Exercise physiology is more than a discipline of exercise science

Historically speaking, exercise science developed from the discipline of physical education sometime around the 1980s, depending on the country of reference. The initial development of exercise science was framed around five subdisciplines: exercise physiology, motor learning/control/skill acquisition, growth and development, biomechanics, and sport sociology/psychology. These subdisciplines became known as the ‘5 pillars of exercise science’.8 Today, thanks to decades of continued research and knowledge creation, one could argue, at least in the Australian context, that added pillars have now emerged. From the exercise physiology pillar alone has grown emphases in strength and conditioning, clinical exercise physiology, sports physiology and perhaps others.

The emergence of exercise physiology as a discipline of exercise science does not mean that exercise physiology is confined to exercise science. For more than 100 years, researchers have investigated human and animal physiological responses to exercise stress. Such research has been pursued by metabolic biochemists, specialist physiologists (eg, cardiovascular, pulmonary, neural, endocrine, muscle, bone and so on), clinicians (cardiologists, pulmonologists, physiotherapists, nutritionists, oncologists, epidemiologists and so on) and of course exercise scientists. The rich multidisciplinary input and applications of exercise physiology have been major factors in the rapid growth of knowledge, and as with all disciplines of science, no topic is owned by any one discipline. The fact that this manuscript is partly in response to work done by a cardiologist interested in a diverse array of human physiological enquiry during exercise stress is testament to this fact.

Thomas Kuhn’s ‘Structure of Scientific Revolutions’

Kuhn (1922–1996) did not write his text with the intention of presenting a model for how the scientific method, or science itself, should function.9 Kuhn commenced his scientific career as a physicist, but was soon drawn to the understandings of the historical development of science. Kuhn was interested in the progression of scientific discovery, and he studied science from a historical perspective, not from a judgement of right or wrong. Kuhn explained this in the Preface to his textbook and concisely expressed the purpose of his text within the Introduction: ‘Its aim is a sketch of the quite different concept of science that can emerge from the historical record of the research activity itself’ (p110).

It is also important to understand that Kuhn published his textbook in 1968. Contemporary science is a much different process to how it was in 1968, and for the centuries that preceded that time and provided the observations that fuelled Kuhn’s passion. Nevertheless, Pires commented that ‘Kuhn’s model’, presumably in reference to a model of scientific progress, was that of extended periods of stagnation, which Kuhn labelled ‘normal science’, separated by bursts of revolutionary discovery.1 9 10 What Pires failed to mention was that Kuhn’s choice of words for labelling the periods of scientific advancement has been widely criticised by his peer scientific philosophers, and his greatest critic was Karl Popper.7 11 Indeed, Kuhn’s work is contrasted by Popper’s more idealistic views of what the scientific method should be. This Popperian lens revealed that the periods of stagnation that characterises so much of the history of science is far from ‘normal science’. Rather, such troughs between infrequent peaks of paradigm shifts to scientific revolutions result from dysfunctional scientific enquiry; so dysfunctional that Popper would label it, both in Kuhn’s time11 and even 30 years before Kuhn published his text,7 as pseudoscience. I find it intriguing that Kuhn did not reference Popper in any part of his text, and that Pires reinforced the same omission.

The omission of the work of Karl Popper

I do not understand why Pires, a person self-proclaimed to be ‘trained in the philosophy of science’ (Pires1, p722), would choose to base a manuscript on the philosophical nature of exercise science (exercise physiology) framed around the views of Kuhn. As previously explained, Kuhn did not write on how science should operate. That topic was masterfully crafted by Popper,7 11 who remains our most influential contemporary philosopher of science.

Popper’s greatest contribution to the development of the continually evolving scientific method was the concept and practice of falsifiability. Popper argued vehemently throughout his long life and career (1902–1994) that no scientific investigation could prove anything. No matter how many similar findings resulted from research on the same topic from different scientists from different institutions, it only requires one contradictory finding, preferably replicated elsewhere, to falsify, or at least seriously question, the result/theory. The preferred evidence of proof for a finding or theory was that of continued failed efforts to falsify the result/theory. However, Popper also went to great efforts to express that the development of theories is central to the scientific method. Furthermore, no theory could ever be perfect, and scientific endeavour through falsification would provide evidence to refine the result/theory, leading to improvement and progress on a path towards the truth.

