Journal Logo

Basic Sciences: Commentary

Maximal oxygen uptake: "classical" versus "contemporary" viewpoints: a rebuttal


Author Information
Medicine& Science in Sports & Exercise: September 1998 - Volume 30 - Issue 9 - p 1381-1398
  • Free


In the written address delivered by his son at the opening of the Institute in Paris that would bear his name in perpetuity, Louis Pasteur implored his audience to "Worship the spirit of criticism" (73). If we are to advance the exercise sciences, then each J. B. Wolffe lecture should, at least in part, excite that spirit of criticism.

I am therefore most thankful to my North American colleagues, Drs. Bassett and Howley (4), for fashioning the contrary arguments to aspects of the 1996 J. B. Wolffe lecture and other of my writings. Such criticisms are an essential ingredient of the scientific process when they identify issues that require more careful analysis and explanation, thereby encouraging a more critical introspection. However, it is my view that Bassett and Howley's rebuttal was based upon an erroneous analysis of the published record, including my own, and contains critical errors of fact and interpretation. Also, the authors appeared to have missed two important subthemes of the J. B. Wolffe lecture.

The first is that data can only be interpreted in the framework of a specific model. In the words of Stephen Hawking (26): "How can we know what is real, independent of a theory or model with which to interpret it?" The truth is "true" only in the context of the specific model through which it is affirmed. Hence the real significance of any experimental study is the wisdom it reveals about the model used to interpret its findings. Throughout their review, Bassett and Howley (4) repeatedly use their model, which I term the cardiovascular/anaerobic model of exercise physiology and athletic performance, to interpret data which they then claim "proves" their model. But a model cannot be used to prove itself. This is precisely why information that refutes the predictions of a model is so valuable.

Second, a model should be considered correct only for as long as it explains all the observed phenomena. When there is information to refute a model, then the scientist's duty is to retire that old and dilapidated model not to fortify it against all criticism.

The central theme of the J. B. Wolffe lecture was an analysis of the cardiovascular/anaerobic model of exercise physiology and athletic performance, which enjoys the support of the majority of exercise physiologists. The lecture included an analysis of the original studies (28,30-32,35-38) that created the model and presented some experimental evidence that refutes three inferences from the model. These inferences are: 1) that progressive muscle hypoxia limits maximal exercise performance, 2) anaerobiosis explains the onset of lactate production by skeletal muscles at the "anaerobic threshold," and 3) that mitochondrial adaptations alone explain changes in performance with endurance training.

In their response, Bassett and Howley (4) argue that I have drawn erroneous conclusions from the studies of Hill and his colleagues and review other studies which, in their assessment, "prove" the cardiovascular/anaerobic model. These studies "prove" that the maximum oxygen consumption (V˙O2max) is indeed limited by the development of anaerobiosis in the active skeletal muscles. So, too, is endurance performance because the V˙O2max is the best predictor of athletic ability. Finally, they wish to refute my earlier writings from other sources that speculate on those factors other than skeletal muscle oxygen delivery that may determine athletic performance; they did not recognize that the J. B. Wolffe lecture included an updated and substantially revised version of those speculations. More important to this debate was their failure to explain why no studies have yet refuted the cardiovascular/anaerobic model. As Einstein wrote of his own theories: "No amount of experimentation can ever prove me right; a single experiment may at any time prove me wrong" (17). Thus the authentic intellectual challenge for Bassett and Howley was to contest the refutations to the cardiovascular/anaerobic model that I presented, not to ignore these and instead to rehash conventional arguments that "prove" the traditional model and which are, in any case, available elsewhere. I present my rebuttal to Bassett and Howley's response to the Wolffe lecture structured according to the topics they have argued.


Did Hill and his colleagues establish the plateau phenomenon? I first became interested in the plateau phenomenon while writing the first edition of Lore of Running(66). I was drawn to the topic by a statement in Costill's book (11): "Since the early work of Hill and Lupton (35), exercise physiologists have associated the limits of human endurance with the ability to consume larger volumes of oxygen during exhaustive exercise" (pp. 25-26). While this statement is factually correct, an analysis of the original publications of Hill et al. (35-38) revealed that none had studied the relationship between oxygen consumption and endurance performance. In fact, the oxygen consumption of seven athletes running at speeds up to about 17 km·h−1 was essentially all of relevance that was measured. From these rather meager data, a series of assumptions, subsequently described and analyzed in detail, generated a hypothesis linking oxygen consumption to endurance performance. The reality was somewhat less than the expectation generated by Costill's text.

During my analysis, I happened upon Hill's circular argument (Fig. 2 in 69) in which he used his personal feelings incorrectly to identify a supposed "plateau phenomenon" when he ran at 16 km·h−1 for 250 s (Fig. 2 in(68); Fig. 3 in (69)) plus supporting text). These findings formed the basis of a 1987 ACSM tutorial lecture, the publication (68) of which has been quite extensively referenced previously without major rebuttal. In response, Bassett and Howley briefly note that"Noakes (68,69) criticizes Hill, contending that he lacked the data to show a plateau in V˙O2max (sic) with increasing running speeds. However, Table 1 and Figure 2 show the original data from A. V. Hill's paper (38). Clearly A. V. Hill did demonstrate a plateau in himself and also in subject J (whose O2 was nearly identical at 4.25 and 4.98 m·s−1 despite a marked increase in oxygen requirement). In his 1988 article (68), Noakes chose to re-fit Hill's velocity versus V˙O2 data using a linear equation. We are puzzled by this reinterpretation of A. V. Hill's data since it appears to be biased toward the view that a plateau does not exist. Hill was not guilty of "circular reasoning" as Noakes suggests but rather conducted experiments that clearly demonstrated the V˙O2 plateau." Unfortunately Bassett and Howley have not applied due care in their analysis, nor was the contentious data correctly referenced. They supplied page numbers from an earlier companion paper by Hill et al. (37).

Figure 2
Figure 2:
Hill and Lupton (35) predicted that the relationship between V˙O2 and running speed was exponential and comprised a limiting real oxygen consumption of 4 L·min−1 (Line A) and an accumulating oxygen debt. Together these reduced the renal oxygen cost of running at different speeds (Line B). The universal V˙O2max of 4 L·min−1 was achieved at a running speed of 13.2 km·h−1. Higher running speeds could be achieved only because of a large capacity for anaerobic metabolism in humans causing high rates of muscle glycogen use and heat accumulation. Because many athletes are able to run faster than 13.2 km·h−1 for prolonged periods, this model predicts that the "plateau phenomenon" can be present for extended periods (minutes to hours) before it causes the termination of exercise. Reference35. Hill, A. V. and H. Lupton. Muscular exercise, lactic acid, and the supply and utilization of oxygen. Q. J. Med. 16:135-171, 1923.
Figure 3
Figure 3:
The logical conflict for the cardiovascular/anaerobic model caused by the finding that the plateau phenomenon does not occur in all subjects.
Cardiorespiratory measures at different altitudes: Operation Everest II.

However, the contentious data (Table 1 and Figure 2 in their response) comprise 28 measurements of oxygen consumption (V˙O2) in seven different athletes. Of these 28 measurements, 14 were collected in Hill himself (subject H). The remaining 14 measurements were collected in six other subjects, three each in four subjects and one each from the final two subjects. Hill et al. (38) admit to modifying four of the 14 results in athletes other than Hill. In their words: "The observations on the two latter have been 'reduced' to the same body-weight as A. V. H. before plotting." They do not explain how these data were transformed, why such transformation was necessary, or whether that transformation influenced their conclusions.

For my original paper (68), I considered that only the 14 unmodified measurements on Hill provided sufficient material for reliable retrospective analysis. I remain unconvinced that one can reasonably expect to identify a "plateau phenomenon" and hence sustain an influential physiological theory from a total of three measurements of V˙O2 in each of four athletes, and one each in another two. However, it appears that Bassett and Howley have no such concerns. They accept these data as absolute proof of the "plateau phenomenon" because: 1) Hill's data showed a "plateau phenomenon," and 2) the V˙O2 of one athlete (J) increased by only 30 mL·min−1 when his running speed increased by 0.7 m·s−1. To arrive at this conclusion, they discarded: 1) the contrary evidence that the "plateau phenomenon" did not occur in the three other subjects (75% of the sample) in whom three data points were measured (subjects S, W, and C.N.H.L) and 2) my analysis of the other set of measurements (35) which irrefutably established that Hill did not exhibit a "plateau phenomenon" when running at 16 km·h−1 for 250 s (Fig. 2 in (68); Fig. 3 in (69)). Such selective analysis, in particular their reluctance to acknowledge, let alone refute, Figures 1-3 in the J. B. Wolffe lecture (69), does not provide adequate data for an observer of this debate to arrive at an informed decision.

Figure 1
Figure 1:
The 14 data points for V˙O2 measured at different running speeds in A. V. Hill (38). There is a linear relationship between V˙O2 and running speed without evidence of a "plateau phenomenon." Note the large (16%) variation in measured V˙O2 at 12 km·h−1 and the discrepantly high V˙O2 measured at 14.5 km·h−1. Notice also the discrepancy between the predicted V˙O2 with increasing running speed (Line A; Fig. 2) with the actually measured V˙O2 in Hill at increasing running speeds (this Figure). Reference 38. Hill, A. V., C. N. H. Long, and H. Lupton. Muscular exercise, lactic acid, and the supply and utilization of oxygen: parts VII-VIII. Proc. R. Soc. B 97:155-176, 1924.

We return to the 14 measurements made on Hill (Table 1 in (4). Those data are reproduced here in Figure 1. The relationship between V˙O2 and running speed is described by the linear equation (y = 0.7x + 0.9; r = 0.89) without evidence for a "plateau phenomenon," as originally argued (68). Therefore, I contend that Bassett and Howley have failed to provide evidence to support their contention that I erred and that Hill and his colleagues were "not guilty of circular reasoning" and "clearly demonstrated the V˙O2 plateau."

But there are additional concerns about the research methods common to Hill and his colleagues and subsequent researchers and should have been acknowledged by Bassett and Howley, given their interest in how a maximal effort is identified (41). Were Hill and his colleagues to submit these data for publication today, an indifferent reviewer would inquire whether the "plateau phenomenon," detected by Bassett and Howley in those classic data, was an artifact of the specific methods used by Hill and his colleagues. For example, to prove the plateau phenomenon with an interrupted protocol in which a single V˙O2 is measured at each exercise load, that sample must capture the peak V˙O2 for that work rate. This is especially relevant at the highest work rates when the duration of exercise is shorter so that a true steady state may never be reached.

Hill and his colleagues sampled expired air for 30 s beginning about 220 s after the onset of exercise. They did not prove that this method captures the peak V˙O2 for each particular work rate (running speed). Hence additional samples measured later than 220 s after the onset of exercise might have revealed yet higher V˙O2 in Hill. A later study (97) suggests that this is unlikely at the range of work rates tested by Hill and his colleagues. Yet Hill and his colleagues did not exclude this possibility, which could certainly apply at the much higher work rates of which modern athletes are capable.

But Hill and Lupton (35) were well aware of an upward drift in V˙O2 during more prolonged steady-state exercise. When Hill ran at 240 m·min−1(14.4 km·h−1), his V˙O2 increased from 3590 to 3910 mL·min−1 from the third to the 26th minute. Hill concluded that this could probably be "attributed to a painful blister on the foot causing inefficient movement" ((35), p. 155) and hence did not consider alternate explanations. Interestingly, Hill's V˙O2 (53 mL·kg−1·min−1) at the end of that exercise was close to his supposed V˙O2max of 57 mL·kg−1·min−1 when he ran at 17 km·h−1 for only 250 s (Figs. 1 in (68, 69)). Hill almost certainly never established his true V˙O2max (Fig. 2 in (68);Fig. 3 in (69)).

