was directly responsible for the
development of the cardiovascular/anaerobic model
of exercise performance, also wrote: ‘‘The fact re-
mains, however, that the chief factor in many forms
of athletic achievement is the supply of energy and
its proper and economic utilization’’ (Hill 1927, p.
237). Dr Peter Snell, Olympic Gold medallist in the
800 and 1500 m and former world record holder has
stated similarly: ‘‘Performance in middle and (long)
distance running ultimately depends upon the run-
ner’s capacity to produce energy for the duration of
the event, and on the efficiency with which that en-
ergy is translated to running velocity. Thus the pur-
pose of training is to improve the energy delivery sys-
tems, according to the demands of the event and to
improve running economy’’ (Snell 1997).
Thus, this model predicts that performance in
events of different durations is determined by the ca-
pacity to produce energy (ATP) by the different meta-
bolic pathways including the phosphagens, oxygen-
independent glycolysis, aerobic glycolysis and aerobic
lipolysis. Superior performance is then explained by
a greater capacity to generate ATP in the specific
metabolic pathway(s) that predominates during that
activity. Thus, the sprinter is assumed to have a
greater capacity to generate ATP from the intramus-
cular phosphagen stores and from oxygen-indepen-
Physiological models to study exercise
dent glycolysis, whereas the ultramarathon runner
has a superior capacity to oxidize fat (aerobic lipoly-
sis) (Hawley & Hopkins 1995).
Whether this hypothesis is true is uncertain as it
has yet to be systematically evaluated. To prove this
model would require (i) that the metabolic capacities
of these different pathways be shown to be causally
related to performance in events lasting the different
durations; (ii) that the specific metabolic pathways be
shown to adapt predictably with specific training, and
(iii) that these adaptations alone explain the changes
in performance that result from training with exer-
cises lasting the different durations. Until these
studies are completed, this model remains hypotheti-
cal, but interesting. It must be remembered that the
truth of this model would need to disprove the op-
posing model, described in the previous section,
which holds that maximal exercise performance is a
regulated process limited by a failure of central neural
recruitment. Fatigue at exhaustion, caused by a fail-
ure of central recruitment, will always appear to be
due to a failure of ATP production unless the alter-
nate possibility is studied (and excluded) simul-
The argument against this model has been intro-
duced elsewhere (Noakes 1997). In short, the predic-
tion of this model is that exercise must terminate
when muscle ATP depletion occurs (Fitts 1994), that
is when the muscle develops rigor. Yet here again, the
evidence appears clear. ATP concentrations, even in
muscles forced to contract under ischaemic con-
ditions, do not drop below about 60% of resting
values (Fitts 1994, Spriet et al. 1987, Hochachka
1994) indicating that muscle ATP concentrations are
‘‘defended’’ in order to prevent the development of
skeletal muscle rigor. As Fitts (1994) has concluded:
‘‘The overriding evidence suggests that (the high en-
ergy phosphates) do not participate in the fatigue
process; that fatigue produced by other factors re-
duces the ATP utilization rate before ATP becomes
limiting. The most compelling evidence for this con-
clusion is that cell ATP rarely falls below 70% of the
pre-exercise level, even in cases of exercise fatigue’’ (p.
Hence, is appears that the rate of ATP demand by
the contracting muscles can never exceed the maxi-
mum rate of ATP supply because of the close
matching of ATP demand to the available ATP sup-
ply (Spriet et al. 1987, Hochachka 1994).
There is an obvious analogy to the centrally situ-
ated neural ‘‘governor’’ that prevents the develop-
ment of myocardial ischaemia during maximal exer-
cise at either sea level or altitude. The difference is
that the ‘‘governor’’ identified by Spriet et al. (1987)
is clearly located in the periphery and acts even in
muscles stimulated to contract with an externally ap-
plied current.
It is of interest that the presence of this peripheral
‘‘governor’’ is an essential component of the cardio-
vascular/anaerobic model as originally conceived
from the work of Hill and his colleagues, and still
widely promoted (Fig. 1). This hypothesis holds that
when the rate of ATP production by oxidative sources
becomes inadequate, high rates of ‘‘anaerobic’’ glyco-
lytic ATP production produce metabolites, particu-
larly H
, which interfere with energy production and
cross-bridge cycling causing fatigue and a failure of
muscle contraction (Fitts 1994). In this way, muscle
contraction fails not because of a failure of central
recruitment (as predicted by the Cardiovascular/Neu-
ral Recruitment Model – previous section), but be-
cause of a peripherally located inhibition of muscular
contraction. Proponents of this model can cite a large
body of evidence showing that a number of metabo-
lites can interfere with muscle cross-bridge cycling
measured in vitro in isolated muscle fibres (Fitts
1994). The necessary assumption is that skeletal
muscle contracting in vitro in the absence of an intact
neural system behaves exactly as it would in vivo
when the influences from the central nervous system
are intact. But there is a body of evidence that is not
compatible with these assumptions and conclusions.
For example, one of the few studies to evaluate
critically this hypothesis that metabolites, particu-
larly H
, can induce skeletal muscle fatigue, is that
of Mannion et al. (1995). They found that there is a
wide range of muscle pH concentrations reached at
exhaustion during intense exercise showing that if an
accumulation of H
limits high intensity exercise in
vivo, then ‘‘considerable interindividual differences
must exist in the pH sensitivities of the various pro-
cesses involved’’ (Mannion et al. 1995, p. 98).
Next, they found that in contrast to the prediction
from in vitro studies, subjects with the highest pro-
portion of type II muscle fibres were able to exercise
to the lowest muscle pH concentrations. In contrast,
in vitro studies have suggested that type I muscle
fibres are more resistant to acidosis than are type II
fibres. Finally, the authors found that subjects with
a greater skeletal muscle buffering capacity did not
accumulate more lactate during maximal exercise;
nor were they able to exercise for longer than did
those with lesser muscle buffering capacity. They con-
cluded that ‘‘if acidosis makes any contribution to the
fatigue during performance of this (high intensity)
type of exercise, it is an indirect one ...’’.
One possibility is that such exercise is terminated
by a central governor responding to factors other
than skeletal muscle pH. Under these circumstances
there would be no relationship between the onset of
fatigue and muscle acidosis. This would not negate
the established finding that in vivo, acidosis inhibits
crossbridge cycling (Fitts 1994). It would mean only
that this mechanism is not relevant in exercising
humans, perhaps because exercise terminates for
other reasons, in particular to prevent the develop-
ment of myocardial ischaemia, before the limiting
skeletal muscle pH is reached.
Other relevant findings include the study of Bog-
danis et al. (1995), who showed that recovery of
muscle function following maximal sprint cycling ex-
ercise was related to recovery of muscle phospho-
creatine concentrations and unrelated to muscle pH
concentrations during recovery. In addition, Vollestad
et al. (1988) showed that the gradual decline in maxi-
mum force generation in subjects performing re-
peated submaximal contractions for 40–70 min was
‘‘not due to lactacidosis or lack of substrates for ATP
resynthesis and must have resulted from excitation/
contraction coupling failure ...’. Yet terminal exhaus-
tion was associated with depletion of intramuscular
phosphagen stores, but without evidence for acidosis.
In summary, a metabolic basis limiting high inten-
sity exercise of short duration is widely assumed but
incompletely documented. There is a need to estab-
lish whether those metabolic factors that appear to
limit muscle function in vitro also play a role in vivo
when the muscle is also under the influence of the
central nervous system. Thus, the possible contri-
bution of neural factors to this form of fatigue needs
to be excluded before results from in vitro studies are
extrapolated, without qualification, to the in vivo
The energy depletion model
The related energy depletion model of exercise per-
formance is specific for exercise lasting more than 2–
3 h. It holds, in essence, that: ‘‘Depletion of endoge-
nous carbohydrate stores has been shown to be a lim-
iting factor in the ability to perform long term exer-
cise’’ (Costill et al. 1973). The findings that support
this conclusion are (i) that fatigue during prolonged
exercise is associated with depletion of liver (causing
hypoglycaemia) or muscle glycogen stores (Fitts 1994,
Bosch et al. 1993, Coggan & Coyle 1987, Coyle et al.
1986, Tsintzas et al. 1996), or both; (ii) that reversal
of hypoglycaemia allows exercise to continue
(Coggan & Coyle 1987, Coyle et al. 1986, Tsintzas et
al. 1996, Christensen & Hansen 1939) and (iii) that
pre-exercise muscle glycogen supercompensation
(carbohydrate-loading) (Hawley et al. 1997) or carbo-
hydrate ingestion during exercise (Coyle et al. 1986),
or both (Bosch et al. 1996), delays the onset of fatigue
and improves exercise performance.
However, it must be remembered that relatively few
carbohydrate-loading studies have been conducted
with an adequate placebo control group. It seems
highly improbable that neither athletes nor re-
searchers are completely unaware of the widely re-
ported benefits of carbohydrate-loading and that
such knowledge is without effect on the findings of
these trials. Indeed, two of the first such trials which
included adequate placebo-controlled groups have
both failed to find any ergogenic effect of pre-exercise
carbohydrate loading (Burke et al. 1999, Hawley et
al. 1997) under experimental conditions when such
an effect might have been expected.
The finding that the reversal of hypoglycaemia
alone allows exercise performance to continue
(Coggan & Coyle 1987, Christensen & Hansen 1939)
proves conclusively that liver glycogen depletion is
one form of energy (carbohydrate) depletion that can
definitely limit exercise performance. Interestingly,
the rapidity with which the reversal of hypoglycaemia
restores exercise performance indicates that a central
neural ‘‘governor’’ must be active, similar to that acti-
vated during high intensity exercise at altitude. How-
ever, this control would be activated by changes in
blood glucose concentrations and would act to pre-
vent continuing high rates of muscle contraction and
blood glucose oxidation that would further reduce
the blood glucose concentration, risking hypogly-
caemic cerebral damage.
But, in as much as no technique has yet been de-
vised that will instantly reverse muscle glycogen de-
pletion, in the same way that intravenous glucose in-
fusion or oral glucose ingestion rapidly reverses hy-
poglycaemia, so it is impossible to prove conclusively
that muscle glycogen depletion alone limits prolonged
exercise performance. It needs to be remembered that
there are many physiological changes besides muscle
glycogen depletion that develop during exercise, and
that any or all of these could contribute to, or cause
fatigue during prolonged exercise. In addition, rela-
tively little attention has been paid to the possible
role of central (neural) fatigue (Davis & Bailey 1997)
as the factor limiting prolonged exercise when muscle
glycogen concentrations are also very low. Future
studies of the energy depletion model need to show
that central neural factors do not cause the fatigue
currently ascribed to the development of muscle gly-
cogen depletion during prolonged exercise.