Popper even went further to distinguish science from pseudoscience. To Popper, researchers who deviated from the tenant of falsification were not following the most fundamental trait of the scientific method, and hence were pursuing pseudoscience. Indeed, Popper proclaimed that the tenant of falsification is a ‘criterion of demarcation’ between science and pseudoscience.7 In this regard, and of importance to the commentary by Pires, Popper’s writing on Kuhn’s work from the published proceedings of the International Colloquium in the Philosophy of Science, London 1965, is an interesting read.11 Remember, Kuhn’s text9 was published 30 years after Popper’s.7 Popper viewed Kuhn’s reported periods of stagnation as evidence of the acceptance of current theories and dogma. Such practice (Kuhn’s ‘normal science’) risks reinforcement of false knowledge and derails the progress of science from a path towards the truth. Popper’s writing revealed disgust of Kuhn’s views of ‘normal science’,9 10 as he has labelled the scientists involved in this practice as ‘to be sorry for’, ‘been badly taught’ and ‘a victim of indoctrination’.11 For Pires1 to assign an understanding of Kuhn’s views to exercise physiology and the CGM is to cast a murky shadow for how exercise physiology has and perhaps remains largely pseudoscientific.

Nevertheless, contemporary science is very different from Popper’s views, first published way back in 1934.7 As such, it is important to explain what Popper’s pseudoscience is so that our current interpretation can be placed in perspective. Popper had a simplistic view of the delineation between science and pseudoscience. Any non-critical application of science, that not based on efforts at falsification, was pseudoscientific. This is a very broad brush from which to define pseudoscience. Debate continues today within the field of the philosophy of science as to how relevant Popper’s delineation of this transition to pseudoscience is. Yet an awareness of pseudoscience for any given attempt at research may not be apparent in current time. A researcher may try to be totally ethical, meticulous in research design and data analysis, and detailed in the writing of the discussion. Some readers may wonder how such initiative and effort could ever be pseudoscientific. This is where Popper’s greatest contribution becomes apparent. Popper has challenged us all to be self-aware of our purpose or intent in research. What are the reasons for conducting the research? Are we following a mindset based on being aware of limitations in prior research, or current paradigms, and are we challenging this? Alternatively, is our research based on an extension of prior accepted paradigms? If so, has there been sufficient critical enquiry of the paradigm? Is it factual? If not, then Popper would argue that research should continue to challenge the paradigm, and research that supports it is at great risk of contributing to knowledge generation that is of no long-term value; or worse, is detrimental, or false knowledge. This then would be pseudoscience, even if the efforts in research were positive in intent. Similarly, and this applies to commentary on Noakes’ CGM, being critical but not following sustained efforts at remaining critical, even of a new theory, is also pseudoscientific, as a researcher should never favour one paradigm over another, even if they propose it. This also fits Popper’s definition of pseudoscience. This will be discussed in great length in the section 'Lessons from Popper on the theory derived from the CGM'.

Curious interpretations of citation numerics

Pires devoted considerable attention to the belief that the number of citations can be used to infer acceptance of a topic/theory, and presented numerics of such citations for the CGM in text, as well as two figures. Does Pires not understand that his methods for compiling these numerics contain both citations supportive and critical of the CGM? To fully reveal this flaw, I adopted a similar method to Pires and counted the peer-reviewed publications on the topic of ‘intelligent design’ for the years 2000–2016 through PubMed. Just from this one database I retrieved 200 published manuscripts where ‘intelligent design’ was stated in the title or abstract. Based on Pires’ interpretations, intelligent design must be a widely accepted construct in evolutionary science. Obviously, this is an incorrect interpretation. Intelligent design is a topic that infers divine intervention, which clearly cannot be experimentally tested and then by definition cannot be falsified. In extension, the topic therefore fails Popper’s criterion of demarcation and is pseudoscience. Journal citations or topical preference cannot be interpreted as evidence of acceptance or agreement.

