See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/12473881

Physiological models to understand exercise

fatigue and the adaptations that predict or
enhance athletic performance

Article in Scandinavian Journal of Medicine and Science in Sports · July 2000

DOI: 10.1034/j.1600-0838.2000.010003123.x · Source: PubMed

CITATIONS READS

310 788

1 author:

Timothy Noakes
University of Cape Town
639 PUBLICATIONS 18,794 CITATIONS

SEE PROFILE

Available from: Timothy Noakes

Retrieved on: 21 September 2016
Scand J Med Sci Sports 2000: 10: 123–145 COPYRIGHT C MUNKSGAARD 2000 ¡ ISSN 0905-7188
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- there is full activation of the total skeletal muscle mass,
tigue develops during exercise of moderate to high inten- and (iv) cardiovascular and other measures believed to re-
sity, when the capacity of the cardiorespiratory system to late to skeletal muscle anaerobiosis, including the maxi-
provide oxygen to the exercising muscles falls behind their mum oxygen consumption (VO2 max) and the ‘‘anaerobic
demand inducing ‘‘anaerobic’’ metabolism. But this car- threshold’’, are indifferent predictors of exercise capacity
diovascular/anaerobic model is unsatisfactory because (i) in athletes with similar abilities. This review considers four
a more rigorous analysis indicates that the first organ to additional models that need to be considered when factors
be affected by anaerobiosis during maximal exercise limiting either short duration, maximal or prolonged sub-
would likely be the heart, not the skeletal muscles. This maximal exercise are evaluated. These additional models
probability was fully appreciated by the pioneering exer- are: (i) the energy supply/energy depletion model; (ii) the
cise physiologists, A. V. Hill, A. Bock and D. B. Dill, but muscle power/muscle recruitment model; (iii) the bio-
has been systematically ignored by modern exercise physi- mechanical model and (iv) the psychological model. By
ologists; (ii) no study has yet definitely established the reviewing features of these models, this review provides a
presence of either anaerobiosis, hypoxia or ischaemia in broad overview of the physiological, metabolic and bio-
skeletal muscle during maximal exercise; (iii) the model is mechanical factors that may limit exercise performance
unable to explain why exercise terminates in a variety of under different exercise conditions. A more complete
conditions including prolonged exercise, exercise in the understanding of fatigue during exercise, and the relevance
heat and at altitude, and in those with chronic diseases of of the adaptations that develop with training, requires that
the heart and lungs, without any evidence for skeletal the potential relevance of each model to fatigue under dif-
muscle anaerobiosis, hypoxia or ischaemia, and before ferent conditions of exercise must be considered.

The nature of the physiological and biochemical ad- in performance that occur with different types of
aptations that occur in response to physical training physical training (Acevedo & Goldfarb 1989, Daniels
has been extensively studied in humans and other et al. 1978, Hawley et al. 1997, Houston et al. 1979,
mammals. This information is readily available and Moore et al. 1997, Ramsbottom et al. 1989,
is likely to be well known to most exercise scientists Westgarth-Taylor et al. 1997, Weston et al. 1997) and
(Saltin & Gollnick 1983, Holloszy & Coyle 1984). which presumably result from changes that delay the
Similarly there is an extensive literature on the cellu- onset or development of fatigue. There are at least
lar mechanisms believed to cause the fatigue that de- three probable reasons for this.
velops during exercise (Fitts 1994). First, many exercise physiologists may consider
In contrast, fewer studies have evaluated the extent this to be the work of the coach, not of the scientist.
to which these adaptations explain the improvements Or, accustomed to the tightly controlled conditions
of laboratory research, some scientists may be reluc-
tant to undertake field-based studies of performance
* Based on Keynote addresses presented at the Fourth IOC in which all the different variables influencing human
World Congress on Sports Sciences, Monte-Carlo, Monaco,
October 22–26, 1997 and the IV Scandinavian Congress on performance are not easily controlled. Human per-
Medicine and Science in Sports, Lahti, Finland, 5–8 Novem- formance is influenced by many variables, not least
ber 1998. those involving the psyche. Many scientists may feel,

123
Noakes
perhaps justifiably, that these variables cannot be suf- might explain superior athletic performance and en-
ficiently well controlled in field studies for there to be hanced resistance to the development of fatigue?
meaningful findings.
Second, there is a dearth of tools to measure accu-
Current physiological models to understand the
rately human performance in the laboratory. If sports
physiology of training for enhanced endurance
performance cannot be measured frequently with a
performance
high degree of precision in the laboratory, then train-
ing-induced changes in exercise performance are not Table 1 lists five different models that are commonly
quantifiable. As a result, most studies use physiologi- used to study and explain the likely physiological and
cal surrogates to predict changes in exercise perform- other training-induced changes that may improve,
ance. The most widely used performance surrogate is especially, endurance performance, probably by de-
the maximum oxygen consumption (VO2 max). But, laying or preventing the onset of fatigue. Each model
the very use of this specific measure has helped to has its own proponents, usually those with a special
entrench a particular and, perhaps, unquestioning expertise in the specific areas embraced by the model.
dogma of the factors that likely determine human ex- Thus, the cardiovascular/anaerobic model is pro-
ercise performance (Noakes 1988, 1997, 1998). moted usually by cardiovascular and respiratory
As a result, most of the training studies reported physiologists; the energy supply/energy depletion
in the literature have measured the physiological and model is favoured by the exercise biochemists; the
biochemical responses of the human to training and muscle power/muscle recruitment model is advocated
have paid less attention (i) to the extent to which hu- by muscle physiologists, and some biomechanists and
man exercise performance is altered by different neuro-physiologists; the biomechanical model by bio-
training programmes and (ii) to the specific physio- mechanists, and the psychological/motivational
logical adaptations which explain training-induced model by sports psychologists.
changes in athletic performance. Yet it is highly improbable that the factors explain-
Indeed, an important weakness in our current ing human exercise performance under all conditions
thinking in exercise physiology is that we lack certain are restricted to one physiological system or to one
knowledge of the precise factors that determine fa- scientific discipline. Thus, human performance is un-
tigue and hence limit performance in different types likely to be adequately defined by any of these uni-
of exercise under a range of environmental con- tary models that are often presented as if they are
ditions. In part, this is because some scientists remain mutually exclusive. The complexity of the physiologi-
unaware that their research is based on the (subcon- cal and other factors determining human perform-
scious) acceptance usually of one specific model of ance is emphasized when the limitations of each of
human exercise physiology (Noakes 1997, 1998). But these models are exposed.
it would be very surprising if one single physiological
model adequately explains human exercise perform-
The cardiovascular/anaerobic model
ance under all conditions.
Accordingly, the aim of this review is not to de- Maximal exercise
scribe how the body adapts to physical training. This This model holds that endurance performance is de-
information is freely available, largely descriptive and termined by the capacity of the athlete’s large heart
not particularly contentious, so that its review is un- to pump unusually large volumes of blood and oxy-
likely to challenge how we think about our science. gen to the muscles. This allows the muscles to achieve
Rather I will use this opportunity to pose two ques- higher work rates before they outstrip the available
tions: What physiological models have exercise scien- oxygen supply, developing skeletal muscle anaerobi-
tists developed (and subconsciously accepted) for the osis (Fig. 1) (Noakes 1988, 1997, 1998, Bassett &
study of the physiological and biochemical determi- Howley 1997). This model remains the most popular
nants of fatigue during exercise? And which specific for explaining why fatigue develops during exercise;
physiological, metabolic or biomechanical attributes how the body adapts to training; how these adap-
tations enhance performance and, as a consequence,
how effective exercise training programmes should be
structured.
Table 1. Current physiological models to understand the physiology of This model predicts that training increases ‘‘cardio-
training for enhanced endurance performance
vascular fitness’’ especially by increasing the body’s
(a) The cardiovascular/anaerobic model maximum capacity to consume oxygen, measured as
(b) The energy supply/energy depletion model the maximum oxygen consumption (VO2 max). This
(c) The muscle recruitment (central fatigue)/muscle power model effect results from an increased maximum capacity of
(d) The biomechanical model
the heart to pump blood (the cardiac output) and
(e) The psychological/motivational model
an enhanced capacity of the muscles to consume that

124
 Physiological models to study exercise
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
Fig. 1. The Cardiovascular/Anaerobic Model of Exercise Physi- mythical interpretation of the work of Hill and his
ology and Athletic Performance holds that the heart has a limit- colleagues, failed also to record what Hill considered
ing maximum cardiac output that is reached at the onset of a to be the physiological cause (and equally the conse-
‘‘plateau phenomenon’’, ascribed incorrectly to the work of Hill quence) of the fatigue that develops during maximal
and colleagues (Hill et al. 1924, Hill 1927). As a result, blood
and hence oxygen flow to the exercising muscles falls behind exercise.
demand, inducing anaerobic metabolism. Metabolites of anaer- For the interpretation of Hill and his colleagues
obic metabolism, in particular hydrogen ions, then inhibit was unequivocal: ‘‘Certain it is that the capacity of
muscle contraction, inducing fatigue. the body for muscular exercise depends largely, if not
mainly, on the capacity and output of the heart. It
oxygen, the latter by increasing skeletal muscle capil- would obviously be very dangerous for the organ to
larization and mitochondrial mass. It is argued that be able, as the skeletal muscle is able, to exhaust itself
these adaptations delay the onset of skeletal muscle very completely and rapidly, to take exercise far in
anaerobiosis during vigorous exercise, thereby reduc- excess of its capacity for recovery ... When the oxygen
ing blood lactate concentrations in muscle and blood supply becomes inadequate, it is probable that the
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-
Fig. 2. The weakness of the Cardiovascular/Anaerobic Model
mation that these changes, with the exception of a of Exercise Physiology and Athletic Performance is that the at-
greater coronary flow which is inferred, not proven, tainment of a maximum cardiac output has more serious conse-
do indeed result from training, as fully documented quences for the heart than it does for the skeletal muscles as
in the literature. The key question is whether these the first organ to be affected by a maximum cardiac output
would be the heart itself. Continuing to exercise with a fixed
changes are causally linked; that is, do these changes
(maximum) cardiac output would cause myocardial ischaemia.
cause the change in exercise performance or do they The heart, unable to increase coronary flow (dependant on an
occur pari-passu with other adaptation(s) that are the increase in cardiac output), would be unable to balance the in-
real cause of changes in exercise performance. For creased myocardial oxygen demand caused by the very increase
there are important deficiencies in this model which in workrate, theoretically necessary to define the mythical ‘‘pla-
teau phenomenon’’ of Hill and his colleagues. Many early re-
are fully argued (Noakes 1988, 1997) and counter- searchers believed that the heart did indeed develop ischaemia
argued (Noakes 1998, Bassett & Howley 1997) else- during maximal exercise, leading to a fall in cardiac output (Hill
where and will not be repeated here. et al. 1924, Bainbridge 1931, Hill 1927, Dill 1938).