In their study Taylor et al. (97) identified two other factors that could interfere with the V˙O2max measured with the methods of Hill and his colleagues. First, any reduction in the rate at which V˙O2 increased at the onset of exercise "may result in apparent declines in the maximal oxygen intake ... which are in reality only the reflection of the fact (that) the rate of increase in oxygen intake is smaller than normal" (p. 79).

The second factor was improvements in running economy that would reduce the V˙O2 at any workload and hence the V˙O2max and that could also theoretically induce an artifactual "plateau phenomenon," as discussed subsequently. But the major criticism, raised previously(68,69), was that Hill should have run at speeds faster than 17 km·h−1 if he wished to prove that his V˙O2 had indeed leveled off at either 16 km·h−1 (Figs. 1 in (68, 69)) or at 17 km·h−1(Fig. 1). The reasons why Hill considered this unnecessary are discussed subsequently.

The question that begs an answer is: Why did Hill and his colleagues include all 28 measurements for their analysis and not just the 14 that I have included in Figure 1? The answer provides the astonishing realization, not previously identified, of what exactly Hill and his colleagues did conclude. For their conclusions were quite different from those that all others (11,53-55,67,82,83,97,101,103) have presented retrospectively.

For Hill and his colleagues did not search for the plateau phenomenon, as do modern exercise physiologists, by analyzing the successive measurements of V˙O2 in individual subjects as first suggested by Taylor et al. (97). Rather they considered the data as a collective and based their conclusion on the finding that no measured V˙O2 exceeded 4 L·min−1 in any of their studies. Hence their interpretation was that there is an absolute V˙O2max of 4 L·min−1 in humans: "The oxygen intake attains its maximum value, which in athletic individuals of about 73 kilograms body weight is strikingly constant (in the case of running) at about 4 L per minute" (author's emphasis) ((38), p. 157); "The form, however, of the oxygen intake curve of Figure 1, approaching a constant level of 4 L per minute, makes it obvious that no useful purpose would be served by investigating higher speeds in this way"((38), p. 157); and "The amount of work which the heart has to do is enormous, and it seems to reach its limit, in the case of athletic people, when about 4 L of oxygen are taken in per minute" ((31), pp. 230-231).

So I conclude that Hill et al. (38) interpreted the 28 measurements in their Figure 1as if all the data had come either from a single individual or from separate but physiologically identical individuals. As no subject achieved a V˙O2 greater than 4 L·min−1, they concluded that this was the universal limit (plateau) for all humans. Thus, paradoxically the authors of the very study that spawned the concept of the individual plateau phenomenon and the individuality of the V˙O2max believed neither.

There were two contributory reasons for this fixation with a universal human V˙O2max of 4 L·min−1. The first was the belief that V˙O2max rises exponentially with increasing running speed as shown in Figure 2, reproduced from Figure 3 in Hill and Lupton (35). The figure incorporates their theory that the total oxygen demand of running (Line B) exceeds the measured V˙O2(Line A) at speeds above 220 m·s−1 (13.2 km·h−1), this difference increased exponentially with further increments in speed. Hence they identified that the original "anaerobic threshold" or what became known as the "plateau phenomenon" began at a running speed of 13.2 km·h−1. Line A confirms their belief that the V˙O2max in humans is 4 L·min−1. Interestingly, another publication (29) indicates that Hill knew that humans can run faster than 13.2 km·h−1 for many hours. The implication is that Hill and his colleagues must have concluded that skeletal muscle anaerobiosis can be present for hours before it causes the termination of exercise.

Interestingly, Bassett and Howley appear to misrepresent the original figure and the interpretation of Hill and his colleagues, because they write: "By measuring steady-state V˙O2 values at different speeds, (Hill and Lupton) demonstrated a linear relationship between running speed and V˙O2." In fact, their acceptance of the cardiovascular/anaerobic model made it impossible for Hill and his colleagues ever to consider a linear relationship between running speed and V˙O2, a logical inconsistency missed by Bassett and Howley. A similar inconsistency identified in the J. B. Wolffe lecture, concerns the absence of a plateau in V˙O2 at those submaximal work rates above the "anaerobic threshold"(Fig. 7 in (69)). Figure 2 shows that Hill and his colleagues were not similarly inconsistent because their line identifying the measured V˙O2 (Line A) does not increase at running speeds greater than their postulated "anaerobic threshold" at 13.2 km·h−1. But V˙O2 does increase linearly with increasing work rate (running speed)(14,68). As a result, the V˙O2 predicted by Hill and his colleagues becomes gravely inaccurate at faster running speeds and was close to 100 mL O2·kg−1·min−1 at 18 km·h−1(Fig. 2) compared with the real value of about 62 mL O2·kg−1·min−1 (Fig. 2 in (69)).

The effect of this error would be to implant the subconscious conclusion that it mattered not whether any individual's V˙O2max was 4 (55 mL O2·kg−1·min−1 for the 73 kg Hill), 5 (68 mL O2·kg−1·min−1), 6 (82 mL O2·kg−1·min−1), or even 10(134 mL O2·kg−1·min−1) L·min−1. For the unmeasured "anaerobic" contribution would exceed even that highest value at all running speeds above 20 km·h−1(Fig. 2). Thus it would be pointless to measure V˙O2 (and hence V˙O2max) at faster running speeds, as the measured component would account for less than 50% of the "real" V˙O2. As Hill, Long and Lupton ((37), p. 135) wrote: "The maximum recorded oxygen intake in man, breathing air, and the total energy made available by the use of this amount in oxidizing foodstuffs is relatively so small that no severe exercise of any kind could be taken were man limited, in his movements, to energy provided through the contemporary supply of oxygen". Interestingly this is the precise model currently used to explain energy metabolism in exercising reptiles (5). But, unlike Hill and his colleagues, Bennett realizes that this system does not allow activity other than of high intensity and short duration: "Anaerobic metabolism (in reptiles) appears to be particularly useful in and used for fueling intense activities of short duration but of great ecological importance (e.g., territorial defense, nest emergence" (p. 131).

Hill and his colleagues derived these incorrect conclusions exactly because their cardiovascular/anaerobic model predicts a running speed at which oxygen demand outstrips supply. Accordingly, they had to develop an elaborate method to predict the "anaerobic" contribution to energy consumption during running. They termed this anaerobic contribution the "oxygen debt" (30,35,37). More recently the term "excess postexercise oxygen consumption" (EPOC) has been preferred(20). As they found that the "oxygen debt" so measured increased as an exponential function of the work rate, so they were bound to conclude that the "real" oxygen cost of exercise rose exponentially with increasing running speed(Fig. 2). With the wrong model, they were bound to reach false conclusions.

Interestingly, Hill should have realized the enormity of his error and perhaps at some time later in his career he might have. For he used this model to extrapolate the durations for which different running speeds could be maintained by athletes with the universal V˙O2max of 4 L·min−1 and an estimated "oxygen debt" capacity of 10-20 L. Whereas the model produced reasonable results for ordinary athletes running short distances (35), its predictions become progressively more inaccurate when applied to longer distances or to the world's best runners.

For example, his model forced Hill to predict that no athlete could sustain a running speed of 348 yards·min−1(19.2 km·h−1) for much more than 10 min (31, p. 239). Yet another of his papers (29) showed that the world records for middle and long distance running events were held by athletes able to run faster and for longer than the limits imposed by Hill's model. For example, that model predicts that when running 10,000 m in 30 min, the then world record holder would have consumed 330 L of oxygen (Fig. 2) of which a maximum of 120 L (30 min at a V˙O2max of 4 L·min−1) would have come from oxygen consumed during the event. This would leave an oxygen debt of 210 L as against their estimated maximum human oxygen debt of 10-20 L. Thus the model could not account for 200 L of oxygen, 60% of the total. The reptilian form of energy metabolism, surmised by Hill and his colleagues, was clearly unable to explain the performances of world-class human athletes of the 1920s.

This revelation could explain why Hill and his colleagues did not attempt to predict running speeds for races longer than 2 miles (35, p. 159) and why they did not discover the accepted physiological principle that athletes maintain different%V˙O2max at different racing distances (66,67). Rather they had to predict that speeds faster than 13.2 km·h−1 always required 100%V˙O2max(Fig. 2).

Thermodynamic and metabolic considerations also indicate that, according to Hill's model, no world record holder could ever finish any running race lasting more than a few minutes. Standard calculations reveal that the body temperature of an athlete running at 18 km·h−1 would rise by 1°C every 2.8 min if the V˙O2 were 94 mL 02 kg−1·min−1 at that running pace. The rate of muscle glycogen use at that V˙O2 would have to be 46 g·min−1. Thus after 15 min of exercise, the athlete's core temperatures would have reached 42.5°C and his total body carbohydrate stores would have been depleted. The point is that these calculations could have alerted Hill to the magnitude of his error, particularly as he was conversant with the thermodynamics of muscle contraction (30,33,34). But Hill seems to have been more interested in formulating models than in verifying, or better refuting, their predictions (25). Perhaps the identification of these errors might have forced Hill to question his cardiovascular/anaerobic model of exercise physiology and human performance.

Figure 2 identifies a second discrepancy in the logic of Hill and his colleagues. For they had previously concluded that Hill's oxygen consumption equaled the oxygen demand at all running speeds below 16 km·h−1 (Figs. 1 in (68, 69) plus accompanying text). Indeed in Figure 1, Hill and his colleagues actually measured a linear increase in measured V˙O2 when Hill ran at speeds up to 17 km·h−1. Yet in their conceptualFigure 2, the oxygen demand exceeds supply at all running speeds above 13.2 km·h−1 so that at 16 km·h−1, the predicted "anaerobic" contribution contributes close to 50% of the expected V˙O2. These discrepancies result from the inconsistent logic used by Hill and his colleagues to identify the onset of "anaerobiosis" in their different studies.

Other factors that contributed to Hill and his colleagues' erroneous belief in a universal V˙O2max of 4 L·min−1 resulted from the small size of the track on which their testing was performed, as well as the equipment they used. As reported previously(68,69), Hill acknowledged that their testing methods limited the peak running speeds they could study: "Greater speeds were not comfortable on our small track" ((38), p. 157); "We have been unable hitherto to continue the observations on H at the higher speeds, owing to the smallness of the track making faster running on it impossible" ((35), p. 159). Notable in their apparatus used to measure V˙O2 during exercise(Fig. 4 in (37)) is the small inlet aperture of the respiratory valve. The resistance to flow in that valve might have contributed to the feelings of breathlessness reported by Hill when he ran at 16 km·h−1 and which he erroneously considered to indicate that he had reached his V˙O2 plateau (Fig. 3 in (69)). Finally and facetiously, Hill may personally have believed in this universal maximum because his was the highest V˙O2max that he and his colleagues ever reported (4172 mL·min−1; 57 mL·kg−1·min−1) (35).

Figure 4
Figure 4:
Small (∼ 3%) changes in running economy induced by prior treadmill training could explain an apparent plateau phenomenon. The development of on-line, real time measurements of V˙O2 led to the introduction of progressive exercise testing which appears to produce the "plateau phenomenon" less frequently than does interrupted testing conducted on successive days (16,95). References: 16. Duncan, G. E., E. T. Howley, and B. N. Johnson. Applicability of V˙O2max criteria: discontinuous versus continuous protocols. Med. Sci. Sports Exerc. 29:273-278, 1997; 95. Taylor, C. Studies in exercise physiology. Am. J. Physiol. 135:27-42, 1941.