Lessons from Popper on the theory derived from the CGM

An important contribution from Popper was his application of falsifiability to the development and testing of theories. In this regard, let me clarify that Noakes’ CGM is not itself a theory. Models are expressions of the connections between components of a system, or systems, often presented in diagrammatic form, from which theories can be worded or vice versa. Nevertheless, it has been a difficult process to explore the prior commentary and research on the CGM to find the exact wording of the theory. I have never understood why Noakes has not clearly stated the CGM as a theory , for as explained in great detail by Popper,7 the theory requires the researcher to communicate essential traits of the model that allow them and other researchers to attempt to falsify the theory. It is through failed attempts at falsification that a theory is best affirmed. In 1997, Noakes stated his ‘alternative model’ to understanding the limits of muscle contractile function (the beginnings of the CGM) to be as follows: ‘…skeletal muscle contractile function is regulated by a hierarchy of controls specifically to prevent damage to any of a number of different organs’.2 Yet, of course, this is neither a model nor theory; it is simply a statement. While theories are also statements, a theory must be worded carefully to adhere to a few simple rules, which I will detail below.

Is the textual explanation of the original CGM a well-worded theory? Let us evaluate Popper’s teachings to us. Popper argued that a theory must also be stated in such a way that it is open to falsification. Indeed, as previously explained, Popper proclaimed that it is the process of falsification, as well as the invitation for it, that is the foundation of the scientific method. Consequently, for a theory to be scientific, it must be worded in such a way to allow other researchers to identify one or more features that can be used to make predictions and for these predictions to be able to be directly challenged by experimentation. If the wording of a theory hampers efforts at falsification, then there is risk for the absence of structured attempts at falsification, and by definition, retarded progress. Furthermore, if a theory is worded in such a way that it cannot be falsified, then based on Popper’s criterion of demarcation, that act qualifies as pseudoscience and renders all research that the pseudoscientific theory has spawned to also be open to the label of pseudoscience.

It is important to note that the potential for falsification is the key, as some theories that are valid can be crafted even though there is no means yet to experimentally test them. The best example of this was Einstein’s general relativity theory. Einstein drafted this theory in 1905. It was not until 1938 that the first empirical evidence was produced to provide proof of the theory based on measurements of the bending of light in a gravitation field.12

What can be tested from the first version of the CGM? There are three components that are raised: (1) a hierarchy of controls, (2) organ damage and (3) for the purpose of preventing organ damage. The wording of items 1 and 2 are vague and adhere to Popper’s concerns of being unfalsifiable. Item 3 is teleological, as it infers that physiological systems have a singular purpose, which then explains why it exists. This cannot be falsifiable and as such is pseudoscientific.

As previously explained, the CGM has developed since this initial expression. How has it changed and how suitable are these new expressions as theories?

In 2000, Noakes further discussed the importance of AV Hill’s original proposition that a neural governor must regulate function during fatiguing exercise in order to prevent organ damage.13 At this time, Noakes favoured a more task-dependent approach to explaining limits to exercise tolerance. However, no version of the CGM was presented as a graphic model or in text as a theory.

The first presentation of the CGM was provided in figure 1 of Noakes’14 (p3232) manuscript on the evidence for a central governor during exercise in acute hypoxia and hyperoxia.14 Noakes referred to the CGM first as a theory and synonymously as a model. However, from now on, I will refer to it as the CGM to avoid any confusion. The CGM was expressed as follows:

‘…the reduced fHmax, Vsmax and Qmax during acute and chronic hypoxia are due to central regulation, the goal of which is to prevent the development of hypoxia in one or more of the vital organs including the brain, heart or perhaps the respiratory muscles. The action of this central governor is to regulate the mass of muscle that can be recruited during exercise under conditions in which the oxygen delivery to these vital organs is threatened. By limiting the muscle mass that can be activated, the central governor limits the peripheral peak VO₂ to a level that will not induce hypoxia in any of the vital organs’.

In this context, and at this time, the CGM was specific to cardiovascular and neuromuscular regulation by a specific region of the central nervous system (CNS) that acted as a governor to process afferent input to fine-tune (dampen) motor output to prevent continued increases in VO₂, oxygen demand and the consequent cardiovascular stress that would be associated with these events.