125
Noakes
heart rapidly begins to diminish its output, so avoid- coronary flow. This limiting coronary blood flow in-
ing exhaustion ...’’. duces myocardial ‘‘fatigue’’, causing the plateau in
The point identified by Hill and his colleagues, and cardiac output and hence in the VO2 max leading,
since ignored by all subsequent generations of exer- finally, to skeletal muscle anaerobiosis. Thus, by this
cise physiologists, is that the heart is also a muscle, logic, the coronary blood flow must be the first
dependent for its function on an adequate blood and physiological function to show a ‘‘plateau phenom-
oxygen supply. But, unlike skeletal muscle, the heart enon’’ during progressive exercise to exhaustion (Fig.
is dependent for its blood supply on its own pumping 3). All subsequent physiological ‘‘plateaus’’ must re-
capacity. Hence any intervention that reduces the sult from this limiting coronary flow (Noakes 1998).
pumping capacity of the heart, or demands the heart Whereas the most influential modern exercise
somehow to sustain an increased work output by the physiologists have enthusiastically embraced this
exercising muscles without any increase in cardiac mythical basis for a ‘‘plateau phenomenon’’ for the
output and coronary flow (as theoretically occurs past 75 years, none seems to have grasped this logical
when the ‘‘plateau phenomenon’’ develops), imperils prediction of the ‘‘plateau phenomenon’’, which is
the heart’s own blood supply. Any reduction in coro- that the ‘‘plateau phenomenon’’ requires the heart to
nary blood flow will consequently reduce the heart’s fatigue first before skeletal muscle fatigue can de-
pumping capacity, thereby inducing a vicious cycle of velop. But this was clearly a concept with which the
progressive and irreversible myocardial ischaemia pioneering exercise physiologists were entirely
(Fig. 2). It would seem logical that human design comfortable. Thus, in addition to the conclusion of
should include controls to protect the heart from ever Hill and his colleagues, already quoted, both Bock
entering this vicious circle. and Dill (Bainbridge 1931) also believed that myocar-
Hence if (skeletal) muscle function fails when its dial hypoxia causes a fall in the cardiac output at the
oxygen demand exceeds supply then, for logical con- point of fatigue during high intensity exercise:
sistency, the inability of the pumping capacity of the ‘‘The blood supply to the heart, in many men, may
heart to ‘‘raise the cardiac output’’ at the VO2 max be the weak link in the chain of circulatory adjust-
(Rowell 1993), must also result from an inadequate ments during muscular exercise, and as the intensity
(myocardial) oxygen supply caused by a plateau in 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’’.
Fig. 3. According to the logic of the Cardiovascular/Anaerobic In summary, the early physiologists who believed
Model of Exercise Physiology and Athletic Performance, coro- that skeletal muscle anaerobiosis limits maximal exer-
nary flow must be the first physiological variable to ‘‘plateau’’
cise clearly understood that any plateau in cardiac
during progressive exercise to exhaustion. The peak in coronary
flow would then induce a plateau in cardiac output as a result output, necessary for there to be a limiting skeletal
of a progressive myocardial ischaemia. Continuing to exercise muscle blood flow, must result from a plateau in coro-
with a fixed cardiac output and coronary flow would rapidly nary blood flow which would expose the heart to a
cause an ischaemia-induced fall in cardiac output and in coro- progressive myocardial ischaemia that would worsen
nary flow, and hence in whole body oxygen consumption. This
logic was accepted by the early proponents of this model (Hill as exercise was prolonged.
et al. 1924, Bainbridge 1931, Hill 1927, Dill 1938) but has since Perhaps the reluctance of modern physiologists to
been overlooked by exercise physiologists for the past 75 years. acknowledge these concepts stems from the current

126
 Physiological models to study exercise
appreciation that progressive myocardial ischaemia mum exercise, its appearance will likely be more eas-
does not occur during maximal exercise in healthy ily identifiable during exercise at altitude under con-
athletes (Raskoff et al. 1976), even though there is ditions of hypobaric hypoxia. Furthermore, such
good evidence that it is a limiting cardiac output that experiments should identify in which organ – heart
probably determines the VO2 max (Rowell 1993). or skeletal muscle – anaerobiosis first becomes ap-
Thus, one postulate might be that even if cardiac out- parent; the heart, according to the ideas of the pion-
put limits maximal exercise as seems likely (Noakes eering British and North American exercise physiol-
1997), termination of exercise must occur before the ogists, or the skeletal muscles, according to the influ-
heart actually reaches that maximum and hence well ential group of modern exercise physiologists
before skeletal muscle anaerobiosis can develop (Noakes 1998).
(Noakes 1998). Hence for 75 years, exercise physiol- The original studies of exercise at altitude were
ogists may have focused on the incorrect organ as the undertaken by a research group co-ordinated by Dill
site of any potential anaerobiosis that may develop and his colleagues from the Harvard fatigue labora-
during maximal exercise (Hill et al. 1924, Bainbridge tory. This research established two crucial findings.
1931, Hill 1927, Hill et al. 1924). First, that peak blood lactate concentrations during
How might a maximal cardiac output be reached maximum exercise fell with increasing altitude (Ed-
without the development of myocardial ischaemia? wards 1936), a phenomenon since labelled the ‘‘lac-
The argument that the rate of cardiac filling, due tate paradox’’ (Hochachka 1989). Second, that maxi-
either to a limiting venous return or the effects of a mum heart rate and cardiac output likewise fell dur-
restrictive pericardium (Stray-Gundersen et al. 1986) ing exercise at increasing altitude (Christensen 1938,
may limit the maximal cardiac output, whilst super- Dill 1938).
ficially attractive, is still unable satisfactorily to ex- Edwards (1936) interpreted the ‘‘lactate paradox’’
plain which physiological events terminates exercise. at altitude accordingly: ‘‘The inability to accumulate
Such an argument fails for the reason that the con- large amounts of lactate at high altitudes suggests a
tinuation of exercise beyond that (however limited) protective mechanism preventing an already low ar-
maximal cardiac output must still cause a progressive terial saturation from becoming markedly lower ... It
myocardial ischaemia to develop (Fig. 2). Hence, even may be that the protective mechanism lies in an inad-
if the cardiac output is limited by factors unrelated equate oxygen supply to essential muscles, e.g. the
to the development of myocardial ischaemia (for ex- diaphragm or the muscles’’.
ample, a limiting venous return), the continuation of The existence of the ‘‘lactate paradox’’ was con-
exercise beyond that point of limitation must induce firmed during the epic laboratory experiment of exer-
myocardial ischaemia and the development of chest cise and acclimatization at simulated high altitude,
pain (angina pectoris) that would terminate exercise. Operation Everest II (Green et al. 1989). That study
Perhaps it is more logical to speculate that maxi- found that muscle lactate concentrations achieved
mal exercise terminates as part of a regulated process during maximal exercise at the highest equivalent alti-
before the absolute maximum cardiac output and tude achieved during that experiment (8848 m –
coronary blood flow are achieved. Interestingly Hill equivalent to the summit of Mount Everest) were no
and his colleagues seem to have been the first to sug- higher than when at rest at sea level.
gest a solution to this dilemma as early as 1924: Hence, in as much as high muscle lactate concen-
‘‘From the point of view of a well co-ordinated mech- trations would have to be present if the exercising
anism, ... it would clearly be useless for the heart to muscles were contracting ‘‘anaerobically’’, this study
make an excessive effort if by doing so it merely pro- proves that exercise at extreme altitude terminates
duced a far lower degree of saturation of the arterial when the exercising muscles are contracting in fully
blood; and we suggest that, in the body (either in the aerobic conditions.
heart muscle itself or in the nervous system), there is Similarly Operation Everest II (Sutton et al. 1988)
some mechanism which causes a slowing of the circu- confirmed these original and subsequent studies
lation as soon as a serious degree of unsaturation oc- (Pugh 1964, Vogel et al. 1974) showing that heart rate
curs, and vice versa. This mechanism would tend to and cardiac output are substantially reduced during
act as a governor maintaining a high degree of satu- exercise at extreme altitude. The key observation is
ration of the blood’’ (Hill et al. 1924, p. 161–162). that the peak cardiac output falls with increasing alti-
Clearly no such governor has yet been discovered, tude. This response is equally paradoxical for those
perhaps because no physiologists have yet searched who believe that the delivery of an adequate oxygen
for it. But there is clear physiological evidence for the supply to the exercising muscles is the cardinal prior-
existence of such a governor. The evidence comes ity during exercise (Noakes 1998). For logic demands
from studies of skeletal and cardiac muscle function that if the principal responsibility of the cardiovascu-
at altitude. For if oxygen deficiency really does de- lar system during exercise is the achievement of an
velop in either heart or skeletal muscle during maxi- (ultimately inadequate) oxygen supply to skeletal

