Neural Regulation of Respiration During Exercise -Beyond the Conventional Central Command and Afferent Feedback Mechanisms

J Phys Fitness Sports Med, 1(2): 235-245 (2012)

JPFSM: Review Article Neural regulation of respiration during exercise -Beyond the conventional central command and afferent feedback mechanisms-

Koji Ishida1* and Miharu Miyamura2

1 Research Center of Health, Physical Fitness and Sports, Nagoya University, E5-2(130), Fro-cho, Chikusa-ku, Nagoya 464-8601, Japan 2 Faculty of Human Science, Kanazawa Seiryo University, Gosho-Machi, Kanazawa 920-8620, Japan

Received: April 16, 2012 / Accepted: June 3, 2012

Abstract Ventilation increases rapidly and significantly in proportion to workload or metabolic rate during dynamic exercise. This increase is called “exercise hyperpnea.” During light to moderate step load exercise, ventilation increases from the first breath and reaches a plateau within 20 s (Phase I), during which metabolites do not reach chemoreceptors; thus Phase I is solely caused by neurogenic drives. It is worthwhile to clarify the aspects of Phase I in order to identify the mechanism of neurally mediated exercise hyperpnea. Until 2000, the mechanisms of exercise hyperpnea during light to moderate step load exercise were assumed to have been derived from two conventional neurogenic drives, “central command,” coming from the motor cortex or the hypothalamus, and “peripheral neural reflex,” originating mainly from the mechanoreceptors in muscles through group III afferents. For about a century there have been a large number of experiments trying to illuminate which mechanism is the cause of exercise hyperpnea. Although central command is thought to be the more likely key source, the consensus is that both central and peripheral neurogenic drives operate ventilation redundantly, building a multiple regulation system during exercise. Recent advantages in technology have enabled us to examine exercise hyperpnea in novel ways. Peripheral neurogenic drive through group III and IV afferents again enters into the limelight by using selective blockers for these afferents without augmenting central command. The vascular distension hypothesis has advocated that a rapid increase in peripheral blood flow is sensed as a plethysmometric change by the mechanoreceptors around the venule near the contracting muscles, stimulating the respiratory center through group IV afferents so as to match ventilation with metabolic rate. On the other hand, “learning” is attracting a growing interest from a central neurogenic point of view. Two types of learning have been proposed: “long term modulation (LTM),” serotonin mediated synaptic adaptation to repeated combined exercise and other stimuli such as an increase in dead space, and “volitional control,” a behavioral and learned response with cognitive function by way of the cerebrum and cerebellum. Nevertheless, these two pathways were derived from, not direct, but circumstantial evidence. The question, “What causes ventilation to increase during exercise?” is not likely to be solved in the near future. Keywords : Exercise hyperpnea, Central command, Peripheral neural reflex, Learning, Phase I

Two representative pathways are humoral inputs from the Introduction peripheral and/or central chemoreceptors (chemorecep- Ventilation increases rapidly and greatly from the first tor reflex), and neural inputs from the higher center and/ breath at the onset of dynamic exercise, and this ventila- or peripheral receptors. Arterial CO2 partial pressure, a tory increase is tightly coupled with a change in workload powerful stimulator for the chemoreceptors, shows little or metabolic rate. This increase is called “exercise hy- change during light to moderate exercise; so it is assumed perpnea,” and the mechanisms that control it have been that the chemoreceptors don’t provide the primary drive a matter of debate in the field of exercise physiology for for exercise hyperpnea, but rather a fine tuning of venti- about a century. A change in breathing occurs when out- lation during exercise1). Neural regulation of ventilation put signals are sent to the respiratory muscles from the thus plays a dominant role during light to moderate ex- respiratory center in the medulla oblongata, and is regu- ercise. In this review, the aspects of ventilatory response lated by various input signals to the respiratory center. to exercise, produced by neural factors, are mentioned, after which comes an overview of the conventional well- *Correspondence: [email protected] accepted mechanisms for exercise hyperpnea. In addition, 236 JPFSM: Ishida K, et al. several novel concepts regarding exercise hyperpnea are Phase I than exercise load. For example, faster treadmill demonstrated. speed produced greater Phase I response, compared to steeper treadmill grade5). As for the subject’s posture, Phase I was more attenuated in the supine position than Characteristics of the neural components of exercise that in an upright one6). It is interesting to note that pas- hyperpnea sive limb movement exerted by electrically-induced mus- Before reviewing the mechanisms for the neural com- cular contraction, or mimic movement by the experiment- ponents of exercise hyperpnea, we need to elucidate ers without muscular contraction (e.g. tandem cycling) the characteristics of the neurally mediated ventilatory also produce a rapid increase in ventilation at the start of response to exercise, because it should shed light on its movement, even during sleep7). Our research group has mechanisms. It is well known that ventilation increases been demonstrating Phase I responses to both voluntary abruptly at the onset of light to moderate step load exer- exercise and passive movement. Phase I response to vol- cise from the first breath and shows a brief plateau until untary exercise and passive movement by the arms were about 20 s (Phase I), thereafter increasing exponentially larger than those by the legs8). Delayed onset muscle (Phase II), and finally reaching a steady state within 3-5 soreness caused by eccentric exercise induced augmented + min (Phase III). Since metabolites (e.g. CO2, H ) cannot Phase I response to voluntary exercise and passive move- reach the peripheral chemoreceptor (carotid bodies) with- ment9). As for the effect of deconditioning, Phase I re- in 20 s at the onset of exercise, and Phase I accounts for sponse to voluntary exercise was attenuated after bed rest almost half of the full response as shown in Fig.1, eluci- for 20 days; but showed no change after unilateral lower dation of the features of Phase I enables us to simplify the limb suspension for 20 days, while passive movement mechanism for exercise hyperpnea solely into the neural exhibited no difference in Phase I after both decondition- factor. ing10). As for different subject groups, our research group Several researchers have observed Phase I under vari- revealed that Phase I response to voluntary exercise and ous conditions. As compared with dynamic step load ex- passive movement was attenuated in the elderly, as shown ercise from rest, static exercise2), ramp and sinusoidal load in Fig. 111), pre-teenage children12), endurance runners13), exercise3), and exercise starting from pre-exercise such as and sprinters (only passive)14) as compared with normal unloaded cycling4), produced little or no Phase I response. adults around twenty years in age. There was no gender Limb movement frequency was a greater determinant for difference in Phase I response to voluntary exercise or

䞉 VI Voluntary Passive

18 (l/min) 18 (l/min)

15 15

12 12

䞉䞉

9 9

Rest Exercise Recovery Rest Movement Recovery 6 6 -20 0 20 40 -20 0 20 40 Time (s) Time (s)

