Recent advances in understanding mechanisms regulating breathing during exercise

Claude Bernard's concept of “d'homeostasie du milieu interieur” developed more than a century ago assumes that living organisms tend to maintain constant the composition of their internal environment (Bernard, 1974). According to this concept, the increase in O₂ extracted by the muscles from the blood must be replenished, and the CO₂ and H⁺ produced by the muscles must be eliminated during exercise. If alveolar ventilation and alveolar–capillary diffusion increase in proportion to the increase in metabolic rate during exercise, these requirements will be met. The respiratory system is also vital in many species to eliminate at least a portion of the heat generated by exercise, which is achieved by increasing ventilation of the conducting airways (dead space ventilation). To maximize the gas exchange potential available for the locomotor muscles during exercise, the increase in pulmonary ventilation and alveolar ventilation must be achieved efficiently with minimal oxygen consumed by the respiratory muscles. To meet these requirements of the respiratory system, neurons in the brain that regulate breathing must receive information from multiple sources during exercise.

The sources of the information and the mechanisms that regulate breathing during exercise have fascinated physiologists for well over a century. Indeed, one of the hallmarks of the exercise hyperpnoea was documented early in the last century by Douglas et al. (2014). Figure 1 is a photograph of Douglas walking on a street breathing into a bag which to this day is known as a ‘Douglas Bag’. After such walks and/or exercise on a treadmill at different work intensities, the volume of air and the O₂ and CO₂ concentration in the bag were measured to calculate metabolic rate. The data obtained were then plotted as shown in Fig. 2. As can be seen at low levels of exercise, pulmonary ventilation increased in proportion to the increase in metabolic rate which results in homeostasis of arterial blood gases as documented by subsequent investigators including the noted exercise physiologists Erling Asmussen and Marius Nielsen (Asmussen, 1974). At higher workloads, ventilation increased proportionally more than metabolic rate (Fig. 2), which leads to arterial hypocapnia Another major hallmark of the exercise hyperpnoea is that the hyperpnoea begins simultaneously with the onset of exercise (Fig. 3), which was documented in 1913 by the pioneering exercise physiologists, August Krogh and Johnas Lindhard (Krogh & Lindhard, 2014). Based on this rapid response, they concluded “that the mechanism which shall produce the abrupt changes must be a nervous mechanism” and “we think that the evidence is in favor of an irradiation of impulses from the motor cortex”.

Figure 1.

Photograph of C. Gordon Douglas walking on street outside of laboratory at Oxford

Figure 2.

Ventilation () as a function of CO₂ excretion (+) or O₂ consumption (•) while walking at various speeds on the level or uphill (final data points). Continuous lines, Oxford; dashed line, Pike's Peak, Colorado, 4250 m. Data from Douglas et al. (2014).

Figure 3.

Recording of inspiration (upward deflection) and expiration in a human subject at rest and during moderate exercise

The onset of exercise is indicated by the downward arrow. Volume change indicated on right side in litres; each bar represents 1 litre. Modified from Krogh & Lindhard (2014).

Accordingly, for decades it has been known that in healthy humans, the exercise hyperpnoea is not associated with an increase in or a decrease in . Mammals do not possess any known receptive mechanisms capable of directly sensing the rate at which CO₂ and O₂ are exchanged in the peripheral tissues or in the lungs; thus, one of the most puzzling challenges in integrative physiology is to resolve how respiratory neurons adjust their output to variables which cannot be directly monitored or do not change during exercise. Particularly perplexing is that there is no known mechanism to account for an over twentyfold increase in breathing representing the largest non-volitional drive to breathe (Fig. 2). During the last few decades of the 20th century, several hypotheses were extensively tested and impressive data were obtained supportive of different mechanisms that seemingly could mediate the hyperpnoea. However, there were in most cases contradictory data or reasons to question the validity of each hypothesized mechanism. For example, on the basis of cross-circulation data obtained in anaesthetized dogs, Kao concluded “there is certainly a peripheral neurogenic drive which must be considered as the, or one of the mechanisms of exercise hyperpnea” (Kao et al. 2014). In contrast, on the basis of spinal lesioning studies, Weissman et al. concluded that “reflex discharge of afferent nerves from the exercising limbs was not requisite for the matching of ventilation to metabolic demand during exercise” (Weissman et al. 1963). As a result, there has never been a consensus among investigators that any of the proposed mechanisms mediates the exercise hyperpnoea.

Over the last two decades, there have been relatively few studies attempting to gain insight into the exercise hyperpnoea. As a result, at national and international meetings, there has been a virtual absence of sessions dedicated to this topic. However, a few investigators have continued pursuit of the elusive mechanism and each has provided information worthy of discussion among investigators with differing perspectives. Accordingly, a symposium was organized for Experimental Biology 2013 at which new data were presented and discussed. The intent of that symposium was to renew interest leading to the development of ideas and acquiring of resources for further study of the exercise hyperpnoea.

The four presentations at this symposium are summarized herein. Dr David Paterson describes data obtained using neuroimaging and functional neurosurgical techniques in humans to provide evidence that specific parts of the periaqueductal grey “appear to be key communicating circuitry for central command” mediation of the exercise hyperpnoea (Paterson, 1913). Dr James Duffin's summary makes the point that “limb movement frequency is effective in influencing ventilation during exercise as well as at the start and end of exercise” (Duffin, 1965). Dr Philippe Haouzi emphasizes that “a population of III and IV muscle afferent fibres located in the adventitia of the small vessels (in the muscles) appear to respond to the change in venular distention” to provide a signal that contributes to “the apparent matching between alveolar ventilation and pulmonary gas exchange” during submaximal exercise (Haouzi, 1912). Finally, Dr Jerome Dempsey summarizes data obtained on healthy humans during exercise after attenuating type III–IV muscle afferents with intrathecal fentanyl that lead to the hypothesis “that type III–IV muscle afferents are obligatory to the hyperpnoea of mild through moderate intensity rhythmic, large muscle steady-state exercise” (Dempsey et al. 1974).

In spite of recent progress, several major issues need to be addressed in future studies. One major issue is whether the strong evidence obtained for central command mediation of the exercise hyperpnoea in animal studies is representative of a neurally intact animal with a normal control system. A second major issue is why responses obtained from sinusoidal exercise studies are in apparent conflict with central command/motor activity. A third major issue is to establish whether indeed venule blood flow in exercising muscle is a major contributor to the exercise hyperpnoea during in vivo spontaneous exercise. Fourth, with the multiple physiological systems changing during exercise, it seems integration occurs somewhere in the brain, but this issue has not yet been adequately addressed.