In summary, Hill and his colleagues did not describe the "plateau phenomenon" as currently understood. Rather they believed there to be a universal V˙O2max of 4 L·min−1. They drew this erroneous deduction because their original intellectual model was based on the belief that skeletal muscle anaerobiosis develops during exercise (Fig. 2 in (69)). As a result they concluded that an "oxygen debt" develops during exercise. Measurement of that"oxygen debt" led to the erroneous conclusion that V˙O2 rises exponentially with increasing running speed. Accordingly they found no logical reason ever to test any athlete running faster than 17 km·h−1. For at that speed the measured oxygen intake comprised only 66% of the erroneously predicted total oxygen requirement (Fig. 2). Hill was the main subject for these studies. As Hill weighed 73 kg and as his V˙O2 at 17 km·h−1 was 57.7 mL O2·kg−1·min−1(Fig. 1), so the conclusion was drawn circuitously that there is a universal V˙O2max of 4 L·min−1 (73 kg × 57.7 mL O2·kg−1·min−1).

The true relevance of the study of Myers et al.(60). In this response to Bassett and Howley, I provide another interpretation of the studies of Myers et al. (57,59,60). The interpretation of that study by Bassett and Howley was the following: "This was a rather unique and intricate approach to analyzing data, which had not previously been used. If the slope of the change in V˙O2were equal to zero, it meant a perfectly linear increase in V˙O2over time. A positive value meant that the rate of increase in V˙O2 was less than it had been previously, though V˙O2 could still be climbing. During submaximal exercise the slope of the change in V˙O2 varied about zero, suggesting a fairly linear increase in V˙O2.... However, in the final 1-2 min of the ramp test, some subjects had values that differed significantly from zero, indicative of a leveling off in V˙O2. Myers et al. (60) noted that only 33% of their subjects (2 of 6) met the criteria for a plateau established by Taylor et al. (97). This failure to demonstrate a plateau in all subjects is consistent with previous studies using more conventional open-circuit spirometry techniques. The authors concluded that the plateau is not a reliable physiological marker for maximal effort in all subjects. This is precisely the reason that investigators have spent considerable effort in finding other objective physiological criteria for the determination of V˙O2max(41)."

However, Myers(57), the author of this research, provides the exact opposite explanation for a change in V˙O2 slope of zero: ".... If oxygen uptake were no longer increasing (while work increases continuously), the slope of the relationship between the two variables would not differ statistically from zero." Hence a zero slope indicates a plateau in oxygen consumption, not a "perfectly linear increase in V˙O2," as Bassett and Howley wish us to believe.

Correctly interpreted, the real value of the studies of Myers et al. (59,60) were that they have established for the first time that the "plateau phenomenon" occurs also during submaximal exercise. Myers (57) concludes:"It appears that the slope of the change in oxygen uptake throughout progressive exercise varies greatly despite a constant, consistent change in external work and the use of large, averaged samples. This degree of variability would appear to preclude the determination of a plateau by most definitions... Recent data from our laboratory(59,60) and others (68) suggest that the plateau concept has limitations for general application during standard exercise testing" ((57) pp. 102-103).

Thus I understand the work of Myers et al. (59,60) to show that the appropriate technology detects many "plateau phenomena" occurring randomly throughout progressive exercise to exhaustion. If the "plateau phenomenon" also occurs during submaximal exercise, it cannot indicate the onset of anaerobiosis as demanded by the cardiovascular/anaerobic model. Hence the really novel contribution of this work was to question further the physiological meaning of the "plateau phenomenon" and hence the essential foundation on which the cardiovascular/anaerobic model has been constructed. To suggest that this work simply rediscovers old knowledge that supports the cardiovascular/anaerobic model is not consistent with deductive reasoning.

The finding that the "plateau phenomenon" occurred randomly during both the submaximal and maximal segments of progressive ramp exercise identified another issue recently addressed by Duncan et al. (16). Interrupted testing of the kind used by Taylor et al. (97) and all the other "classical" workers (1,2,53,55,79,80,84,97,103) may be more likely to identify the "plateau phenomenon" than the progressive exercise protocols of the type used by Myers et al. (59,60) and others (14) which became popular when on-line, real-time measurements of V˙O2 first became feasible. As early as 1940, Taylor (95) noted: "The trend of oxygen consumption at maximal levels is of great significance because of the prevailing view that the ability to absorb oxygen is a limiting factor in individual physical performance. As summarized in Figure 3, oxygen consumption is by no means always deficient at exhaustion levels. In fact, in 50% of cases no deviation in the linear increase of oxygen intake occurs at exhaustion and in the remaining cases this value may accelerate more often than fall off."

One obvious difference between progressive and interrupted testing is that the latter occurs over a few days, as many as five in the original study of Taylor et al. (97). This invites the question: Is the "plateau phenomenon" an artifact, resulting from rapid physiological adaptations that develop in response to three to five exercise bouts on successive days? For example, rapid but quite small (∼3%) adaptations in running economy in response especially to uphill treadmill running or walking on the previous test days could produce an artifactual "plateau" on the final test day when the highest work rate was achieved (Fig. 4).

Interestingly, Taylor et al. (97) acknowledged that changes in running economy did occur and might impact on the maximum oxygen consumption so measured: "Men have been able to complete the (same) work task in the presence of a 17% loss of maximal oxygen intake. It should be kept in mind, however, that there is good reason to believe that repeated bouts of running on the treadmill result in improved skill for performing the task and reduced oxygen cost(author's emphasis). If the workload chosen is close to workloads which fail to elicit a maximal oxygen intake in a given subject, then a substantial improvement in skill in running on the treadmill could result in a reduction of the apparent maximal oxygen intake ... Such a situation could result in a lower oxygen intake which was not the result of any loss of performance by the cardiorespiratory system" (p. 77).

Thus, while Taylor et al. (97) concerned themselves with the effects of long-term changes in running economy that would apparently lower the V˙O2max measured at a previously determined maximal workload, they did not consider the effects of the previous exercise bouts on running economy on the final test day and hence the possibility that this might produce an artifactual "plateau phenomenon" (Fig. 4). But this remains a possibility that needs to be refuted.

Others have noted the converse difficulty, that of determining the "plateau phenomenon" when there are clear and progressive improvements in running economy, resulting from such interrupted testing on successive days. For example, Seeherman et al. (88) observed that the treadmill running economy of animals improved for 2-6 wk. Only thereafter could the "plateau phenomenon" be identified: "An average of 6 to 8 wk of daily treadmill exercise was required for each determination of V˙O2max"(p. 22). If limiting oxygen consumption always terminates maximal exercise, it is surprising that such a protracted process is required to establish that relationship. Finally, although Bassett and Howley (4) vigorously defend the existence of the "plateau phenomenon," a previous publication by Duncan et al. (16) concluded that: "... a V˙O2 plateau is not a prerequisite for defining V˙O2max and is of limited use as a primary objective criterion for evaluating the quality of a graded exercise test." One is left to question what really is the significance of the "plateau phenomenon," if any.


As they now admit that the "plateau phenomenon" is an occasional feature of maximal exercise (and also of submaximal exercise according to the findings of Myers et al. (59,60)) and hence of"limited" value, Bassett and Howley must present an alternate argument to sustain the cardiovascular/anaerobic model of exercise physiology and athletic performance. They thus argue that the "plateau phenomenon" is not critical to their model as "there is a tremendous body of evidence suggesting (author's emphasis) that oxygen transport is the limiting factor for V˙O2max" so that any contrary argument, more particularly that proposed by myself, is necessarily "flawed."

Accordingly, they begin their defense of this section with what has become the standard textbook statement: "Physiologists have closely examined a number of factors that may limit V˙O2max, including: 1) pulmonary diffusion capacity for oxygen, 2) maximal CO, 3) the peripheral circulation, and 4) the metabolic capacity of skeletal muscle.... Today most physiologists believe that the capacity of the central cardiovascular system to transport oxygen to the tissues is the principal determinant of V˙O2max(82). The evidence for this view is not based on a plateau in oxygen intake, as Noakes asserts, but on a large number of scientific experiments that will be discussed."

But is it really true that the "plateau phenomenon" is irrelevant to the belief that limitations in oxygen transport determine the V˙O2max?

My readings suggests that Taylor et al. (97) were the first to interpret the "plateau phenomenon" as understood and described in the majority of current exercise physiology textbooks: "The classic work of Hill (30) has demonstrated that there is an upper limit to the capacity of the combined respiratory and cardiovascular system to transport oxygen to the muscles. There is a linear relationship between oxygen intake and workload until the maximum oxygen intake is reached. Further increases in workload beyond this point merely result in an increase in oxygen debt and a shortening of the time in which the work can be performed" (p. 78). But Figure 2 shows that Hill actually believed the relationship between running speed and V˙02 to be exponential so that the universal V˙O2max of 4 L·min−1 was reached at a running speed of 13.2 km·h−1 although, as shown in Figure 1, Hill reached this universal V˙O2max at 17 km·h−1. Perhaps Taylor (97) are responsible for establishing the historic myth that Hill and his colleagues both searched for and identified a "plateau phenomenon." Taylor et al. (97) also concluded that the existence of the "plateau phenomenon" proves that the cardiovascular system limits oxygen delivery to the periphery and that skeletal muscle anaerobiosis limits maximal exercise performance.

The next substantive paper of Mitchell et al. (55) begins with the statement: "According to Hill, maximal oxygen intake is reached when oxygen intake per unit time has attained... its maximum and remains constant... owing to the limitation of the circulatory and respiratory systems." In their subsequent classic review, Mitchell and Blomqvist (53) extend this interpretation: "If a person is subjected to progressively increasing workloads, there is a linear relationship between workload and oxygen uptake until the maximal oxygen uptake is reached. Heavier workloads can usually be achieved, but oxygen uptake levels off or may even decline .... Maximal oxygen uptake is the greatest amount of oxygen a person can take in during physical work and is a measure of his maximal capacity to transport oxygen to the tissues of the body. It is an index of maximal cardiovascular function.... and, therefore is valuable in the evaluation of abnormal cardiovascular function." Clearly Mitchell and Blomqvist (53) also use the existence of the plateau phenomenon to conclude: 1) that exercise capacity is limited by factors relating exclusively to oxygen use and 2) that cardiovascular function limits the V˙O2max.

Other influential physiologists, Wyndham et al. (103) began their study of the "plateau phenomenon" with the following statement: "In a classic study of the physiology of exercise, Hill and Lupton (35) noted that: 1) each individual has a maximum level of oxygen uptake per minute, 2) the level varies from one individual to another, and 3) the extra work done above the maximum level of O2 intake is by means of "anaerobic" metabolism, i.e., the oxygen requirement comprises O2 intake plus an "O2 debt."

Interestingly, these authors had great difficulty identifying the plateau phenomenon with repeated testing in their subjects. When they applied curve fitting equations to identify the "plateau," they were consistently left with calculated V˙O2max values that were lower than the maximum values they had actually measured, clearly an impossible conclusion. Harris (25) has warned of the need to ensure that the facts fit the theory. The "plateau" exercise of Wyndham et al. (103) is a clear example of forcing the facts to fit a preconceived theory. Wyndham et al. (103) also concluded that the methods of neither Astrand (1) nor of Taylor et al.(97) could establish "the level of maximum oxygen uptake." Perhaps at the same time they might have asked whether they were studying something that does not exist.