120 Exercise Movement (%) 50 (%) Steady state value of 100 3 min Exercise 40 80 30 60 20 40 * * * * 10 * 20 * * * * * * * * * * * * 0 * 0 * * Rest 5 10 15 20 Rest 5 10 15 20 Time (s) Time (s)

Fig. 1 Elderly (66.8yr) Young (22.9yr)

・ Fig. 1 Absolute (upper) and relative (lower) changes of inspiratory minute ventilation (VI) responses to voluntary leg extension exercise (left) and passive movement (right) for 20 s in the elderly (●) and the young (△). As for the relative change, 100% indicates the steady state value of the same exercise for 3 min in voluntary exercise and 0% indicates resting value. * repre- sents significant difference between the elderly and the young. Data modified from Ishida et al.11) JPFSM: Neural regulation for exercise hyperpnea 237 passive movement15). These results will either be suc- rogenic drive,” “peripheral neural reflex,” and “afferent cessfully explained by the mechanisms mentioned in the feedback.” The latter is incorrect as feedback is defined following sections, or suggest a new concept for exercise as a self-regulating system where the output signal loops hyperpnea. back to an input circuit, which is not the case here. As mentioned in the Phase I section above, passive limb movement produces a rapid increase in ventilation at the Classic neurogenic mechanisms for exercise hyperpnea initial stage of movement, indicating the involvement In 1996, the classic concepts of central and peripheral of afferent neurogenic drives from moving limbs as the neural regulation of ventilation during exercise were el- cause of exercise hyperpnea. This mechanism is strongly egantly and thoroughly summarized by Waldrop et al.16) supported by experiments showing that increased ventila- and Kaufman and Foster1), respectively, in the “Handbook tion induced by electrical stimulation or manual passive of Physiology.” Additionally, in the 1990s, Mateika and limb movement disappeared in partial spinal cord resect- Duffin17), Miyamura18), Turner19), and Whipp and Ward20) ed21) or anesthetized22) animals, or human patients with published fruitful reviews on exercise hyperpnea. In this a spinal cord transection23), as shown in Fig. 3. It is well review, these classic concepts are explained briefly. established that the pathways from the peripheral receptors to the respiratory center go through group III and/or 1. Peripheral neurogenic mechanisms IV thin afferent fibers24). The endings of group III fibers Afferent neurogenic drive from moving limbs (peripheral are mainly connected to mechanoreceptors. They respond neural reflex) quickly at the onset of muscle contraction (within 200 Mechanical and/or chemical information concerning ms) and begin to fire from a lower level of contraction muscle contraction such as distension of the muscles, force. On the contrary, the endings of group IV fibers are production of metabolites, etc. is sensed by the mechano- located near blood and/or lymphatic vessels, and mainly receptors and/or metaboreceptors in the muscles, tendons, connected to metabo- or chemo-sensitive receptors. They or joints and reflexively stimulates the respiratory center respond later between roughly 5-20 s from the onset of in the medulla oblongata through afferent sensory nerve contraction, and have a higher threshold. Some investiga- fibers, thus producing increased ventilation1) as shown in tors25,26) have proven that trapping metabolites after ex- Fig. 2➊. This mechanism has been called “afferent neu- ercise by arterial occlusion of the exercised limb induces

Primary motor cortex Cortical motor related area 䐢䐢䐢 Limb Volitional Premotor cortex movement ventilation Sensory Supplementary motor area area Prefrontal cortex Basal Limbic ganglia system Thalamus Hypothalamus 䐟䐢 Midbrain Cortico-spinal pathways 䐠䠍㻟㻌 Cerebellum Irradiation 䐟 Respiratory center Heart 䠎䐢 Group III afferents Serotonin Baroreceptor 䠍 Venula Group IV 㻟㻌 Mechanoreceptor 䠏 afferents (metaboreceptor) Spinal cord Lung Motor Muscle nerve Phrenic Diaphragm nerve

Fig. 2 Schematic view of the pathways involving mechanisms for neurogenic exercise hyperpnea. Gray arrows mean indirectly related pathways for exercise hyperpnea. There are a great number of connections in the brain, so this figure shows limited pathways related to exercise hyperpnea. 238 JPFSM: Ishida K, et al.

ments intentionally increasing central command were Control voluntary performed using a neuromuscular blockade31) and patients passive with unilateral weakness32), and the results supported cor- ᵏᵎᴾ ࣭ tical irradiation. Central command has been defined as a VE feedforward mechanism involving parallel activation of (l/min) motor, respiratory and cardiovascular centers during exercise. Nevertheless, some investigators have been opposed