Thus, the belief that muscle glycogen depletion
causes fatigue is an interesting hypothesis that is sup-
ported logically by the findings that subjects who are
exhausted during prolonged exercise develop very low
muscle glycogen content (Fitts 1994, Bosch et al.
1993, Tsintzas et al. 1996, Burke et al. 1999), and that
muscle glycogen is the metabolic fuel required for sus-
tained high intensity exercise (Bosch et al. 1993). It
would seem logical to assume that the two are caus-
ally linked at exhaustion during prolonged exercise;
namely, that the near absence of muscle glycogen in
exhausted subjects explains why they are unable to
maintain, let alone increase, their exercise intensity at
exhaustion. The finding that the vast majority of the
modern and historical carbohydrate-loading studies
show that this technique improves endurance per-
Physiological models to study exercise
formance, presumably by increasing muscle glycogen
utilization and delaying the onset of terminal muscle
glycogen depletion, strongly supports the theory
(Hawley et al. 1997). However, the possibility that
part or all of these findings could also result from
a placebo effect, acting through the central nervous
system, needs to be considered (Burke et al. 1999). In
addition, it is unclear how the inability to produce
ATP at sufficiently high rates from one fuel source
can explain this form of fatigue, given that skeletal
muscle ATP concentrations remain high at exhaus-
tion (as they do in all other forms of exhaustion
(Fitts 1994)).
Nevertheless, there is a body of evidence that con-
flicts with the predictions of this hypothesis. For ex-
ample, the classic study of Coyle et al. (1986) showed
that athletes ingesting carbohydrate terminated exer-
cise after 4 h when their muscle glycogen concen-
trations and rates of carbohydrate oxidation were the
same as values measured 1 h earlier when the athletes
were not exhausted. Another study found that ath-
letes who adapted to a high fat diet were able to exer-
cise to significantly lower muscle glycogen concen-
trations at exhaustion than when they were carbo-
hydrate adapted (Lambert et al. 1994).
Conversely, Helge et al. (1996) showed that pre-
viously untrained subjects who trained on a high fat
diet for 7 weeks before switching to a high carbo-
hydrate diet for one week increased their pre-exercise
muscle glycogen concentrations by 44% with only a
small further increase in performance between the
seventh and eighth weeks of the trial. Furthermore,
performance was still substantially worse in fat-
adapted subjects than it was in subjects who trained
for 8 weeks on a high carbohydrate diet and whose
pre-exercise muscle glycogen concentrations were
32% lower than fat-adapted athletes exposed to a
high carbohydrate diet for one week.
In addition, exercise performance evaluated on 3
occasions in both dietary groups terminated before
there was marked muscle glycogen depletion. The
authors concluded: ‘‘Factors other than carbohydrate
availability are responsible for the differences in en-
durance time between the groups’’ (p. 303); and
‘‘These observations also indicate that fatigue during
prolonged moderately intense exercise does not al-
ways seem to be closely related to glycogen depletion,
as is usually stated’’ (Christensen & Hansen 1939,
Bergstrom et al. 1967). In his extensive review, Fitts
(1994) similarly concludes: ‘‘It seems unlikely that
muscle glycogen depletion, low blood glucose, and
the resultant decline in carbohydrate oxidation is an
exclusive fatigue factor during prolonged exercise’’.
He does, however, acknowledge that ‘‘a possibility
exists that muscle glycogen depletion is causative in
fatigue via a mechanism independent of its role in
energy production’’ (p. 83).
In addition, to my knowledge, no study has yet es-
tablished that training improves endurance perform-
ance exclusively by increasing body carbohydrate
stores and by delaying the onset of carbohydrate de-
pletion during prolonged exercise in humans, al-
though this finding has been reported in rats (Fitts et
al. 1975) whose metabolism is substantially different
from that of humans.
Similarly, it is currently difficult to explain perform-
ance in ultra-endurance events, especially the final 42
km running leg of the 226 km Ironman triathlon
events according to this model, which holds that exer-
cise of moderately high intensity is not possible once
there is marked muscle glycogen depletion. After cyc-
ling at 40 km ¡h
for 4.5 h, the lead cyclists would be
expected to have near total muscle glycogen depletion
according to data from laboratory studies (Bosch et al.
1993). Yet the best performers in that event are able to
run at close to 16 km ¡h
for a further 160 min. This
probably represents an exercise intensity of ±66% VO
max. The studies of Rauch et al. (1998) and O’Brien
et al. (1993) suggest that total carbohydrate oxidation
during very prolonged exercise of up to 6 h duration
exceeds the estimated carbohydrate stores in liver and
active muscle by up to 100%. Either these calculations
are incorrect, or other sources of carbohydrate, in ad-
dition to those in the active muscles and liver, must
contribute to fuel oxidation in events lasting more than
4–6 h. One possibility is that lactate oxidation of glyco-
gen stored in the inactive skeletal muscles contributes
a substantial additional amount to fuel use during very
prolonged exercise (Rauch et al. 1998). How increased
lactate oxidation contributes to performance is not
Similarly, provision of carbohydrate at high rates
intravenously (Coggan & Coyle 1987) cannot extend
exercise performance indefinitely. Whilst this could
support the argument that muscle glycogen is the im-
portant carbohydrate source limiting exercise per-
formance, the alternate possibility is that another
factor, unrelated to depletion of body carbohydrate
stores, perhaps a rising body temperature discussed
subsequently, or central neural fatigue induced by
other factors, may also limit endurance performance.
Perhaps there is a necessary rate of carbohydrate
oxidation that is required to sustain a specific exercise
intensity and that progressive whole body carbo-
hydrate depletion lowers that rate, inducing fatigue.
If this is correct, then fatigue resistance during pro-
longed exercise could be due to the capacity to sus-
tain a higher rate of carbohydrate oxidation, and
hence a higher respiratory quotient (RQ) during pro-
longed exercise. Again, this model suffers from the
persisting logical impasse that a failure to generate
ATP sufficiently rapidly must cause exercise to ter-
minate because of muscle ATP depletion and rigor, a
phenomenon which does not occur (Fitts 1994).
Fig. 8. Laboratory simulation suggests that an elite cyclist cyc-
ling 180 km in 4 h 30 min would cycle at an oxygen consump-
tion of 57 ml/kg/min. Based on the measured contribution of
fat and carbohydrate oxidation to this energy requirement (col-
umns on the left of the figure), this would require the oxidation
of about 700 g of carbohydrate and 175 g of fat. This compares
to the predicted maximum body stores of 520 g of carbohydrate
and 5000 g of fat in an elite triathlete. Hence this model predicts
that the elite triathlete must commence the running leg of the
triathlon with very low or absent whole body carbohydrate
An equally plausible alternate theory postulates
that superior endurance capacity may be determined
by the exact opposite; by a superior capacity to oxid-
ize fat and hence maintain a lower RQ during pro-
longed exercise. The latter possibility is supported by
at least some evidence. In the studies of Bosch et al.
(1993), those athletes unable to complete 3 h of exer-
cise at 70% VO
max after carbohydrate-loading had
significantly higher RQ during exercise and were
therefore characterized by an inability to sustain high
rates of fat oxidation during prolonged exercise. In-
deed, simulated metabolic balance studies for the 226
km Hawaiian Ironman triathlon suggest it to be very
likely that the capacity to oxidize fat at high rates will
influence running speed late in the race when calcu-
lations suggest that muscle glycogen stores are likely
to be depleted.
Figure 8 shows the expected energy metabolism
during 4.5 h cycling at an oxygen consumption of 57
ml/kg/min. This is equivalent to cycling at 40 km/h
and completing the 180 km cycle leg of the Ironman
triathlon in the time necessary to be amongst the race
leaders. The data for this simulation come from lab-
oratory data measured on elite South African cyclists
(I. Rodger. Unpublished data).
The simulation predicts that after 4.5 h of cycling
an elite male Ironman triathlete would be expected to
have oxidized about 700 g of carbohydrate and 175 g
of fat. This compares to predicted whole body carbo-
hydrate and fat stores of 520 g and 5000 g, respec-
tively. Hence this model predicts that, at the end of
the cycle leg, an elite athlete would have depleted his
body carbohydrate stores, yet must still run 42.2 km
at close to 16 km/h if he wishes to be successful.
Our other laboratory data suggest that after 4.5 h
of such exercise, the carbohydrate contribution to
whole body energy metabolism would comprise a
blood glucose oxidation rate of 1.2 g/min (21 kJ/min)
and a lactate oxidation rate of 0.6 g/min (10.5 kJ/
min). Together with the average maximum rate of fat
oxidation that we have measured after 6 h of labora-
tory cycling (0.76 g/min; 28 kJ/min), this provides a
total rate of energy production of 59.5 kJ/min. This
would provide energy at a rate sufficient to sustain a
running speed of approximately 12 km/h, sufficient
to complete the 42 km marathon leg of the Ironman
triathlon in3h30min(Fig. 9). To equal the best
marathon time yet run in that race, the athlete would
be required to oxidize fat at a rate of 1.15 g/min (Fig.
10). This rate is approximately 50% faster than we
have measured in cyclists in our laboratory.
Accordingly, if this metabolic model of fatigue in
the Ironman triathlon is correct, then the difference
between running the final marathon in 2 h 40 min
versus 3 h 30 min may simply be a 51% (0.4 g/min)
Fig. 9. The top panel of this figure shows the oxygen require-
ment (VO
), the exercise intensity (% VO
max) and the rates
of energy expenditure that would be sustained by a world-class
athlete completing the final 42.2 km marathon running leg of
the Ironman triathlon in times of either 2 h 40 min; 3 h 00 min;
or 3 h 30 min. Laboratory studies suggest that after a 3.8 km
swim and a 180 km cycle, the athlete’s body carbohydrate stores
would be depleted so that energy for the running leg would
come from oxidation of (mainly ingested) blood glucose, blood
lactate and from circulating free fatty acids derived from muscle
and adipose tissue triglyceride. To complete the marathon run-
ning leg of the Ironman triathlon in 3 h 30 min, the athlete
would have to sustain a VO
of 42 ml/kg/min (52% VO
equivalent to an energy expenditure of 59.5 kJ/min. Laboratory
simulations (Fig. 8) suggest that under these conditions of near
total carbohydrate depletion, peak glucose oxidation rates are
1.2 g/min and peak lactate (from glycogen) oxidation rates are
0.6 g/min. If these data for the maximum capacity to oxidize
glucose and lactate in the carbohydrate-depleted state are cor-
rect, then to sustain the rate of energy expenditure necessary to
run the marathon in 03:30:00, the carbohydrate-depleted tri-
athlete must oxidize fat at a rate of 0.76 g/min.
