Ganong’s Review of Medical Physiology, 24th Edition

CHAPTER 36 Regulation of Respiration


After studying this chapter, you should be able to:

image Locate the pre-Bötzinger complex and describe its role in producing spontaneous respiration.

image Identify the location and probable functions of the dorsal and ventral groups of respiratory neurons, the pneumotaxic center, and the apneustic center in the brain stem.

image List the specific respiratory functions of the vagus nerves and the respiratory receptors in the carotid body, the aortic body, and the ventral surface of the medulla oblongata.

image Describe and explain the ventilatory responses to increased CO2 concentrations in the inspired air.

image Describe and explain the ventilatory responses to decreased O2 concentrations in the inspired air.

image Describe the effects of each of the main nonchemical factors that influence respiration.

image Describe the effects of exercise on ventilation and O2 exchange in the tissues.

image Define periodic breathing and explain its occurrence in various disease states.


Spontaneous respiration is produced by rhythmic discharge of motor neurons that innervate the respiratory muscles. This discharge is totally dependent on nerve impulses from the brain; breathing stops if the spinal cord is transected above the origin of the phrenic nerves. The rhythmic discharges from the brain that produce spontaneous respiration are regulated by alterations in arterial PO2, PCO2, and H+ concentration, and this chemical control of breathing is supplemented by a number of nonchemical influences. The physiological bases for these phenomena are discussed in this chapter.



Two separate neural mechanisms regulate respiration. One is responsible for voluntary control and the other for automatic control. The voluntary system is located in the cerebral cortex and sends impulses to the respiratory motor neurons via the corticospinal tracts. The automatic system is driven by a group of pacemaker cells in the medulla. Impulses from these cells activate motor neurons in the cervical and thoracic spinal cord that innervate inspiratory muscles. Those in the cervical cord activate the diaphragm via the phrenic nerves, and those in the thoracic spinal cord activate the external intercostal muscles. However, the impulses also reach the innervation of the internal intercostal muscles and other expiratory muscles.

The motor neurons to the expiratory muscles are inhibited when those supplying the inspiratory muscles are active, and vice versa. Although spinal reflexes contribute to this reciprocal innervation, it is due primarily to activity in descending pathways. Impulses in these descending pathways excite agonists and inhibit antagonists. The one exception to the reciprocal inhibition is a small amount of activity in phrenic axons for a short period after inspiration. The function of this postinspiratory output appears to be to brake the lung’s elastic recoil and make respiration smooth.


The main components of the respiratory control pattern generator responsible for automatic respiration are located in the medulla. Rhythmic respiration is initiated by a small group of synaptically coupled pacemaker cells in the pre-Bötzinger complex (pre-BÖTC) on either side of the medulla between the nucleus ambiguus and the lateral reticular nucleus (Figure 36–1). These neurons discharge rhythmically, and they produce rhythmic discharges in phrenic motor neurons that are abolished by sections between the pre-Bötzinger complex and these motor neurons. They also contact the hypoglossal nuclei, and the tongue is involved in the regulation of airway resistance.


FIGURE 36–1 Pacemaker cells in the pre-Bötzinger complex (pre-BÖTC). Top: Anatomical diagram of the pre-BÖTC from a neonatal rat. Bottom: Sample rhythmic discharge tracing of neurons in the pre-BÖTC complex from a brain slice of a neonatal rat. IO, inferior olive; LRN, lateral reticular nucleus; NA, nucleus ambiguus; XII, nucleus of 12th cranial nerve; 5SP, spinal nucleus of trigeminal nerve. (Modified from Feldman JC, Gray PA: Sighs and gasps in a dish. Nat Neurosci 2000;3:531.)

Neurons in the pre-Bötzinger complex discharge rhythmically in brain slice preparations in vitro, and if the slices become hypoxic, discharge changes to one associated with gasping. Addition of cadmium to the slices causes occasional sigh-like discharge patterns. There are NK1 receptors and μ-opioid receptors on these neurons, and, in vivo, substance P stimulates and opioids inhibit respiration. Depression of respiration is a side effect that limits the use of opioids in the treatment of pain. However, it is now known that 5HT4 receptors are present in the pre-Bötzinger complex and treatment with 5HT4 agonists blocks the inhibitory effect of opiates on respiration in experimental animals, without inhibiting their analgesic effect.