As the authors also refer to the work of Rowell, it is appropriate also to review his opinions (82,83): "Hill and Lupton(1923) showed that oxygen uptake does not continuously increase with increasing work intensity until intolerable levels of effort are reached. Rather a plateau of oxygen uptake is eventually reached well before the effort becomes intolerable. Hill and his colleagues went on to show that individuals could exercise for brief periods at levels exceeding those required to elicit this plateau, that is, beyond V˙O2max. Our ability to define the functional limits of the cardiovascular system has provided a powerful scientific base from which to investigate many aspects of regulation. This functional limit occurs at the maximal oxygen uptake (V˙O2max) which, by definition, equals the product of maximal values of heart rate (HR), stroke volume, and arteriovenous oxygen difference (Fick principle)" ((82), p. 213). Finally Rowell concludes: "In summary, the capacity to increase oxygen uptake during exercise is limited in normal humans by the ability to raise the CO" (82, p. 246).

Thus Rowell uses the historically erroneous interpretation of Hill and his colleagues' work also to conclude that: 1) oxygen delivery limits maximal exercise because of a limited CO and 2) the V˙O2max defines the functional limits of the cardiovascular system.

Similarly in a recent review, Wagner (99) writes: "The fundamental point is that a maximal rate of O2 utilization can in fact be shown experimentally to exist in normal skeletal muscle.... Thus, in intact mammals at very high exercise levels, it can generally be shown that even if external power is increased, there is no significant further rise in V˙O2(2,30,88). The positive linear relationship between external power and V˙O2 characteristic of submaximal exercise flattens out (or asymptotes) to define V˙O2max...."

Thus Wagner argues that the "plateau phenomenon" proves that there is a maximal rate of oxygen consumption by muscle. Interestingly, his commitment to the "plateau phenomenon" is not entirely unconditional for he includes the enigmatic observation: "Even limiting oneself to a discussion of V˙O2max does not preclude confusion because V˙O2max is a somewhat elusive variable both conceptually and experimentally." Clearly something has happened more recently to produce doubts about the certainty of the "plateau phenomenon," at least in Wagner's mind. Yet he does not allow these doubts to undermine his certainty that an oxygen limitation can be identified during maximal exercise in humans. This begs the question: What additional direct (not inferential) information is available to sustain his belief in the face of such uncertainty?

Bassett and Howley's attempt to discount the pivotal importance of the plateau phenomenon as the foundation for the cardiovascular/anaerobic model is disingenuous in the extreme. Indeed the plateau phenomenon probably qualifies as the single most influential concept in modern exercise physiology. As shown by the extremely influential opinions of Taylor et al. (97), Rowell (82,83), Mitchell and Blomqvist (53), Wyndham et al. (103), and Wagner (99), it is the singular physiological observation that underpins the cardiovascular/anaerobic model of exercise physiology and its predictions. Indeed, had Hill and his colleagues not advocated the plateau phenomenon, there would have been no logical basis for the brace of studies quoted by Bassett and Howley as "proof" of the cardiovascular/anaerobic model of exercise physiology and athletic performance. For the simple reason that these is still no published evidence proving that muscles become anaerobic during exercise and that this anaerobiosis then limits maximal exercise performance. Were such direct evidence available, it would have been quoted extensively by these authors in preference to the indirect evidence provided by the "plateau phenomenon."

Hence there is no direct proof for the cardiovascular/anaerobic model of exercise physiology and athletic performance. Nor, indeed, as Mitchell et al. (55) warned four decades ago:"The view that cardiac capacity is the determinant of maximal oxygen intake is surmise, not established fact."

The studies reviewed by Bassett and Howley in this section of their response have produced results that are compatible with the cardiovascular/anaerobic model but which do not prove its truth. For none excludes the possibility that exercise performance and V˙O2max might be limited by other factors that are also influenced by interventions that improve exercise performance or the V˙O2max, seemingly by increasing exclusively oxygen delivery to the exercising muscles. This represents the classic scientific dilemma of proving causation and excluding association between apparently related phenomena. Unequivocal proof of the cardiovascular/anaerobic model requires data showing that the oxidative capacity of the mitochondria in the maximally exercising muscles is limited by an inadequate oxygen supply caused by a failure of the pumping capacity of the heart.

The cardinal point is that, without the "plateau phenomenon," the cardiovascular/anaerobic model has no greater claim to be the sole and authentic explanation of exercise physiology and athletic performance than does any other competing model. But such is the primacy enjoyed by the plateau phenomenon and the cardiovascular/anaerobic model of exercise physiology and athletic performance (1,2,30-32,35-38,41,53-55,82-85,97,99,101,103), that all other competing models have been rigorously ignored, perhaps even suppressed. The J. B. Wolffe lecture tried to encourage a broader, more inclusive perspective.

For example, the obvious point ignored by many is that muscles do not work by oxygen alone; they contract in response to activation of central and peripheral neural pathways with coupling of excitation to contraction by the complex processes leading to calcium release from the sarcoplasmic reticulum and its binding by the thin filament proteins. Nor is it prudent to ignore the possibility that the activity of the thick and thin filament proteins might contribute to skeletal muscle contractile activity during exercise, as occurs in the heart (43). The question that has to be asked is: Why have so few modern exercise physiologists championed these possibilities, preferring rather to accept the cardiovascular/anaerobic model as the sole reasonable explanation for human exercise physiology and athletic performance?

The only explanation I can offer is that most exercise physiologists have ignored any other possibility simply because of the oppressive hegemony of the cardiovascular/anaerobic model and the unusual vigor of its proponents in protecting their particular model. Consider Bassett and Howley's certitude: "(Noakes') view assumes that the cross-bridge cycling rate is the limiting factor in distance running. However, cross-bridge cycling cannot occur without adequate quantities of ATP supplied by aerobic metabolism. The majority of the evidence points to the cardiorespiratory system as being the limiting factor for oxygen uptake (which is the quantitative measure of aerobic metabolism)". Thus the only "evidence" these authors can provide to "prove" that cross-bridge cycling does not limit maximal exercise performance is not a body of published findings disproving a specific conjecture, for example, that myosin ATPase activity is related to exercise performance, but rather an uncritical devotion to an unproven model that, as I have now shown for the third time, is based on a mythical foundation (68,69).

Thus the J. B. Wolffe lecture included a series of examples in which there is clear evidence that maximal exercise is definitely not limited by skeletal muscle anaerobiosis and hence by a failure of the cardiac pumping capacity. These examples include exercise at extreme altitude (40,44), in the heat (63-65), and in persons with renal (45) and cardiac disease (102), especially following cardiac transplantation (15). In none of these conditions does the defect appear to be in either oxygen delivery to or oxygen utilization by the exercising muscles. Rather the primary defect is either in central recruitment of muscle contraction (definitely in exercise at moderate to extreme altitude (44) and possibly in other conditions) or because of impaired skeletal muscle contractile function resulting from changes either in excitation/contraction coupling or in the function of the thick or thin filament proteins. The crucial distinction is that the defect in these conditions is not in oxygen provision and energy production, but in the processes of skeletal muscle recruitment and contraction.

That many do not appreciate this crucial distinction is apparent in the traditional explanation that the poor exercise tolerance of patients with metabolic muscle diseases results from impaired oxygen utilization by their diseased muscles. Rather the primary defect, as suggested by the "lactate paradox" (69) is likely an inability of their diseased muscles to generate sufficient force for there to be a normal oxygen requirement. Thus the low V˙O2max in these patients is the result, not the cause of their poor exercise tolerance, as also found in patients with renal(45) or cardiac disease (15,101). Abnormalities in skeletal muscle force generation, perhaps resulting also from a reduced total muscle mass, and not in oxygen delivery or utilization, explain the poor exercise tolerance of these patients.

Despite these arguments, there is indeed good reason to believe that the V˙O2max must be regulated by the central cardiovascular function in healthy individuals. I contend that in contrast to the conventional arguments, the proof of this hypothesis requires the absence of a plateau in CO and in V˙O2. Furthermore if cardiovascular function regulates or limits the V˙O2max, then skeletal muscle anaerobiosis cannot develop during maximal exercise. The questions to be addressed are: 1) Does the pumping capacity of the heart limit the V˙O2max? and, if so, 2) what are the consequences for the heart?

Bassett and Howley(4) present another common conclusion in this section: "However, in the vast majority of healthy subjects the evidence points to the pumping capacity of the heart (CO) as being the major limiting factor for V˙O2max." Other highly influential exercise physiologists have concluded similarly (6,82,83,85). But the pumping capacity of the heart is seldom directly measured in healthy subjects during maximum exercise. Hence this conclusion must be based not on directly measured evidence but on inferences drawn from other information, perhaps from the measurement of the V˙O2 and from the (theoretical) existence of the "plateau phenomenon." Figure 5 has been redrawn from the classic paper of Mitchell and Blomqvist (53). It shows the clearly defined plateau in oxygen consumption in this subject when the work rate exceeds stage 4 of the test corresponding, interestingly, to Hill's universal V˙O2max of 4 L·min−1. During the submaximal portions of the exercise test (stages 1 to 4), increases in work rate cause essentially linear increases in CO, stroke volume, and HR. When the V˙O2 plateaus at stages 4 and 5, these variables must also level off, hence proving that the pumping capacity of the heart limits the V˙O2max.

Figure 5
Figure 5:
The classical diagram of Mitchell and Blomqvist (53) depicting the expected changes in V˙O2, stroke volume (SV), HR, and CO during progressive exercise to exhaustion. Note that whereas the "plateau" in V˙O2 is clearly identified, no plateau in SV, HR, and CO is shown as data for these cardiovascular variables were not included at workloads beyond the V˙O2max. Reference 53. Mitchell, J. H. and G. Blomqvist. Maximal oxygen uptake. New Engl. J. Med. 284:1018-1022, 1971.

Surprisingly, Mitchell and Blomqvist (53) do not provide data for these cardiovascular measures once the V˙O2max has been reached (Fig. 5). A more recent review by Mitchell of that paper (54) also lacks such data. Perhaps as serious scientists they were correctly reluctant to speculate on what they had not measured. Or perhaps they realized the dilemma that such a conclusion posed, especially for the heart.

For the issue that must ultimately be addressed is: What are the consequences of the cardiovascular/anaerobic model for the heart itself? Or, alternatively, if the CO limits the V˙O2max, what limits the CO? For the heart is also a muscle. The cardiovascular/anaerobic model predicts that (skeletal) muscle function fails when its oxygen supply is inadequate. Hence if logic is to be preserved, the failure of the heart's pumping capacity, in Rowell's words "to raise the CO" at the V˙O2max must result from an inadequate (myocardial) oxygen supply. Hence coronary blood flow must plateau sometime before the CO levels off at the V˙O2max. This limiting coronary blood flow induces myocardial "fatigue," causing the plateau in CO and hence in the V˙O2max, leading only thereafter to skeletal muscle anaerobiosis. Thus by this logic, the coronary blood flow must be the first physiological function to show a "plateau phenomenon" during progressive exercise to exhaustion (Fig. 6).

Figure 6
Figure 6:
Diagram showing the expected cardiovascular changes that produce the "plateau phenomenon," if the cardiovascular/anaerobic model is correct. Note that a plateau in coronary blood flow (CF) causing myocardial ischemia must occur before the CO levels off, and well before skeletal muscle anaerobiosis develops at the plateau in V˙O2. Furthermore, continuing exercise after the plateau in CF and V˙O2 must cause a precipitous fall in CO as a result of progressively increasing myocardial ischemia.