Physiological models to study exercise
Fig. 10. To complete the marathon running leg of the Ironman
triathlon in 2 h 40 min, currently the fastest running time yet
recorded in the Hawaiian Ironman triathlon, the athlete would
have to sustain a VO
of 53 ml/kg/min (66% VO
max), equiva-
lent to an energy expenditure of 74 kJ/min. If the maximum
capacity to oxidize glucose and lactate in the carbohydrate-de-
pleted state is unchanged from values given in Fig. 9, then to
sustain such a high rate of energy expenditure, the athlete must
oxidize fat at a rate of 1.15 g/min. This model predicts that the
superior ability of the elite Ironman triathlete may result from
a much greater (approximately 50%) capacity to oxidize fat
than has been measured in our laboratory experiments of very
prolonged laboratory exercise involving sub-elite athletes
(Rauch et al. 1998).
greater capacity to oxidize fat when body carbo-
hydrate and, especially, muscle glycogen stores are
depleted. Of course, this model does not negate the
requirement that such high rates of fat oxidation
can only be achieved if the central nervous system
continues to recruit an appropriately large number
of muscle fibres able to produce an appropriate
In summary, the human body has a limited ca-
pacity to store carbohydrates. In addition, high rates
of carbohydrate oxidation are necessary to sustain
high rates of energy expenditure (in the fed state).
Furthermore, studies of very prolonged exercise (6 h)
show that rates of carbohydrate oxidation remain
high in athletes who ingest appropriate amounts of
carbohydrate during exercise (Rauch et al. 1998). As
both muscle and liver glycogen depletion occur in the
fatigued state, it has popularly been assumed that
there is a direct causal relationship between, espe-
cially, muscle glycogen depletion and the develop-
ment of fatigue during prolonged exercise. Yet some
findings suggest that this relationship may not be
strictly causal under all circumstances. In addition is
the logical impasse which requires that any energy
depletion model predicts that exercise must terminate
when muscle ATP depletion occurs, leading to muscle
rigor. In the absence of such evidence, it would seem
that factors in addition to depletion of body carbo-
hydrate store may contribute to, or even cause, fa-
tigue during prolonged exercise.
At present, no study has conclusively established
that training-induced changes in the capacity to store
and metabolize carbohydrate during prolonged exer-
cise are causally related to training-induced changes
in performance in humans, although this relationship
is frequently assumed. The alternate possibility is
that the capacity to oxidize fat at high rates when
body carbohydrate stores are depleted, may delay fa-
tigue and determine performance during exercise of
moderately high intensity that lasts more than 4 h
and is typified by ultradistance running and triathlon
events (Fig. 8–10).
In this regard, the ‘‘crossover’’ concept of Brooks
and Mercier (1994) is of particular interest. These
authors argue that fuel choice during exercise
‘‘crosses over’’ from predominantly fat to exclusively
carbohydrate at exercise intensities above about 80%
max. They conclude that training produces dif-
ferent effects depending on the intensity of the exer-
cise being evaluated. At exercise intensities below
their ‘‘crossover’’ point, training increases fat oxi-
dation whereas at higher exercise intensities, training
increases the capacity to burn carbohydrates. Accord-
ing to their concept, exercise at higher intensities
(±70% VO
max) would be limited as an example of
energy supply limitations (an inability to sustain high
rates of carbohydrate oxidation), whereas exercise at
lower intensities would be limited by muscle glycogen
depletion. In fact the models are identical – according
to the energy (muscle glycogen) depletion model, fa-
tigue results from an inability to supply ATP suffi-
ciently rapidly from fat oxidation (failure of energy
supply from fat oxidation). The opposite pertains
during high intensity exercise. In fact, as argued here
and elsewhere (Fitts 1994), both the energy supply
and the energy depletion models predict that muscle
ATP depletion limits exercise. This does not occur;
hence, the models must be too simplistic to explain
what has been found.
The muscle recruitment (central fatigue)/muscle power
The two previous models are based on the assump-
tion that it is either the delivery of substrate either
in blood (oxygen) or via the glycolytic and oxidative
pathways (ATP) that limits exercise performance. The
steps of (il)logic that have influenced these assump-
tions have been described (Noakes 1997, 1998). It re-
mains difficult to prove whether or not either of these
models is correct. Yet both continue to dominate, per-
haps subconsciously, research and teaching in the ex-
ercise sciences, often to the exclusion of competing
An alternate view is that it is not the rate of supply
of substrate, either oxygen or fuel, to muscle that lim-
its its performance but rather the processes involved
in skeletal muscle recruitment, excitation and con-
A failure of central nervous system recruitment of
skeletal muscle forms the basis for the ‘‘central (ner-
vous system) fatigue’’ hypothesis (Davis & Bailey
1997). This model holds that the brain concentration
of serotonin (and perhaps other neurotransmitters,
including dopamine and acetylcholine) alters the den-
sity of the neural impulses reaching the exercising
muscles, thereby influencing the rate at which fatigue
develops, especially during exercise. Alternatively,
there may be inhibitory reflexes arising from the exer-
cising muscles and which feedback to the spinal cord,
reducing skeletal muscle recruitment at the level of
the a-motoneuron. The evidence for both mechan-
isms has been extensively reviewed (Davis & Bailey
In brief, a number of studies indicate that manipu-
lation of central nervous system neurotransmitter
concentrations, in particular increasing dopamine
and reducing serotonin concentrations, can enhance
exercise performance whereas the opposite impairs
performance. In addition, there is direct evidence for
reduced central neural drive to muscle after fatiguing
muscle contractions (Behm & St-Pierre 1997, Baker
et al. 1993, Newham et al. 1991). This evidence is
sufficiently persuasive to believe that central nervous
system fatigue contributes to fatigue during pro-
longed exercise lasting tens of minutes to hours.
The clear evidence that fatigue at high altitude is
caused by reduced central nervous system recruit-
ment of the exercising muscles has been described. It
is very likely that the fatigue that occurs during exer-
cise in the heat is also likely limited by a failure of
central recruitment as this form of fatigue cannot be
explained by any other model. Thus, there is now
substantive evidence that each athlete can store only
some limiting amount of heat before being forced to
reduce the exercise intensity or alternatively cease ex-
ercising altogether (Nielsen et al. 1990, 1993, 1997).
There is no evidence from these studies that exhaus-
tion under these conditions is associated with either
skeletal muscle ‘‘anaerobiosis’’ or energy depletion.
Rather, that it is either the total heat accumulated or
its rate of accumulation that limits exercise is con-
firmed by intervention studies which show that pre-
cooling improves performance (Nielsen et al. 1997,
Booth et al. 1997) whereas pre-heating has the op-
posite effect (Gongalez-Alonso et al. 1999).
In summary, there is sufficient evidence to suggest
that a reduced central nervous system recruitment of
the active muscles terminates maximum exercise at
high altitudes (and probably also at sea level). The
same mechanism likely terminates exercise in the heat
when the body temperature reaches some limiting
maximum and also when hypoglycaemia develops. In
all these examples, reduced central recruitment of
muscle would function to prevent organ damage
(Noakes 1997).
However, the contrasting finding that skeletal
muscle recruitment, measured as skeletal muscle elec-
tromyographic activity, rises during exercise at the
same workload (Takaishi et al. 1994), is usually inter-
preted as evidence for a progressive failure of muscle
fibre contractile function requiring additional fibre
recruitment if the required force is to be sustained.
Thus, this model holds that there is a progressive pe-
ripheral fatigue for which the central nervous system
makes an appropriate adjustment.
Yet these studies usually impose a constant work-
load on the subject for the duration of the activity.
But competitive sport does not usually involve such
constant workrates; workrates tend to vary in a
random, stochastic way (Palmer et al. 1994). We
found that during prolonged exercise, which in-
cludes bouts of self-chosen high intensity exercise
designed to simulate stochastic exercise, there is a
progressive reduction in power output during the
bouts of high intensity exercise (Burke et al. 1999).
This strongly suggests central fatigue in which, dur-
ing bouts of exercise requiring a near maximum ef-
fort, there is an inadequate increase in central neu-
ral drive to compensate for the expected reduction
in the power output of the fatiguing skeletal muscle
fibres (peripheral fatigue). That a relatively small
percentage of the available muscle mass is ever re-
cruited, even during maximal exercise (Sloniger et
al. 1997), remains a perplexing but relatively under-
recognized enigma. Proponents of any model of pe-
ripheral limitations for exercise performance need
to explain why the body does not recruit all its
available muscle mass to produce the necessary
force under varying exercise conditions as so-called
‘‘peripheral fatigue’’ develops.
In summary, one interpretation of the muscle re-
cruitment (central fatigue) model is that changes in
central neurotransmitters induce fatigue simply as a
natural consequence of prolonged exercise and
changes in the relative balance of the different (ergo-
genic and ergolytic) neurotransmitters in the brain.
No specific physiological value or importance is as-
signed to this phenomenon.
Alternatively, I have argued that a reduced central
activation of the exercising muscles may be necessary
to protect the human under specific conditions
(Noakes 1997, 1998). It is postulated that these con-
trol mechanisms are necessary (i) to prevent myocar-
dial ischaemia during exercise at high intensity; (ii) to
prevent the development of muscle ATP depletion
and muscle rigor during high intensity exercise; (iii)
to prevent myocardial ischaemia or cerebral hypoxia
during exercise at altitude; (iv) to prevent a fall in
blood pressure during exercise in patients with
chronic heart failure; (v) to prevent heatstroke during
Physiological models to study exercise
prolonged exercise in the heat, and (vi) to prevent glu-
copaenic brain damage during prolonged exercise
when hypoglycaemia results from liver glycogen de-
pletion. The likely mechanism of control is through
the regulation either of skeletal muscle recruitment or
of excitation/contraction coupling in the muscle.