127
Noakes
muscle, then the maximum cardiac output during ex- under conditions of hypobaric hypoxia. The sole con-
ercise at increasing altitude must either stay the same clusion must be that some type of ‘‘governor’’, as
or even increase at increasing altitude in order to limit originally proposed by A.V. Hill, must limit maxi-
the effects of the progressive reduction in the arterial mum exercise at altitude. Furthermore, it would be
oxygen content. difficult to explain why the same control mechanism
Yet the evidence is absolutely clear. The heart should not act similarly during maximum exercise at
makes the exactly opposite adjustment – maximum sea level.
cardiac output falls with increasing altitude (Sutton In summary, a number of famous studies have
et al. 1988). The reduction is due to the reduction in shown that under the precise conditions likely to in-
heart rate; stroke volume and myocardial contrac- duce anaerobiosis in either the heart or skeletal
tility are, if anything, enhanced during peak exercise muscles – maximal exercise at altitude – neither the
at altitude (Reeves et al. 1987, Suarez et al. 1987). heart nor the skeletal muscle show any evidence
Hence the conclusion must be that some currently whatsoever for ‘‘anaerobic’’ metabolism. This unex-
unrecognized mechanism must exist to insure that the pected finding can be explained only if there is a
heart does not become ‘‘anaerobic’’ during maximal ‘‘governor’’, probably in the central nervous system,
exercise at any altitude – from sea level to the summit whose function is likely to prevent the development
of Mount Everest – in healthy humans. of myocardial ischaemia. The same governor could
Interestingly Christensen, but not Dill, interpreted also serve the identical function also at sea level,
this phenomenon correctly: ‘‘Christensen and I dif- thereby preventing the development of myocardial
fered in our interpretation of his measurements of ischaemia during maximum exercise at sea level, ac-
respiratory and circulatory function in exercise (at cording to Fig. 2. As Dill (1938) concluded, probably
altitude). In his opinion, the chief limiting factor is correctly: ‘‘The capacity of the heart, as has already
the ventilation of the lungs. In the hardest grade of been suggested, is restricted at high altitude because
work at any station, the pulmonary ventilation of the deficiency in supply of oxygen to it’’ (p. 15).
reached about as high a value as at sea level, while the But the important point is that the heart never actu-
maximal cardiac output became less as the altitude ally develops an oxygen deficiency at altitude or at
increased. He thinks this means that the heart has an sea level; the governor acts to terminate exercise be-
untapped reserve; it is circulating blood fast enough fore that deficiency becomes apparent.
to carry to the tissues all the oxygen supplied by the The final confirmation for the presence of this the-
lungs’’ (Dill 1938, p. 170–171). oretical governor comes from the study of Kayser et
These studies invite two precise conclusions. First, al. (1994). They showed that skeletal muscle recruit-
that the oxygen demands of the skeletal muscles are ment, measured as skeletal muscle EMG activity at
not the cardinal priority and hence are not ‘‘pro- peak exercise, falls with increasing altitude, but in-
tected’’ during maximum exercise, at least at extreme creases acutely with oxygen administration. They
altitude. Second, neither the skeletal muscles nor the conclude: ‘‘during chronic hypobaric hypoxia, the
heart becomes ‘‘anaerobic’’ during maximal exercise 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-
Fig. 4. The ‘‘governor’’ postulated by Hill could be activated by
a limiting myocardial oxygen delivery. As a result, there would ebral hypoxia whilst their skeletal muscles were exer-
be a reduced activation of the exercising skeletal muscle by the cising vigorously and unrestrainedly, in pursuit of an-
cerebral motor cortex. There is compelling evidence for the aerobiosis and fatigue, according to the model de-
presence of this reflex during exercise at altitude (Kayser et al. picted in Fig. 1.
1994) with the result that neither the heart (Sutton et al. 1988,
Reeves et al. 1987, Suzrez et al. 1987) nor the skeletal muscles Figure 4 therefore summarizes the hypothetical
(Green et al. 1989) develop ‘‘anaerobiosis’’ during maximal ex- existence and action of the ‘‘governor’’, first proposed
ercise at extreme altitude. by A.V. Hill. It is postulated that receptor(s) exist in

128
 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 Fig. 5. According to the Hill/Noakes Cardiovascular/Neural
the case for skeletal muscle, as discussed subsequently Model of Exercise Physiology and Athletic Performance, per-
(Fitts 1994, Spriet et al. 1987). Reduction of myocar- formance during maximal exercise is ultimately limited by the
dial ATP concentrations could lead directly to a re- peak coronary blood flow. However, the actual workrate
achieved at that peak coronary blood flow would be determined
duction in myocardial contractile force as occurs in by the efficiency and contractility of both the heart and the
‘‘myocardial stunning’’ (Braunwald & Kloner 1982). active skeletal muscles.
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 adapted Nepalese Sherpas is not different from that
and shows no evidence for ‘‘fatigue’’. of acclimatized Caucasian climbers (Kayser et al.
Accordingly, it is proposed that maximal exercise is 1991) except that (iii) the volume density of skeletal
limited by a regulated process that terminates exercise muscle mitochondria is significantly smaller in Sherp-
before the development of a progressive myocardial as than in untrained sedentary subjects (Kayser et al.
ischaemia, that would precede the development of 1991), can best be explained if exercise performance
skeletal muscle anaerobiosis. This model further pre- at altitude is more likely determined by factors pro-
dicts that peak coronary blood flow is an important ducing superior oxygenation of the heart, than of the
determinant of maximum exercise performance, and skeletal muscles. This model therefore predicts that
that interventions, including exercise training, that in- coronary blood flow and perhaps cardiac mitochon-
crease the maximum cardiac output probably also in- drial mass would be higher in high altitude natives
crease the maximum coronary blood flow, as their and superior performers at high altitude. This would
more important effect. explain why high altitude natives achieve almost simi-
But this model does not exclude the possibility that lar cardiac outputs at sea level and at altitude (Vogel
interventions could also improve exercise perform- et al. 1974), indicating a lesser activation of the ‘‘gov-
ance by altering either skeletal muscle or myocardial ernor’’ during exercise at altitude.
contractile function or the efficiency of oxygen util- Indeed the ability of high altitude natives to
ization, or both (Fig. 5). achieve high heart rates and cardiac outputs during
Interestingly, the presence of a ‘‘governor’’ prevent- exercise at altitude is associated with ‘‘relatively high
ing the development of anaerobiosis in either heart oxygen tension and saturation’’ (Vogel et al. 1974)
or skeletal muscle during exercise at altitude has in- compatible with this postulate that superior myocar-
teresting implications for theories of the value of ex- dial oxygenation might be an important factor deter-
ercise training at altitude. For its presence means that mining exercise capacity at altitude.
any beneficial effect of altitude training cannot result
from repeated exposure of either the heart or exercis- Prolonged submaximal (endurance) exercise
ing skeletal muscle to greater levels of ‘‘anaerobiosis’’ Many physiologists, notably in the past (Bainbridge
than can be achieved during maximal exercise at sea 1931, Hill 1927, Dill 1938) but even today (Bassett &
level. This might explain why there remains consider- Howley 1997), have used this cardiovascular/anaer-
able controversy about the proven value of high in- obic model also to explain the fatigue that develops
tensity training at altitude (Boning 1997). during prolonged submaximal exercise and conse-
The additional paradoxes that (i) adaptation to ex- quently have evoked changes in cardiovascular func-
treme altitude is associated with reduced skeletal tion to explain the mechanisms by which exercise
muscle mitochondrial volume and enzyme content training improves (endurance) performance during
(Green et al. 1989, Oelz et al. 1986, Hoppeler et al. prolonged submaximal exercise.
1990); (ii) the skeletal muscle morphology of altitude- Yet it is not entirely apparent why changes in the

129
Noakes
Table 2. World rankings of male Kenyan runners in 1996 physiological explanation for the Kenyan’s success
Distance Rankings in world top 10
must be able also to explain why these physiological
attributes, uniquely common in Kenyan runners, of-
800 m 1*, 4, 6, 7, 10 fer an even greater advantage in cross-country events
1500 m 4, 5, 7, 8 and in the steeplechase, rather than at other distances
5000 m 1, 7, 8, 9, 10
10 000 m 3, 4, 5, 6
in which repeated jumping and changes in speed do
3000 m steeplechase 1, 2, 3, 4, 5, 6, 8, 9, 10 not occur.
3000 m 1, 3, 4, 5 Two studies of Kenyan runners failed to provide a
Marathon 4, 6 definitive physiological answer for their manifest su-
* The naturalized Dane, Wilson Kipketer, is considered a Kenyan for the
periority as distance runners although two of the best
purposes of this analysis. 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’ VO2
Table 3. Performances of Kenyans in the I.A.A.F. world cross-country max values were not inordinately high; hence, a su-
championships (1986–1997) perior capacity for oxygen consumption during maxi-
Kenyan Senior Men, 1st for the last 12 years (1986–1997)
mum exercise did not explain the Kenyans’ manifest
Kenyan Junior Men, 1st for the last 10 years (1988–1997) superiority during more prolonged submaximal exer-
Kenyan Senior Women, 1st 5 times in the last 7 years cise. In the words of Bengt Saltin, the senior author:
Kenyan Junior Women, 1st 8 times in the last 9 years ‘‘A comparison of some data on some of the very
Total: 35 Championships in 49 competitions including 24 individual
best runners in Kenya during the last decades and
championships.
world class runners in Scandinavia does not reveal
much that was not already known or could be antici-
pated’’ (Saltin 1996).
maximum capacity to transport and utilize oxygen The only other study of elite (South) African dis-
must also explain alterations in performance during tance runners is that of Coetzer et al. (1993). That
submaximal exercise when oxygen transport cannot study, which reported physiological data in the best
be limiting. An early proponent of this (il)logic was group of distance runners yet evaluated anywhere in
Sir Roger Bannister who wrote in 1956 that: ‘‘The the world, has been largely ignored, for reasons that
muscular effort in long-distance running appears to are not immediately clear. The sole weakness of the
be limited by cardio-respiratory failure as a whole study was that the physiological characteristics of el-
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 Fig. 6. In comparison to a group of elite Scandinavian runners
measures of that dominance are provided by the (the highest running curve), the running economy (VO2 at any
world rankings of male Kenyan track runners in 1996 running speed) of former world marathon record holder, Derek
Clayton, and two of the greatest Kenyan runners, Julius Korir
(Tanser 1997) (Table 2) and of the performances of
and John Ngugi, is substantially better. The Biomechanical
the men and women’s team, both senior and junior, Model of Exercise Physiology and Athletic Performance pre-
in the World Cross-Country championships over the dicts that superior running economy would aid elite athletic
past 12 years (Tanser 1997) (Table 3). Of particular performance by reducing both the rate at which heat is pro-
interest is the almost total dominance of the 3000 m duced during exercise as well as the extent to which Stretch
Shortening Cycle Fatigue develops during both training and
steeplechase by Kenyans (Table 2). Indeed in excess racing. This theory predicts that the superior running ability of
of 90% of the 100 fastest-ever 3000 m steeplechase the Kenyans may relate, at least in part, to characteristics of
times in the world have been set by Kenyans. Any their skeletal muscle and tendon elasticity (see also Fig. 12).