In addition, dorsal and ventral groups of respiratory neurons are present in the medulla (Figure 36–2). However, lesions of these neurons do not abolish respiratory activity, and they apparently project to the pre-Bötzinger pacemaker neurons.


FIGURE 36–2 Respiratory neurons in the brain stem. Dorsal view of brain stem; cerebellum removed. The effects of various lesions and brain stem transections are shown; the spirometer tracings at the right indicate the depth and rate of breathing. If a lesion is introduced at D, breathing ceases. The effects of higher transections, with and without vagus nerves transection, are shown (see text for details). CP, middle cerebellar peduncle; DRG, dorsal group of respiratory neurons; IC, inferior colliculus; NPBL, nucleus parabrachialis (pneumotaxic center); VRG, ventral group of respiratory neurons; 4th vent, fourth ventricle. The roman numerals identify cranial nerves. (Modified from Mitchell RA, Berger A: State of the art: Review of neural regulation of respiration. Am Rev Respir Dis 1975;111:206.)


Although the rhythmic discharge of medullary neurons concerned with respiration is spontaneous, it is modified by neurons in the pons and afferents in the vagus from receptors in the airways and lungs. An area known as the pneumotaxic center in the medial parabrachial and Kölliker–Fuse nuclei of the dorsolateral pons contains neurons active during inspiration and neurons active during expiration. When this area is damaged, respiration becomes slower and tidal volume greater, and when the vagi are also cut in anesthetized animals, there are prolonged inspiratory spasms that resemble breath holding (apneusis; section B in Figure 36–2). The normal function of the pneumotaxic center is unknown, but it may play a role in switching between inspiration and expiration.

Stretching of the lungs during inspiration initiates impulses in afferent pulmonary vagal fibers. These impulses inhibit inspiratory discharge. This is why the depth of inspiration is increased after vagotomy (Figure 36–2) and apneusis develops if the vagi are cut after damage to the pneumotaxic center. Vagal feedback activity does not alter the rate of rise of the neural activity in respiratory motor neurons (Figure 36–3).


FIGURE 36–3 Afferent vagal fibers inhibit inspiratory discharge. Superimposed records of two breaths: A) with and B) without feedback vagal afferent activity from stretch receptors in the lungs. Note that the rate of rise in phrenic nerve activity to the diaphragm is unaffected but the discharge is prolonged in the absence of vagal input.

When the activity of the inspiratory neurons is increased in intact animals, the rate and the depth of breathing are increased. The depth of respiration is increased because the lungs are stretched to a greater degree before the amount of vagal and pneumotaxic center inhibitory activity is sufficient to overcome the more intense inspiratory neuron discharge. The respiratory rate is increased because the after-discharge in the vagal and possibly the pneumotaxic afferents to the medulla is rapidly overcome.


A rise in the PCO2 or H+ concentration of arterial blood or a drop in its PO2 increases the level of respiratory neuron activity in the medulla, and changes in the opposite direction have a slight inhibitory effect. The effects of variations in blood chemistry on ventilation are mediated via respiratory chemoreceptors—the carotid and aortic bodies and collections of cells in the medulla and elsewhere that are sensitive to changes in the chemistry of the blood. They initiate impulses that stimulate the respiratory center. Superimposed on this basic chemical control of respiration, other afferents provide nonchemical controls that affect breathing in particular situations (Table 36–1).


TABLE 36–1 Stimuli affecting the respiratory center.


The chemical regulatory mechanisms adjust ventilation in such a way that the alveolar PCO2 is normally held constant, the effects of excess H+ in the blood are combated, and the PO2 is raised when it falls to a potentially dangerous level. The respiratory minute volume is proportional to the metabolic rate, but the link between metabolism and ventilation is CO2, not O2. The receptors in the carotid and aortic bodies are stimulated by a rise in the PCO2 or H+concentration of arterial blood or a decline in its PO2. After denervation of the carotid chemoreceptors, the response to a drop in PO2 is abolished; the predominant effect of hypoxia after denervation of the carotid bodies is a direct depression of the respiratory center. The response to changes in arterial blood H+ concentration in the pH 7.3–7.5 range is also abolished, although larger changes exert some effect. The response to changes in arterial PCO2, on the other hand, is affected only slightly; it is reduced no more than 30–35%.