Rest Exercise to this mechanism because their experiments revealed no Ჯ difference in ventilatory response between voluntary and Patients ᵏᵎᴾ passive electrically induced exercise during either the transient phase33,34) or steady state35). Additionally, supportive ex- ࣭ periments have only provided circumstantial evidence, so VE (l/min) it has not been proven directly or anatomically whether irradiation indeed occurs; and if it does, where it occurs. Rest Movement Ჯ 2-2. Hypothalamic central command -Ჯ -Ხ -Ჭ -Წ -Ძ Ძ Წ Ჭ Ხ Ჯ Breath Number Eldridge et al.36) observed that electrical or chemical Exercise onset stimulation of the hypothalamic locomotor region (HTLR) ・ Fig. 3 Changes of minute ventilation (VE) during voluntary ex- to anesthetized cats with an intact brain and to unanes- △ ercise (●) and passive movement ( ) in control subjects thetized decorticate (intact hypothalamus) cats induced (upper) and paraplegic patients (lower; passive only). X axis indicates breath number from the start of exercise. actual limb locomotion and the preceding increase in res- ★ indicates significant difference between voluntary and piration and blood pressure. They also demonstrated that passive movement (★ p<0.05, ★★ p<0.01), * indicates HTLR stimulation to paralyzed cats (indicating no affer- significant difference from resting value (* p<0.05, ** ent drives) produced fictive locomotion and an increase in 23) p<0.01). Data modified from Morikawa et al. respiratory activity. Accordingly, Eldridge and Waldrop’s group16,36) hypothesized that central command should originate not from the motor cortex but from HTLR (Fig. a rapid decrease in ventilation, having a minor effect of 2,➁) although their experiments were performed only on muscle metaboreflex activity on exercise hyperpnea. Ac- animals, not humans. It appears that the hypothalamus cordingly, it is assumed that peripheral neurogenic drive has reciprocal connections to the cerebrum and cerebel- is mainly sensed by mechanoreceptors and goes through lum, and it also receives sensory afferents from peripheral group III afferents from the onset of dynamic exercise. organs, e.g. muscles, and projects to the brainstem and However, some investigators refused to accept this drive spinal cord. Consequently, these inputs may stimulate as the causal mechanism due to observations that ventila- the hypothalamus so that efferent signals, such as central tory response to voluntary or electrically induced exercise command, are produced from there, inducing the parallel was not affected by blocking afferent nerve fibers with activation of locomotion and respiration. In conclusion, epidural anesthesia in human subjects27) or by transection it has been accepted that central command, presumably of the spinal cord in animals28); although these responses originating from HTLR, should play a primary role in ex- were observed only during steady state. The general con- ercise hyperpnea16). sensus has been that the peripheral neural reflex mechanism has a role in exercise hyperpnea1). 3. Multiple regulations Although other afferent inputs to the respiratory center 2. Central neurogenic mechanisms exist, for example, from the heart, lungs, respiratory mus- 2-1. Cortical central command (cortical irradiation) cles, and thermo-receptors, it is assumed that their role In 1913, Krogh and Lindhard29) first advocated that an is not a primary one in relation to exercise hyperpnea1). abrupt increase in ventilation at the onset of exercise was Short-term potentiation, persistent increased neural ac- caused by the “irradiation of motor impulses” from the tivity after removal of the stimulation, also plays not a motor cortex to the respiratory center (Fig. 2,➀). When primary, but a supplementary role by preventing inappro- the intensity of exercise increases, the motor command priate over- and undershoot of ventilation during the tran- and resultant irradiation should also proportionately in- sition phase of exercise16). It is accepted that central com- crease so that ventilation increases in accordance with mand originated from the motor cortex or HTLR, and/ the intensity of exercise. This cortical irradiation was or afferent drives mainly from mechanoreceptors in the afterwards named “central command” by Goodwin et muscles, transmitted through group III afferents, should al.30) in the 1970s, who demonstrated the involvement of be the main causes of exercise hyperpnea. For over 100 cortical irradiation by controlling central command, using years, whether central command or peripheral neurogenic vibration reflex with a fixed external load. Similar experi- drive plays a predominant role in exercise hyperpnea has JPFSM: Neural regulation for exercise hyperpnea 239 been a matter of debate for exercise and respiratory physi- pnea. To challenge this notion, Amann et al.38) pointed out ologists. There has been a lot of research both for and that local anesthesia (e.g. lidocaine), used in the above against experiments to each drive. It should be reasonable experiments in order to attenuate afferent drives, should to consider that both are involved in exercise hyperpnea also simultaneously reduce the activity of efferent nerves though a single mechanism should be able to elicit nearly to the limbs like a neuromuscular blockade, inducing the whole ventilatory response during exercise. It can “weakened” limbs and activating central command and be likened to a backup system that allows the system to ventilation, as similar experiments mentioned before31,32). function in the event that one of them is impaired. How- Taking the opposite stance, Amann et al.38) employed lum- ever, if that were true, with both drives intact, ventilatory bar intrathecal fentanyl, impairing the central projection response should increase twice as much as expected. Ac- of the spinal opioid receptor-sensitive muscle afferents cordingly, this system would also need a masking system. (group III and IV afferents) during arm and leg cycling These functions are called “redundancy”. Waldrop et al.37) exercise, and found that a selective blockade of the affer- observed that simultaneous activation of both HTLR and ent drives produced attenuated ventilatory, cardiovascular, muscle afferents in cats evoked increases in ventilation and perceptual responses to leg cycling exercise from a that were less than the sum of the increase evoked by sep- low workload level, compared with the placebo condi- arate activation. They also demonstrated that additional tion, demonstrating the essential contribution of group HTLR stimulation during muscular contraction caused III and IV afferents to exercise hyperpnea. Following this a large increase in ventilation while additional muscular finding, Kaufman39) suggested that afferent input onto contraction during HTLR stimulation showed a small hypothalamic and midbrain neurons should evoke central increase, indicating the predominance of hypothalamic command. In other words, thin fiber muscle afferents may central command. Kaufman and Foster1) suggested that amplify central command. This is a novel concept that afferent neurogenic drives should influence exercise hy- greatly differs from the traditional redundant multiple perpnea in a supplementary manner so that they may pro- regulation system mentioned before. However, some in- vide information to the respiratory center about conditions vestigators have thrown doubt on Amann’s experiments, in the limbs that enable central mechanisms to function continuing the debate (see Journal of Applied Physiology optimally. This redundancy mechanism involves three 110: 860-863 and 1499-1500, 2011). mechanisms1): neural occlusion, presynaptic inhibition, and chemoreceptor reflex feedback (fine tune). Although 1-2. Vascular distension hypothesis it is not yet clear whether or how these three redundant Several investigators have insisted that ventilation mechanisms actually operate together, many researchers must be tightly coupled with, not the workload, but the now support the idea that a redundant multiple regulation metabolic (or gas exchange) rate with respect to oxygen system is the most reasonable way to explain exercise hy- uptake and especially carbon dioxide production (≒mixed perpnea induced by two neurogenic mechanisms. venous CO2 content × cardiac output) during dynamic exercise. In order to explain this, in the mid 1970s, Was- serman et al.40) proposed the cardiodynamic hypothesis Current concepts relating to exercise hyperpnea which states that exercise hyperpnea is second to increase In 1996, Kaufman and Foster1) and Waldrop et al.16) re- in cardiac output (thus cardiovascular change). They viewed almost all previous studies on exercise hyperpnea speculated that the change of cardiac output (or venous and summarized them in the “Handbook of Physiology.” return) would be sensed at the baroreceptor in the right At that point, it seemed as though the controversy about ventricle and stimulate ventilation (Fig. 2,➋). However, the mechanisms for exercise hyperpnea would come to a this hypothesis was disproved, recently, by experiments certain end. Nevertheless, new approaches regarding the where the changes in ventilation did not always coincide mechanisms causing exercise hyperpnea continue to be with that of cardiac output during exercise, especially in advocated. In the section below, some of the novel con- the initial phase23,41), and where heart and lung transplan- cepts presented after the publication of the 1996 Hand- tation patients with attenuated cardiac responses showed book of Physiology are summarized. normal ventilatory response during exercise34,42). On the other hand, since the latter half of the 1990s, Haouzi’s 1. Peripheral neural reflex update group has been proposing an interesting and novel con- 1-1. New approach to afferent neurogenic drive cept explaining how ventilation matches metabolic rate As mentioned before, the peripheral neural reflex mech- during exercise. It is well known that an anatomically anism was disproven by experiments employing epidural significant portion of group III and IV afferents have anesthesia to blockade afferents which produced a similar been recognized in the vicinity of the adventitia of the ventilatory response to voluntary dynamic exercise as arterioles and the venules; and both of these afferents can compared with the control condition without anesthesia27). respond to mechanical stimuli43). In addition, Haouzi et This is a fundamental point suggesting the predominance al.44) observed a slowed ventilatory response to dynamic of the central command mechanism for exercise hyper- walking exercise in patients with severe peripheral blood 240 JPFSM: Ishida K, et al. flow limitation to the lower limbs as compared with classified as “central command,” their origins are quite normal age-matched subjects. Taking this into consider- different and sometimes have no “motor command.” So ation, Haouzi et al.45) hypothesized that group III and IV we need to use the term “central command” discriminate- afferents with endings in the skeletal muscle must signal ly. Brain neural networks and functions of the individual the distension of the vascular network. As shown in Fig. organs are very complicated and unsettled, but the attempt 4, they observed that pharmacologically induced vaso- is made to summarize them in Fig. 2 and Table 1. dilation and acute obstruction of venous drainage of the hindlimbs in anesthetized cats, which induced venous 2-1. Modulation and Plasticity distension, increased the activity of group III and IV af- Recent advances in neurobiology have introduced a ferents - especially group IV afferents - which, for the novel concept, describing how respiration is controlled most part, respond to muscular contraction. Accordingly, during exercise, called “modulation” and “plasticity.” Haouzi et al.43) advocated that the volume change at the Mitchell and Johnson46) defined “modulation” as a neu- vascular and venule level of contracting muscles is sensed rochemically induced alteration in synaptic strength or by group III and/or IV afferents and reflexively stimulates cellular properties, and “plasticity” as a persistent change the respiratory center (Fig. 2,➌). This plethysmometric in the neural control system based on prior experience. mechanism was later called the “vascular distension hy- Applying this to ventilatory response to exercise, Babb et pothesis.” It is assumed that blood flow to the muscles al.47) explained “modulation” as a ventilatory response due shows a rapid and proportional change with metabolic to neural alternations that are rapidly (according to one rate at the onset of exercise so that a change of ventilation trial) reversed in subsequent exercise trials if the inducing should also coincide with that of the metabolic rate. This stimulus is removed. They also explained “plasticity” as hypothesis is very interesting in that, although it must be a ventilatory response that persists after the stimulus has an updated peripheral neural reflex mechanism, it should ended, a type of “learning,” so to speak. Some investiga- also be a peripherally derived and modified version of tors have alternatively used the terms “short term modu- the cardiodynamic hypothesis. However, it is not clear lation (STM)” and “long term modulation (LTM).” STM whether the afferent signal caused by vascular distension is equivalent to modulation, while LTM is an example is strong enough to explain exercise hyperpnea, or how of plasticity. Repeated activation of STM would produce venule blood flow is increased and tightly coupled with LTM 47). Mitchell’s group has advocated that LTM should arterial blood flow or cardiac output. Experimental sup- contribute to exercise hyperpnea46-48). For example, Mar- port both for and against this hypothesis is limited; thus tin and Mitchell49) observed in goats that the repeated (20 further investigation is needed to confirm it. trials) association of exercise and increased respiratory dead space produced an increase in future ventilatory re- 2. Central neurogenic drive update sponse to the same exercise without dead space. Mitchell At first, although most of the proposed centrally medi- et al.46,47) postulated that LTM and STM may be caused ated mechanisms of exercise hyperpnea are sometimes by an augmented serotonin release from brainstem raphe