Muscle power model
This model holds that muscle contractile capacity,
that is the ability of individual muscle cross-bridges
to generate force, is not the same in the muscles of
all humans, so that those with superior athletic abil-
ity have muscles with a superior capacity to generate
force (superior contractility) by the individual cross-
bridges of the different muscle fibres. This model is
well accepted by cardiac physiologists, the majority
of whom would argue that calcium delivery to the
myofibres and the activity of the enzyme involved in
ATP hydrolysis, myosin ATPase, rather than sub-
strate supply, determine the contractile state of the
myocardium in both health and disease (Opie 1998).
I could find only one recent statement using this
model to explain superior athletic performance,
specifically in swimming: ‘‘First, the strength of the
muscles used in swimming is a major determinant of
success in events from 50 m to 1500 m. Though this
may not seem surprising, it must be remembered that
strength per se does not dictate fast swimming. The
forces generated by the muscle must be effectively ap-
plied to the water if they are to propel the body. Thus,
strength specificity is the key to swimming success’’
(Costill et al. 1992).
There are rather few studies of the contractility of
skeletal muscle isolated from athletes. These studies
generally show that endurance training reduces skel-
etal muscle contractility (Fitts et al. 1989, Widrick et
al. 1996). This establishes that skeletal muscle con-
tractility is not an immutable characteristic of the dif-
ferent muscle fibre types (Fitss & Widrick 1996). By
extension, one might speculate that the contractility
of the specific muscle fibre types might differ between
athletes of different abilities in different sporting
disciplines, compatible with this muscle power model.
In summary, these two models of exercise perform-
ance predict that changes in exercise performance
may result from increased skeletal muscle recruitment
resulting from enhanced central neural drive, or from
increased muscle contractile function resulting from
biochemical adaptations in muscle that increase
either force production or the rate of sarcomere
shortening, or both.
However, the increase in performance resulting
from these adaptations would occur only to the ex-
tent that the cardiovascular limits for exercise per-
formance were not exceeded, according to the Car-
diovascular/Neural Model.
The biomechanical model
There is growing interest in the role of muscles as
elastic energy return systems which function both as
springs and torque producers during exercise (Pennisi
1997, Roberts et al. 1997). Central to this model is
the prediction that the greater the muscle’s capacity
to act as a spring, the less torque it must produce and
hence the more efficient it is. The more efficient, more
elastic muscle will enhance exercise performance,
especially in weight-bearing activities, by slowing (i)
the rate of accumulation of those metabolites that
may cause fatigue during exercise, and (ii) the rate of
rise of body temperature, thereby delaying the
achievement of the core temperature that prevents the
continuation of exercise.
This new information underscores another import-
ant logical weakness of the cardiovascular/anaerobic
model for explaining enhanced endurance perform-
ance. For that model predicts that superior perform-
ance during prolonged exercise results from an in-
creased oxygen delivery to muscle and an increased
rate of energy and hence heat production. Thus, ac-
cording to that model, the price of running faster is
that more heat must be produced. But a higher rate
of heat production would induce fatigue prematurely
due to excessive heat accumulation, according to the
findings of Nielsen and colleagues (Nielsen et al.
1993, 1997). A more logical biological adaptation
would be to reduce the rate of oxygen consumption
and hence the rate of heat production by increasing
the athlete’s efficiency (economy) of movement.
Indeed, if the rate of heat accumulation limits exer-
cise performance under specific conditions, then fac-
tors that slow the rate at which heat accumulates
when running fast should enhance performance. Two
such factors are small size (Dennis & Noakes 1999)
and superior running economy. A smaller size reduces
the amount of heat produced when running at any
speed. When environmental conditions limit the ca-
pacity for heat loss, smaller runners will be favoured
(Dennis & Noakes 1999).
Further evidence supporting this argument that
heat accumulation is a factor limiting endurance per-
formance, is the finding that race times in both the
marathon (Noakes 1992) and the longer distance
track races including the 3000 m steeplechase and the
10 000 m (McCann & Adams 1997) deteriorate as the
environmental heat load increases. Thus, there is an
inverse relationship between the environmental heat
load, measured as the Wet Bulb Globe Temperature
Index, and the reduction in race performance.
Therefore, according to this model, the more econ-
omical the athlete, the faster he or she will be able to
run before reaching a limiting body temperature. A
number of studies indicate that the best endurance
athletes are also frequently the most economical
(Noakes 1992; Fig. 6). Indeed, most training studies
show that improvements in running economy are per-
haps the most likely response to training, especially
in those who are already well-trained (Svedenhag &
Sjodin 1985). This adaptation allows the athlete to
run faster at the same oxygen consumption; thus, he
or she completes a given distance more rapidly for
the same average rate of heat accumulation but a re-
duced overall heat expenditure. This would be advan-
tageous under conditions in which the heat load on
the athlete increases (during the day).
Figure 11 shows that this adaptation may indeed
exist. In a cross-sectional study of recreational (not
elite) ultramarathon runners, it was found that those
who trained more were more economical and hence
could run faster at the same oxygen consumption or %
max. During competition, the better trained
athletes ran at the same or a slightly lower % VO
but completed the races in a shorter time (Scrimgeour
et al. 1986). Hence being more economical, not having
a higher VO
max, appears to be a more logical tech-
nique to enhance endurance performance.
In contrast, a high aerobic capacity, often a marker
of poor running economy (Noakes 1988, 1992),
would likely cause more rapid rates of heat accumu-
lation and hence the more rapid onset of fatigue dur-
ing prolonged exercise. This finding alone could ex-
plain why the best marathon runners usually have
max values in the range of 63–74 ml/kg/min.
Less economical runners with higher VO
max values
(Noakes 1988) have not necessarily been more suc-
cessful (Noakes et al. 1990, Noakes 1992).
Thus, this model predicts that success in endurance
Fig. 11. Ultramarathon athletes who trained more than 60
km ¡week
(the lower running curve) outperformed less well
trained athletes with similar VO
max values because their su-
perior running economy allowed them to race faster but at a
lower % VO
max, thereby finishing races of 10–90 km in
shorter times. Hence changes in running economy with training
allow athletes to run faster and at a lower exercise intensity, the
opposite of what is usually assumed.
events is not likely to result from training that makes
the athlete ever more powerful with a larger muscle
mass and greater VO
max. A more likely adaptation
would be to reduce the athlete’s size and increase his
or her running efficiency. That runners believe they
run better when lighter, is well known.
Another African analogy for this prediction is pro-
vided by the physiological strategy that the cheetah
has evolved to survive as a successful predator. The
cheetah, whose chase is terminated by an elevated
rectal temperature after running at up to 100 km ¡h
for less than a minute (Taylor & Rowntree 1973), suc-
ceeds because of the animal’s small size and probably
a high degree of running economy (due to elasticity
provided by the flexible spine). Thus, laboratory ex-
periments showed that when the cheetah’s rectal tem-
perature reached 40.5–41æC, ‘‘the cheetahs refused to
run ... They would simply turn over with their feet in
the air and slide on the tread(mill) surface’’ (Taylor &
Rowntree 1973).
The small size of the cheetah and its likely high
running economy slows its rate of heat accumulation
just sufficiently for it to outrun the smaller gazelles
(25 kg) on which it preys and whose escape is also
restrained by a rising body temperature (Taylor & Ly-
man 1972). Thus the chase between the gazelle and
the cheetah is probably decided by which individual
animal accumulates heat more slowly during the
chase. In contrast, the heavier, more muscular lion
has evolved a different co-operative, hunting strategy,
targeting larger but slower mammals.
Perhaps the point is that smallness and greater run-
ning economy would seem to be a technique used to
increase endurance capacity in one animal, the chee-
tah. Logic suggests that this technique may also be
applicable to elite human athletes.
A second component of the biomechanical models
stems from the accumulating evidence that repeated
high velocity, short duration eccentric muscle con-
tractions, as occur during running, induce a specific
form of fatigue that develops during running races
and is measurable for at least 7 days after a marathon
race (Komi & Nicol 1998; Nicol et al. 1991).
Characteristics of this fatigue are a failure of the
contractile capacity of the exercised muscles with a
reduced tolerance to muscle stretch and a delayed
transfer from muscle stretch to muscle shortening in
the stretch/shortening cycle. As a result, the durations
of both the braking and push-off phases in the run-
ning stride are increased, leading to mechanical
changes in the stride with landing occurring on a
more extended leg but with greater subsequent knee
As these abnormalities persist for many days after
the race (Fig. 12), they cannot be explained by acute
changes in oxygen or substrate delivery to the
muscles, or by the elevated body temperature during
Physiological models to study exercise
exercise, as required by the first 3 models. Rather,
Komi and Nicol (1998) conclude that: ‘‘Stretch short-
ening fatigue results usually in a reversible muscle
damage process and has considerable influence on
muscle mechanics, joint and muscle stiffness as well
as on reflex intervention’’. Thus any evaluation of fa-
tigue resistance, especially in weight-bearing activities
like running, needs to consider this specific form of
stretch/shortening cycle fatigue.
To return to the African analogy, empirical obser-
vation of the running stride and the anatomical struc-
ture of the lower limb of Kenyan runners suggests, at
least to this author, that an evaluation of the elastic
elements of the legs of elite Kenyan runners and their
resistance to stretch/shortening cycle fatigue would
likely be very rewarding.
For example, it appears that African athletes gen-
erally train harder than do Caucasian runners (Tans-
er 1997, Coetzer et al. 1993). Especially the training
volumes and intensities of the Kenyan runners (Tans-
er 1997) are unmatched by other athletes. But to
achieve such training volumes, there must be superior
resistance to the stretch/shortening cycle damage pro-
posed by Komi & Nicol (1998), both in training and
in marathon racing.
Hence, another possibility is that the more elastic
muscles of elite distance runners are better able to
resist eccentrically induced damage in training. This
may allow more intensive daily training and hence
superior adaptations to training. That same superior-
ity would also enhance performance during competi-
tive racing by delaying the onset of the stretch/short-
ening cycle fatigue that is an inevitable consequence
of repeated eccentric muscle contractions.