130
 Physiological models to study exercise
Table 4. Characteristics of elite South African distance runners cise than is the VO2 max alone. Furthermore, it has
Middle Long
not been appreciated that the % VO2 max sustained
distance distance during exercise is a measure of the athlete’s resistance
to fatigue.
Height (cm) 181 169* Hence, the important finding of that study was to
Weight (kg) 70 56*
Aerobic capacity (ml/kg/min) 72 71
show that the cardiovascular/anaerobic model may be
Fatigue resistance (% VO2 max for 21 km) 82 90* unable to discriminate between very good and su-
Muscle fibre composition (% Type I) 63 53 perior performance in events lasting more than a few
minutes and which constitute the bulk of sporting
*PΩ∞0.05.
Data from Coetzer et al. 1993.
events. It is consistent with the finding that the VO2
max is a relatively poor predictor of endurance per-
formance in athletes whose abilities are relatively
homogenous (Noakes 1988, 1997, 1998, Davies &
ite black South African distance runners were, per- Thompson 1979, Noakes et al. 1990). The failure
force, compared to those of white South African stems from the inability of this model to measure or
middle distance runners. This comparison group was predict fatigue resistance during prolonged submaxi-
not ideal but was necessary because no other readily mal exercise on the basis of physiological variables
available group could match the performances of the and performance measured during a single bout of
black distance runners at any distance over 5 km. But progressive, maximal exercise to exhaustion. This
the performances of the black and white runners in confirms that the VO2 max test does not measure all
that study were at least equal in races of 1–3 km. the physiological variables determining success dur-
Table 4 lists the important findings of that study. ing more prolonged exercise.
The black distance runners were lighter and Further support for this explanation can be sur-
smaller, as also reported by Saltin (1996), with a mised from other information in Fig. 7, which shows
slightly lower proportion of type I muscle fibres. But that these athletes run at 100% or greater of their
the key finding was that the black runners were able VO2 max in race distances of 1–2 km. Yet it is not at
to run substantially faster at all distances beyond 5 those distances that the Kenyans’ dominance is most
km despite VO2 max values that were the same as apparent. If the Kenyans’ success was due to their
those of the middle distance runners. Hence the car- unusually high VO2 max values, one would expect
diovascular/anaerobic model failed to explain the su- Kenyans also to be dominant at race distances of 800
perior endurance capacity of the black distance run- m to the mile, which is not the case (Table 2).
ners in that study and perhaps also in the studies of Indeed, comparison of the performances of the
Saltin and colleagues (Saltin 1996, Saltin et al.
1995a,b).
Rather, the important difference was that the black
runners were able to sustain a substantially higher
proportion of their VO2 max when racing. This is
shown graphically in Fig. 7, which compares the %
VO2 max sustained by the black and white runners
at different racing distances. At distances beyond 5
km, black runners sustained a significantly higher %
VO2 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 (VO2 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 % VO2 max that athletes can Fig. 7. The % VO2 max sustained by elite South African black
sustain during exercise is an important predictor of and white distance runners falls with increasing racing distance.
performance (Costill et al. 1973, Davies & Thompson However, black runners sustain a significantly higher % VO2
1979). Yet we have perhaps failed to emphasize, prob- max at race distances of 10 km and 21 km, indicating superior
fatigue resistance. Such superiority cannot be explained by the
ably because of a devotion to the cardiovascular/an- Cardiovascular/Anaerobic Model of Exercise Physiology and
aerobic model, that this is likely to be a more import- Athletic Performance as VO2 max values of black and white
ant determinant of performance in prolonged exer- athletes in this study (Coetzer et al. 1993) were the same.

131
Noakes
great British runner, Sebastian Coe, with those of a ercise is determined by either the presence or absence
current Kenyan champion, Daniel Komen, provides of skeletal muscle anaerobiosis.
further evidence for this interpretation. Remarkably, Accordingly, changes in exercise performance that
Komen’s best time for the mile is 1 s faster than Coe’s result from endurance training are unlikely to be de-
best. Yet the real performance difference occurs at 5 termined solely by changes in cardiovascular func-
km: Komen’s best time is 83 s faster than Coe’s best, tion, with the exception that increases in maximum
a performance difference of 10%. coronary blood flow would likely be crucial for any
In summary, there are serious theoretical flaws in increases in maximal cardiac output and hence in
the proposed cardiovascular/anaerobic model of exer- VO2 max. It is of interest that the vasodilator ca-
cise physiology and athletic performance (Noakes pacity of the major epicardial coronary vessel is
1998), not least because the model predicts that a greatly increased in veteran long distance runners
‘‘plateau’’ in cardiac output must develop before skel- (Haskell et al. 1993). Perhaps this indicates that an
etal muscle anaerobiosis can begin to occur. But any important effect of endurance training, possibly at
‘‘plateau’’ in cardiac output requires that myocardial some critical growth periods, may be to increase
ischaemia be present either to cause that plateau (ac- maximum coronary blood flow as shown in animal
cording to the theory that anaerobiosis limits muscle models (Scheuer & Tipton 1997).
function) or as a result of it, as the cardiac output
determines both coronary and skeletal muscle blood
The energy supply/energy depletion model
flow. As myocardial ischaemia has never been shown
to develop during maximal exercise in healthy The energy supply model
humans, so it would seem unlikely that skeletal The central premise in the cardiovascular/anaerobic
muscle anaerobiosis can develop during progressive model is that it is the provision of a substrate (oxy-
exercise to exhaustion (Noakes 1998). Rather, it gen) to muscle that limits exercise performance so
would seem that ‘‘fatigue’’ during maximal exercise that fatigue is a direct consequence of a failure of
of short duration is part of a regulated neural process oxygen delivery to the exercising muscles. A subtle
that prevents the development of myocardial isch- extension of this idea produces a second model which
aemia during maximal exercise. proposes that fatigue during high intensity exercise
Whilst this mechanism is designed to protect the may, alternatively, result from the inability to supply
heart from myocardial ischaemia, only indirectly does another substrate (ATP) at rates sufficiently fast to
it determine the actual peak work rate achieved dur- sustain exercise. Nobel Laureate A. V. Hill, whose re-
ing maximal exercise (Fig. 5). The actual peak work- search in the 1920s was directly responsible for the
rate achieved will depend on the ‘‘quality’’ of the skel- development of the cardiovascular/anaerobic model
etal and cardiac muscle. Superior myocardial contrac- of exercise performance, also wrote: ‘‘The fact re-
tility and efficiency of oxygen use would increase the mains, however, that the chief factor in many forms
maximum cardiac output achieved at any maximum of athletic achievement is the supply of energy and
(limiting) coronary flow. Similarly at any maximum its proper and economic utilization’’ (Hill 1927, p.
skeletal muscle blood flow, superior contractility and 237). Dr Peter Snell, Olympic Gold medallist in the
efficiency of skeletal muscle contraction would in- 800 and 1500 m and former world record holder has
crease the peak workrate achieved at that maximum stated similarly: ‘‘Performance in middle and (long)
cardiac output. This hypothesis forms what might be distance running ultimately depends upon the run-
called the Cardiovascular/Neural Recruitment Model ner’s capacity to produce energy for the duration of
of Exercise Physiology and Athletic Performance. the event, and on the efficiency with which that en-
Thus, this analysis of the traditional cardiovascu- ergy is translated to running velocity. Thus the pur-
lar/anaerobic model of exercise performance leads to pose of training is to improve the energy delivery sys-
the alternate hypothesis that superior fatigue resis- tems, according to the demands of the event and to
tance, determined perhaps by the central nervous sys- improve running economy’’ (Snell 1997).
tem or skeletal muscle contractile function, might ex- Thus, this model predicts that performance in
plain superior performance in events lasting more events of different durations is determined by the ca-
than a few minutes. This superior fatigue resistance pacity to produce energy (ATP) by the different meta-
cannot be predicted by the cardiovascular/anaerobic bolic pathways including the phosphagens, oxygen-
model which uses exercise tests of short duration and independent glycolysis, aerobic glycolysis and aerobic
in which the fatigue resistance component of endur- lipolysis. Superior performance is then explained by
ance performance is not measured. By extension, it a greater capacity to generate ATP in the specific
would seem that fatigue resistance is not causally de- metabolic pathway(s) that predominates during that
termined by the magnitude of the athlete’s cardio- activity. Thus, the sprinter is assumed to have a
vascular capacity. There is also no logical reason to greater capacity to generate ATP from the intramus-
believe that fatigue resistance during submaximal ex- cular phosphagen stores and from oxygen-indepen-