Vasodilator Injection Group III afferents Venous Occlusion Group IV afferents Injection Occlusion 䠅 s 䠅 s al Activity al Activity 䠄 impulses/2 䠄 impulses/2 Neur Neur

䠅 rr 䠅䠄 To 䠄 ml/min nous Pressure Mean Blood Flow Ve

Time 䠄s䠅 Time 䠄s䠅

Fig. 4 The effect of vasodilator injection (left) and venous occlusion (right) on the activity of group III and IV afferents. Lower left panel indicates mean popliteal blood flow change and lower right panel indicates mean venous pressure change. Data modified from Haouzi et al.45) JPFSM: Neural regulation for exercise hyperpnea 241

Table 1. Anatomy and functions of the brain associated with exercise hyperpnea Area Subdivision Region Function Cerebrum (Telencephalon䋩 Movement, Sensory processing, Learning and memory Primary motor cortex Execution of movements (Origin of motor command) Premotor cortex Motor programming and planning Frontal lobe Cerebral cortex Supplementary motor area Motor programming and planning, Motor coordination Prefrontal cortex Learning, Memory, Cognition, Thought, Motivation, etc. Insular cortex Emotion, Perception, Autonomic nervous control Autonomic nervous control, Memory, Emotion, Motivation, Hippocampus, Septal nuclei Instinct Limbic system Cingulate cortex Autonomic nervous control, Learning, Memory, Emotion, Limbic lobe 䋨Cingulate gyrus䋩 Cognition Subthalamic nucleus (STN), Substantia Basal ganglia Motor control, Cognition, Motivation, Emotion, Learning nigra, Globus pallidus, etc. Autonomic nervous center, Endocrine system, Emotion, Hypothalamus Instinct Diencephalon Locomotor Region (HTLR) Origin of central command? Thalamus Relaying sensory and motor signals, Consciousness Mesencephalon Relaying sensory and motor signals, Smooth movement, (Midbrain) Arousal Periaqueductal gray area Cardiovascular activity, Behavioral response to stress (pain), Brainstem 䋨PAG䋩 Emotion Pons Relaying between the cerebrum and cerebellum Respiratory center, Cardiovascular center, Relaying between Medulla oblongata the pons and spinal cord Cerebellar cortex, Deep nuclei (Fastigial Motor control (modification, coordination), Motor learning, Cerebellum nucleus; CFN) Cognition,