In summary, the biomechanical model predicts that
superior performance, especially in a weight-bearing
activity like running, may be influenced by the ca-
pacity of the muscles to act as elastic energy return
systems. Changes in the efficiency and durability of
this process would (i) enhance movement economy
and reduce the rate of heat production during exer-
cise, thereby enhancing exercise capacity by slowing
the rate at which the body temperature rises when
environmental conditions are severe; (ii) enhance the
quality of training by allowing more rapid recovery
from stretch/shortening cycle fatigue so that more
frequent bouts of intensive training can be under-
taken and (iii) enhance fatigue resistance during com-
petition by increasing resistance to that form of
muscle damage that develops during repeated cycles
of stretch/shortening contractions.
The psychological/motivational model
This model holds that the ability to sustain exercise
performance results from a conscious effort and is
often included as a component of the central fatigue
Fig. 12. According to the Biomechanical Model of Exercise
Physiology and Athletic Performance, repetitive eccentric
muscle contractions, as occurs in the quadriceps and calf
muscles during running, produces altered muscle function with
a loss of elastic energy production, requiring an increased work
during the push-off phase of the running stride. The studies of
Komi & Nicol (1998) and Nicol et al. (1991) show that there is a
persistent reduction in peak torque and endurance performance
time of the quadriceps muscle of subjects after they completed
a 42.2 km marathon race. Integrated electromyographic activity
(IEMG) in both the vastus medialis (VM) and the vastus lat-
eralis (VL) was also reduced after the marathon. These studies
show that there is a long-term effect of marathon running on
skeletal muscle performance that persists after the muscle’s en-
ergy stores are replaced, indicating that this abnormality is not
explained by the Energy Depletion Model. It seems probable
that the marathon ‘‘wall’’ (Noakes 1992) is probably explained
by these alterations in skeletal muscle function.
hypothesis (Davis & Bailey1997). But it conflicts with
one proposal of the muscle recruitment model, which
holds that exercise performance is regulated at a sub-
conscious level and which exists, in part, to prevent
conscious override that might damage the human.
It would seem that exercise performance must in-
clude at least some component that can be influenced
by conscious effort. The dichotomy of physiology
and psychology has generally prevented adequate lab-
oratory evaluation of this model. Any studies show-
ing an ergogenic effect of any placebo intervention
on exercise performance would prove that this model
contributes, in part, to athletic performance. Any de-
tailed discussion of this model is beyond the scope of
this author’s expertise.
This paper has reviewed some of the models currently
promoted to describe how exercise performance is
limited by fatigue and enhanced by training. The ar-
gument advanced here is that until the factors deter-
mining both fatigue and athletic performance are es-
tablished more definitely, it remains difficult to define
which training adaptations are the most important
for enhancing exercise performance, or how training
should be structured to maximize those adaptations.
This remains a serious weakness for the practical ap-
plication of much research in the exercise sciences.
However, within the constraints provided by these
models, it would seem that the following training ad-
aptations would contribute to enhanced exercise per-
formance with training.
(a) Cardiovascular/neural recruitment model: Rel-
evant training adaptations would be those that
result in an increased VO
max and skeletal
muscle blood flow during both maximal and pro-
longed submaximal exercise.
According to the model I have presented here,
a plateau in coronary flow would appear to be
the factor that limits the cardiac output, and
hence the VO
max. Thus, an essential compo-
nent of training would be to increase the maximal
coronary flow.
However, even if the maximal coronary blood
flow limits the maximal cardiac output, the actual
peak workrate or peak VO
max that is achieved
will depend on the contractile state of the myo-
cardium and the efficiency with which the heart
is able to convert that maximum coronary flow
into a peak cardiac output. Similarly, the actual
maximum work of which the muscles are capable
at the VO
max, will equally depend on the econ-
omy and contractility of the skeletal muscles.
Thus, optimum training would not only maxim-
ize coronary flow but also the efficiency and con-
tractility of both the heart and skeletal muscles.
Incidentally, according to this model, inter-
ventions such as EPO therapy, blood re-infusion,
or the administration of oxygen, all of which im-
prove performance, would act in part by increas-
ing oxygenation of the myocardium at maximal
exercise, thereby allowing a greater cardiac output
and hence a higher VO
(b) Energy supply model: The important training ad-
aptation would be an increased capacity to store
and utilize metabolic substrates during exercise
with a greater capacity to produce ATP and pre-
vent a reduction in muscle phosphagen concen-
trations under all exercise conditions. Adap-
tations in different metabolic pathways would be
necessary for optimizing performance in activi-
ties of different durations and intensities.
(c) Energy depletion model: A reduced rate of carbo-
hydrate utilization during prolonged exercise
would enhance performance by delaying the on-
set of whole body carbohydrate depletion. This
model predicts that an increased capacity to burn
fat during prolonged exercise would enhance en-
durance performance. That this model can also
be considered as an ‘‘energy supply’’ model was
described above.
(d) Muscle recruitment and muscle power models:
These models predict that increased skeletal
muscle contractile function, a peripheral effect,
or increased neural recruitment, a central nervous
system effect, would be advantageous for per-
formance during exercise.
(e) The biomechanical model: A key predictor of the
biomechanical model is that increased movement
economy would improve performance by reduc-
ing the rate of heat accumulation during exercise.
This model also explains why a reduced body
mass would improve performance during pro-
longed exercise as it slows the rate of heat ac-
The importance of elastic return energy, espe-
cially in weight-bearing sports, and the identifi-
cation of stretch/shortening cycle fatigue suggests
that training may improve elasticity and delay
stretch/shortening cycle fatigue, perhaps by al-
tering the elastic component of skeletal muscle,
tendons and ligaments.
The importance of these conceptual models is that they
indicate that different physiological systems may de-
termine performance under different exercise con-
ditions. Hence, exercise physiologists need to consider
all these models when they design studies to determine
which are the most important physiological, biochem-
ical, neural and other factors determining the changes
in exercise performance that result from training.
More importantly, this review shows that many
findings are incompatible with the predictions of one
or more of these models. Rather than simply continu-
ing to accept these inconsistencies uncritically, the
modern generation of exercise physiologists should
challenge old dogmas and so approach more closely
the unattainable truth (Noakes 1997, 1998).
Key words: fatigue; VO
max; heart; skeletal muscle;
glycogen; fat metabolism; muscle recruitment; con-
The author’s research is funded by the dedicated financial sup-
port of the University of Cape Town, the Medical Research
Council of South Africa, the Liberty Life Insurance Company,
the Foundation for Research and Development, the Founding
Donors of the Sports Science Institute of South Africa, Bromor
Foods, and a variety of nutritional and pharmaceutical com-
Physiological models to study exercise
Acevedo EO, Goldfarb AH. Increased Clair Gibson A, Mujika I, Noakes TD. respiratory capacity, endurance, and
training intensity effects on plasma Carbohydrate loading failed to improve glycogen utilization. Am J Physiol 1975:
lactate, ventilatory threshold, and 100-km cycling performance in a 228: 1029–33.
endurance. Med Sci Sports Exerc 1989: placebo-controlled trial. J Appl Physiol Fitts RH, Costill DL, Gardetto PR.
21: 563–8. in press. Effect of swim exercise training on
Bainbridge FA. The physiology of Christensen EH, Forbes WH. Skand human muscle fiber function. J Appl
muscular exercise. 3rd edn. Bock AV, Arch Physiol 1937: 76: 75. Quoted in: Physiol 1989: 66: 465–75.
Dill DB, eds. London: Longmans, Dill DB. Life, heat and altitude. Fitts RH. Cellular mechanisms of muscle
Green and Co. 1931: 1–272. Cambridge, MA: Harvard University fatigue. Physiol Rev 1994: 74: 49–94.
Baker AJ, Kostov KG, Miller RG, Press, 1938: 168. Fitts RH, Widrick JJ. Muscle mechanics:
Weiner MW. Slow force recovery after Christensen EH, Hansen O. adaptations with exercise-training.
long-duration exercise: metabolic and Arbeitsfahigkeit und Erna
¨hrung. Skand Exerc Sports Sci Rev 1996: 24: 427–73.
activation factors in muscle fatigue. J Arch Physiol 1939: 81: 160–71. Gongalez-Alonso J, Teller C, Andersen
Appl Physiol 1993: 74: 2294–300. Christensen EH, Hansen O. SL, Jensen FB, Hyldig T, Nielsen B.
Bale J, Sang J. Kenyan running. London, Hypoglykamie, arbeitsfahigkeit und Influence of body temperature on the
U.K.: Frank Cass, 1996: 1–209. Erna
¨hrung. Skand Arch Physiol 1939: development of fatigue during
Bannister RG. Muscular effort. Br Med 81: 172–9. prolonged exercise in the heat. Eur J
Bull 1956: 12: 222–5. Coetzer P, Noakes TD, Sanders B, et al. Appl Physiol 1999: 86: 1032–9.
Bassett DR, Howley ET. Maximal Superior fatigue resistance of elite black Green HJ, Sutton JR, Young P,
oxygen uptake: ‘‘classical’’ versus South African distance runners. J Appl Cymerman A, Houston CS. Operation
‘‘contemporary’’ viewpoints. Med Sci Physiol 1993: 75: 1822–7. Everest II: muscle energetics during
Sports Exerc 1997: 29: 591–603. Coggan AR, Coyle EF. Reversal of maximal exhaustive exercise. J Appl
Behm DG, St-Pierre DMM. Effects of fatigue during prolonged exercise by Physiol 1989: 66: 142–50.
fatigue duration and muscle type on carbohydrate infusion or ingestion. J Green HJ, Sutton JR, Cymerman A,
voluntary and evoked contractile Appl Physiol 1987: 63: 2388–95. Young PM, Houston CS. Operation
properties. J Appl Physiol 1997: 82: Costill DL, Thomason H, Roberts E. Everest II: adaptations in human
1654–61. Fractional utilization of the aerobic skeletal muscle. J Appl Physiol 1989: 66:
Bergstrom JB, Hermansen L, Hultman capacity during distance running. Med 2454–61.
E, Saltin B. Diet, muscle glycogen and Sci Sports Exerc 1973: 5: 248–52. Haskell WL, Sims C, Myll J, Bortz WM,
physical performance. Acta Physiol Costill DL, Bennett A, Branam G, Eddy Goar FGS, Alderman EL. Coronary
Scand 1967: 71: 140–50. D. Glucose ingestion at rest and during arter y size and dilating capacity in
Bogdanis GC, Nevill ME, Boobis LH, prolonged exercise. J Appl Physiol ultradistance runners. Circulation 1993:
Kalomy HK, Nevill AM. Recovery of 1973: 34: 764–9. 87: 1076–82.
power output and muscle metabolites Costill DL, Maglischo EW, Richardson Hawley JA, Hopkins WG. Aerobic
following 30 s of maximal sprint cycling AB. Swimming. London: Blackwell glycolytic and aerobic lipolytic power
in man. J Physiol (London) 1995: 482: Science Ltd, 1992: 1–214. systems. Sports Med 1995: 19: 240–50.