132
 Physiological models to study exercise
dent glycolysis, whereas the ultramarathon runner It is of interest that the presence of this peripheral
has a superior capacity to oxidize fat (aerobic lipoly- ‘‘governor’’ is an essential component of the cardio-
sis) (Hawley & Hopkins 1995). vascular/anaerobic model as originally conceived
Whether this hypothesis is true is uncertain as it from the work of Hill and his colleagues, and still
has yet to be systematically evaluated. To prove this widely promoted (Fig. 1). This hypothesis holds that
model would require (i) that the metabolic capacities when the rate of ATP production by oxidative sources
of these different pathways be shown to be causally becomes inadequate, high rates of ‘‘anaerobic’’ glyco-
related to performance in events lasting the different lytic ATP production produce metabolites, particu-
durations; (ii) that the specific metabolic pathways be larly Hπ, which interfere with energy production and
shown to adapt predictably with specific training, and cross-bridge cycling causing fatigue and a failure of
(iii) that these adaptations alone explain the changes muscle contraction (Fitts 1994). In this way, muscle
in performance that result from training with exer- contraction fails not because of a failure of central
cises lasting the different durations. Until these recruitment (as predicted by the Cardiovascular/Neu-
studies are completed, this model remains hypotheti- ral Recruitment Model – previous section), but be-
cal, but interesting. It must be remembered that the cause of a peripherally located inhibition of muscular
truth of this model would need to disprove the op- contraction. Proponents of this model can cite a large
posing model, described in the previous section, body of evidence showing that a number of metabo-
which holds that maximal exercise performance is a lites can interfere with muscle cross-bridge cycling
regulated process limited by a failure of central neural measured in vitro in isolated muscle fibres (Fitts
recruitment. Fatigue at exhaustion, caused by a fail- 1994). The necessary assumption is that skeletal
ure of central recruitment, will always appear to be muscle contracting in vitro in the absence of an intact
due to a failure of ATP production unless the alter- neural system behaves exactly as it would in vivo
nate possibility is studied (and excluded) simul- when the influences from the central nervous system
taneously. are intact. But there is a body of evidence that is not
The argument against this model has been intro- compatible with these assumptions and conclusions.
duced elsewhere (Noakes 1997). In short, the predic- For example, one of the few studies to evaluate
tion of this model is that exercise must terminate critically this hypothesis that metabolites, particu-
when muscle ATP depletion occurs (Fitts 1994), that larly Hπ, can induce skeletal muscle fatigue, is that
is when the muscle develops rigor. Yet here again, the of Mannion et al. (1995). They found that there is a
evidence appears clear. ATP concentrations, even in wide range of muscle pH concentrations reached at
muscles forced to contract under ischaemic con- exhaustion during intense exercise showing that if an
ditions, do not drop below about 60% of resting accumulation of Hπ limits high intensity exercise in
values (Fitts 1994, Spriet et al. 1987, Hochachka vivo, then ‘‘considerable interindividual differences
1994) indicating that muscle ATP concentrations are must exist in the pH sensitivities of the various pro-
‘‘defended’’ in order to prevent the development of cesses involved’’ (Mannion et al. 1995, p. 98).
skeletal muscle rigor. As Fitts (1994) has concluded: Next, they found that in contrast to the prediction
‘‘The overriding evidence suggests that (the high en- from in vitro studies, subjects with the highest pro-
ergy phosphates) do not participate in the fatigue portion of type II muscle fibres were able to exercise
process; that fatigue produced by other factors re- to the lowest muscle pH concentrations. In contrast,
duces the ATP utilization rate before ATP becomes in vitro studies have suggested that type I muscle
limiting. The most compelling evidence for this con- fibres are more resistant to acidosis than are type II
clusion is that cell ATP rarely falls below 70% of the fibres. Finally, the authors found that subjects with
pre-exercise level, even in cases of exercise fatigue’’ (p. a greater skeletal muscle buffering capacity did not
82). accumulate more lactate during maximal exercise;
Hence, is appears that the rate of ATP demand by nor were they able to exercise for longer than did
the contracting muscles can never exceed the maxi- those with lesser muscle buffering capacity. They con-
mum rate of ATP supply because of the close cluded that ‘‘if acidosis makes any contribution to the
matching of ATP demand to the available ATP sup- fatigue during performance of this (high intensity)
ply (Spriet et al. 1987, Hochachka 1994). type of exercise, it is an indirect one ...’’.
There is an obvious analogy to the centrally situ- One possibility is that such exercise is terminated
ated neural ‘‘governor’’ that prevents the develop- by a central governor responding to factors other
ment of myocardial ischaemia during maximal exer- than skeletal muscle pH. Under these circumstances
cise at either sea level or altitude. The difference is there would be no relationship between the onset of
that the ‘‘governor’’ identified by Spriet et al. (1987) fatigue and muscle acidosis. This would not negate
is clearly located in the periphery and acts even in the established finding that in vivo, acidosis inhibits
muscles stimulated to contract with an externally ap- crossbridge cycling (Fitts 1994). It would mean only
plied current. that this mechanism is not relevant in exercising

133
Noakes
humans, perhaps because exercise terminates for these trials. Indeed, two of the first such trials which
other reasons, in particular to prevent the develop- included adequate placebo-controlled groups have
ment of myocardial ischaemia, before the limiting both failed to find any ergogenic effect of pre-exercise
skeletal muscle pH is reached. carbohydrate loading (Burke et al. 1999, Hawley et
Other relevant findings include the study of Bog- al. 1997) under experimental conditions when such
danis et al. (1995), who showed that recovery of an effect might have been expected.
muscle function following maximal sprint cycling ex- The finding that the reversal of hypoglycaemia
ercise was related to recovery of muscle phospho- alone allows exercise performance to continue
creatine concentrations and unrelated to muscle pH (Coggan & Coyle 1987, Christensen & Hansen 1939)
concentrations during recovery. In addition, Vollestad proves conclusively that liver glycogen depletion is
et al. (1988) showed that the gradual decline in maxi- one form of energy (carbohydrate) depletion that can
mum force generation in subjects performing re- definitely limit exercise performance. Interestingly,
peated submaximal contractions for 40–70 min was the rapidity with which the reversal of hypoglycaemia
‘‘not due to lactacidosis or lack of substrates for ATP restores exercise performance indicates that a central
resynthesis and must have resulted from excitation/ neural ‘‘governor’’ must be active, similar to that acti-
contraction coupling failure ...’’. Yet terminal exhaus- vated during high intensity exercise at altitude. How-
tion was associated with depletion of intramuscular ever, this control would be activated by changes in
phosphagen stores, but without evidence for acidosis. blood glucose concentrations and would act to pre-
In summary, a metabolic basis limiting high inten- vent continuing high rates of muscle contraction and
sity exercise of short duration is widely assumed but blood glucose oxidation that would further reduce
incompletely documented. There is a need to estab- the blood glucose concentration, risking hypogly-
lish whether those metabolic factors that appear to caemic cerebral damage.
limit muscle function in vitro also play a role in vivo But, in as much as no technique has yet been de-
when the muscle is also under the influence of the vised that will instantly reverse muscle glycogen de-
central nervous system. Thus, the possible contri- pletion, in the same way that intravenous glucose in-
bution of neural factors to this form of fatigue needs fusion or oral glucose ingestion rapidly reverses hy-
to be excluded before results from in vitro studies are poglycaemia, so it is impossible to prove conclusively
extrapolated, without qualification, to the in vivo that muscle glycogen depletion alone limits prolonged
condition. exercise performance. It needs to be remembered that
there are many physiological changes besides muscle
The energy depletion model glycogen depletion that develop during exercise, and
The related energy depletion model of exercise per- that any or all of these could contribute to, or cause
formance is specific for exercise lasting more than 2– fatigue during prolonged exercise. In addition, rela-
3 h. It holds, in essence, that: ‘‘Depletion of endoge- tively little attention has been paid to the possible
nous carbohydrate stores has been shown to be a lim- role of central (neural) fatigue (Davis & Bailey 1997)
iting factor in the ability to perform long term exer- as the factor limiting prolonged exercise when muscle
cise’’ (Costill et al. 1973). The findings that support glycogen concentrations are also very low. Future
this conclusion are (i) that fatigue during prolonged studies of the energy depletion model need to show
exercise is associated with depletion of liver (causing that central neural factors do not cause the fatigue
hypoglycaemia) or muscle glycogen stores (Fitts 1994, currently ascribed to the development of muscle gly-
Bosch et al. 1993, Coggan & Coyle 1987, Coyle et al. cogen depletion during prolonged exercise.
1986, Tsintzas et al. 1996), or both; (ii) that reversal Thus, the belief that muscle glycogen depletion
of hypoglycaemia allows exercise to continue causes fatigue is an interesting hypothesis that is sup-
(Coggan & Coyle 1987, Coyle et al. 1986, Tsintzas et ported logically by the findings that subjects who are
al. 1996, Christensen & Hansen 1939) and (iii) that exhausted during prolonged exercise develop very low
pre-exercise muscle glycogen supercompensation muscle glycogen content (Fitts 1994, Bosch et al.
(carbohydrate-loading) (Hawley et al. 1997) or carbo- 1993, Tsintzas et al. 1996, Burke et al. 1999), and that
hydrate ingestion during exercise (Coyle et al. 1986), muscle glycogen is the metabolic fuel required for sus-
or both (Bosch et al. 1996), delays the onset of fatigue tained high intensity exercise (Bosch et al. 1993). It
and improves exercise performance. would seem logical to assume that the two are caus-
However, it must be remembered that relatively few ally linked at exhaustion during prolonged exercise;
carbohydrate-loading studies have been conducted namely, that the near absence of muscle glycogen in
with an adequate placebo control group. It seems exhausted subjects explains why they are unable to
highly improbable that neither athletes nor re- maintain, let alone increase, their exercise intensity at
searchers are completely unaware of the widely re- exhaustion. The finding that the vast majority of the
ported benefits of carbohydrate-loading and that modern and historical carbohydrate-loading studies
such knowledge is without effect on the findings of show that this technique improves endurance per-