neurons to nearby respiratory motor neurons in the spinal speed change during treadmill walking in sheep57), was cord, thereby increasing motor neuron excitability (Fig. dissociated from work rate and speed change, especially 2,➂). LTM, in this case, has been supported by the hu- at faster oscillations. In addition, he insisted that HTLR man experiments. Repeated exercise with dead space50,51) is not always required because a lesion of HTLR had no or inspiratory resistive loading52) produced an increase in influence on the cardiovascular and locomotor responses ventilation during the transient phase at the onset of post in beagles58). They have argued these points heatedly in normal exercise. Wood et al.53) found that repeated hyper- a journal56,59) and, thereafter, several famous investiga- capnic exercise (70 trials) induced an increase in ventila- tors joined the debate (see Journal of Applied Physiology tion during subsequent steady state normocapnic exercise, 100:1417-1418 and 1743-1747, 2006). In addition, recent and suggested the involvement of learning and memory. brain functional imaging techniques, mentioned below, On the contrary, some investigators54,55) found little effect showed no involvement of the hypothalamus during or of repeated hypercapnic exercise with dead space on ven- after exercise60), or during imagined exercise61). A definite tilatory response to subsequent normal exercise. Wood et conclusion has not yet been reached. al.53) rebutted that this was because the number of trials in More recently, direct evidence of the importance of sub- their experiments was too few to elicit LTM. Research on cortical regions in human subjects has come from patients LTM and the specificity of the situations examined, e.g. with Parkinson’s disease or chronic pain undergoing deep hypercapnic or inspiratory resistive load exercise, is still brain stimulation by Paterson’s group. Thorton et al.62) limited. It is not yet known whether or how LTM affects reported that electrical stimulation of the thalamus, and normal ventilatory response to exercise. the subthalamic nucleus (STN) and substantia nigra in the basal ganglia, which is associated with motor control, 2-2. Subcortical central command update learning, cognitive, and emotional functions, produced Waldrop and Iwamoto56) have emphasized the predomi- an increase in heart rate and blood pressure in conscious nance of hypothalamic central command as the mecha- humans. However, it produced no actual movement, per- nism of exercise hyperpnea, based on the premise that haps because stimulation intensity was too low to elicit ventilation should be tightly correlated with the intensity movement. Green et al.63) revealed that the periaqueductal of exercise. On the contrary, Haouzi56) opposed this no- grey area (PAG) in the midbrain was involved in central tion because ventilatory response to sinusoidal work command mediated cardiorespiratory responses to an- rate change in human cycling exercise3) and sinusoidal ticipatory and actual exercise by using directly recorded 242 JPFSM: Ishida K, et al. electrical activity around the related regions of the brains the left and right superolateral primary motor cortex, of the patients. It is assumed that PAG has a role in inte- which is known to be related to motor control of the grated behavioral responses to stressors, and has connec- respiratory muscles during volitional breathing. After tions from the limbic system and prefrontal cortex, and to exercise, the former showed no activation though the lat- the brainstem. These results indicate that these subcorti- ter remained activated. In addition, relative rCBF was cal areas may not be the source of central command but increased in the supplementary motor area, ventrolateral rather the integration areas between central command and thalamus, and cerebellum, which are also known to be in- afferent neurogenic drive64). Nonetheless, at the moment, volved in volitional breathing. Furthermore, the cingulate it is still unclear whether these areas play a specific role cortex, parietal and frontal cortex, and globus pallidus, in exercise hyperpnea. which are associated with motor control, were also activated. Fink et al.60) concluded that the motor cortex should 2-3. Cortical and cerebellar mechanisms update be involved in exercise hyperpnea and suggested that the For the sake of activating conventional central com- control of breathing in the motor cortex might be a behav- mand, motor command to the contracting muscles is re- ioral or learned phenomenon. Similarly, Thornton et al.61), quired to induce parallel activation of ventilation during also using PET, revealed that ventilation and the dorsolat- exercise. However, ventilation can be increased without eral prefrontal cortex, supplementary motor area, premo- motor command or actual movement when exercise is tor cortex, sensorimotor cortex, thalamus and cerebellum anticipated65), imagined in conscious athletes66), or imag- were activated while imagining rigorous exercise (uphill ined under hypnosis67). Williamson68) emphasized that the cycling) under hypnosis without muscle contraction. They perception of effort, or effort sense, should be closely as- suggested that respiratory response to exercise should be sociated with central command, independent of workload generated by behavioral response learned in development. or force production; and this effort-induced central com- In the cerebrum, it is assumed that the prefrontal cortex, mand should play a predominant role in cardiovascular basal ganglia, and limbic system are related to movement response to exercise. It is assumed that perception of ef- based on learning and memory. These results suggest that fort is derived from some neurogenic information, for ex- volitional breathing by way of learning and memory in ample, somatosensory signals (e.g. from the muscles and the cerebrum should affect exercise hyperpnea. heart), neurocognitive mechanisms (e.g. cognitive ability On the other hand, Forster71) suggested that the cerebel- and experience), malaise (e.g. discomfort and pain), and lum may play a prominent role in exercise hyperpnea. As psychological factors (e.g. depression and neuroticism)68). mentioned above, Fink et al.60) and Thornton et al.61) re- Furthermore, Williamson et al.69) contended the existence vealed that the cerebellum was also involved in ventilato- of an independent “central cardiovascular command” and ry response during actual or imaged exercise. Previously, “central motor command”, both of which do not always Panda et al.72) observed an attenuated ventilatory response work simultaneously, for example, during imagined exer- to muscular stimulation with ablation or cooling of the cise. In this context, the term “cardiovascular” could be anterior lobe of the cerebellum in dogs. Williams et al.73) substituted for “respiratory” (i.e. “central respiratory com- found an increase in respiratory activity in response to mand”) because it is well accepted that input mechanisms stimulating the cerebellar fastigial nucleus (CFN) in cats. for cardiovascular and ventilatory responses to exercise While Martino et al.74) demonstrated that lesioning CFN should be quite similar1,16,19). Concerning the associated had a modest effect on exercise ventilation and suggested regions in the brain, Williamson et al.69,70) postulated that the cerebellum had a small effect on exercise hyperpnea the insular cortex (insula) in the cerebral cortex and the in goats, the redundancy mechanism to compensate may anterior cingulate cortex in the limbic system should be play a role as well. Until now, it is not yet clear whether the key structures for central cardiovascular command be- the cerebellum is actually involved in exercise hyperpnea cause they found that these structures are activated during in humans. It is well known that the cerebellum is closely imagined exercise when examined using single-photon- related to motor control (e.g. correct and smooth move- emission computed tomography. However, it is yet un- ment) by integrating information from the cerebral cortex known whether these structures are also associated with and peripheral organs such as muscles. The cerebellum is respiration during exercise. also assumed to be associated with automated movement Recent developments in medical engineering have through motor learning (memory). Accordingly, it is rea- pushed brain research forward enabling it to identify the sonable to suppose that the cerebellum plays a role in fine regions involved in exercise and ventilation. Fink et al.60) tuning ventilation in proportion to exercise intensity or used positron emission tomography (PET) to do brain effort sense, and is involved in learning and automating imaging and observed the activated regions during and respiratory movement. To our regret, there is no evidence immediately after dynamic leg exercise. Their research to confirm this notion. revealed that relative regional cerebral blood flow (rCBF) significantly increased in the left and right superomedial 2-4. Learning and cognition hypothesis primary motor cortex (the motor cortical leg areas), and Putting all these arguments together, an updated central JPFSM: Neural regulation for exercise hyperpnea 243 neurogenic “learning and cognition” hypothesis using the which induce parallel activation of locomotion and res- cerebrum and cerebellum can be proposed. An individual piration, have now become questionable because ventila- may have repeatedly experienced and learned a combina- tion is dissociated from workload, but is tightly coupled tion of the effort sense and ventilation at the same time with metabolic rate or the effort sense. Moreover, recent during exercise, and have it saved in the cerebrum as advances in brain research indicate the involvement of memory. Afterward, when the individual performs a simi- the cerebrum and cerebellum on exercise hyperpnea. Ac- lar type of exercise, he/she may sense or imagine the ef- cordingly, the authors propose the learning and cognition fort required in light of that memory, and calculate and/or hypothesis as the source of exercise hyperpnea. Quite a speculate the required level of ventilation. In this system, few investigators have new research interests regarding a human needs to recognize (e.g. interpret and anticipate) the effects of learning on exercise hyperpnea at the supra- the effort sense based on a previous experience so that spinal60,61,76) and spinal levels47,48,50,52,53) although it is still exercise hyperpnea should require a certain level of cog- difficult to observe the activity of the human brain during nitive ability. Wuyam et al.66) observed that highly trained exercise. The mechanisms for exercise hyperpnea have athletes, who recognize the efforts of exercise well, were been the focus of debate in the field of exercise physiol- more likely to show a greater ventilatory response to ogy for almost a century; even though the Handbook imagined exercise than control subjects. Bell et al.75) re- of Physiology, published in 1996, seemed to help put a vealed that initial ventilatory increase during exercise was certain end to the controversy. However, the true mecha- depressed while solving a puzzle (i.e. added obstructive nism remains a mystery, even though new technology is cognitive task). Learning/memory and cognition are in- shining more light on the subject. Another century may be volved in the cerebrum (e.g. the prefrontal cortex, limbic needed to bring this unsettled debate to a final close. system, and basal ganglia) and cerebellum. Accordingly, during exercise, an individual should recognize the inten- Acknowledgements sity of exercise by means of effort sense and/or memory in the cerebrum, and the information is then integrated This work is partly supported by a Grant-in-Aid for Scientific and transmitted to the primary motor cortex associated Research from the Japan Society for the Promotion of Science with volitional respiratory control, stimulating the respi- (C:23500777). ratory motoneuron in the spinal cord directly through the cortico-spinal pathways, resulting in increased ventilation References ➃ (Fig. 2, ). This drive is also transmitted to the cerebel- 1) Kaufman MP, Forster HV. 1996. Reflexes controlling circula- lum so that ventilation is regulated automatically and pre- tory, ventilatory and airway responses to exercise. In: Hand- cisely as mentioned before. This learning and cognition book of Physiology Sect.12, Exercise: Regulation and inte- hypothesis can reasonably explain exercise hyperpnea gration of multiple systems (Rowell LB, Shepherd JT, eds.), especially during Phase I. For example, Phase I in chil- 381-447, Oxford University Press, New York. dren was less than in adults12) because they are in the de- 2) Asmussen E. 1973. Ventilation at transition from rest to exer- velopmental learning and cognitive stage where they are cise. Acta Physiol Scand 89: 68-78. beginning to obtain the appropriate response. The elderly 3) Casaburi R, Whipp BJ, Wasserman K, Beaver WL, Koyal SN. also showed a slowed Phase I response11) because their 1977. Ventilatory and gas exchange dynamics in response to sinusoidal work. J Appl Physiol 42: 300-311. movements are more stereotyped and cognitive function 4) Whipp BJ, Ward SA, Lamarra N, Davis JA, Wasserman K. is delayed due to aging. Further, normal Phase I response 1982. Parameters of ventilatory and gas exchange dynamics 34,42) to exercise in heart and lung transplant patients may during exercise. J Appl Physiol 52: 1506-1513. be attributed to the learned response after surgery. This 5) Casey K, Duffin J, Kelsey CJ, McAvoy GV. 1987. The effect hypothesis seems very attractive and a reasonable mecha- of treadmill speed on ventilation at the start of exercise in nism to explain exercise hyperpnea, but to our regret, man. J Physiol 391: 13-24. direct evidence has not yet been revealed. 6) Weiler-Ravell D, Cooper DM, Whipp BJ, Wasserman K. 1983. Control of breathing at the start of exercise as influ- enced by posture. J Appl Physiol 55: 1460-1466. Closing remarks 7) Ishida K, Yasuda Y, Miyamura M. 1993. Cardiorespiratory Although the conventional peripheral neural reflex response at the onset of passive leg movements during sleep mechanism has been disregarded as the source of exercise in humans. Eur J Appl Physiol 66: 507-513. 8) Ishida K, Takaishi T, Miyamura M. 1994. Ventilatory re- hyperpnea, by investigators who have maintained the im- sponses at the onset of passive movement and voluntary ex- portance of the hypothalamic central command since the ercise with arms and legs. Acta Physiol Scand 151: 343-352. 38) 1990s, the recent new approach by Amann et al. , and 9) Hotta N, Yamamoto K, Katayama K, Ishida K. 2009. The the novel and attractive vascular distension hypothesis respiratory response to passive and active arm movements by Haouzi et al.43) has again shed light on the peripheral is enhanced in delayed onset muscle soreness. Eur J Appl neural mechanism. On the other hand, the conventional Physiol 105: 483-491. central cortical and hypothalamic command mechanisms, 10) Ishida K, Katayama K, Akima H, Iwase S, Sato K, Hotta N, 244 JPFSM: Ishida K, et al.