67–480. Coyle EF, Coggan AR, Hemmert MK, Hawley JA, Myburgh KH, Noakes TD,
Boning D. Altitude and hypoxia Ivy JL. Muscle glycogen utilization Dennis SC. Training techniques to
training–ashortreview.IntJSports during prolonged strenuous exercise improve fatigue resistance and enhance
Med 1997: 18: 565–70. when fed carbohydrate. J Appl Physiol endurance performance. J Sports Sci
Bosch AN, Dennis SC, Noakes TD. 1986: 61: 165–72. 1997: 15: 325–33.
Influence of carbohydrate loading on Daniels JT, Yarbough RA, Foster C. Hawley JA, Schabort EJ, Noakes TD,
fuel substrate turnover and oxidation Changes in VO
max and running Dennis SC. Carbohydrate-loading and
during prolonged exercise. J Appl perfor mance with training. Eur J Appl exercise performance. An update.
Physiol 1993: 74: 1921–7. Physiol 1978: 39: 249–54. Sports Med 1997: 24: 73–81.
Bosch AN, Weltan SM, Dennis SC, Davies CTM, Thompson MW. Aerobic Helge JW, Richter EA, Kiens B.
Noakes TD. Fuel substrate kinetics of performance of female marathon and Interaction of training and diet on
carbohydrate loading differs from that male ultramarathon athletes. Eur J metabolism and endurance during
of carbohydrate ingestion during Appl Physiol 1979: 61: 611–7. exercise in man. J Physiol 1996: 492:
prolonged exercise. Metabolism 1996: Davis JM, Bailey SP. Possible 293–306.
45: 415–23. mechanisms of central nervous system Hill AV, Long CNH, Lupton H.
Booth J, Marino F, Ward JJ. Improved fatigue during exercise. Med Sci Sports Muscular exercise, lactic acid, and the
running performance in hot humid Exerc 1997: 29: 45–57. supply and utilization of oxygen: parts
conditions following whole body Dennis SC, Noakes TD. Advantages of I–III. Proc Royal Soc Bri 1924: 97: 438–
precooling. Med Sci Sports Exerc 1997: smaller body mass for distance running 75.
29: 943–9. performances in warm, humid Hill AV. Living machinery. London: G
Braunwald E, Kloner RA. The stunned conditions. Eur J Appl Physiol 1999: 79: Dell and Sons Ltd, 1927: 1–241.
myocardium: prolonged, postischemic 280–4. Hill AV, Long CHN, Lupton H.
ventricular dysfunction. Circulation Dill DB. Life, heat and altitude. Muscular exercise, lactic acid and the
1982: 66: 1146–9. Cambridge, MA: Harvard University supply and utilisation of oxygen: parts
Brooks GA, Mercier J. Balance of Press, 1938: 1–211. VII–VIII. Proc Royal Soc Bri 1924: 97:
carbohydrate and lipid utilization Edwards HT. Lactic acid in rest and work 155–76.
during exercise: the ‘‘crossover’ at high altitude. Am J Physiol 1936: Hochachka PW. The lactate paradox:
concept. J Appl Physiol 1994: 76: 116: 367–75. analysis of underlying mechanisms.
2253–61. Fitts RH, Booth FW, Winder WW, Ann Sports Med 1989: 4: 184–8.
Burke LM, Hawley JA, Schabort EJ, St Holloszy JO. Skeletal muscle Hochachka PW. Muscles as molecular
and metabolic machines. Boca Raton, dr y environment. J Physiol 1993: 460: turkeys: The economy of minimizing
Florida, USA: CRC Press, 1994. 467–85. work. Science 1997: 275: 1113–5.
Holloszy JO, Coyle EF. Adaptations of Nielsen B, Strange S, Christensen NJ, Rowell LB. Human cardiovascular
skeletal muscle to endurance exercise Warberg J, Saltin B. Acute and adaptive control. New York: Oxford University
and their metabolic consequences. J responses in humans to exercise in a Press, 1993: 1–500.
Appl Physiol 1984: 56: 831–8. warm, humid environment. Pflu
¨gers Saltin B, Gollnick PD. Skeletal muscle
Hoppeler H, Kleinert E, Schlegel C, et al. Arch 1997: 434: 49–56. adaptability: Significance for
Morphological adaptations of human Noakes TD. Implications of exercise metabolism and performance. In:
skeletal muscle to chronic hypoxia. Int testing for prediction of athletic Peachey LD, Adrian RH, Geiger SR
J Sports Med 1990: 1: S3–9. performance: a contemporary (Eds). Handbook of Physiology,
Houston ME, Bentzen H, Larsen H. perspective. Med Sci Sports Exerc 1988: Section 10, Skeletal Muscle. Bethesda,
Interrelationships between skeletal 20: 319–30. Maryland: American Physiological
muscle adaptations and performance Noakes TD, Myburgh KH, Schall R. Society, 1983.
as studied by detraining and retraining. Peak treadmill running velocity during Saltin B, Kim CK, Terrados N, Larsen
Acta Physiol Scand 1979: 105: 163–70. the VO
max test predicts running H, Svedenhag J, Rolf C. Morphology,
Kayser B, Hoppeler H, Claassen H, performance. J Sports Sci 1990: 8: 35– enzyme activities and buffer capacity in
Cerretelli P. Muscle structure and 45. leg muscles of Kenyan and
performance capacity of Himalayan Noakes TD. Lore of Running. 2rd edn. Scandinavian runners. Scand J Med
Sherpas. J Appl Physiol 1991: 70: 1938– Cape Town: Oxford University Press, Sci Sports 1995a: 5: 222–30.
42. 1992: 1–535. Saltin B, Larsen H, Terrados N, et al.
Kayser B, Narici M, Binzoni T, Grassi B, Noakes TD. Challenging beliefs: ex Aerobic exercise capacity at sea level
Cerretelli P. Fatigue and exhaustion in Africa semper aliquid novi. Med Sci and at altitude in Kenyan boys, junior
chronic hypobaric hypoxia: influence Sports Exerc 1997: 29: 571–90. and senior runners compared with
of exercising and muscle mass. J Appl Noakes TD. Maximal oxygen uptake: Scandinavian runners. Scand J Med
Physiol 1994: 76: 634–40. ‘‘classical’’ versus ‘‘contemporary’’ Sci Sports 1995b: 5: 209–21.
Komi PV, Nicol C. Stretch-shortening viewpoints. A rebuttal. Med Sci Sports Saltin B. Adaptive responses to training
cycle fatigue. Biomechanics and Biology Exerc 1998: 30: 1381–98. at medium altitude; with a note on
of Movement 1998. O’Brien MJ, Viguie CA, Mazzeo RS, Kenyan runners and a proposal for a
Lambert EV, Speechley DP, Dennis SC, Brooks GA. Carbohydrate dependence multi-centre study. Revised and
Noakes TD. Enhanced endurance in during marathon running. Med Sci extended version of Saltin B. Exercise
trained cyclists during moderate Sports Exerc 1993: 25: 1009–17. and the Environment: focus on altitude.
intensity exercise following 2 weeks Oelz O, Howald H, Di Prampero PE, et Res Quart Exerc Sport 1996: 67: 1–10.
adaptation to a high fat diet. Eur J al. Physiological profile of world-class Scheuer J, Tipton CM. Cardiovascular
Appl Physiol 1994: 69: 287–93. high altitude climbers. J Appl Physiol adaptations to physical training. Annu
Mannion AF, Jakeman PM, Willan PLT. 1986: 60: 1734–42. Rev Physiol 1997: 39: 221–51.
Skeletal muscle buffer value, fibre type Opie LH. The Heart: Physiology from Scrimgeour AG, Noakes TD, Adams B,
distribution and high intensity exercise cell to circulation. 3rd edn. New York: Myburgh K. The influence of weekly
performance in man. Exp Physiol 1995: Lipincott-Raven, 1998: 1–637. training distance on fractional
80: 89–101. Palmer GS, Hawley JA, Dennis SC, utilization of maximum aerobic
McCann DJ, Adams WC. Wet bulb globe Noakes TD. Heart rate response during capacity in marathon and
temperature index and performance in a 4-d cycle stage race. Med Sci Sports ultramarathon runners. Eur J Appl
competitive distance runners. Med Sci Exerc 1994: 26: 1278–83. Physiol 1986: 55: 202–9.
Sports Exerc 1997: 29: 955–61. Pennisi E. A new view of how leg muscles Sloniger MA, Cureton KJ, Prior BM,
Moore RL, Thacker EM, Kelley GA, et operate on the run. Science 1997: 275: Evans EM. Lower extremity muscle
al. Effect of training/detraining on 1067. activation during horizontal and uphill
submaximal exercise responses in Pugh LGCE. Cardiac output in muscular running. J Appl Physiol 1997: 83: 2073–
humans. J Appl Physiol 1987: 63: 1719– exercise at 5,800 m (19,000 ft). J Appl 9.
24. Physiol 1964: 19: 441–7. Snell P. Training of the distance runners.
Newham DJ, McCarthy T, Turner J. Ramsbottom R, Williams C, Fleming N, Second Annual Congress of the
Voluntary activation of human Nute MLG. Training induced European College of Sport Science.
quadriceps during and after isokinetic physiological and metabolic changes Copenhagen, August 20–23, 1997.
exercise. J Appl Physiol 1991: 71: 2122– associated with improvements in Abstract book, p. 57.
6. running performance. Br J Sports Med Spriet LL, Soderlund K, Bergstrom M,
Nicol C, Komi PV, Marconnet P. Fatigue 1989: 23: 171–6. Hultman E. Anaerobic energy release in
effects of marathon running on Raskoff WJ, Goldman S, Cohn K. The skeletal muscle during electrical
neuromuscular performance. I. ‘‘Athletic Heart’’. Prevalence and stimulation in men. J Appl Physiol
Changes in muscle force and stiffness physiological significance of left 1987: 62: 611–5.
characteristics. Scand J Med Sci Sports ventricular enlargement in distance Stray-Gundersen J, Musch TI, Haidet
1991: 1: 10–7. runners. JAMA 1976: 236: 158–62. GC, Swain DP, Ordway GA, Mitchell
Nielsen B, Savard G, Richter EA, Rauch LHG, Hawley JA, Noakes TD, JH. The effect of pericardiectomy on
Hargreaves M, Saltin B. Muscle blood Dennis SC. Fuel metabolism during maximal oxygen consumption and
flow and muscle metabolism during ultra-endurance exercise. Pflu
¨gers Arch maximal cardiac output in untrained
exercise and heat stress. J Appl Physiol 1998: 436: 211–9. dogs. Circulation Res 1986: 58: 523–30.