134
 Physiological models to study exercise
formance, presumably by increasing muscle glycogen In addition, to my knowledge, no study has yet es-
utilization and delaying the onset of terminal muscle tablished that training improves endurance perform-
glycogen depletion, strongly supports the theory ance exclusively by increasing body carbohydrate
(Hawley et al. 1997). However, the possibility that stores and by delaying the onset of carbohydrate de-
part or all of these findings could also result from pletion during prolonged exercise in humans, al-
a placebo effect, acting through the central nervous though this finding has been reported in rats (Fitts et
system, needs to be considered (Burke et al. 1999). In al. 1975) whose metabolism is substantially different
addition, it is unclear how the inability to produce from that of humans.
ATP at sufficiently high rates from one fuel source Similarly, it is currently difficult to explain perform-
can explain this form of fatigue, given that skeletal ance in ultra-endurance events, especially the final 42
muscle ATP concentrations remain high at exhaus- km running leg of the 226 km Ironman triathlon
tion (as they do in all other forms of exhaustion events according to this model, which holds that exer-
(Fitts 1994)). cise of moderately high intensity is not possible once
Nevertheless, there is a body of evidence that con- there is marked muscle glycogen depletion. After cyc-
flicts with the predictions of this hypothesis. For ex- ling at 40 km ¡ hª1 for 4.5 h, the lead cyclists would be
ample, the classic study of Coyle et al. (1986) showed expected to have near total muscle glycogen depletion
that athletes ingesting carbohydrate terminated exer- according to data from laboratory studies (Bosch et al.
cise after 4 h when their muscle glycogen concen- 1993). Yet the best performers in that event are able to
trations and rates of carbohydrate oxidation were the run at close to 16 km ¡ hª1 for a further 160 min. This
same as values measured 1 h earlier when the athletes probably represents an exercise intensity of ±66% VO2
were not exhausted. Another study found that ath- max. The studies of Rauch et al. (1998) and O’Brien
letes who adapted to a high fat diet were able to exer- et al. (1993) suggest that total carbohydrate oxidation
cise to significantly lower muscle glycogen concen- during very prolonged exercise of up to 6 h duration
trations at exhaustion than when they were carbo- exceeds the estimated carbohydrate stores in liver and
hydrate adapted (Lambert et al. 1994). active muscle by up to 100%. Either these calculations
Conversely, Helge et al. (1996) showed that pre- are incorrect, or other sources of carbohydrate, in ad-
viously untrained subjects who trained on a high fat dition to those in the active muscles and liver, must
diet for 7 weeks before switching to a high carbo- contribute to fuel oxidation in events lasting more than
hydrate diet for one week increased their pre-exercise 4–6 h. One possibility is that lactate oxidation of glyco-
muscle glycogen concentrations by 44% with only a gen stored in the inactive skeletal muscles contributes
small further increase in performance between the a substantial additional amount to fuel use during very
seventh and eighth weeks of the trial. Furthermore, prolonged exercise (Rauch et al. 1998). How increased
performance was still substantially worse in fat- lactate oxidation contributes to performance is not
adapted subjects than it was in subjects who trained known.
for 8 weeks on a high carbohydrate diet and whose Similarly, provision of carbohydrate at high rates
pre-exercise muscle glycogen concentrations were intravenously (Coggan & Coyle 1987) cannot extend
32% lower than fat-adapted athletes exposed to a exercise performance indefinitely. Whilst this could
high carbohydrate diet for one week. support the argument that muscle glycogen is the im-
In addition, exercise performance evaluated on 3 portant carbohydrate source limiting exercise per-
occasions in both dietary groups terminated before formance, the alternate possibility is that another
there was marked muscle glycogen depletion. The factor, unrelated to depletion of body carbohydrate
authors concluded: ‘‘Factors other than carbohydrate stores, perhaps a rising body temperature discussed
availability are responsible for the differences in en- subsequently, or central neural fatigue induced by
durance time between the groups’’ (p. 303); and other factors, may also limit endurance performance.
‘‘These observations also indicate that fatigue during Perhaps there is a necessary rate of carbohydrate
prolonged moderately intense exercise does not al- oxidation that is required to sustain a specific exercise
ways seem to be closely related to glycogen depletion, intensity and that progressive whole body carbo-
as is usually stated’’ (Christensen & Hansen 1939, hydrate depletion lowers that rate, inducing fatigue.
Bergstrom et al. 1967). In his extensive review, Fitts If this is correct, then fatigue resistance during pro-
(1994) similarly concludes: ‘‘It seems unlikely that longed exercise could be due to the capacity to sus-
muscle glycogen depletion, low blood glucose, and tain a higher rate of carbohydrate oxidation, and
the resultant decline in carbohydrate oxidation is an hence a higher respiratory quotient (RQ) during pro-
exclusive fatigue factor during prolonged exercise’’. longed exercise. Again, this model suffers from the
He does, however, acknowledge that ‘‘a possibility persisting logical impasse that a failure to generate
exists that muscle glycogen depletion is causative in ATP sufficiently rapidly must cause exercise to ter-
fatigue via a mechanism independent of its role in minate because of muscle ATP depletion and rigor, a
energy production’’ (p. 83). phenomenon which does not occur (Fitts 1994).

135
Noakes
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
Fig. 8. Laboratory simulation suggests that an elite cyclist cyc- to complete the 42 km marathon leg of the Ironman
ling 180 km in 4 h 30 min would cycle at an oxygen consump- triathlon in 3 h 30 min (Fig. 9). To equal the best
tion of 57 ml/kg/min. Based on the measured contribution of marathon time yet run in that race, the athlete would
fat and carbohydrate oxidation to this energy requirement (col-
umns on the left of the figure), this would require the oxidation
be required to oxidize fat at a rate of 1.15 g/min (Fig.
of about 700 g of carbohydrate and 175 g of fat. This compares 10). This rate is approximately 50% faster than we
to the predicted maximum body stores of 520 g of carbohydrate have measured in cyclists in our laboratory.
and 5000 g of fat in an elite triathlete. Hence this model predicts Accordingly, if this metabolic model of fatigue in
that the elite triathlete must commence the running leg of the the Ironman triathlon is correct, then the difference
triathlon with very low or absent whole body carbohydrate
stores. between running the final marathon in 2 h 40 min
versus 3 h 30 min may simply be a 51% (0.4 g/min)

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% VO2 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- Fig. 9. The top panel of this figure shows the oxygen require-
ment (VO2), the exercise intensity (% VO2 max) and the rates
lations suggest that muscle glycogen stores are likely
of energy expenditure that would be sustained by a world-class
to be depleted. athlete completing the final 42.2 km marathon running leg of
Figure 8 shows the expected energy metabolism the Ironman triathlon in times of either 2 h 40 min; 3 h 00 min;
during 4.5 h cycling at an oxygen consumption of 57 or 3 h 30 min. Laboratory studies suggest that after a 3.8 km
ml/kg/min. This is equivalent to cycling at 40 km/h swim and a 180 km cycle, the athlete’s body carbohydrate stores
would be depleted so that energy for the running leg would
and completing the 180 km cycle leg of the Ironman come from oxidation of (mainly ingested) blood glucose, blood
triathlon in the time necessary to be amongst the race lactate and from circulating free fatty acids derived from muscle
leaders. The data for this simulation come from lab- and adipose tissue triglyceride. To complete the marathon run-
oratory data measured on elite South African cyclists ning leg of the Ironman triathlon in 3 h 30 min, the athlete
(I. Rodger. Unpublished data). would have to sustain a VO2 of 42 ml/kg/min (52% VO2 max),
equivalent to an energy expenditure of 59.5 kJ/min. Laboratory
The simulation predicts that after 4.5 h of cycling simulations (Fig. 8) suggest that under these conditions of near
an elite male Ironman triathlete would be expected to total carbohydrate depletion, peak glucose oxidation rates are
have oxidized about 700 g of carbohydrate and 175 g 1.2 g/min and peak lactate (from glycogen) oxidation rates are
of fat. This compares to predicted whole body carbo- 0.6 g/min. If these data for the maximum capacity to oxidize
glucose and lactate in the carbohydrate-depleted state are cor-
hydrate and fat stores of 520 g and 5000 g, respec- rect, then to sustain the rate of energy expenditure necessary to
tively. Hence this model predicts that, at the end of run the marathon in 03:30:00, the carbohydrate-depleted tri-
the cycle leg, an elite athlete would have depleted his athlete must oxidize fat at a rate of 0.76 g/min.

136
 Physiological models to study 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
Fig. 10. To complete the marathon running leg of the Ironman and Mercier (1994) is of particular interest. These
triathlon in 2 h 40 min, currently the fastest running time yet
recorded in the Hawaiian Ironman triathlon, the athlete would
authors argue that fuel choice during exercise
have to sustain a VO2 of 53 ml/kg/min (66% VO2 max), equiva- ‘‘crosses over’’ from predominantly fat to exclusively
lent to an energy expenditure of 74 kJ/min. If the maximum carbohydrate at exercise intensities above about 80%
capacity to oxidize glucose and lactate in the carbohydrate-de- VO2 max. They conclude that training produces dif-
pleted state is unchanged from values given in Fig. 9, then to ferent effects depending on the intensity of the exer-
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 cise being evaluated. At exercise intensities below
superior ability of the elite Ironman triathlete may result from their ‘‘crossover’’ point, training increases fat oxi-
a much greater (approximately 50%) capacity to oxidize fat dation whereas at higher exercise intensities, training
than has been measured in our laboratory experiments of very increases the capacity to burn carbohydrates. Accord-
prolonged laboratory exercise involving sub-elite athletes
(Rauch et al. 1998).
ing to their concept, exercise at higher intensities
(±70% VO2 max) would be limited as an example of
energy supply limitations (an inability to sustain high
rates of carbohydrate oxidation), whereas exercise at
greater capacity to oxidize fat when body carbo- lower intensities would be limited by muscle glycogen
hydrate and, especially, muscle glycogen stores are depletion. In fact the models are identical – according
depleted. Of course, this model does not negate the to the energy (muscle glycogen) depletion model, fa-
requirement that such high rates of fat oxidation tigue results from an inability to supply ATP suffi-
can only be achieved if the central nervous system ciently rapidly from fat oxidation (failure of energy
continues to recruit an appropriately large number supply from fat oxidation). The opposite pertains
of muscle fibres able to produce an appropriate during high intensity exercise. In fact, as argued here
force. and elsewhere (Fitts 1994), both the energy supply
In summary, the human body has a limited ca- and the energy depletion models predict that muscle
pacity to store carbohydrates. In addition, high rates ATP depletion limits exercise. This does not occur;
of carbohydrate oxidation are necessary to sustain hence, the models must be too simplistic to explain
high rates of energy expenditure (in the fed state). what has been found.
Furthermore, studies of very prolonged exercise (6 h)
show that rates of carbohydrate oxidation remain
The muscle recruitment (central fatigue)/muscle power
high in athletes who ingest appropriate amounts of
model
carbohydrate during exercise (Rauch et al. 1998). As
both muscle and liver glycogen depletion occur in the The two previous models are based on the assump-
fatigued state, it has popularly been assumed that tion that it is either the delivery of substrate either
there is a direct causal relationship between, espe- in blood (oxygen) or via the glycolytic and oxidative
cially, muscle glycogen depletion and the develop- pathways (ATP) that limits exercise performance. The
ment of fatigue during prolonged exercise. Yet some steps of (il)logic that have influenced these assump-
findings suggest that this relationship may not be tions have been described (Noakes 1997, 1998). It re-
strictly causal under all circumstances. In addition is mains difficult to prove whether or not either of these
the logical impasse which requires that any energy models is correct. Yet both continue to dominate, per-
depletion model predicts that exercise must terminate haps subconsciously, research and teaching in the ex-
when muscle ATP depletion occurs, leading to muscle ercise sciences, often to the exclusion of competing
rigor. In the absence of such evidence, it would seem possibilities.
that factors in addition to depletion of body carbo- An alternate view is that it is not the rate of supply
hydrate store may contribute to, or even cause, fa- of substrate, either oxygen or fuel, to muscle that lim-
tigue during prolonged exercise. its its performance but rather the processes involved