There is a carotid body near the carotid bifurcation on each side, and there are usually two or more aortic bodies near the arch of the aorta (Figure 36–4). Each carotid and aortic body (glomus) contains islands of two types of cells, type I and type II cells, surrounded by fenestrated sinusoidal capillaries. The type I or glomus cells are closely associated with cuplike endings of the afferent nerves (Figure 36–5). The glomus cells resemble adrenal chromaffin cells and have dense-core granules containing catecholamines that are released upon exposure to hypoxia and cyanide. The cells are excited by hypoxia, and the principal transmitter appears to be dopamine, which excites the nerve endings by way of D2 receptors. The type II cells are glia-like, and each surrounds four to six type I cells. The function of type II cells is not fully defined.


FIGURE 36–4 Location of carotid and aortic bodies. Carotid bodies are positioned near a major arterial baroreceptor, the carotid sinus. Aortic bodies are shown near the aortic arch.


FIGURE 36–5 Organization of the carotid body. Type I (glomus) cells contain catecholamines. When exposed to hypoxia, they release their catecholamines, which stimulate the cuplike endings of the carotid sinus nerve fibers in the glossopharyngeal nerve. The glia-like type II cells surround the type I cells and probably have a sustentacular function.

Outside the capsule of each body, the nerve fibers acquire a myelin sheath; however, they are only 2–5 μm in diameter and conduct at the relatively low rate of 7–12 m/s. Afferents from the carotid bodies ascend to the medulla via the carotid sinus and glossopharyngeal nerves, and fibers from the aortic bodies ascend in the vagi. Studies in which one carotid body has been isolated and perfused while recordings are being taken from its afferent nerve fibers show that there is a graded increase in impulse traffic in these afferent fibers as the PO2 of the perfusing blood is lowered (Figure 36–6) or the PCO2 is raised.


FIGURE 36–6 Effect of PCO2 on afferent nerve firing. The rate of discharge of a single afferent fiber from the carotid body (circles) is plotted at several PO2 values and fitted to a line. A sharp increase in firing rate is observed as PO2 falls below normal resting levels (ie, near 100 mm Hg). (Courtesy of S Sampson.)

Type I glomus cells have O2-sensitive K+ channels, whose conductance is reduced in proportion to the degree of hypoxia to which they are exposed. This reduces the K+ efflux, depolarizing the cell and causing Ca2+ influx, primarily via L-type Ca2+ channels. The Ca2+ influx triggers action potentials and transmitter release, with consequent excitation of the afferent nerve endings. The smooth muscle of pulmonary arteries contains similar O2-sensitive K+ channels, which mediate the vasoconstriction caused by hypoxia. This is in contrast to systemic arteries, which contain adenosine triphosphate-(ATP) dependent K+ channels that permit more K+ efflux with hypoxia and consequently cause vasodilation instead of vasoconstriction.

The blood flow in each 2 mg carotid body is about 0.04 mL/min, or 2000 mL/100 g of tissue/min compared with a blood flow of 54 mL or 420 mL per 100 g/min in the brain and kidneys, respectively. Because the blood flow per unit of tissue is so enormous, the O2 needs of the cells can be met largely by dissolved O2 alone. Therefore, the receptors are not stimulated in conditions such as anemia or carbon monoxide poisoning, in which the amount of dissolved O2 in the blood reaching the receptors is generally normal, even though the combined O2 in the blood is markedly decreased. The receptors are stimulated when the arterial PO2 is low or when, because of vascular stasis, the amount of O2 delivered to the receptors per unit time is decreased. Powerful stimulation is also produced by cyanide, which prevents O2 utilization at the tissue level. In sufficient doses, nicotine and lobeline activate the chemoreceptors. It has also been reported that infusion of K+ increases the discharge rate in chemoreceptor afferents, and because the plasma K+ level is increased during exercise, the increase may contribute to exercise-induced hyperpnea.

Because of their anatomic location, the aortic bodies have not been studied in as great detail as the carotid bodies. Their responses are probably similar but of lesser magnitude. In humans in whom both carotid bodies have been removed but the aortic bodies left intact, the responses are essentially the same as those following denervation of both carotid and aortic bodies in animals: little change in ventilation at rest, but the ventilatory response to hypoxia is lost and the ventilatory response to CO2 is reduced by 30%.