In 2004, Noakes then presented further evidence for the relevance of the CGM to all forms of exercise likely to stress organ function.15 The CGM was stated as follows:

‘…during self-paced exercise, which is the natural, as opposed to laboratory, form of human physical activity, the CNS continuously modifies the pace as part of a complex, non-linear dynamic system. In this model, the power output (work rate) during exercise is continuously adjusted on the basis of metabolic calculations performed at a subconscious CNS level. These calculations take into account prior knowledge acquired during previous exercise bouts, the planned end point of the current exercise bout, and the current metabolic rate, among many other potential variables. These subconscious calculations create a continuously adjusting power output, and hence pace, during exercise’.

This was a very different explanation to the prior version, as it now pertained only to self-paced exercise. However, the key evidence for supporting the CGM remains an altered motor unit recruitment. In fact, at this phase of the development of the CGM, there was clear direction for the need to measure motor unit recruitment (at least estimate it as best as possible from electromyography (EMG) based on current limitations in instrumentation) as part of any experimentation aimed at falsification (see next section). Nevertheless, it is important to acknowledge that there is controversy regarding the validity of surface EMG signals to be reflective of motor unit recruitment.16–19 As such, it remains dubious at best to even infer that changes in human motor unit recruitment can be accurately quantified during changes in exercise intensity and/or physiological transitions to fatigue at constant exercise intensities. This fact also raises concern for how Noakes has interpreted research of EMG signals during exercise to support a reduced motor unit recruitment component within the CGM.

Also in 2004, St Clair Gibson and Noakes presented another version of the CGM (referred to as the ‘central integrative model’, and also the ‘central governor theory of fatigue’) in their figure 6 (St Clair Gibson20, p801). This version was focused on exercise and fatigue (a term which I have avoided by using the phrase ‘exercise tolerance’). This version of the CGM was stated as follows:

‘… the subconscious brain sets the exercise intensity by determining the number of motor units that are activated and hence the mass of the skeletal muscle that is recruited throughout the exercise bout. The extent of motor neurone activity and hence skeletal muscle motor unit recruitment can then be influenced by sensory feedback from a variety of peripheral organs, yet to be characterised and which probably include the skeletal muscles, the respiratory muscles, the heart, and central and peripheral chemoreceptors and mechanoreceptors. From the beginning of the exercise bout, the subconscious brain informs the conscious brain of an increasing neural effort, perhaps related to an increased difficulty in maintaining homeostasis at the exercise intensity and this is interpreted by the brain as the increased sensation of fatigue, which may itself control further subconscious brain control processes’.

This version of the CGT now pertains to all forms of exercise.

In 2005, the CGM was further modified to include added details of the roles of the subconscious brain in the perception of fatigue, as well as an integration of both central and peripheral factors in the regulatory scheme.21 22 The model was now referred to as ‘the central governor model of fatigue incorporating teleo-anticipation during exercise’ (p5822) or more simply as the ‘integrated complex governor model of fatigue’ (Lambert et al 22, p61). Unfortunately, the description of the new CGM was excessively long, but its inclusion is needed in this rebuttal so there can be a critical reflection of the content in the next section. The new CGM was stated as follows:

‘In the context of a complex and integrated model of fatigue, during exercise, there is no single regulatory component. Rather, there are multiple levels of regulation, resulting in integrative homeostatic control of the different physiological systems, which is a continuous oscillatory process, and which thereby invokes compensatory responses in the periphery. There is also, however, evidence for central programming and a feed forward component in this complex model, which serves to alter the settings at which gain and the time constant of the gain of the entire system are achieved, and which thereby alters the perception of fatigue.

Furthermore, in this complex model, there are redundancies in control mechanisms that can be altered by antecedent exposures to exercise or metabolic perturbations and external environmental factors. These multiple controls and redundancies produce a robust system, providing for changes in gain and the time constant of gain as part of a general homeostatic process. These homeostatic controls maintain system integrity and ensure that any system is not overwhelmed or used to absolute maximal capacity using continuous feed forward and feedback control. The integrated complex governor model of fatigue therefore recognises multiple interdependent control systems, the importance of peripheral metabolic factors as signallers, and that the teleoanticipatory central nervous system is mutable and can be ‘reset’ or ‘hard wired’ by various stimuli such as prior experience, antecedent exposures, and training.’