137
Noakes
in skeletal muscle recruitment, excitation and con- muscle would function to prevent organ damage
traction. (Noakes 1997).
A failure of central nervous system recruitment of However, the contrasting finding that skeletal
skeletal muscle forms the basis for the ‘‘central (ner- muscle recruitment, measured as skeletal muscle elec-
vous system) fatigue’’ hypothesis (Davis & Bailey tromyographic activity, rises during exercise at the
1997). This model holds that the brain concentration same workload (Takaishi et al. 1994), is usually inter-
of serotonin (and perhaps other neurotransmitters, preted as evidence for a progressive failure of muscle
including dopamine and acetylcholine) alters the den- fibre contractile function requiring additional fibre
sity of the neural impulses reaching the exercising recruitment if the required force is to be sustained.
muscles, thereby influencing the rate at which fatigue Thus, this model holds that there is a progressive pe-
develops, especially during exercise. Alternatively, ripheral fatigue for which the central nervous system
there may be inhibitory reflexes arising from the exer- makes an appropriate adjustment.
cising muscles and which feedback to the spinal cord, Yet these studies usually impose a constant work-
reducing skeletal muscle recruitment at the level of load on the subject for the duration of the activity.
the a-motoneuron. The evidence for both mechan- But competitive sport does not usually involve such
isms has been extensively reviewed (Davis & Bailey constant workrates; workrates tend to vary in a
1997). random, stochastic way (Palmer et al. 1994). We
In brief, a number of studies indicate that manipu- found that during prolonged exercise, which in-
lation of central nervous system neurotransmitter cludes bouts of self-chosen high intensity exercise
concentrations, in particular increasing dopamine designed to simulate stochastic exercise, there is a
and reducing serotonin concentrations, can enhance progressive reduction in power output during the
exercise performance whereas the opposite impairs bouts of high intensity exercise (Burke et al. 1999).
performance. In addition, there is direct evidence for This strongly suggests central fatigue in which, dur-
reduced central neural drive to muscle after fatiguing ing bouts of exercise requiring a near maximum ef-
muscle contractions (Behm & St-Pierre 1997, Baker fort, there is an inadequate increase in central neu-
et al. 1993, Newham et al. 1991). This evidence is ral drive to compensate for the expected reduction
sufficiently persuasive to believe that central nervous in the power output of the fatiguing skeletal muscle
system fatigue contributes to fatigue during pro- fibres (peripheral fatigue). That a relatively small
longed exercise lasting tens of minutes to hours. percentage of the available muscle mass is ever re-
The clear evidence that fatigue at high altitude is cruited, even during maximal exercise (Sloniger et
caused by reduced central nervous system recruit- al. 1997), remains a perplexing but relatively under-
ment of the exercising muscles has been described. It recognized enigma. Proponents of any model of pe-
is very likely that the fatigue that occurs during exer- ripheral limitations for exercise performance need
cise in the heat is also likely limited by a failure of to explain why the body does not recruit all its
central recruitment as this form of fatigue cannot be available muscle mass to produce the necessary
explained by any other model. Thus, there is now force under varying exercise conditions as so-called
substantive evidence that each athlete can store only ‘‘peripheral fatigue’’ develops.
some limiting amount of heat before being forced to In summary, one interpretation of the muscle re-
reduce the exercise intensity or alternatively cease ex- cruitment (central fatigue) model is that changes in
ercising altogether (Nielsen et al. 1990, 1993, 1997). central neurotransmitters induce fatigue simply as a
There is no evidence from these studies that exhaus- natural consequence of prolonged exercise and
tion under these conditions is associated with either changes in the relative balance of the different (ergo-
skeletal muscle ‘‘anaerobiosis’’ or energy depletion. genic and ergolytic) neurotransmitters in the brain.
Rather, that it is either the total heat accumulated or No specific physiological value or importance is as-
its rate of accumulation that limits exercise is con- signed to this phenomenon.
firmed by intervention studies which show that pre- Alternatively, I have argued that a reduced central
cooling improves performance (Nielsen et al. 1997, activation of the exercising muscles may be necessary
Booth et al. 1997) whereas pre-heating has the op- to protect the human under specific conditions
posite effect (Gongalez-Alonso et al. 1999). (Noakes 1997, 1998). It is postulated that these con-
In summary, there is sufficient evidence to suggest trol mechanisms are necessary (i) to prevent myocar-
that a reduced central nervous system recruitment of dial ischaemia during exercise at high intensity; (ii) to
the active muscles terminates maximum exercise at prevent the development of muscle ATP depletion
high altitudes (and probably also at sea level). The and muscle rigor during high intensity exercise; (iii)
same mechanism likely terminates exercise in the heat to prevent myocardial ischaemia or cerebral hypoxia
when the body temperature reaches some limiting during exercise at altitude; (iv) to prevent a fall in
maximum and also when hypoglycaemia develops. In blood pressure during exercise in patients with
all these examples, reduced central recruitment of chronic heart failure; (v) to prevent heatstroke during

138
 Physiological models to study exercise
prolonged exercise in the heat, and (vi) to prevent glu-
The biomechanical model
copaenic brain damage during prolonged exercise
when hypoglycaemia results from liver glycogen de- There is growing interest in the role of muscles as
pletion. The likely mechanism of control is through elastic energy return systems which function both as
the regulation either of skeletal muscle recruitment or springs and torque producers during exercise (Pennisi
of excitation/contraction coupling in the muscle. 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
Muscle power model hence the more efficient it is. The more efficient, more
This model holds that muscle contractile capacity, elastic muscle will enhance exercise performance,
that is the ability of individual muscle cross-bridges especially in weight-bearing activities, by slowing (i)
to generate force, is not the same in the muscles of the rate of accumulation of those metabolites that
all humans, so that those with superior athletic abil- may cause fatigue during exercise, and (ii) the rate of
ity have muscles with a superior capacity to generate rise of body temperature, thereby delaying the
force (superior contractility) by the individual cross- achievement of the core temperature that prevents the
bridges of the different muscle fibres. This model is continuation of exercise.
well accepted by cardiac physiologists, the majority This new information underscores another import-
of whom would argue that calcium delivery to the ant logical weakness of the cardiovascular/anaerobic
myofibres and the activity of the enzyme involved in model for explaining enhanced endurance perform-
ATP hydrolysis, myosin ATPase, rather than sub- ance. For that model predicts that superior perform-
strate supply, determine the contractile state of the ance during prolonged exercise results from an in-
myocardium in both health and disease (Opie 1998). creased oxygen delivery to muscle and an increased
I could find only one recent statement using this rate of energy and hence heat production. Thus, ac-
model to explain superior athletic performance, cording to that model, the price of running faster is
specifically in swimming: ‘‘First, the strength of the that more heat must be produced. But a higher rate
muscles used in swimming is a major determinant of of heat production would induce fatigue prematurely
success in events from 50 m to 1500 m. Though this due to excessive heat accumulation, according to the
may not seem surprising, it must be remembered that findings of Nielsen and colleagues (Nielsen et al.
strength per se does not dictate fast swimming. The 1993, 1997). A more logical biological adaptation
forces generated by the muscle must be effectively ap- would be to reduce the rate of oxygen consumption
plied to the water if they are to propel the body. Thus, and hence the rate of heat production by increasing
strength specificity is the key to swimming success’’ the athlete’s efficiency (economy) of movement.
(Costill et al. 1992). Indeed, if the rate of heat accumulation limits exer-
There are rather few studies of the contractility of cise performance under specific conditions, then fac-
skeletal muscle isolated from athletes. These studies tors that slow the rate at which heat accumulates
generally show that endurance training reduces skel- when running fast should enhance performance. Two
etal muscle contractility (Fitts et al. 1989, Widrick et such factors are small size (Dennis & Noakes 1999)
al. 1996). This establishes that skeletal muscle con- and superior running economy. A smaller size reduces
tractility is not an immutable characteristic of the dif- the amount of heat produced when running at any
ferent muscle fibre types (Fitss & Widrick 1996). By speed. When environmental conditions limit the ca-
extension, one might speculate that the contractility pacity for heat loss, smaller runners will be favoured
of the specific muscle fibre types might differ between (Dennis & Noakes 1999).
athletes of different abilities in different sporting Further evidence supporting this argument that
disciplines, compatible with this muscle power model. heat accumulation is a factor limiting endurance per-
In summary, these two models of exercise perform- formance, is the finding that race times in both the
ance predict that changes in exercise performance marathon (Noakes 1992) and the longer distance
may result from increased skeletal muscle recruitment track races including the 3000 m steeplechase and the
resulting from enhanced central neural drive, or from 10 000 m (McCann & Adams 1997) deteriorate as the
increased muscle contractile function resulting from environmental heat load increases. Thus, there is an
biochemical adaptations in muscle that increase inverse relationship between the environmental heat
either force production or the rate of sarcomere load, measured as the Wet Bulb Globe Temperature
shortening, or both. Index, and the reduction in race performance.
However, the increase in performance resulting Therefore, according to this model, the more econ-
from these adaptations would occur only to the ex- omical the athlete, the faster he or she will be able to
tent that the cardiovascular limits for exercise per- run before reaching a limiting body temperature. A
formance were not exceeded, according to the Car- number of studies indicate that the best endurance
diovascular/Neural Model. athletes are also frequently the most economical