Neuroepithelial bodies composed of innervated clusters of amine-containing cells are found in the airways. These cells have an outward K+ current that is reduced by hypoxia, and this would be expected to produce depolarization. However, the function of these hypoxia-sensitive cells is uncertain because, as noted above, removal of the carotid bodies alone abolishes the respiratory response to hypoxia.


The chemoreceptors that mediate the hyperventilation produced by increases in arterial PCO2 after the carotid and aortic bodies are denervated are located in the medulla oblongata and consequently are called medullary chemoreceptors. They are separate from the dorsal and ventral respiratory neurons and are located on the ventral surface of the medulla (Figure 36–7). Recent evidence indicates that additional chemoreceptors are located in the vicinity of the solitary tract nuclei, the locus ceruleus, and the hypothalamus.


FIGURE 36–7 Rostral (R) and caudal (C) chemosensitive areas on the ventral surface of the medulla. Cranial nerves, pyramid and pons are labeled for reference.

The chemoreceptors monitor the H+ concentration of cerebrospinal fluid (CSF), including the brain interstitial fluid. CO2 readily penetrates membranes, including the blood–brain barrier, whereas H+ and image penetrate slowly. The CO2 that enters the brain and CSF is promptly hydrated. The H2CO3 dissociates, so that the local H+ concentration rises. The H+ concentration in brain interstitial fluid parallels the arterial PCO2. Experimentally produced changes in the PCO2 of CSF have minor, variable effects on respiration as long as the H+ concentration is held constant, but any increase in spinal fluid H+ concentration stimulates respiration. The magnitude of the stimulation is proportional to the rise in H+ concentration. Thus, the effects of CO2 on respiration are mainly due to its movement into the CSF and brain interstitial fluid, where it increases the H+ concentration and stimulates receptors sensitive to H+.


In metabolic acidosis due, for example, to the accumulation of acid ketone bodies in the circulation in diabetes mellitus, there is pronounced respiratory stimulation (Kussmaul breathing). The hyperventilation decreases alveolar PCO2(“blows off CO2”) and thus produces a compensatory fall in blood H+ concentration. Conversely, in metabolic alkalosis due, for example, to protracted vomiting with loss of HCl from the body, ventilation is depressed and the arterial PCO2 rises, raising the H+ concentration toward normal. If there is an increase in ventilation that is not secondary to a rise in arterial H+ concentration, the drop in PCO2 lowers the H+ concentration below normal (respiratory alkalosis); conversely, hypoventilation that is not secondary to a fall in plasma H+ concentration causes respiratory acidosis.


The arterial PCO2 is normally maintained at 40 mm Hg. When arterial PCO2 rises as a result of increased tissue metabolism, ventilation is stimulated and the rate of pulmonary excretion of CO2 increases until the arterial PCO2 falls to normal, shutting off the stimulus. The operation of this feedback mechanism keeps CO2 excretion and production in balance.

When a gas mixture containing CO2 is inhaled, the alveolar PCO2 rises, elevating the arterial PCO2 and stimulating ventilation as soon as the blood that contains more CO2 reaches the medulla. CO2 elimination is increased, and the alveolar PCO2 drops toward normal. This is why relatively large increments in the PCO2 of inspired air (eg, 15 mm Hg) produce relatively slight increments in alveolar PCO2 (eg, 3 mm Hg). However, the PCO2 does not drop to normal, and a new equilibrium is reached at which the alveolar PCO2 is slightly elevated and the hyperventilation persists as long as CO2 is inhaled. The essentially linear relationship between respiratory minute volume and the alveolar PCO2is shown in Figure 36–8.


FIGURE 36–8 Responses of normal subjects to inhaling O2 and approximately 2, 4, and 6% CO2. The relatively linear increase in respiratory minute volume in response to increased CO2 is due to an increase in both the depth and rate of respiration. (Reproduced with permission from Lambertsen CJ in: Medical Physiology, 13th ed. Mountcastle VB [editor]. Mosby, 1974.)

Of course, this linearity has an upper limit. When the PCO2 of the inspired gas is close to the alveolar PCO2, elimination of CO2 becomes difficult. When the CO2 content of the inspired gas is more than 7%, the alveolar and arterial PCO2 begin to rise abruptly in spite of hyperventilation. The resultant accumulation of CO2 in the body (hypercapnia) depresses the central nervous system, including the respiratory center, and produces headache, confusion, and eventually coma (CO2 narcosis).