In 2007, Noakes applied the CGM to better understand marathon running.23 However, Noakes used the CGM without clarification of the exact version of the model he was applying. Based on the descriptions used, it seems that Noakes was referring to the ‘integrated complex governor model of fatigue’.

In 2008 Noakes24 presented a very similar description of the CGM to what was presented in 2005. Nevertheless, the wording was different, and in science, words matter. Noakes referred to this theory or model as the ‘anticipatory (central governor) model’, and stated it to be as follows:

‘…human exercise performance is regulated (not limited) by a complex, intelligent system, the goal of which is the maintenance of homeostasis in all body systems. According to this model, subjects begin exercise at an intensity that is determined by their physiological capacity including their state of training, the expected duration of the exercise bout, their previous experience and the relative importance of the exercise bout, among many other influences. This established the feed-forward component of the central motor command, which determines the initial pace by recruiting the appropriate number of motor units in the muscles in the exercising limbs. Feedback form a variety of organs that monitor both the internal and external environment, then modify the pace by altering the number of motor units recruited in the exercising limbs. The goal of this control is to ensure that the exercise bout terminates before there is damage to any organ system as a result of a failure of homeostasis. The symptoms of fatigue become progressively more severe during exercise in order to ensure that exercise terminates within a predetermined and safe duration’.

Since 2004, the theory has developed to read more like a thesis than a theory. However, more on that in the next section. Later manuscripts from Noakes25 26 refer back to the prior versions of the CGM as stated since 2004, but given that the wording, components and function of each iteration are very different, it is unclear which specific version he was applying his current thinking and written expression to. Noakes had the opportunity to update everybody about the latest version of the CGM in his 2012 synopsis of the CGM and recent research that might support it; unfortunately, he did not.26

The central governor theory: a paradigm shift or pseudoscience?

What is wrong about the CGM? Or perhaps a better question is, how can it be improved? There is value for detailing specific aspects of each iteration of the CGM to decipher the presence of content that is unfalsifiable, versus content that can be gleamed from the words to focus attempts at falsification.

Concerns about specialisation

As an academic scientist initially educated within a 4-year physical education undergraduate degree, followed by postgraduate education in the USA consisting of a Master’s in Exercise and Sport Science and Cardiac Rehabilitation, followed by a PhD in Human Bioenergetics, it is reasonable to conclude that I have a broad education and research training. Such formal education has occurred across the basic and applied sciences spanning disciplines as diverse as organic chemistry, biochemistry, physics, biomechanics, biophysics, research methods, statistics, all physiological systems, pharmacology, molecular biology, health behaviour, computer programming and so on.

Throughout my career it has become increasingly evident that research of physiology, and especially exercise physiology, is handicapped by the subconscious bias of too many researchers to establish the single most important physiological explanation for a variety of topics of interest. I fear that this behaviour is spawned by our increasingly obvious trend of overspecialisation. Such a fear is a daunting reality when reading the following words from Popper’s Preface,7 first published in 1938!

‘…only a revival of interest in these riddles [man’s knowledge of the world in which he lives] can save the sciences and philosophy from narrow specialization and from an obscurantist faith in the expert’s special skill and in his personal knowledge and authority; a faith that so well fits our ‘post-rationalist’ and ‘post-critical’ age, proudly dedicated to the destruction of the tradition of rational philosophy, and of rational thought itself’. (p23)

Shouldn’t the most important theories involving human physiology be integrative, not exclusionary to other physiological regulatory systems? I wonder whether the most serious obstacle facing the physiological sciences, and for that matter, all applications of science, is our modern rigid educational and training emphasis on specialisation! In this context, the evolution of the CGM to a more integrative model is encouraging.