Hill et al. (36) anticipated this dilemma: "Certain it is that the capacity of the body for muscular exercise depends largely, if not mainly, on the capacity and output of the heart. It would obviously be very dangerous for that organ to be able, as the skeletal muscle is able, to exhaust itself very completely and rapidly, to take exercise far in excess of its capacity for recovery... The enormous output of the heart of an able-bodied man, maintained for considerable periods during vigorous exercise, requires a large contemporary supply of oxygen to meet the demand for energy... When the oxygen supply becomes inadequate, it is probable that the heart rapidly begins to diminish its output, so avoiding exhaustion; the evidence for this, however, is indirect, and an important field of research lies open in the study of the recovery process in heart muscle, on the lines of which it has been developed in skeletal muscle" (p. 443). As a result: "It would seem possible that a deciding factor in the capacity of a man for severe prolonged exercise may often be the efficiency of his coronary circulation" ((30), p. 108).

Hence the interpretation of Hill and his colleagues is unambiguous and internally consistent. Their model predicts that the first event leading to exhaustion during maximal exercise is an inadequate oxygen supply to the heart. This impairs myocardial function causing a leveling of the CO, a reduced skeletal muscle blood flow, and ultimately skeletal muscle fatigue as depicted in Figure 6. They postulate that the heart must reduce its CO specifically to prevent the development of a large "oxygen debt."

But even this explanation is unsatisfactory as it still fails to explain what physiological events terminate maximal exercise. For it is now known that the heart has no capacity to generate an "oxygen debt." Rather, an inadequate myocardial oxygen supply at the V˙O2max would induce the symptoms of angina pectoris(43), which would terminate maximal exercise. But it is now also known that this does not occur, at least in healthy humans without coronary artery disease. Thus all the evidence shows that none of the markers of myocardial ischemia, including electrocardiographic ST-segment displacement (77), myocardial release of lactate (27), and increases in both the end-diastolic and end-systolic ventricular volumes with a reduction in stroke volume(82,83), has been measured in healthy athletes at maximum effort. Rather myocardial lactate uptake occurs across the myocardium at those high exercise intensities (21,27) which induce the highest rates of lactate release from the exercising muscles.

No modern physiologist has been sufficiently brave to tackle this intellectual dilemma inherited from Hill and his colleagues. For it is not appealing to suggest that myocardial ischemia limits maximal exercise in healthy individuals. Furthermore, all the published evidence refutes these logical predictions of the cardiovascular/anaerobic model.

Conclusions. There is an obvious discrepancy in the logic presented in the previous two sections by Bassett and Howley. For they initially argue with great conviction that the "plateau phenomenon" was indeed identified by Hill and his colleagues. Next they argue the opposite; that the "plateau phenomenon" is not universally present and can be identified in a variable proportion of subjects as shown in their own studies (16). Although this is an important admission, it is hardly novel. Taylor et al. (97) who were the first researchers after Hill and his colleagues specifically to analyze the "plateau phenomenon" wrote that "there are conditions under which a maximal oxygen uptake cannot be elicited"(p. 143). One wonders why it has taken so long for this admission to be forthcoming from some devotees of the cardiovascular/anaerobic model. Perhaps it is because this admission unearths another logical nightmare: If a subject terminates exercise without demonstrating a "plateau phenomenon," does this indicate that skeletal muscle anaerobiosis limited their exercise, according to the cardiovascular/anaerobic model? Proponents of the cardiovascular/anaerobic model must either respond in the affirmative or abandon their model. Yet in confirming their model, they commit the logical error revealed in Figure 3.

Statements 1-4 form the logical basis of the cardiovascular/anaerobic model of exercise physiology. Yet if statement 1 is true, then so too must be its converse, that is, that the absence of the plateau phenomenon indicates that skeletal muscle is adequately perfused and oxygenated during maximal exercise (statement 2).

Hence the finding (statement 5) that the plateau phenomenon is present in as few as 50% of subjects during maximal exercise (1,4,14,16,41,55,57,59,60,95-97,99,103) leads to the logical conclusion (statement 7) that factors other than skeletal muscle anaerobiosis must limit exercise in a sizable proportion of athletes or indeed in all athletes if one accepts the findings of Myers et al. (59,60). This was the essential conclusion of the J. B. Wolffe lecture which developed what might be termed "the regulated skeletal muscle contraction model of exercise physiology and athletic performance."

Denying this conclusion introduces the logical conflict between statements 2 and 6. But as I have argued earlier, it is the very absence of the plateau phenomenon that more likely proves that central cardiovascular function regulates or "limits" the V˙O2max.


In this section, Bassett and Howley set out to show that physiological variables relating to oxygen transport also predict endurance performance. They wish to prove that oxygen delivery determines endurance performance as, in their opinion, it also determines maximal exercise performance.

I would argue that it is essential to undertake research in the scientific area in which one professes expertise. Perhaps it is bold of me to suggest that the quotation of Thomas Huxley (42) is particularly appropriate to their presentation:"Science is organized common sense where many a beautiful theory was killed by an ugly fact." This section of Bassett and Howley's response produced a beautiful theory despite a host of ugly facts that disprove it. For example, the conceptual error that a statistical relationship proves causation was used as justification of their position. Furthermore, in proving statistical relationships it appeared that they selectively ignored a large body of evidence showing the opposite, namely, that these variables are either unrelated or only poorly related to endurance performance. For example, the studies Bassett and Howley quoted to "prove" that V˙O2max and running economy predict racing performance included athletes whose performances differed by 18 min (38%) in a 16-km race (Fig. 10 in (4)) and by 9 min (20%) in a 10-km race (Fig. 12 in (4)). But in studies of athletes whose performances are more similar, neither V˙O2max(9,10,12,13,71,74,75,87) nor running economy (71) was a good predictor of running performance. Even in their Figure 12, the running economy of the four fastest 10-km runners was identical, yet their performances differed by 4 min (9%). By their logic, this absence of statistical association must therefore disprove the theory that oxygen delivery determines endurance performance in elite athletes. In athletes with similar abilities the best predictor of performance is usually a measure of the peak workload achieved during exercise(19,56,68,71,86,87). Indeed, the study of Morgan et al. (56) showed a significant negative correlation between V˙O2max and 10-km racing speed in a group of runners with similar best 10-km times.

The second disagreement between Bassett and Howley and myself in this section is the absence of a valid conceptual framework to explain how oxygen delivery determines endurance performance. Bassett and Howley believe a link exists among oxygen delivery, exercise intensity, and changes in blood lactate concentrations:"However, what was needed was a single measurement that would incorporate all three variables and simplify the prediction of performance. The answer was provided by the classic study of Farrell et al. (18) dealing with the link between lactate accumulation and performance." In fact, that "classic" study (18) suffered from the very limitation, already identified, since it also concerned athletes whose performances varied substantially. The range of finishing times for 42 km in the 13 marathon runners in that study varied from 2 h 17 min to 3 h 49 min, a variation of 67%. Similarly the range of V˙O2max values in the group ranged from 36-59 mL O2·kg−1·min−1. When athletes of similar abilities are studied, the lactate threshold, however defined, and indeed any measurement derived from the blood lactate concentration, is of "negligible" value in predicting athletic performance (39).

But even if the blood lactate concentration is always an excellent predictor of endurance performance, this does not "prove" that oxygen delivery determines endurance performance. For the cardiovascular/anaerobic model provides the only conceptual link between oxygen consumption, exercise intensity and blood lactate concentration. Thus their Figure 13 provides another example of how Bassett and Howley interpret data exclusively according to the cardiovascular/anaerobic model, specifically to "prove" that model. The authors do not explain how they interpret their Figure 13. Presumably they would argue that the greater the running economy and the higher the V˙O2max, the faster the running speed at which "anaerobiosis" first develops and, hence, the greater the running speed that can be achieved before the blood lactate concentration rises. The rise in blood lactate concentration then determines endurance performance in an unspecified way.

However, this interpretation ignores the substantive section in the J. B. Wolffe lecture which dissociates skeletal muscle lactate production from skeletal muscle oxygen delivery (7,8,20,46-48,50,58,91). The onset of blood lactate accumulation is more likely related to the rate of carbohydrate oxidation (7,14,48) than the V˙O2, as required by the cardiovascular/anaerobic model. Thus, there is no longer any logical reason to interpret blood lactate concentrations as an index of skeletal muscle anaerobiosis according to the cardiovascular/anaerobic model. As already argued, it was this error that led Hill to draw all his erroneous conclusions. Modern exercise physiologists should not continue to repeat his errors if they wish to advance knowledge (25).

Bassett and Howley concluded this section with a critique of my view that a factor related to running speed over a shorter (sprint) distance may also predict performance at much longer distances of 5-50 km in groups of athletes who specialize in those long distances. I did not conclude that this means that sprinters would be the best distance runners. Rather I concluded that the best distance runners must have muscles that are both very powerful (and hence can generate a fast sprinting speed) and have superior fatigue resistance. The accuracy of this prediction is shown by our more recent work (9) which shows that the best distance runners (5-42 km) yet studied are almost as fast as superior milers at distances of 1-2 km. But the muscles of the distance runners have substantially greater fatigue resistance.


I have argued that proponents of the cardiovascular/anaerobic model face key logical dilemmas when they propose factors that limit maximal exercise performance (Fig. 3). Not least is the logical inference that if CO limits the V˙O2max, then the limitation in CO must result from myocardial ischemia. But their challenge becomes even more daunting when the same model is used to explain endurance performance, as do Bassett and Howley: "In contrast to Noakes' assertion, the research literature provides considerable support for the hypothesis that performance in endurance events is limited by oxygen delivery (mL·kg−1·min−1), which is set by the subject's V˙O2max and the percent of V˙O2max that can be maintained."

First, why should prolonged endurance exercise in which the oxygen consumption is not maximal and therefore not limiting be determined by oxygen delivery to the active muscles? Besides their argument provided in the previous section, Bassett and Howley also quote this statement of Hill et al. (38) to support their belief: "Some individuals can naturally run or walk uphill for long periods without distress. This may be partly a matter of diffusion constant of the lungs for oxygen; largely, however, it is probably one of the capacity of the heart itself." But this interpretation invites the question: If the ability to exercise without fatigue during prolonged exercise results from the "capacity of the heart itself," then it follows that the onset of fatigue during prolonged exercise must be caused by heart fatigue, that is, by the onset of cardiac failure with a reduced capacity to pump blood to the active muscles. Hence the cardiovascular/anaerobic model can only explain endurance performance if the heart itself also limits that performance. Yet there is no logical physiological explanation for this proposed dependence. Rather one is forced to postulate that the progressive development of cardiac failure during prolonged exercise reduces skeletal muscle blood flow inducing anaerobiosis and skeletal muscle fatigue. But the development of cardiac failure during prolonged exercise is the exception, not the rule (52), and produces a unique set of symptoms and signs quite unlike those present in normally fatigued athletes. Furthermore, blood lactate concentrations are low at exhaustion following very prolonged exercise (51), a further example of the "lactate paradox" discussed previously. I would note that Bassett and Howley's proposition that changes in muscle H+ and glycogen concentrations affect exercise performance is extraneous to this discussion. These represent a different model of fatigue (70,72) separate from their cardiovascular/anaerobic model.