When the O2 content of the inspired air is decreased, respiratory minute volume is increased. The stimulation is slight when the PO2 of the inspired air is more than 60 mm Hg, and marked stimulation of respiration occurs only at lower PO2 values (Figure 36–9). However, any decline in arterial PO2 below 100 mm Hg produces increased discharge in the nerves from the carotid and aortic chemoreceptors. There are two reasons why this increase in impulse traffic does not increase ventilation to any extent in normal individuals until the PO2 is less than 60 mm Hg. First, because Hb is a weaker acid than HbO2, there is a slight decrease in the H+ concentration of arterial blood when the arterial PO2 falls and hemoglobin becomes less saturated with O2. The fall in H+ concentration tends to inhibit respiration. In addition, any increase in ventilation that does occur lowers the alveolar PCO2, and this also tends to inhibit respiration. Therefore, the stimulatory effects of hypoxia on ventilation are not clearly manifest until they become strong enough to override the counterbalancing inhibitory effects of a decline in arterial H+ concentration and PCO2.


FIGURE 36–9 Top: Average respiratory minute volume during the first half hour of exposure to gases containing various amounts of O2. Marked changes in ventilation occur at PO2 values lower than 60 mm Hg. The horizontal line in each case indicates the mean; the vertical bar indicates one standard deviation. Bottom: Alveolar PO2 and PCO2 values when breathing air at various barometric pressures. The two graphs are aligned so that the PO2 of the inspired gas mixtures in the upper graph correspond to the PO2 at the various barometric pressures in the lower graph. (Courtesy of RH Kellogg.)

The effects on ventilation of decreasing the alveolar PO2 while holding the alveolar PCO2 constant are shown in Figure 36–10. When the alveolar PCO2 is stabilized at a level 2–3 mm Hg above normal, there is an inverse relationship between ventilation and the alveolar PO2 even in the 90–110 mm Hg range; but when the alveolar PCO2 is fixed at lower than normal values, there is no stimulation of ventilation by hypoxia until the alveolar PO2 falls below 60 mm Hg.


FIGURE 36–10 Ventilation at various alveolar PO2 values when PCO2 is held constant at 49, 44, or 37 mm Hg. Note the dramatic effect on the ventilatory response to PAO2 when PACO2 is increased above normal. (Data from Loeschke HH and Gertz KH).


When the converse experiment is performed—that is, when the alveolar PO2 is held constant while the response to varying amounts of inspired CO2 is tested—a linear response is obtained (Figure 36–11). When the CO2 response is tested at different fixed PO2 values, the slope of the response curve changes, with the slope increased when alveolar PO2 is decreased. In other words, hypoxia makes the individual more sensitive to increases in arterial PCO2. However, the alveolar PCO2 level at which the curves in Figure 36–11 intersect is unaffected. In the normal individual, this threshold value is just below the normal alveolar PCO2, indicating that normally there is a very slight but definite “CO2 drive” of the respiratory area.


FIGURE 36–11 Fan of lines showing CO2 response curves at various fixed values of alveolar PO2. Decreased PAO2 results in a more sensitive response to PACO2.


The stimulatory effects of H+ and CO2 on respiration appear to be additive and not, like those of CO2 and O2, complexly interrelated. In metabolic acidosis, the CO2 response curves are similar to those in Figure 36–11, except that they are shifted to the left. In other words, the same amount of respiratory stimulation is produced by lower arterial PCO2 levels. It has been calculated that the CO2 response curve shifts 0.8 mm Hg to the left for each nanomole rise in arterial H+. About 40% of the ventilatory response to CO2 is removed if the increase in arterial H+ produced by CO2 is prevented. As noted above, the remaining 60% is probably due to the effect of CO2 on spinal fluid or brain interstitial fluid H+ concentration.