Nevertheless, I am concerned that the CGM has morphed into a theory of everything such that any organism with a brain and extended CNS is presumed to adhere to it. I was excited when I first read the manuscript from Lambert et al 22 as I contemplated evidence of an awareness of the need for a model of multidisciplinary inclusion. However, there was still evidence of bias in support of the original CGM in this work. For example, Lambert et al proposed a new model involving different (multiple) physiological systems, but then stated ‘evidence for central programming’. Furthermore, these authors advocated the presence and function of a new brain location, which they referred to as the ‘teleoanticipatory central nervous system’, but provide no empirical evidence of this centre.

Have they simply created this name because it aligned well with their model? If so, then such a development is pseudoscience.

Given the prior explanations of the errors of the CGM, I fear that all Noakes has accomplished is to replace one bias towards understanding exercise tolerance, that of cardiopulmonary limitations, with another bias, that of CNS control. I also fear that unless scientific enquiry begins to move from a paradigm of simplicity towards one of multidisciplinary complexity through integrative models, it will be a long time until we actually do witness a paradigm shift to a real scientific revolution of physiological understanding.

Publication biases in editorial peer review

This journal has been supportive of Noakes’ commentaries on the CGM.15 20–22 24 26 This has been refreshing to see, for many scientific journals remain closed minded to threats to current dogmas. Nevertheless, any manuscript that eventually receives acceptance for publication in peer-reviewed journals must undergo peer review. I am not sure how to interpret the content of the journal within the acknowledgements of Pires’ manuscript,1 where it is described that the manuscript underwent a non-commissioned external review. What does this mean? This is relevant, as I fail to comprehend how a blinded, impartial peer review would allow Pires’ use and misinterpretation of citation numerics. Further, I do not see the value of a manuscript that only presents one side, a supportive side, of a controversial theory. Such journal conduct is inconsistent to the tenants of falsification, where quality scientific discourse would be to present attempts to repudiate a theory. In fact, I would go even further on this issue. I think it is the requirement of a scientist to be the most energetic ambassador of attempts to oppose their own theory. In this context, there would be no greater achievement in science than for a scientist to propose a new theory, and then be the one to most convincingly disprove it. The Popperian view would have all scientists be their own greatest critics.

I am also disturbed by the unequal presentation of content on the CGM that the journal has shown over the last 16 years. The task of falsification requires deliberate support by scientific journals for evidence from all views of a theory. If the journal wishes to support critical and creative scientific thinking supported by empiricism, then all interpretations need to be supported. As such, empirically supported criticisms of the CGM should be just as welcomed as manuscripts that support it, if not more so based on the tenant of falsification. Simply because the CGM may have revealed and removed a bias does not mean this alternate view, in any of its prior versions, is closer to the truth! Balanced, unbiased, critical commentary is essential and must continue if there is to be further progress on understanding the determinants to limitations of exercise tolerance. Scholarly journals are essential to this process in their role of fulfilling the dissemination component of the scientific method. Conversely, if scholarly journals fail in this endeavour, they inflict ripples of bias as feedback through the entire scientific method, exacerbating the incidence and severity of scientific dysfunction: pseudoscience.

What needs to be done

It is a shame that Noakes’ brilliant critical enquiry of the physiology of constraints to exercise tolerance has ended in the absence of a well-worded theory. It is also confusing to see how the wording of the CGM has changed so dramatically across a 10-year span and remained so incompletely communicated despite a wealth of scholarly publication across this time interval.

Based on the prior content, it is reasonable to request the following from Noakes:

1. to change the name of the CGM to better reflect the latest iteration; specifically, the model should not reflect any connection to a central neural governor

2. to express the CGM as a theory, and in so doing be far more concise and adherent to the need to provide clear content to allow researchers to identify aspects of the theory that offer prediction of outcomes that can be used to develop research questions and methods that support falsification

3. to remove all teleological terminology within the theory.

For researchers who have chosen to express support of the CGM through commentary and research scholarship, it would be more beneficial to science if they were to adhere to the following changes:

1. to develop research to critically challenge the theory resulting from the CGM and attempt to falsify it

2. to improve their efforts at the item above, such researchers need to invest effort to remove their personal biases, allowing themselves to seek and detect additional limitations of the model or theory and design research to test such limitations.

For scholarly journals, regardless of the area of emphasis, there is a need to restructure their journal format, as well as editorial-peer review system, to promote and nurture a more critical process of research enquiry based on the process of falsification.