Second, the model fails to explain why exercise is impaired in the heat when the oxygen and blood supply and metabolism of the active muscles are normal(63-65) or at altitude when the "lactate paradox," as fully detailed in the J. B. Wolffe lecture, reveals itself as it does also in patients with various chronic diseases or in those exhausted by prolonged exercise. Third, this model cannot explain why treatment with dichloroacetate (DCA), a metabolic stimulator of the pyruvate dehydrogenase complex, increases V˙O2max and maximal exercise performance (49), reduces the rise in blood lactate concentrations during submaximal (49,98) and maximal exercise (49), and maintains higher muscle PCr concentrations during exercise (98). These studies indicate that metabolic events in skeletal muscle can alter exercise performance and metabolism independent of changes in skeletal muscle oxygenation.

Fourth, the model cannot explain why elite black South African distance runners, the fastest group of distance runners yet studied, performed significantly better at distances beyond 5 km than did a group of middle distance runners whose V˙O2max values were the same (9). The discrepancy in performance increased with increasing race distance. Nor can this model explain why women with either lower measured V˙O2max values (89) or who are slower over racing distances of 10-42 km (3) begin to outperform men in races longer than 56 km.

Fifth, this model cannot explain why the form of muscle weakness which develops during the course of prolonged exercise like marathon racing (61,62,76) persists for at least 7 d after exercise and is characterized by impaired skeletal muscle contractile activity (62) and altered electromyographic activity(92). It would seem probable that this weakness is in some way linked to the development and expression of fatigue during prolonged exercise. Its persistence after exercise proves that it is not caused by the transient metabolic changes occurring during exercise.

Similarly, this model cannot explain the personal observations (68,69), common to all athletes, that the inability to train vigorously every day or to increase one's pace near the finish of an endurance running event is associated with discomfort in the leg muscles. Measurement of lower than maximum HRs under both conditions suggests that the symptoms result from factors other than cardiac fatigue and are most likely related to the changes in skeletal muscle function described by Nicol et al. (61,62), Pullinen et al. (76), and Strojnick and Komi (92).

It was precisely because of these discrepancies that I concluded that the cardiovascular/anaerobic model cannot explain fatigue and performance under all conditions. Hence another model was necessary. Initially I argued that a muscle factor alone must predict exercise performance (66-68). That a localized muscle factor is involved in the form of fatigue studied by Spriet et al. (90) is clear. They showed that force production by ischemic skeletal muscle contracting in response to external stimulation falls progressively, ceasing before the development of irreversible ATP depletion and skeletal muscle rigor.

But my original model was not entirely satisfactory, not least because it ignored the role of central neural factors which limit exercise performance at altitude and explain that form of the"lactate paradox" (22,44). The same mechanism might also explain why exercise terminates in the heat even though skeletal muscle blood flow and metabolism are unaffected (63,64). It also seems unlikely that precooling enhances endurance performance exclusively by a peripheral, skeletal muscular effect. Hence by the time of the J. B. Wolffe lecture, the model had been refined substantially. The model that I have now proposed is one in which "skeletal muscle contractile activity is regulated by a series of central, predominantly neural, and peripheral, predominantly chemical, regulators that act to prevent the development of organ damage or even death during exercise in both health and disease and under demanding environmental conditions" ((69), p. 571). Unfortunately, this model was not acknowledged by Bassett and Howley.


Bassett and Howley list the contributions made by Hill. I contend that the majority of Hill's ideas are now antiquated as argued here. Nor am I the first so to argue.

In 1969, Harris (25) published a classic paper describing how the "oxygen debt" also attributed to Hill and his colleagues achieved a mythical status in the exercise sciences even though the facts consistently showed that, like phlogiston, it did not exist: "Excess lactate represents the last major attempt on the part of physiologists to fit the facts into the classical conception that 'oxygen debt' is due to the metabolism of lactate. Two generations have been mesmerized by this conception and each has failed to fit the facts to the theory." He continued: "Much of the present-day exercise physiology is still based on the misconceptions of the past; and the time has come to call a halt to the over-earnest examination of largely fortuitous relationships between the concentrations of lactate or pyruvate in the blood and the recovery volume of oxygen ... The theory of oxygen debt looks backwards from the premise that the muscle cell has to make good something which is lost during work, endowing the human cell with the all-to-human property of hindsight. We should now begin at the other end, start with the chemical disturbances produced by contraction and see how (not why) they cause the muscle cell and the body to increase its uptake of oxygen" (p. 389).

Perhaps Harris' sentiments about the "oxygen debt" apply equally to: 1) Hill and Lupton's 1923 concept of the universal V˙O2max, 2) the description by Taylor et al. (97) of a mythical "plateau phenomenon" that was neither looked for nor ever identified by Hill and his colleagues, 3) Wasserman and McIlroy's(101) description of the "anaerobic threshold"; and 4) the cardiovascular/anaerobic model of exercise physiology and athletic performance (69), that all these erroneous ideas have spawned. Perhaps, as Harris concludes, "we have been looking ... for something which is not there."

Indeed I continue to emphasize that it is the absence, not the presence of the plateau phenomenon, that is the stronger evidence that cardiovascular function regulates or "limits" the V˙O2max.

Conclusion: the common theory of Hill and Noakes. This more extended analysis of the work of Hill and his colleagues has discovered why the "plateau phenomenon," as currently understood, is implausible precisely because it predicts that exercise terminates when the oxygen demands of the exercising skeletal muscles exceed their supply.

For the inherent weakness in Hill's cardiovascular/anaerobic model is simply stated: The heart is a muscle, just as is skeletal muscle, dependent on an adequate oxygen supply for its contraction. But unlike skeletal muscle the heart is dependent on its own pumping capacity for its oxygen supply. Hence if the CO reaches a peak and can rise no further, the immediate cause will have been a plateau in (coronary) blood flow and the immediate effect will be a developing myocardial ischemia. The knock-on effect of this will be a further reduction in myocardial contractile function, a further fall in CO, and more severe myocardial ischemia. Thus, the inability of the heart "to raise its CO" at the V˙O2max(83) must be both the result and the cause of a developing myocardial ischemia, according to the model that holds that oxygen deficiency impairs muscle function.

Hence according to the cardiovascular/anaerobic model, the logical end point of vigorous exercise must be a progressive and irreversible myocardial ischemia with the development of angina pectoris. As this does not occur in healthy human athletes, some other mechanism must intervene to prevent the CO reaching a maximum while the oxygen demands of the tissues continue to rise.

Proponents of the cardiovascular/anaerobic model have not acknowledged that Hill had himself realized that his model was fatally flawed. So he introduced modifications, unrecognized until again revealed here: "It would seem probable ... that the heart is able to regulate its output, to some extent, in accordance with the degree of saturation of the arterial blood, either of that which reaches it through the coronary vessels or by some reflex in other organs produced by a deficient oxygen supply(author's italics). From the point of view of a well coordinated mechanism, some such arrangement is eminently desirable; it would clearly be useless for the heart to make an excessive effort if by so doing it merely produced a far lower degree of saturation of the arterial blood; and we suggest that, in the body(either in the heart muscle itself or in the nervous system), there is some mechanism which causes a slowing of the circulation as soon as a serious degree of unsaturation occurs, and vice versa. This mechanism would tend ... to act as a "governor," maintaining a high degree of saturation of the blood" ((38), pp. 161-162).

Hence Hill and his colleagues realized that even if the ultimate limits for maximal exercise are set by the cardiovascular system, some mechanism (other than skeletal muscle anaerobiosis) must be present to terminate exercise before the heart is itself damaged by the very plateau in CO that is theoretically necessary to explain the "plateau phenomenon" and the development of skeletal muscle anaerobiosis (Fig. 6). Thus they proposed the existence of a "governor" in either the myocardium or nervous system, specifically to terminate exercise before myocardial damage developed.

Contrast this to my model proposed in the J. B. Wolffe lecture (69):"The alternate model proposes that skeletal muscle contractile function is regulated by a hierarchy of controls specifically to prevent damage to any of a number of different organs. Severe anaerobiosis is one specific end point that must be thwarted so that irreversible rigor and necrosis in the active muscles is prevented. The challenge for exercise scientists is to understand how the body anticipates potential for organ damage and how skeletal muscle contractile function is regulated specifically to preclude any such calamities" (p. 587).

Thus the more logical analysis of Hill and his colleagues, ignored by all modern exercise scientists, has exposed the one critical weakness of my model. It is that skeletal muscle anaerobiosis can only ever occur after myocardial ischemia has developed (Fig. 6).

Perhaps the strengths of my model and that of Hill and his colleagues can now be combined into an even more contemporary hypothesis: Skeletal muscle recruitment and contractile function are regulated by a hierarchy of controls, specifically to prevent damage to any of a number of different organs. During maximal exercise, progressive myocardial ischemia preceding skeletal muscle anaerobiosis must be thwarted so that neither the heart nor the skeletal muscle develops irreversible rigor and necrosis with fatal consequences.


An important test of this combined model is provided by studies of the cardiovascular response to exercise at increasing altitude in which the progressive reduction in the arterial oxygen tension increases the probability that myocardial ischemia will develop during maximal exercise. If skeletal muscle function at altitude is limited by an inadequate oxygen delivery and is not regulated to prevent the development of myocardial ischemia, then the peak CO during maximum exercise must be the same or higher at altitude than at sea level. But if the heart is to be protected from myocardial ischemia during maximal exercise at increasing altitude, then the falling arterial oxygen content and hence the reduced potential for oxygen delivery to the myocardium must cause a progressive reduction in maximum CO with increasing altitude. Furthermore, the absence of myocardial ischemia during maximal exercise at extreme altitude when the arterial oxygen tension is barely sufficient to maintain cerebral function would be consistent with the presence of the governor proposed by Hill and his colleagues.

Table 1 lists peak cardiorespiratory measures at different altitudes measured during the epic Operation Everest II (24,40,78,81,93,94,100). As the arterial oxygen tension falls with increasing altitude, peak achieved work rate, HR, and, most significantly, CO fall substantially whereas maximum ventilation rises (94). No electrocardiographic evidence of myocardial ischemia was detected during maximum exercise at any altitude up to 8840 m (81,94). Indeed left ventricular systolic function was maintained and perhaps even enhanced during maximal exercise at a simulated altitude of 7620 m (78,93). Hence the authors concluded that "limitations of performance may have been caused by other factors, and reduced HRs and COs at maximum effort (at altitude) were the result rather than the cause of the reduced maximal O2 uptake".

But a more insightful observation is that at any work rate CO was the same at sea level and at any higher altitude up to the highest tested (8840m)(40,94).

Hence at altitude CO is not increased during exercise to compensate for the reduced arterial oxygen content. Thus the body makes no attempt to increase CO to optimize oxygen delivery to the tissues at increasing altitude. The sole conclusion must be that the oxygen demands of skeletal muscle are not pre-eminent during exercise at altitude. Rather, skeletal muscle recruitment must be regulated, perhaps by the governor first proposed by Hill, probably to protect the heart from an hypoxia that would be inevitable were the oxygen demands of the skeletal muscles, not the heart, the priority. This governor is presumably responsible for reduced skeletal muscle recruitment and the lower than expected blood and muscle lactate concentrations during maximum exercise at altitude (44).

If skeletal muscle function is indeed regulated to prevent myocardial hypoxia at altitude, the question becomes: what physiological variable is sensed by the governor? Logic suggests that the variable must be related to oxygen delivery to, or use by, the myocardium. Table 1 shows that the oxygen tension in the mixed venous blood draining from the systemic circulation including skeletal muscle and from the heart in the coronary sinus is the only variable that is the same at maximal exercise at all altitudes. Hence one possibility is that the venous oxygen tension in either mixed venous or coronary sinus blood is the variable that is sensed.