Miyamura M. 2010. Effects of deconditioning on the initial Thomas SN. 1990. Cardiovascular and ventilatory responses ventilatory and circulatory responses at the onset of exercise to dynamic exercise during epidural anaethesia in man. J in man. Adv Exp Med Biol 669: 319-322. Physiol 420: 281-293. 11) Ishida K, Sato Y, Katayama K, Miyamura M. 2000. Initial 28) Cross BA, Davey A, Guz A, Katona PG, Maclean M, Mur- ventilatory and circulatory responses to dynamic exercise are phy K, Semple SJG, Stidwill R. 1982. The role of spinal slowed in the elderly. J Appl Physiol 89: 1771-1777. cord transmission in the ventilatory response to electrically 12) Sato Y, Katayama K, Ishida K, Miyamura M. 2000. Ventila- induced exercise in the anaesthetized dog. J Physiol 329: 37- tory and circulatory responses at the onset of voluntary exer- 55. cise and passive movement in children. Eur J Appl Physiol 29) Krogh A, Lindhard J. 1913. The regulation of respiration 83: 516-523. and circulation during the initial stages of muscular work. J 13) Miyamura M, Ishida K, Hashimoto I, Yuza N. 1997. Ventila- Physiol 47: 112-136. tory response at the onset of voluntary exercise and passive 30) Goodwin GM, McCloskey DI, Mitchell JH. 1972. Cardio- movement in endurance runners. Eur J Appl Physiol 76: 221- vascular and respiratory responses to changes in central com- 229. mand during isometric exercise at constant muscle tension. J 14) Sato K, Matsuo H, Katayama K, Ishida K, Honda Y, Katsu- Physiol 226: 173-190. mata K, Miyamura M. 2004. Ventilatory and circulatory re- 31) Galbo H, Kjær M, Secher NH. 1987. Cardiovascular, ven- sponses at the onset of voluntary exercise and passive move- tilatory and catecholamine responses to maximal dynamic ment in sprinters. Eur J Appl Physiol 92: 196-203. exercise in partially curarized man. J Physiol 389: 557-568. 15) Matsuo H, Katayama K, Ishida K, Muramatsu T, Miyamura 32) Innes JA, De Cort SC, Evans PJ, Guz A. 1992. Central com- M. 2003. Effect of menstrual cycle and gender on ventilatory mand influences cardiorespiratory response to dynamic ex- and heart rate responses at the onset of exercise. Eur J Appl ercise in humans with unilateral weakness. J Physiol 448: Physiol 90: 100-108. 551-563. 16) Waldrop TG, Eldridge FL, Iwamoto GA, Mitchell JH. 1996. 33) Adams L, Garlick J, Guz A, Murphy K, Semple SJG. 1984. Central neural control of respiration and circulation during Is the voluntary control of exercise in man necessary for the exercise. In: Handbook of Physiology Sect.12, Exercise: ventilatory response? J Physiol 355: 71-83. Regulation and integration of multiple systems (Rowell LB, 34) Banner N, Guz A, Heaton R, Innes JA, Murphy K, Yacoub Shepherd JT, eds.), 333-380, Oxford University Press, New M. 1988. Ventilatory and circulatory responses at the onset of York. exercise in man following heart or heart-lung transplantation. 17) Mateika JH, Duffin J. 1995. A review of the control of breath- J Physiol 399: 437-449. ing during exercise. Eur J Appl Physiol 71: 1-27. 35) Brice AG, Forster HV, Pan LG, Funahashi A, Lowry TF,