1990: 69: 1040–6. Reeves JT, Groves BM, Sutton JR, et al. Suarez J, Alexander JK, Houston CS.
Nielsen B, Hales JRS, Strange S, Operation Everest II: preservation of Enhanced left ventricular systolic
Christensen NJ, Warberg J, Saltin BN. cardiac function at extreme altitude. J performance at high altitude during
Human circulatory and Appl Physiol 1987: 63: 531–9. Operation Everest II. Am J Cardiol
thermoregulatory adaptations with heat Roberts JR, Marsh RL, Weyand PG, 1987: 60: 137–42.
acclimatization and exercise in a hot, Taylor CR. Muscular force in running Sutton JR, Reeves JT, Wagner PD, et al.
Physiological models to study exercise
Operation Everest II: oxygen transport Tsintzas O, Williams C, Boobish L, adaptations to interval training in
during exercise at extreme simulated Greenhaff P. Carbohydrate ingestion endurance-trained cyclists. Eur J Appl
altitude. J Appl Physiol 1988: B64: and single muscle fiber glycogen Physiol 1997: 75: 298–304.
1309–21. metabolism during prolonged running Weston AR, Myburgh KH, Lindsay FH,
Svedenhag J, Sjodin B. Physiological in men. J Appl Physiol 1996: 81: 801– Dennis SC, Noakes TD, Hawley JA.
characteristics of elite male runners in 9. Skeletal muscle buffering capacity and
and out of season. Can J Appl Sport Vogel JA, Hartley LH, Cruz JC, Hogan endurance performance after high-
Sci 1985: 10: 127–33. RP. Cardiac output during exercise in intensity interval training by well-
Takaishi T, Yasuda Y, Moritani T. sea-level residents at sea level and high trained cyclists. Eur J Appl Physiol
Neuromuscular fatigue during altitude. J Appl Physiol 1974: 36: 169– 1997: 75: 7–13.
prolonged pedalling exercise at 72. Widrick JJ, Trappe SW, Blaser CA,
different pedalling rates. Eur J Appl Vogel JA, Hartley LH, Cruz JC. Cardiac Costill DL, Fitts RH. Isometric force
Physiol 1994: 69: 154–8. output during exercise in altitude and maximal shortening velocity of
Tanser T. Train hard, win easy. Mountain natives at sea level and high altitude. J single muscle fibers from elite master
Views, California: Tafnews Press Inc., Appl Physiol 1974: 36: 173–6. runners. Am J Physiol 1996: 271(2 Pt
1997: 1–198. Vollestad NK, Sejersted OM, Bahr R, 1): C666–75.
Taylor CR, Lyman CP. Heat storage in Woods JJ, Bigland-Ritchie B. Motor Widrick JJ, Trappe SW, Costill DL, Fitts
running antelopes: independence of drive and metabolic responses during RH. Force-velocity and force-power
brain and body temperatures. Am J repeated submaximal contractions in properties of single muscle fibers from
Physiol 1972: 22: 114–7. humans. J Appl Physiol 1988: 64: elite master runners and sedentary men.
Taylor CR, Rowntree VJ. Temperature 1421–7. Am J Physiol 1996: 271(2 Pt 1): C676–
regulation and heat balance in running Westgarth-Taylor C, Hawley JA, Rickard 83.
cheetahs: a strategy for sprinters? Am S, Myburgh KH, Noakes TD, Dennis
J Physiol 1973: 224: 848–51. SC. Metabolic and performance
... Krogh and Lindhard clearly favoured an exercise model in which the high-fat diet causes skeletal muscular acidosis, which is then the direct cause of their high levels of fatigue both during and after exercise. Their model was probably based on the dominant model of fatigue then popular; one developed by English Nobel Laureate A.V. Hill in the early 1920s to explain how fatigue develops during exercise of high intensity [5]-what has since been termed Hill's Cardiovascular-Anaerobic model [6]. Hill's model continues to enjoy a committed following nearly a century later [7,8]. ...
... Essentially his only options, other than heart rate, also measured with difficulty, were blood lactate concentrations and rates of oxygen consumption. Within these constraints, Hill's mind conceived a model of exercise performance based solely on these two variables [6,[9][10][11]. ...
... An analogy is the control exerted by the brain and central nervous system in terminating exercise, particularly at extreme altitudes, to prevent the development of a catastrophic brain hypoxia [6,9,11]-the so-called 'lactate paradox of high altitude' [32]. ...
Full-text available
The introduction of the needle muscle biopsy technique in the 1960s allowed muscle tissue to be sampled from exercising humans for the first time. The finding that muscle glycogen content reached low levels at exhaustion suggested that the metabolic cause of fatigue during prolonged exercise had been discovered. A special pre-exercise diet that maximized pre-exercise muscle glycogen storage also increased time to fatigue during prolonged exercise. The logical conclusion was that the athlete’s pre-exercise muscle glycogen content is the single most important acutely modifiable determinant of endurance capacity. Muscle biochemists proposed that skeletal muscle has an obligatory dependence on high rates of muscle glycogen/carbohydrate oxidation, especially during high intensity or prolonged exercise. Without this obligatory carbohydrate oxidation from muscle glycogen, optimum muscle metabolism cannot be sustained; fatigue develops and exercise performance is impaired. As plausible as this explanation may appear, it has never been proven. Here, I propose an alternate explanation. All the original studies overlooked one crucial finding, specifically that not only were muscle glycogen concentrations low at exhaustion in all trials, but hypoglycemia was also always present. Here, I provide the historical and modern evidence showing that the blood glucose concentration—reflecting the liver glycogen rather than the muscle glycogen content—is the homeostatically-regulated (protected) variable that drives the metabolic response to prolonged exercise. If this is so, nutritional interventions that enhance exercise performance, especially during prolonged exercise, will be those that assist the body in its efforts to maintain the blood glucose concentration within the normal range.
... However, skeletal muscle fibers are never fully recruited during exercise; muscle adenosine triphosphate (ATP) never falls below 60% of resting levels, and glycogen concentration decreases but is not depleted during exercise [39]. Even more, in many circumstances, fatigue occurs before high concentrations of metabolites, such as lactate, H + , extracellular K + , without disturbances in muscle Ca 2+ kinetics and without high core temperatures or significant hypohydration [40]. ...
... The control of the metabolic demand regulated by the brain is called teleanticipation. This central governor evidenced by Ulmer [49] has been analyzed in several research studies [40,41,45,50,51]. These studies were the starting point for the appearance of the so-called anticipatory feedback model of exercise regulation [4], which constitutes a psychophysiological approach of the fatigue phenomenon. ...
... These studies were the starting point for the appearance of the so-called anticipatory feedback model of exercise regulation [4], which constitutes a psychophysiological approach of the fatigue phenomenon. This model assumes that the exercise is self-regulated from the beginning by the athletes based on previous experiences, knowledge of the expected distance, duration of the current exercise, and afferent physiological feedback regarding some variables (i.e., muscle glycogen levels, skin and body temperature) [40,41]. Processing this information allows the brain to predict and regulate the most appropriate exercise intensity allowing an optimal performance without serious homeostatic disruptions [51]. ...
Over the years, there has been a growing interest in the study of issues related to the psychophysiological processes underlying sports performance. A relatively recent perspective is supported by the concept that the brain acts as a central regulator of performance during exercise. This phenomenon is called pacing and is based on the premise that prior knowledge about the activity plays a fundamental role for individuals to self-regulate their efforts throughout the exercise. However, knowledge regarding this topic remains scarce, and further clarification is needed. This chapter reports new perspectives in relation to the existing evidence regarding the role of the brain as a central regulator of performance, questioning the complex interdependencies and interrelations between fatigue and physical exercise in the light of a psychophysiological perspective. A broader understanding of the cogni-tive basis of the psychophysiological phenomenon during the exercise is needed, bringing together concepts such as pacing behavior, decision-making, self-regulation of effort, prior knowledge of the duration of the task, and perception of effort.
... It means that the encephalon stops the maximal effort after it has integrated several somatic information, to avoid a 'homeostasis catastrophic damage'. 3 The question is: Which theory is better? Before answering this question, we need to settle a potential criterion for knowledge progress. ...
... Noakes has raised the below list of evidence to refute Hill's theory: 1,3,43 There is a failure in Hill's theory because only fifty percent of all subjects shows the oxygen uptake plateau during the maximal effort. Noakes has criticised Taylor et al.'s ergometric protocol, 44 since they showed V_O 2 plateau in almost all of their subjects. ...
Full-text available
An important epistemological problem has been faced by Exercise Physiologists. On one hand, one theory explains the fatigue through a ceiling effect of oxygen uptake. On the other hand, the new theory proposes that an encephalon mechanism would stop the effort before a catastrophic homeostasis failure. Many physiologists have looked for evidence to support their favourite theory even though the induction logic problem does not allow to prove whether truth is discovered; however, it is possible to prove that it does not occur. When some researchers fail to test their hypotheses, they use relativism to bring up their theories again. Noakes and his colleagues have based their theory on relativism, because it is impossible to refute by empirical observation. It also doesn’t explain all phenomena that the oldest Hill’s theory is able to explain. Noakes’s theory isn’t more accurate in its previsions. Noakes did not check whether the oxygen uptake plateau occurs in suitable tests to measure on the mouth what happens in the muscles. Finally, it doesn’t propose new tests for the encephalon role during maximal effort, as that is expected in scientific work. For all of these reasons, it is possible to conclude there are no advantages in switching to the “Central Governor” theory. [206 words]
... Es conocido que los cambios fisiológicos y metabólicos parecen ser los causantes de esta fatiga [8,9]. En respuesta, existen varios métodos para controlar y medir la carga interna del entrenamiento, entre ellos está la variabilidad de la frecuencia cardiaca (VFC) que se ha descrito como indicador de estrés, fatiga, recuperación y adaptación al entrenamiento a través de la actividad del sistema nervioso autónomo y su interacción con el corazón [10][11][12]. ...