But since myocardial and not skeletal muscle oxygenation is at risk during exercise at altitude, it would be more likely that the oxygen tension in the coronary vascular bed would be the monitored variable.

Indeed the study of Grover et al. (23) showed that the coronary sinus PO2 is the same during submaximal exercise at all altitudes from sea level to 3100 m, suggesting that "the coronary circulation is autoregulated to maintain coronary sinus blood O2 tension constant" (23). This would mean that, at altitude, coronary flow must increase in proportion to the reduction in the arterial PO2 to maintain a constant coronary sinus oxygen tension.

Hence, the rate of either coronary flow or of myocardial oxygen delivery would provide the necessary input for the appropriate functioning of Hill's governor.

In summary, the evidence from these altitude studies shows that the heart is prevented from developing myocardial ischemia during maximal exercise even at extreme altitude in keeping with Hill and Noakes' model that skeletal muscle function is regulated to prevent organ, especially heart damage during exercise. The evidence further disproves the model which holds that the oxygen demands of the exercising muscle are preeminent and that exercise is limited only after oxygen delivery to the exercising skeletal muscles becomes inadequate (4).

As the ultimate source of all those controversial ideas that have so entranced his scientific progeny, Nobel Laureate A. V. Hill would hopefully have no option but to agree with this updated version of his original contributions. It would be in keeping with his pithy conclusion derived late in his career: "I have long believed, and am still inclined to believe, that all theories of muscle contraction are wrong. But they have been very useful in stimulating new research. In fact, many of the best theories are self destructive, by provoking fresh inquiry and leading to new facts which they cannot explain. The only useless theories are those than cannot be tested and can explain everything" ((34), pp. 362-363).


1. Astrand, P. Experimental Studies of Physical Work Capacity in Relation to Sex and Age. Copenhagen: Munksgaard, 1952, pp. 1-171.
2. Astrand, P. O. and B. Saltin. Maximal oxygen uptake and heart rate in various types of muscular activity. J. Appl. Physiol. 16:977-981, 1961.
3. Bam, J., T. D. Noakes, J. Juritz, and S. C. Dennis. Could women outrun men in ultramarathon races? Med. Sci. Sports Exerc. 29:244-247, 1997.
4. Bassett, D. R. and E. T. Howley. Maximal oxygen uptake: "classical" versus "contemporary" viewpoints. Med. Sci. Sports Exerc. 29:591-603, 1997.
5. Bennett, A. F. Exercise performance of reptiles. Adv. Vet. Sci. Comparat. Med. 38B:113-138, 1994.
6. Blomqvist, C. G. and B. Saltin. Cardiovascular adaptations to physical training. Ann. Rev. Physiol. 45:169-189, 1983.
7. Brooks, G. A. The lactate shuttle during exercise and recovery. Med. Sci. Sports Exerc. 18:360-368, 1986.
8. Brooks, G. A. and G. A. Gaesser. End points of lactate and glucose metabolism after exhausting exercise. J. Appl. Physiol. 49:1057-1069, 1980.
9. Coetzer, P., T. D. Noakes, B. Sanders, et al. Superior fatigue resistant of elite black South African distance runners. J. Appl. Physiol. 75:1822-1827, 1993.
10. Conley, D. L. and G. Krahenbuhl. Running economy and distance running performance of highly trained athletes. Med. Sci. Sports Exerc. 12:357-360, 1980.
11. Costill, D. L. A Scientific Approach to Distance Running. Los Altos, CA: Tefnews, 1979, pp. 1-128.
12. Costill, D. L. and E. Winrow. Maximal oxygen intake among marathon runners. Arch. Physiol. Med. Rehab. 51:317-320, 1970.
13. Costill, D. L. and E. Winrow. A comparison of two middle-aged ultramarathon runners. Res. Q. 41:135-139, 1970.
14. Dennis, S. C., T. D. Noakes, and A. Bosch. Ventilation and blood lactate increase exponentially during incremental exercise. J. Sport Sci. 10:437-449, 1992.
15.Derman, K. L. Exercise tolerance and skeletal muscle structure and function in patients with severe chronic heart failure. M.Sc. thesis, University of Cape Town, Cape Town, South Africa, 1995.
16. Duncan, G. E., E. T. Howley, and B. N. Johnson. Applicability of V˙O2max criteria: discontinuous versus continuous protocols. Med. Sci. Sports Exerc. 29:273-278, 1997.
17. Einstein, A. In. The Mind of God and Other Musing: the Wisdom of Science. S. A. Jones (Ed.). San Rafael, CA: New World Library, 1994; pp. 1-135.
18. Farrell, P. A., J. H. Wilmore, E. F. Coyle, J. E. Billing, and D. L. Costill. Plasma lactate accumulation and distance running performance. Med. Sci. Sports 11:338-344, 1979.
19. Florence, S-L. J. P. and Weir. Relationship of critical velocity to marathon running performance.Eur. J. Appl. Physiol. 75:274-278, 1997.
20.Gaesser, G. A. and G. A. Brooks. Metabolic bases of excess post-exercise oxygen consumption: a review. Med. Sci. Sports. Exerc. 16:29-43, 1984.
21. Gertz, E. W., J. A. Wisneski, R. Neese, J. D. Bristow, and G. L. Searle. Myocardial lactate metabolism: evidence of lactate release during net chemical extraction in man. Circulation 63:1273-1279, 1981.
22. Green, H. J., J. R. Sutton, P. Young, A. Cymerman, and C. S. Houston. Operation Everest II. muscle energetics during maximal exhaustive exercise.J. Appl. Physiol. 66:142-150, 1989.
23. Grover, R. F., R. Lufschanowski, and J.K. Alexander. Alterations in the coronary circulation of man following ascent to 3,100 m altitude. J. Appl. Physiol. 41:832-838, 1976.
24. Groves, B. M., J. T. Reeves, J. R. Sutton, P. D. Wagner, Al. Cymerman, M. K. Malconian, P. B. Rock, P. M. Young, and C. S. Houston. Operation Everest II: elevated high-altitude pulmonary resistance unresponsive to oxygen. J. Appl. Physiol. 63:521-530, 1987.
25.Harris, P. Lactic acid and the phlogiston debt. Cardiovasc. Res. 3:381-390, 1969.
26. Hawking, S. Black Holes and Baby Universes and other Essays. London: Bantam Press, 1993, pp. 1-173.
27.Heiss, H. W., J. Barmeyer, K. Wink, G. Hell, F. J. Cerny, J. Keul, and H. Reindell. Studies on the regulation of myocardial blood flow in man. Basic Res. Cardiol. 71:658-675, 1976.
28. Hill, A. V. The oxidative removal of lactic acid. J. Physiol. 48:x-xi, 1914.
29.Hill, A. V. The physiological basis of athletic records. Lancet 2:481-486, 1925.
30. Hill, A. V. Muscular Activity. London. Bailliere, Tindall and Cox, 1925, pp. 1-115.
31.Hill, A. V. Living Machinery. London. G. Bell and Sons Ltd, 1927, pp. 1-241.
32. Hill, A. V. Muscular movement in man: The factors governing speed and recovery from fatigue. McGraw-Hill, New York, 1927.
33.Hill, A. V. Myothermic apparatus. Proc. R. Soc. B. 103:117, 1928.
34. Hill, A. V. Trails, and Trials in Physiology. London. Edward Arnold, 1965, pp. 1-374.
35. Hill, A. V. and H. Lupton. Muscular exercise, lactic acid, and the supply and utilization of oxygen. Q. J. Med. 16:135-171, 1923.
36. Hill, A. V., C. N. H. Long, and H. Lupton. Muscular exercise, lactic acid, and the supply and utilization of oxygen: parts I-III. Proc. R. Soc. B 96:438-475, 1924.
37. Hill, A. V., C. N. H. Long, and H. Lupton. Muscular exercise, lactic acid, and the supply and utilization of oxygen: parts IV-VI. Proc. R. Soc. B 97:84-138, 1924.
38. Hill, A. V., C. N. H. Long, and H. Lupton. Muscular exercise, lactic acid, and the supply and utilization of oxygen: parts VII-VIII. Proc. R. Soc. B 97:155-176, 1924.
39. Hoogeveen, A. R., and G. Schep. The plasma lactate response to exercise and endurance performance: Relationships in elite triathletes. Int. J. Sports Med. 18:526-530, 1997.
40. Houston, C. S., J. R. Sutton, A. Cymerman, and J. T. Reeves. Operation Everest II: man at extreme altitude. J. Appl. Physiol. 63:877-882, 1987.
41.Howley, E. T., D. R. Bassett, and H. G. Welch. Criteria for maximal oxygen uptake: review and commentary. Med. Sci. Sports Exerc. 27:1292-1301, 1995.
42. Huxley, T. H. In: The Mind of God and Other Musing: The Wisdom of Science. S. A. Jones (Ed.). San Rafael, CA: New World Library, 1994, pp. 1-135.
43. Katz, A. M. Physiology of the Heart 2nd Ed. New York: Raven Press, 1992, pp.1-687.
44.Kayser, B., M. Narici, T. Binzoni, B. Grassi, and P. Cerretelli. Fatigue and exhaustion in chronic hypobaric hypoxia: influence of exercising muscle mass. J. Appl. Physiol. 76:634-640, 1994.
45. Kempeneers, G., T. D. Noakes, R. van Zyl-Smit, et al. Skeletal muscle limits the exercise tolerance of renal transplant recipients: effects of a graded exercise training programme.Am. J. Kidney Dis. 16:57-65, 1990.
46. MacRae, H. S-H., S. C. Dennis, A. N. Bosch, and T. D. Noakes. Effects of training on lactate production and removal during progressive exercise in humans. J. Appl. Physiol. 72:1649-1656, 1992.
47. MacRae, H. S-H., T. D. Noakes, and S. C. Dennis. Effects on endurance-training on lactate removal by oxidation and gluconeogenesis during exercise. Pflugers Arch. 430:964-970, 1995.
48. MacRae, H. S-H., T. D. Noakes, and S. C. Dennis. Role of decreased carbohydrate oxidation on slower rises in ventilation with increasing exercise intensity after training. Eur. J. Appl. Physiol. 71:523-529, 1995.
49. Mayer, L. B., G. Stifter, S. Putz, D. Barnas, and H. Graf. Effects of dichloroacetate on exercise performance in healthy volunteers. Pflugers Arch. 423:251-254, 1993.
50. Mazzeo, R. S., G. A. Brooks, D. A. Schoeller, and T. F. Budinger. Disposal of blood [1-13C] lactate in humans during rest and exercise. J. Appl. Physiol. 60:232-241, 1986.
51. McKechnie, J. K., W. P. Leary, and T. D. Noakes. Metabolic responses to a 90-km running race. S. Afr. Med. J. 61:482-484, 1982.
52. McKechnie, J. J., W. P. Leary, T. D. Noakes, J. C. Kallmeyer, E. T. M. Macsearraigh, and L. R. Olivier. Acute pulmonary oedema in two athletes during a 90-km running race. S. Afr. Med. J. 56:261-265, 1979.
53. Mitchell, J. H. and G. Blomqvist. Maximal oxygen uptake. New Engl. J. Med. 284:1018-1022, 1971.
54.Mitchell, J. H. and P. B. Raven. Cardiovascular adaptation to physical activity. In: Physical Activity, Fitness and Health, C. Bouchard, R.J. Shephard and T. Stephens (Eds.). Champaign, IL: Human Kinetics, 1994, pp. 286-301.
55. Mitchell, J. H., B. J. Sproule, and C. B. Chapman. The physiological meaning of the maximal oxygen intake test. J. Clin. Invest. 37:538-546, 1958.
56. Morgan, D. W., F. D. Baldini, P. E. Martin, and W. M. Kohrt. Ten kilometer performance and predicted velocity at V˙O2max among well-trained male runners. Med. Sci. Sports Exerc. 21:78-83, 1989.
57. Myers, J. N. Essentials of Cardiopulmonary Exercise Testing. Champaign, IL: Human Kinetics, 1996, pp. 1-178.
58.Myers, J. and E. Ashley. Dangerous curves: a perspective on exercise, lactate, and the anaerobic threshold. Chest 111:787-795, 1997.
59.Myers, J., D. Walsh, N. Buchanan, and V. F. Froelicher. Can maximal cardiopulmonary capacity be recognized by a plateau in oxygen uptake? Chest 96:1312-1316, 1989.
60. Myers, J., D. Walsh, M. Sullivan, and V. Froelicher. Effect of sampling on variability and plateau in oxygen uptake. J. Appl. Physiol. 68:404-410, 1990.
61. Nicol, C., P. V. Komi, and P. Marconnet. Fatigue effects of marathon running on neuromuscular performance. I. Changes in muscle force and stiffness characteristics. Scand. J. Med. Sci. Sports 1:10-17, 1991.
62. Nicol, C., P. V. Komi, and P. Marconnet. Fatigue effect of marathon running on neuromuscular performance. II. changes in force, integrated electromyographic activity and endurance capacity.Scand. J. Med. Sci. Sports 1:18-24, 1991.
63.Nielsen, B., J. R. S. Hales, S. Strange, N. J. Christensen, J. Warberg, and B. Saltin. Human circulatory and thermoregulatory adaptations with heat acclimation and exercise in a hot, dry environment. J. Physiol. 460:467-485, 1993.
64. Nielsen, B., G. Savard, E. A. Richter, M. Hargreaves, and B. Saltin. Muscle blood flow and muscle metabolism during exercise and heat stress.J. Appl. Physiol. 69:1040-1046, 1990.
65. Nielsen, B., S. Strange, N. J. Christensen, J. Warberg, and B. Saltin. Acute and adaptive responses in humans to exercise in a warm, humid environment. Pflugers Arch. 434:49-56, 1997.
66. Noakes, T. D. Lore of Running, 1st Ed. Cape Town: Oxford University Press, 1985, pp. 1-535.
67. Noakes, T. D. Lore of Running. 3rd Ed. Cape Town: Oxford University Press, 1991, pp. 1-734,
68. Noakes, T. D. Implications of exercise testing for prediction of athletic performance: a contemporary perspective. Med. Sci. Sports Exerc. 20:319-330, 1988.
69. Noakes, T. D. Challenging beliefs: ex Africa semper aliquid novi. Med. Sci. Sports Exerc. 29:571-590, 1997.
70.Noakes, T. D. Physiological models to understand the effects of training for enhanced athletic performance. J. Sports Sci. 1998 (in review).
71. Noakes, T. D., K. H. Myburgh, and R. Schall. Peak treadmill running velocity during the V˙O2max test predicts running performance. J. Sports Sci. 8:35-45, 1990.
72. Noakes, T. D., V. E. Lambert, M. I. Lambert, P. McArthur, K. Myburgh, and A. J. S. Benade. Carbohydrate ingestion and muscle glycogen depletion during marathon and ultramarathon running.Eur. J. Appl. Physiol. 57:482-489, 1988.
73.Pasteur, L. Worship the spirit of criticism. In: Lend Me Your Ears. Great Speeches in History. W. Safire (Ed.). New York: W.W. Norton, 1997, pp. 511-513.
74. Pollock, M. L. Submaximal and maximal working capacity of elite distance runners. Part. 1. Cardiorespiratory aspects. Ann. N. Y. Acad. Sci. 301:310-321, 1977.
75. Pugh, L. G. C. E. Athletes at altitude. J. Physiol. 192:619-646, 1967.
76.Pullinen, T., M. Leynaert, and P. V. Komi. Neuromuscular function after marathon.Abstract Book of XVI Int. Soc. of Biomechanics Congress, Tokyo. August 24-27:1997.
77. Raskoff, W. J., S. Goldman, and K. Cohn. The "athletic heart": prevalence and physiological significance of left ventricular enlargement in distance runners. JAMA 236:158-162, 1976.
78. Reeves, J. T., B. M. Groves, J. R. Sutton, et al. Operation Everest II: preservation of cardiac function at extreme altitude. J. Appl. Physiol. 63: 531-539, 1987.
79. Robinson, S. Experimental studies of physical fitness in relation to age. Arbeitsphysiol. 10:251-323, 1938.
80. Robinson, S., H. T. Edwards, and D. B. Dill. New records in human power. Science 85:409-410, 1937.
81. Rock, P. B., M. K. Malconian, H. Donner, et al. Operation Everest 11: electrocardiography during maximal exercise at extreme altitude (Abstract). Med. Sci. Sports Exerc. 18:S74, 1986.
82. Rowell, L. B. Human Circulation: Regulation During Physical Stress. New York: Oxford University Press, 1986, pp. 1-416.
83. Rowell, L. B. Human Cardiovascular Control. New York: Oxford University Press, 1993, pp. 1-500.
84.Saltin, B. and P. O. Astrand. Maximal oxygen uptake in athletes. J. Appl. Physiol. 23:353-358, 1967.
85. Saltin, B. and L. B. Rowell. Functional adaptations to physical activity and inactivity. Fed. Proc. 39:1506-1513, 1980.
86. Scott, B. K. and J. A. Houmard. Peak running velocity is highly related to distance running performance.Int. J. Sports Med. 15:504-507, 1994.
87. Scrimgeour, A. G., T. D. Noakes, B. Adams, and K. Myburgh. The influence of weekly training distance on fractional utilization of maximum aerobic capacity in marathon and ultramarathon runners. Eur. J. Appl. Physiol. 55:202-209, 1986.
88.Seeherman, H. J., C. R. Taylor, G. M. O. Maloiy, and R. B. Armstrong. Design of the mammalian respiratory system. II. measuring maximal aerobic capacity. Respir. Physiol. 44:11-23, 1981.
89. Speechley, D. P., S. R. Taylor, and G. G. Rogers. Differences in ultra-endurance exercise in performance-matched male and female runners. Med. Sci. Sports Exerc. 28:359-365, 1996.
90. Spriet, L. L., K. Soderlund, M. Bergstrom, and E. Hultman. Anaerobic energy release in skeletal muscle during electrical stimulation in men. J. Appl. Physiol. 62:611-615, 1987.
91. Stanley, W. C., E. W. Gertz, J. A. Wisneski, D. L. Morris, R. A. Neese, and G. A. Brooks. Systematic lactate kinetics during graded exercise in man. Am. J. Physiol. 249:E595-E602, 1985.
92. Strojnik, V. and P. V. Komi. Neuromuscular fatigue after maximal stretch-shortening cycle exercise. J. Appl. Physiol. 84:344-350, 1998.
93. Suarez, J., J. K. Alexander, and C. S. Houston. Enhanced left ventricular systolic performance at high altitude during Operation Everest II. Am. J. Cardiol. 60:137-142, 1987.
94. Sutton, J. R., J. T. Reeves, P. D. Wagner, et al. Operation Everest II: oxygen transport during exercise at extreme simulated altitude. J. Appl. Physiol. 64:1309-1321, 1988.
95. Taylor, C. Studies in exercise physiology. Am. J. Physiol. 135:27-42, 1941.
96. Taylor, H. L. Exercise and metabolism. In: Science and Medicine of Exercise and Sports, Chapt. 8, W. R. Johnson (Ed.). New York: Harper and Brothers, 1960, pp. 123-161.
97. Taylor, H. L., E. Buskirk, and A. Henschel. Maximal oxygen intake as an objective measure of cardio-respiratory performance. J. Appl. Physiol. 8:73-80, 1955.
98.Timmons, J. A., T. Gustafsson, C. J. Sundberg, E. Jansson, and P. L. Greenhaff. Muscle acetyl group availability is a major determinant of oxygen deficit in humans during submaximal exercise. Am. J. Physiol. 274:E1-E4, 1998.
99. Wagner, P. D. Determinants of maximal oxygen transport and utilization.Ann. Rev. Physiol. 58:21-50, 1996.
100. Wagner, P. D., J. R. Sutton, J. T. Reeves, A. Cymerman, B. M. Groves, and M. K. Malconian. Operation Everest II: pulmonary gas exchange during a simulated ascent of Mt. Everest.J. Appl. Physiol. 63:2348-2359, 1987.
101. Wasserman, K. and M. B. McIlroy. Detecting the threshold of anaerobic metabolism in cardiac patients during exercise. Am. J. Cardiol. 14:844-852, 1964.
102. Wilson, J. R. Exercise intolerance in heart failure. Circulation 91:559-561, 1995.
103. Wyndham, C. H., N. B. Strydom, J. S. Maritz, J. F. Morrison, J. Peter, and Z. U. Potgieter. Maximum oxygen intake and maximum HR during strenuous work. J. Appl. Physiol. 14:927-936, 1959.