18) Miyamura M. 1994. Control of ventilation during exercise in Murphy CL, Hoffman MD. 1988. Ventilatory and PaCO2 re- man with special reference to the feature at the onset. Jpn J sponses to voluntary and electrically induced leg exercise. J Physiol 44: 123-139. Appl Physiol 64: 218-225. 19) Turner DL. 1991. Cardiovascular and respiratory control 36) Eldridge FL, Millhorn DE, Kiley JP, Waldrop TG. 1985. mechanisms during exercise: an integrated view. J Exp Biol Stimulation by central command of locomotion, respiration 160: 309-340. and circulation during exercise. Respir Physiol 59: 313-337. 20) Whipp BJ, Ward SA. 1998. Determinants and control of breath- 37) Waldrop TG, Mullins DC, Millhorn DE. 1986. Control of res- ing during muscular exercise. Br J Sports Med 32: 199-211. piration by the hypothalamus and by feedback from contract- 21) Pan LG, Forster HV, Wurster RD, Murphy CL, Brice AG, ing muscles in cats. Respir Physiol 64: 317-328. Lowry TF. 1990. Effect of partial spinal cord ablation on ex- 38) Amann M, Blain GM, Proctor LT, Sebranek JJ, Pegelow DF, ercise hyperpnea on ponies. J Appl Physiol 69: 1821-1827. Dempsey JA. 2010. Group III and IV muscle afferents con- 22) Hida W, Shindoh C, Kikuchi Y, Chonan T, Inoue H, Sasaki H, tribute to ventilatory and cardiovascular response to rhythmic Takishima T. 1986. Ventilatory response to phasic contraction exercise in humans. J Appl Physiol 109: 966-976. and passive movement in graded anesthesia. J Appl Physiol 39) Kaufman MP. 2010. Control of breathing during dynamic 61: 91-97. exercise by thin fiber muscle afferents. J Appl Physiol 109: 23) Morikawa T, Ono Y, Sasaki K, Sakakibara Y, Tanaka Y, 947-948. Maruyama R, Nishibayashi Y, Honda Y. 1989. Afferent and 40) Wasserman K, Whipp BJ, Castagna J. 1974. Cardiodynamic cardiodynamic drives in the early phase of exercise hyper- hyperpnea: hyperpnea secondary to cardiac output increase. pnea in humans. J Appl Physiol 67: 2006-2013. J Appl Physiol 36: 457-464. 24) McCloskey DI, Mitchell JH. 1972. Reflex cardiovascular 41) Adams L, Guz A, Innes JA, Murphy K. 1987. The early circu- and respiratory responses originating in exercising muscle. J latory and ventilatory response to voluntary and electrically Physiol 224: 173-186. induced exercise in man. J Physiol 383: 19-30. 25) Haouzi P, Huszczuk A, Porszasz J, Chalon B, Wasserman K, 42) Grassi B, Ferretti G, Xi L, Rieu M, Meyer M, Marconi C, Whipp BJ. 1993. Femoral vascular occlusion and ventilation Cerretelli P. 1993. Ventilatory response to exercise after heart during recovery from heavy exercise. Respir Physiol 94: 137- and lung denervation in humans. Respir Physiol 92: 289-304. 150. 43) Haouzi P, Chenuel B, Huszczuk A. 2004. Sensing vascular 26) Innes JA, Solarte I, Huszczuk A, Yeh E, Whipp BJ, Wasser- distension in skeletal muscle by slow conducting afferent fi- man K. 1989. Respiration during recovery from exercise: ef- bers: neurophysiological basis and implication for respiratory fects of trapping and release of femoral blood flow. J Appl control. J Appl Physiol 96: 407-418. Physiol 67: 2608-2613. 44) Haouzi P, Hirsch JJ, Marchal F, Huszczuk A. 1997. Venti- 27) Fernandes A, Galbo H, Kjær M, Mitchell JH, Secher NH, latory and gas exchange response during walking in severe JPFSM: Neural regulation for exercise hyperpnea 245

