Noakes, T. D. (2000). Physiological models to understand exercise fatigue and the adaptations that predict or enhance athletic performance. Scandinavian Journal of Medicine and Science in Sports, 10, 123-145.

This review article critically evaluates several physiological models (i.e., explanations) that are supposed to account for exercise responses and improvements. Such models are variously used as the theoretical bases for structuring training programs for athletes. A central theme of the review is that contemporary physiology looks at explanations for responding rather than the accurate prediction of performance improvements. The former is relatively secure from critical evaluation whereas the latter is difficult to research and has an inherent possibility of failure.

A second suggestion is that contemporary physiologists have forgotten the history of the discipline. Many informative, substantive, and valuable principles of exercise response were discovered in the first part of the twentieth century but have gradually have fallen out of the common literature. That omission is one of the reasons that contribute to modern theories of exercise physiology being incomplete and inaccurate.

The accurate measurement of exercise responses in the field has been subverted by laboratory testing. The author offers three reasons why this has occurred.

1. The variables influencing human performance are not easily controlled. A field setting exacerbates that difficulty. This has led to the situation where laboratory measurements are used to infer performance characteristics in the field (e.g., a change in VO2max is used to infer the likelihood of an endurance performance change).
2. There is a dearth of tools to measure accurately human performance in the laboratory. If sports performance cannot be measured frequently with a high degree of precision in the laboratory, then training-induced changes in sports performance are not quantifiable. Direct, accurate testing is rarely possible. Consequently, physiological surrogates (e.g., VO2peak, VO2) are used to predict changes in performance.

3. An important weakness in current exercise physiology is a lack of certain knowledge of the precise factors that determine fatigue and hence, limit performance in different types of exercise under a range of environmental conditions. This is largely due to researchers and teachers advocating only one specific incomplete model of exercise physiology that does not explain performance under all conditions.

The review contemplates five exercise physiology models used popularly to explain and guide physical conditioning programs.

In maximal exercise, this model holds that endurance performance is determined by the capacity of the heart to pump large volumes of blood and oxygen to muscles. That facilitates muscles achieving higher work rates ("cardiovascular fitness") before outstripping the available oxygen supply ("anaerobiosis"). The capacity of the muscles to use fat as fuel ("aerobic lipolysis") is also increased. This is currently the most popular model for guiding the structure of training programs.

An increase in coronary blood flow that is inherent in this model is largely overlooked. However, the pumping capacity of the heart is restricted and limits oxygen utilization. Since that is so, the heart itself will be the first organ affected by the postulated oxygen deficiency. A. V. Hill's pioneering work has been incorrectly interpreted and an error perpetuated. Hill's actual interpretation of the fatigue that develops during maximal exercise was:

The heart is a muscle that is subject to the same functional constraints as skeletal muscle -- it needs oxygen to operate. The cardiovascular/anaerobic model ignores the role of the heart and assumes that all muscles fatigue at the same rate, both heart and skeletal. There must be some form of a central "governor" that stops the heart from reaching dangerous levels of fatigue. No such mechanism has been discovered because no one has looked for it.

Peak blood lactate, maximum heart rate, and cardiac output all fall with increases in altitude. At altitude, exercise terminates when exercising muscles are contracting in fully aerobic conditions. Thus, this model is unsatisfactory when it proposes the delivery of an adequate oxygen supply to exercising muscles is the cardinal priority during exercise. Some unexplained mechanism must exist that prevents the heart from becoming anaerobic during maximal exercise at any altitude. Neither skeletal nor cardiac muscles show any evidence for anaerobic metabolism at altitude.

The model is inconsistent when submaximal work is compared to maximal work. Similar function in maximal work must also exist in submaximal work, but at the lesser level, oxygen transport cannot be limiting. For example, a superior capacity for oxygen consumption during maximum exercise does not explain the manifest superiority of Kenyan runners during more prolonged submaximal exercise. Black runners have been shown to run substantially faster at all distances beyond 5 km despite VO2max values that are similar to middle-distance runners. What they did exhibit was a capability to sustain a substantially higher proportion of their VO2max when racing. They have superior fatigue resistance rather than aerobic capacity [%VO2max is a valid measure of fatigue resistance.] A VO2max test does not measure all the physiological variables determining success during more prolonged exercise.

Implications. Skeletal muscles do not develop anaerobiosis and form the effect that limits exercise. Some governor that protects the human from destructive fatigue level appears to exist. The functioning of the heart seems to be the important factor in any maximum exercise, and should be the emphasis of training programs.