Full-text available
El objetivo del presente estudio fue evaluar el efecto de un entrenamiento intenso en atletas de resistencia sobre el comportamiento de las colinesterasas (ChE) tras condiciones de fatiga y su relación con otros marcadores de carga interna. Participaron 18 atletas de sexo masculino especialistas en pruebas de resistencia. Se evaluó las ChE y dos índices de variabilidad de la frecuencia cardiaca en tres momentos diferentes, previo al protocolo (BASAL), 15 minutos posterior al protocolo (FINAL) y 24 horas después del entrenamiento (24H). Un ANOVA de una vía con post-hoc de Tukey HSD se utilizó para comparar las medias. Se encontraron cambios significativos en las variables analizadas (p < .001) con tamaños de efecto muy grandes (d > 0.9) en los diferentes momentos y correlaciones moderadas entre variables (p < .001). El comportamiento de las ChE muestra un cambio significativo (p < .001) posterior al ejercicio y una relación con otros indicadores de carga interna. Nuestros resultados indican que las ChE tienen relación con la fatiga en el caso de los deportistas estudiados, pudiendo ser una medida para determinar la carga de entrenamiento.
... A certain degree of fatigue, resulting in functional overreaching, is required to mediate adaptations to training, which drive performance enhancement (Noakes, 2000). However, excessive fatigue through insufficient recovery may increase susceptibility to non-functional over-reaching, injury, and illness of the players (Nimmo and Ekblom, 2007). ...
... A certain degree of fatigue, resulting in functional overreaching, is required to mediate adaptations to training, which drive performance enhancement (Noakes, 2000). However, excessive fatigue through insufficient recovery may increase susceptibility to non-functional over-reaching, injury, and illness of the players (Nimmo and Ekblom, 2007). ...
... It is widely known that muscle fatigue is one of the major factors for limiting or completely terminating a physical activity during exercise [8]. There have been many experiments to investigate muscle fatigue during exercise by employing different types of exercise equipment and different exercise protocols. ...
... Competitive sports are known to be demanding and stressful for both human and animal athletes. In particular, endurance is one of the most challenging among equestrian disciplines; endurance horses are susceptible to metabolic imbalance due to dehydration, acid balance and electrolyte abnormalities, substrate depletion and heat accumulation, which can result in life-threatening conditions [6]. Indeed, equine endurance competitions, which have recently gained popularity worldwide, are governed by the rules established by National and International Equestrian Federations (FEI), which implement strict regulations to safeguard and ensure the welfare of animals [7]. ...
Full-text available
Physical exercise has been associated with the modulation of micro RNAs (miRNAs), actively released in body fluids and recognized as accurate biomarkers. The aim of this study was to measure serum miRNA profiles in 18 horses taking part in endurance competitions, which represents a good model to test metabolic responses to moderate intensity prolonged efforts. Serum levels of miRNAs of eight horses that were eliminated due to metabolic unbalance (Non Performer-NP) were compared to those of 10 horses that finished an endurance competition in excellent metabolic condition (Performer-P). Circulating miRNA (ci-miRNA) profiles in serum were analyzed through sequencing, and differential gene expression analysis was assessed comparing NP versus P groups. Target and pathway analysis revealed the up regulation of a set of miRNAs (of mir-211 mir-451, mir-106b, mir-15b, mir-101-1, mir-18a, mir-20a) involved in the modulation of myogenesis, cardiac and skeletal muscle remodeling, angiogenesis, ventricular contractility, and in the regulation of gene expression. Our preliminary data open new scenarios in the definition of metabolic adaptations to the establishment of efficient training programs and the validation of athletes’ elimination from competitions.
Full-text available
This study examined the effects of sustained high-intensity interval training (HIT) on the athletic performances and fuel utilisation of eight male endurance-trained cyclists. Before HIT, each subject undertook three baseline peak power output tests and two simulated 40-km time-trial cycling performance (TT40) tests, of which the variabilities were 1.5 (1.3)% and 1.0 (0.5)%, respectively [mean (SD)]. Over 6 weeks, the cyclists then replaced 15 (2)% of their 300 (66) km · week−1 endurance training with 12 HIT sessions, each consisting of six to nine 5-min rides at 80% of , separated by a 1-min recovery. HIT increased from 404 (40) to 424 (53) W (P < 0.01) and improved TT40 speeds from 42.0 (3.6) to 43.0 (4.2) km · h−1 (P < 0.05). Faster TT40 performances were due to increases in both the absolute work rates from 291 (43) to 327 (51) W (P < 0.05) and the relative work rates from 72.6 (5.3)% of pre-HIT to 78.1 (2.8)% of post-HIT (P < 0.05). HIT decreased carbohydrate (CHO) oxidation, plasma lactate concentration and ventilation when the cyclists rode at the same absolute work rates of 60, 70 and 80% of pre-HIT (P < 0.05), but not when they exercised at the same relative (% post-HIT ) work rates. Thus, the ability of the cyclists to sustain higher percentages of in TT40 performances after HIT was not due to lower rates of CHO oxidation. Higher relative work rates in the TT40 rides following HIT increased the estimated rates of CHO oxidation from ≈ 4.3 to ≈ 5.1 g · min−1.
Cardiac output during muscular exercise was estimated by the acetylene technique on four members of the Himalayan Scientific and Mountaineering expedition 1960–1961 at sea level and 5,800 m (19,000 ft). The output for a given work intensity at 5,800 m (19,000 ft) was comparable with the output at the same work intensity at sea level, but the maximum output was reduced, the mean value being 16 liters/min, compared with 23 liters/min at sea level. Heart rates during light and moderate exercise were higher than the rates observed at the same work intensity at sea level. The maximum heart rate during exercise was limited to 130–150 beats/min compared with 180–196 beats/min at sea level. The stroke volume at altitude was lower than at sea level at each work rate. On breathing oxygen at sea-level pressure, heart rate for a given work intensity was reduced; but the maximum heart rate increased. Indirect evidence suggested that maximum cardiac output increased but probably not to the sea-level values because of the increased hemoglobin and lower heart rate. altitude acclimatization; cardiac function, work and altitude; hypoxia and cardiac output Submitted on July 29, 1963
Studies of exercise performance during hypobaric hypoxia among subjects acclimatized to high altitudes have raised an intriguing metabolic paradox. On the one hand, the maximum aerobic metabolic rate declines as a function of high altitude, reaching values only slightly higher than the resting metabolic rate (RMR) at altitudes equivalent to that of the peak of Everest. On the other hand, when subjects perform incremental exercise tests to fatigue, the amounts of lactate formed also decline as a function of altitude. This effect is so dramatic that it is predicted that no lactate whatsoever can be produced at aerobic fatigue at altitudes equivalent to that of Everest. For practical purposes, the glytolytic pathway in muscle is blocked under these high-altitude conditions. The paradox is that anaerobic glycolysis works perfectly well when subjects acclimatized to high altitudes start from rest, and in fact it is well known that the anaerobic power output of muscle is unaffected by hypobaric hypoxia. Therefore, the problem arises as to why it seems to be impossible to activate the pathway in muscle brought to fatigue during aerobic work at high altitude, where the demands for glycolytic adenosine triphosphate (ATP) synthesis are normally exaggerated. To clarify this problem, it is necessary to focus closely on the nature of the coupling mechanisms between ATP demand and ATP supply during sustained work. I do this for two kinds of systems, both those that remain closely coupled as work rate increases and those that necessarily assume new steady states as work rates change.
The muscle glycogen content of the quadriceps femoris muscle was determined in 9 healthy subjects with the aid of the needle biopsy technique. The glycogen content could be varied in the individual subjects by instituting different diets after exhaustion of the glycogen store by hard exercise. Thus, the glycogen content after a fat ± protein (P) and a carbohydrate-rich (C) diet varied maximally from 0.6 g/100g muscle to 4.7 g. In all subjects, the glycogen content after the C diet was higher than the normal range for muscle glycogen, determined after the mixed (M) diet. After each diet period, the subjects worked on a bicycle ergometer at a work load corresponding to 75 per cent of their maximal O2 uptake, to complete exhaustion. The average work time was 59, 126 and 189 min after diets P, M and C, and a good correlation was noted between work time and the initial muscle glycogen content. The total carbohydrate utilization during the work periods (54–798 g) was well correlated to the decrease in glycogen content. It is therefore concluded that the glycogen content of the working muscle is a determinant for the capacity to perform long-term heavy exercise. Moreover, it has been shown that the glycogen content and, consequently, the long-term work capacity can be appreciably varied by instituting different diets after glycogen depletion.
IntroductionNeuromuscular Fatigue During Isolated Eccentric ActionsFatigue Definition That May Apply to Repeated SSC LoadingExperimental SettingsFunctional InfluencesPotential Mechanisms Related to Performance Deterioration and Recovery in Connection with Exhaustive SSC ExerciseConclusions References
 Acute and repeated exposure for 8–13 consecutive days to exercise in humid heat was studied. Twelve fit subjects exercised at 150 W [45% of maximum O2 uptake (V.O2,max)] in ambient conditions of 35°C and 87% relative humidity which resulted in exhaustion after 45 min. Average core temperature reached 39.9 ± 0.1°C, mean skin temperature (T– sk) was 37.9 ± 0.1°C and heart rate (HR) 152 ± 6 beats min–1 at this stage. No effect of the increasing core temperature was seen on cardiac output and leg blood flow (LBF) during acute heat stress. LBF was 5.2 ± 0.3 l min–1 at 10 min and 5.3 ± 0.4 l min–1 at exhaustion (n = 6). After acclimation the subjects reached exhaustion after 52 min with a core temperature of 39.9 ± 0.1°C, T– sk 37.7 ± 0.2°C, HR 146 ± 4 beats min–1. Acclimation induced physiological adaptations, as shown by an increased resting plasma volume (3918 ± 168 to 4256 ± 270 ml), the lower exercise heart rate at exhaustion, a 26% increase in sweating rate, lower sweat sodium concentration and a 6% reduction in exercise V.O2. Neither in acute exposure nor after acclimation did the rise of core temperature to near 40°C affect metabolism and substrate utilization. The physiological adaptations were similar to those induced by dry heat acclimation. However, in humid heat the effect of acclimation on performance was small due to physical limitations for evaporative heat loss.