peripheral vascular disease. Respir Physiol 107: 181-190. 61) Thornton JM, Guz A, Murphy K, Griffith AR, Pedersen DL, 45) Haouzi P, Hill JM, Lewis BK, Kaufman MP. 1999. Responses Kardos A, Leff A, Adams L, Casadei B, Paterson DJ. 2001. of group III and IV muscle afferents to distension of the pe- Identification of higher brain centres that may encode the ripheral vascular bed. J Appl Physiol 87: 545-553. cardiorespiratory response to exercise in humans. J Physiol 46) Mitchell GS, Johnson SM. 2003. Neuroplasticity in respira- 533: 823-836. tory motor control. J Appl Physiol 94: 358-374. 62) Thornton JM, Aziz T, Schlugman D, Paterson DJ. 2002. Elec- 47) Babb TG, Wood HE, Mitchell GS. 2010. Short- and long- trical stimulation of the midbrain increases heart rate and term modulation of the exercise ventilatory response. Med Sci arterial blood pressure in awake humans. J Physiol 539: 615- Sports Exerc 42: 1681-1687. 621. 48) Mitchell GS, Babb TG. 2006. Layers of exercise hyperpnea: 63) Green AL, Wang S, Purvis S, Owen SL, Bain PG, Stein JF, modulation and plasticity. Respir Physiol Neurobiol 151: Guz A, Aziz TZ, Paterson DJ. 2007. Identifying cardiorespi- 251-266. ratory neurocircuitry involved in central command during 49) Martin PA, Mitchell GS. 1993. Long-term modulation of the exercise in humans. J Physiol 578: 605-612. exercise ventilatory response in goats. J Physiol 470: 601- 64) Green AL, Paterson DJ. 2008. Identification of neurocircuitry 617. controlling cardiovascular function in humans using func- 50) Helbling D, Boutellier U, Spengler CM. 1997. Modulation tional neurosurgery: implications for exercise control. Exp of the ventilatory increase at the onset of exercise in humans. Physiol 93: 1022-1028. Respir Physiol 109: 219-229. 65) Tobin MJ, Perez W, Guenther SM, D’Alonzo G, Dantzker 51) Turner DL, Sumners DP. 2002. Associative conditioning of DR. 1986. Breathing pattern and metabolic behavior during the exercise ventilatory response in humans. Respir Physiol anticipation of exercise. J Appl Physiol 60: 1306-1312. Neurobiol 132: 159-168. 66) Wuyam B, Moosavi SH, Decety J, Adams L, Lansing RW, 52) Turner D, Stewart JD. 2004. Associative conditioning with Guz A. 1995. Imagination of dynamic exercise produced ven- leg cycling and inspiratory resistance enhances the early ex- tilatory responses which were more apparent in competitive ercise ventilatory response in humans. Eur J Appl Physiol 93: sportsmen. J Physiol 482: 713-724. 333-339. 67) Daly WJ, Overly T. 1966. Modification of ventilatory regula- 53) Wood HE, Fatemian M, Robbins PA. 2003. A learned compo- tion by hypnosis. J Lab Clin Med 68: 279-285. nent of the ventilatory response to exercise in man. J Physiol 68) Williamson JW. 2010. The relevance of central command for 553: 967-974. the neural cardiovascular control of exercise. Exp Physiol 95: 54) Cathcart AJ, Herrold N, Turner AP, Wilson J, Ward SA. 2005. 1043-1048. Absence of long-term modulation of ventilation by dead- 69) Williamson JW, Fadel PJ, Mitchell JH. 2006. New insights space loading during moderate exercise in humans. Eur J into central cardiovascular control during exercise in hu- Appl Physiol 93: 411-420. mans: a central command update. Exp Physiol 91: 51-58. 55) Moosavi SH, Guz A, Adams L. 2002. Repeated exercise 70) Williamson JW, McColl R, Mathews D, Mitchell JH, Raven paired with “imperceptible” dead space loading does not al- PB, Morgan WP. 2002. Brain activation by central command ter VE of subsequent exercise in humans. J Appl Physiol 92: during actual and imagined handgrip under hypnosis. J Appl 1159-1168. Physiol 92: 1317-1324. 56) Waldrop TG, Iwamoto GA, Haouzi P. 2006. Point:Counterpoint: 71) Forster HV. 2000. Exercise hyperpnea: where do we go from Supraspinal locomotor centers do/do not contribute signifi- here? Exerc Sport Sci Rev 28: 133-137. cantly to the hyperpnea of dynamic exercise. J Appl Physiol 72) Panda A, Senapati JM, Parida B, Fahim M. 1979. Role of cer- 100: 1077-1083. ebellum on ventilatory change due to muscle-receptor stimu- 57) Haouzi P, Chenuel B, Chalon B. 2004. The control of ventila- lation in the dog. J Appl Physiol 47: 1062-1065. tion is dissociated from locomotion during walking in sheep. 73) Williams JL, Everse SJ, Lutherer LO. 1989. Stimulating fas- J Physiol 559: 315-325. tigial nucleus alters central mechanisms regulating phrenic 58) Ordway GA, Waldrop TG, Iwamoto GA, Gentile BJ. 1989. activity. Respir Physiol 76: 215-227. Hypothalamic influences on cardiovascular response of bea- 74) Martino PF, Davis S, Opansky C, Krause K, Bonis JM, Czer- gles to dynamic exercise. Am J Physiol 257: H1247-H1253. niak SG, Pan LG, Qian B, Forster HV. 2006. Lesions in the 59) Haouzi P. 2006. The Last Word: Point:Counterpoint authors cerebellar fastigial nucleus have a small effect on the hyper- respond to commentaries on “Supraspinal locomotor centers pnea needed to meet the gas exchange requirements of sub- do/do not contribute significantly to the hyperpnea of dynam- maximal exercise. J Appl Physiol 101: 1199-1206. ic exercise in humans”. J Appl Physiol 101: 371. 75) Bell HJ, Feenstra W, Duffin J. 2005. The initial phase of ex- 60) Fink GR, Adams L, Watson JDG, Innes JA, Wuyam B, Ko- ercise hyperpnoea in humans is depressed during a cognitive bayashi I, Corfield DR, Murphy K, Jones T, Frackowiak task. Exp Physiol 90: 357-365. RS, Guz A. 1995. Hyperpnoea during and immediately after 76) Somjen GG. 1992. The missing error signal - regulation be- exercise in man: evidence of motor cortical involvement. J yond negative feedback. News Physiol Sci 7: 184-185. Physiol 489: 663-675.