139
Noakes
(Noakes 1992; Fig. 6). Indeed, most training studies events is not likely to result from training that makes
show that improvements in running economy are per- the athlete ever more powerful with a larger muscle
haps the most likely response to training, especially mass and greater VO2 max. A more likely adaptation
in those who are already well-trained (Svedenhag & would be to reduce the athlete’s size and increase his
Sjodin 1985). This adaptation allows the athlete to or her running efficiency. That runners believe they
run faster at the same oxygen consumption; thus, he run better when lighter, is well known.
or she completes a given distance more rapidly for Another African analogy for this prediction is pro-
the same average rate of heat accumulation but a re- vided by the physiological strategy that the cheetah
duced overall heat expenditure. This would be advan- has evolved to survive as a successful predator. The
tageous under conditions in which the heat load on cheetah, whose chase is terminated by an elevated
the athlete increases (during the day). rectal temperature after running at up to 100 km ¡ hª1
Figure 11 shows that this adaptation may indeed for less than a minute (Taylor & Rowntree 1973), suc-
exist. In a cross-sectional study of recreational (not ceeds because of the animal’s small size and probably
elite) ultramarathon runners, it was found that those a high degree of running economy (due to elasticity
who trained more were more economical and hence provided by the flexible spine). Thus, laboratory ex-
could run faster at the same oxygen consumption or % periments showed that when the cheetah’s rectal tem-
VO2 max. During competition, the better trained perature reached 40.5–41æC, ‘‘the cheetahs refused to
athletes ran at the same or a slightly lower % VO2 max run ... They would simply turn over with their feet in
but completed the races in a shorter time (Scrimgeour the air and slide on the tread(mill) surface’’ (Taylor &
et al. 1986). Hence being more economical, not having Rowntree 1973).
a higher VO2 max, appears to be a more logical tech- The small size of the cheetah and its likely high
nique to enhance endurance performance. running economy slows its rate of heat accumulation
In contrast, a high aerobic capacity, often a marker just sufficiently for it to outrun the smaller gazelles
of poor running economy (Noakes 1988, 1992), (∑25 kg) on which it preys and whose escape is also
would likely cause more rapid rates of heat accumu- restrained by a rising body temperature (Taylor & Ly-
lation and hence the more rapid onset of fatigue dur- man 1972). Thus the chase between the gazelle and
ing prolonged exercise. This finding alone could ex- the cheetah is probably decided by which individual
plain why the best marathon runners usually have animal accumulates heat more slowly during the
VO2 max values in the range of 63–74 ml/kg/min. chase. In contrast, the heavier, more muscular lion
Less economical runners with higher VO2 max values has evolved a different co-operative, hunting strategy,
(Noakes 1988) have not necessarily been more suc- targeting larger but slower mammals.
cessful (Noakes et al. 1990, Noakes 1992). Perhaps the point is that smallness and greater run-
Thus, this model predicts that success in endurance 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
Fig. 11. Ultramarathon athletes who trained more than 60 changes in the stride with landing occurring on a
km ¡ weekª1 (the lower running curve) outperformed less well more extended leg but with greater subsequent knee
trained athletes with similar VO2 max values because their su- flexion.
perior running economy allowed them to race faster but at a As these abnormalities persist for many days after
lower % VO2 max, thereby finishing races of 10–90 km in
shorter times. Hence changes in running economy with training the race (Fig. 12), they cannot be explained by acute
allow athletes to run faster and at a lower exercise intensity, the changes in oxygen or substrate delivery to the
opposite of what is usually assumed. muscles, or by the elevated body temperature during

140
 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 Fig. 12. According to the Biomechanical Model of Exercise
resistance to stretch/shortening cycle fatigue would Physiology and Athletic Performance, repetitive eccentric
likely be very rewarding. muscle contractions, as occurs in the quadriceps and calf
muscles during running, produces altered muscle function with
For example, it appears that African athletes gen- a loss of elastic energy production, requiring an increased work
erally train harder than do Caucasian runners (Tans- during the push-off phase of the running stride. The studies of
er 1997, Coetzer et al. 1993). Especially the training Komi & Nicol (1998) and Nicol et al. (1991) show that there is a
volumes and intensities of the Kenyan runners (Tans- persistent reduction in peak torque and endurance performance
time of the quadriceps muscle of subjects after they completed
er 1997) are unmatched by other athletes. But to a 42.2 km marathon race. Integrated electromyographic activity
achieve such training volumes, there must be superior (IEMG) in both the vastus medialis (VM) and the vastus lat-
resistance to the stretch/shortening cycle damage pro- eralis (VL) was also reduced after the marathon. These studies
posed by Komi & Nicol (1998), both in training and show that there is a long-term effect of marathon running on
in marathon racing. skeletal muscle performance that persists after the muscle’s en-
ergy stores are replaced, indicating that this abnormality is not
Hence, another possibility is that the more elastic explained by the Energy Depletion Model. It seems probable
muscles of elite distance runners are better able to that the marathon ‘‘wall’’ (Noakes 1992) is probably explained
resist eccentrically induced damage in training. This by these alterations in skeletal muscle function.
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- hypothesis (Davis & Bailey1997). But it conflicts with
ening cycle fatigue that is an inevitable consequence one proposal of the muscle recruitment model, which
of repeated eccentric muscle contractions. holds that exercise performance is regulated at a sub-
In summary, the biomechanical model predicts that conscious level and which exists, in part, to prevent
superior performance, especially in a weight-bearing conscious override that might damage the human.
activity like running, may be influenced by the ca- It would seem that exercise performance must in-
pacity of the muscles to act as elastic energy return clude at least some component that can be influenced
systems. Changes in the efficiency and durability of by conscious effort. The dichotomy of physiology
this process would (i) enhance movement economy and psychology has generally prevented adequate lab-
and reduce the rate of heat production during exer- oratory evaluation of this model. Any studies show-
cise, thereby enhancing exercise capacity by slowing ing an ergogenic effect of any placebo intervention
the rate at which the body temperature rises when on exercise performance would prove that this model
environmental conditions are severe; (ii) enhance the contributes, in part, to athletic performance. Any de-
quality of training by allowing more rapid recovery tailed discussion of this model is beyond the scope of
from stretch/shortening cycle fatigue so that more this author’s expertise.
frequent bouts of intensive training can be under-
taken and (iii) enhance fatigue resistance during com-
Summary
petition by increasing resistance to that form of
muscle damage that develops during repeated cycles This paper has reviewed some of the models currently
of stretch/shortening contractions. 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-
The psychological/motivational model
mining both fatigue and athletic performance are es-
This model holds that the ability to sustain exercise tablished more definitely, it remains difficult to define
performance results from a conscious effort and is which training adaptations are the most important
often included as a component of the central fatigue for enhancing exercise performance, or how training

141
Noakes
should be structured to maximize those adaptations. be considered as an ‘‘energy supply’’ model was
This remains a serious weakness for the practical ap- described above.
plication of much research in the exercise sciences. (d) Muscle recruitment and muscle power models:
However, within the constraints provided by these These models predict that increased skeletal
models, it would seem that the following training ad- muscle contractile function, a peripheral effect,
aptations would contribute to enhanced exercise per- or increased neural recruitment, a central nervous
formance with training. system effect, would be advantageous for per-
(a) Cardiovascular/neural recruitment model: Rel- formance during exercise.
evant training adaptations would be those that (e) The biomechanical model: A key predictor of the
result in an increased VO2 max and skeletal biomechanical model is that increased movement
muscle blood flow during both maximal and pro- economy would improve performance by reduc-
longed submaximal exercise. ing the rate of heat accumulation during exercise.
According to the model I have presented here, This model also explains why a reduced body
a plateau in coronary flow would appear to be mass would improve performance during pro-
the factor that limits the cardiac output, and longed exercise as it slows the rate of heat ac-
hence the VO2 max. Thus, an essential compo- cumulation.
nent of training would be to increase the maximal The importance of elastic return energy, espe-
coronary flow. cially in weight-bearing sports, and the identifi-
However, even if the maximal coronary blood cation of stretch/shortening cycle fatigue suggests
flow limits the maximal cardiac output, the actual that training may improve elasticity and delay
peak workrate or peak VO2 max that is achieved stretch/shortening cycle fatigue, perhaps by al-
will depend on the contractile state of the myo- tering the elastic component of skeletal muscle,
cardium and the efficiency with which the heart tendons and ligaments.
is able to convert that maximum coronary flow
into a peak cardiac output. Similarly, the actual
Conclusion
maximum work of which the muscles are capable
at the VO2 max, will equally depend on the econ- The importance of these conceptual models is that they
omy and contractility of the skeletal muscles. indicate that different physiological systems may de-
Thus, optimum training would not only maxim- termine performance under different exercise con-
ize coronary flow but also the efficiency and con- ditions. Hence, exercise physiologists need to consider
tractility of both the heart and skeletal muscles. all these models when they design studies to determine
Incidentally, according to this model, inter- which are the most important physiological, biochem-
ventions such as EPO therapy, blood re-infusion, ical, neural and other factors determining the changes
or the administration of oxygen, all of which im- in exercise performance that result from training.
prove performance, would act in part by increas- More importantly, this review shows that many
ing oxygenation of the myocardium at maximal findings are incompatible with the predictions of one
exercise, thereby allowing a greater cardiac output or more of these models. Rather than simply continu-
and hence a higher VO2 max. ing to accept these inconsistencies uncritically, the
(b) Energy supply model: The important training ad- modern generation of exercise physiologists should
aptation would be an increased capacity to store challenge old dogmas and so approach more closely
and utilize metabolic substrates during exercise the unattainable truth (Noakes 1997, 1998).
with a greater capacity to produce ATP and pre-
vent a reduction in muscle phosphagen concen- Key words: fatigue; VO2 max; heart; skeletal muscle;
trations under all exercise conditions. Adap- glycogen; fat metabolism; muscle recruitment; con-
tations in different metabolic pathways would be tractility.
necessary for optimizing performance in activi-
ties of different durations and intensities.
(c) Energy depletion model: A reduced rate of carbo- Acknowledgements
hydrate utilization during prolonged exercise The author’s research is funded by the dedicated financial sup-
would enhance performance by delaying the on- port of the University of Cape Town, the Medical Research
set of whole body carbohydrate depletion. This Council of South Africa, the Liberty Life Insurance Company,
the Foundation for Research and Development, the Founding
model predicts that an increased capacity to burn Donors of the Sports Science Institute of South Africa, Bromor
fat during prolonged exercise would enhance en- Foods, and a variety of nutritional and pharmaceutical com-
durance performance. That this model can also panies.

142
 Physiological models to study exercise

References

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 Ernährung. 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. Ernährung. 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 artery 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 – a short review. Int J Sports 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 VO2 max and running Dennis SC. Carbohydrate-loading and
during prolonged exercise. J Appl performance 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

143
Noakes
and metabolic machines. Boca Raton, dry 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. Pflügers 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 VO2 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. Pflügers 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.

144
 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

145