This energy supply model predicts that performance in events of different duration is determined by the capacity to produce energy (ATP) by the separate metabolic pathways including the phosphagens, oxygen-independent glycolysis, aerobic glycolysis, and aerobic lipolysis. Superior performance is explained by a greater capacity to generate ATP in the specific metabolic pathways associated with an activity. For example, a common explanation is that a sprinter has a greater capacity to generate ATP from intramuscular phosphagen stores and oxygen-independent glycolysis, as opposed to a marathon runner who has a superior capacity to oxidize fat (aerobic lipolysis). Its basic tenet is that exercise must cease when ATP depletion occurs.

The status of this model's hypotheses is uncertain, as insufficient substantive research has been conducted. To validate this model's reasoning, the following have to be demonstrated.

• The different metabolic pathway capacities need to be causally related to different events.
• Specific metabolic pathways adapt specifically to different forms of training.
• Adaptations alone explain different performances that result from training exercises of different duration.

Research has shown the following, each of which contradicts the implications of this model.

• ATP concentrations in "exhausted" muscles rarely drop below 60% of resting values.
• High-energy phosphates do not participate in fatigue, but other factors reduce the use rate of ATP before ATP becomes limiting.
• There is a wide range of muscle pH concentrations reached at exhaustion (uniform acidosis is not exhibited across muscles).
• ATP demand by contracting muscles never exceeds the maximum rate of ATP supply.
• Muscle recovery is related to recovery of muscle phosphocreatine concentrations and unrelated to muscle pH concentrations.

Some peripheral governor needs to exist to account for these observations because acidosis does not play a direct role in fatigue in maximal exercise. Exercise terminates for reasons other than muscular lactacidosis.

Implication. Exercise is not limited by muscles achieving any critical level of acidosis although the availability of intramuscular phosphagen stores is. The contribution of neural factors that intervene with maximum peripheral muscular exercise has to be considered.

The energy depletion model is specific to exercises lasting longer than two hours. It holds that depletion of endogenous carbohydrate (CHO) stores limits the ability to perform long term exercise.

Much research to support this model has been conducted with inadequate or absent controls. Recently, better experimental designs have been used and placebo effects, as well as less consistent results, have been recorded. Additionally, the phenomenon of CHO-loading is not as evident in women as it is in men (it is hard to grasp why there would be a gender difference in biochemical function).

It is virtually impossible to prove conclusively that muscle glycogen depletion alone limits prolonged exercise performance because so many other factors occur concurrently. To support this model, future research has to show that neural factors are not involved (they seem to be involved in the previous two models).

• It is unclear how an inability to produce ATP at sufficiently high rates from one fuel source can explain this form of fatigue, given that ATP concentrations in muscles remain high in all forms of exhaustion.
• Further contradictory studies have shown that individuals ceased performing in the fourth hour of exercise when their muscle glycogen concentrations were the same as they were at the end of the first hour of exercising.
• Huge increases in muscle glycogen concentrations at the start of exercise only have a minimal impact on performance improvement in some subjects.
• It has not been shown that training improves endurance performance exclusively by increasing body carbohydrate stores and by delaying the onset of carbohydrate depletion.
• Very prolonged exercises (e.g., Ironman triathlons, 100-mile races) oxidize amounts of CHO in volumes that far exceed those existing in the body, without much detriment to performance in their later stages.

The human body is limited in its capacity to store CHO. High rates of CHO oxidation are required to sustain high rates of energy expenditure. Studies of very prolonged exercise show that rates of CHO oxidation remain high in athletes who ingest appropriate CHO during exercise. Because both muscle and liver glycogen depletion occur in fatigue, it is commonly assumed there is a direct causal relationship between muscle glycogen depletion and the development of fatigue in prolonged exercise. However, the relationship might not be causal under all circumstances. There is a logical impasse because any energy depletion model predicts that exercise must terminate when muscle ATP depletion occurs (leading to muscle rigor). Other factors must be involved in causing fatigue in prolonged exercise.

A popular explanation for "sparing" CHO is that with training, the ability to oxidize fats improves and this extends and improves performance without requiring higher CHO utilization. However, that explanation is still inadequate because ultimately it proposes that ATP depletion limits exercise, something that does not occur. This model is too simple to explain the physiology of prolonged exercise.

Implication. The CHO-depletion model does not adequately explain the response to prolonged exercise because at the end of the metabolic chain, ATP is not depleted. Some other reason has to exist for exercising to cease.

The fourth model has two parts that imply it is not the rate of supply of substrate (oxygen or fuel) to muscle that limits performance, but rather the processes involved in skeletal muscle recruitment, excitation, and contraction. The concept of central neural fatigue is invoked.