Section Description

Editor's Note: In 1997, we published the 1996 J. B. Wolffe Memorial Lecture by T. D. Noakes entitled "Challenging Beliefs:Ex Africa Semper Aliquid Novi" (Med. Sci. Sports Exerc. 29:571-590, 1997) and by request a response to the challenge by Basset and Howley entitled "Maximal Oxygen Uptake: 'Classical' versus 'Contemporary' Viewpoints" (Med. Sci. Sports Exerc. 29:591-603, 1997). As expected, Noakes decided to rebut the response of Bassett and Howley, and the following presentation is his rebuttal.

As Editor-in-Chief, I encourage scientific debate, especially when our fundamental concepts are challenged and new concepts are proposed based upon sound principles of scientific inquiry. In my opinion, the arguments of Noakes should be heard, digested, and subsequently investigated as to their scientific fallibility. The aforementioned articles and the following presentation are the initial process in the evaluation of Noakes' concepts. As a follow-up to these presentations, I have recommended a debate to the 1999 ACSM Annual Meeting Program Committee, under the auspices of the Cardiorespiratory Interest Group, on the topic "Why Skeletal Muscles Cannot Become Anaerobic During Progressive Maximal Exercise Testing." If accepted by the Program Committee, the debate will be chaired by Bengt Saltin, M.D., and the discussants will be Pro-Timothy Noakes, M.D., and Con-Brian J. Whipp, Ph.D. This debate will be another part of the scientific inquiry necessary to address the challenge of Noakes. I encourage all our members and readers to become aware of the issues raised and in the future to accept or reject the concepts developed by Noakes based upon the weight of scientific evidence generated.

I am constantly reminding my students that to have one's fundamental concepts challenged requires us, as scientists, to respond with factual reasons as to why the challenge is incorrect and not respond that the challenge is incorrect because it does not fit what one's personal reference base expounds. In my opinion, the challenge by Noakes is one that we as members of the scientific community, which accepts the concept of maximum oxygen consumption (V˙O2max) and its measurement as a parameter of physiological function, must address and one that may lead us to accept a revised definition of V˙O2max.



©1998The American College of Sports Medicine