For the muscle recruitment model, the evidence is sufficiently persuasive to believe that central nervous system fatigue contributes to diminished performance in prolonged exercise, at altitude, and in the heat. In no study observing this phenomenon, is there evidence of anaerobiosis or energy depletion. This model also proposes there is a progressive peripheral fatigue for which the central nervous system makes an appropriate adjustment.

The central neural model does not specify important physiological mechanisms to account for fatigue.

Professor Noakes has argued elsewhere that a reduced central activation of exercising muscles is a protective mechanism. It prevents the following states.

• Myocardial ischaemia.
• Muscle ATP depletion.
• Myocardial ischaemia or cerebral hypoxia at altitude.
• A fall in blood pressure.
• Heatstroke.
• Glucopaenic brain damage during states of hypoglycaemia.

The muscle power model proposes that muscle contractile capacity, the ability to generate force, is not the same in the muscles of all humans. Superior athletes have a superior capacity to generate force. Very little has been researched on this alternative. David Costill reported that endurance training reduces skeletal muscle contractility, which shows that muscle contractility is not an immutable characteristic of the different muscle fiber types.

Implication. Performance increases resulting from this model would only occur within the limits of cardiovascular function within the specific activity.

Performance prediction is based on the greater the muscle's capacity to act as a spring, the less torque it must produce and hence, the more efficient it is. An improvement in performance stems from an increase in elastic muscle efficiency. That efficiency results from slowing the:

• Rate of accumulation of metabolites that cause fatigue, and
• The rate of rise in body temperature.

Reaching a core temperature that prevents continuing exercise is delayed.

This model is in direct contrast to the cardiovascular/anaerobic model, which predicts that superior performance during prolonged exercise results from an increased oxygen delivery to muscle and an increased rate of energy, resulting in increased heat production. A more logical assumption would be to reduce the rate of oxygen consumption and heat production by increasing the economy of movement.

Two factors that reduce heat production are small size and superior running economy. The more economical the athlete, the faster he/she will be able to run before reaching a limiting body temperature. Most training studies show that improvements in running/movement economy result from practice. Thus, being more economical, rather than having a higher VO2max, appears to be a more logical approach for explaining enhanced endurance performance.

A second component of this model stems from observations that repeated high velocity, short duration eccentric muscle contractions, as occur during running, induce a specific form of fatigue that lasts for a considerable time after cessation of the fatiguing activity. Characteristics of that fatigue are reduced contractile capacity, reduced tolerance for muscle stretch, and a delayed transfer from muscle stretch to muscle shortening in the stretch-shortening cycle. These result in the duration of a movement cycle being extended. Since these abnormalities last for several days they cannot be explained by oxygen or substrate delivery models.

In summary, the biomechanical model predicts that superior performance, especially in a weight-bearing activity like running, may be influenced by the capacity of the muscles to act as elastic energy return systems.

Implication. This model will demand that recovery be given as much emphasis as overload in training so that muscle function is preserved as long as possible, thereby facilitating the greatest volume of effective training.

Any demonstration of an ergogenic effect of any placebo intervention on exercise would prove this model contributes, at least in part, to athletic performance. In the field of physiology, this model is rarely considered. It does have some credibility in sport psychology, as well as emerging support from CHO-loading studies that are beginning to show a placebo effect.

Implication. The structure and content of an athlete's thinking could have an effect on the quality of performance.

Many findings are incompatible with the predictions of these models. The traditional tenets of physiology should be challenged until universal predictive validity is established.

New interpretations of training structures and content are warranted. The limited reasons and implications from the restrictive models described in this review will not result in the best form of training. The following are implied [training adaptations are considered to be responses that will transfer to competitive performances].

• The use of laboratory measurements, which are only partially related to laboratory performance, are useless for predicting competitive performances.
• Training programs based on oxygen and substrate supply theories, are likely to result in incorrect stimulation and will not yield maximal fitness adaptation for a specific sport.
• Training that emphasizes the reaction of muscles in the replicated activities of the sport is likely to produce beneficial fitness adaptation. [It should be noted that training with auxiliary activities, such as weight training, will not produce adaptations that generalize to competitive performances.]
• The physiological responses to complicated sporting activities are likely to be caused by a complicated set of physiological processes. Limiting training "theory" to one incomplete physiological model will not result in maximal fitness adaptation for a specific sport.
• It is likely that training programs developed by incorporating principles from psychology, biomechanics, and physiology will stimulate the best training adaptations for a particular sport.

Return to Table of Contents for this issue.