Respiration can be voluntarily inhibited for some time, but eventually the voluntary control is overridden. The point at which breathing can no longer be voluntarily inhibited is called the breaking point. Breaking is due to the rise in arterial PCO2 and the fall in PO2. Individuals can hold their breath longer after removal of the carotid bodies. Breathing 100% oxygen before breath holding raises alveolar PO2 initially, so that the breaking point is delayed. The same is true of hyperventilating room air, because CO2 is blown off and arterial PCO2 is lower at the start. Reflex or mechanical factors appear to influence the breaking point, since subjects who hold their breath as long as possible and then breathe a gas mixture low in O2 and high in CO2 can hold their breath for an additional 20 s or more. Psychological factors also play a role, and subjects can hold their breath longer when they are told their performance is very good than when they are not.



Receptors in the airways and lungs are innervated by myelinated and unmyelinated vagal fibers. The unmyelinated fibers are C fibers. The receptors innervated by myelinated fibers are commonly divided into slowly adapting receptors and rapidly adapting receptors on the basis of whether sustained stimulation leads to prolonged or transient discharge in their afferent nerve fibers (Table 36–2). The other group of receptors presumably consists of the endings of C fibers, and they are divided into pulmonary and bronchial subgroups on the basis of their location.


TABLE 36–2 Airway and lung receptors.

The shortening of inspiration produced by vagal afferent activity (Figure 36–3) is mediated by slowly adapting receptors, as are the Hering–Breuer reflexes. The Hering–Breuer inflation reflex is an increase in the duration of expiration produced by steady lung inflation, and the Hering–Breuer deflation reflex is a decrease in the duration of expiration produced by marked deflation of the lung. Because the rapidly adapting receptors are stimulated by chemicals such as histamine, they have been called irritant receptors. Activation of rapidly adapting receptors in the trachea causes coughing, bronchoconstriction, and mucus secretion, and activation of rapidly adapting receptors in the lung may produce hyperpnea.

Because the C fiber endings are close to pulmonary vessels, they have been called J (juxtacapillary) receptors. They are stimulated by hyperinflation of the lung, but they respond as well to intravenous or intracardiac administration of chemicals such as capsaicin. The reflex response that is produced is apnea followed by rapid breathing, bradycardia, and hypotension (pulmonary chemoreflex). A similar response is produced by receptors in the heart (Bezold–Jarisch reflex or the coronary chemoreflex). The physiologic role of this reflex is uncertain, but it probably occurs in pathologic states such as pulmonary congestion or embolization, in which it is produced by endogenously released substances.


Coughing begins with a deep inspiration followed by forced expiration against a closed glottis. This increases the intrapleural pressure to 100 mm Hg or more. The glottis is then suddenly opened, producing an explosive outflow of air at velocities up to 965 km (600 mi) per hour. Sneezing is a similar expiratory effort with a continuously open glottis. These reflexes help expel irritants and keep airways clear. Other aspects of innervation are considered in a special case (Clinical Box 36–1).


Lung Innervation & Patients with Heart–Lung Transplants

Transplantation of the heart and lungs is now an established treatment for severe pulmonary disease and other conditions. In individuals with transplants, the recipient’s right atrium is sutured to the donor heart, and the donor heart does not reinnervate, so the resting heart rate is elevated. The donor trachea is sutured to the recipient’s just above the carina, and afferent fibers from the lungs do not regrow. Consequently, healthy patients with heart–lung transplants provide an opportunity to evaluate the role of lung innervation in normal physiology. Their cough responses to stimulation of the trachea are normal because the trachea remains innervated, but their cough responses to stimulation of the smaller airways are absent. Their bronchi tend to be dilated to a greater degree than normal. In addition, they have the normal number of yawns and sighs, indicating that these do not depend on innervation of the lungs. Finally, they lack Hering–Breuer reflexes, but their pattern of breathing at rest is normal, indicating that these reflexes do not play an important role in the regulation of resting respiration in humans.


Carefully controlled experiments have shown that active and passive movements of joints stimulate respiration, presumably because impulses in afferent pathways from proprioceptors in muscles, tendons, and joints stimulate the inspiratory neurons. This effect probably helps increase ventilation during exercise. Other afferents are considered in Clinical Box 36–2.


Afferents from “Higher Centers”

Pain and emotional stimuli affect respiration, suggesting that afferents from the limbic system and hypothalamus signal to the respiratory neurons in the brain stem. In addition, even though breathing is not usually a conscious event, both inspiration and expiration are under voluntary control. The pathways for voluntary control pass from the neo-cortex to the motor neurons innervating the respiratory muscles, bypassing the medullary neurons.

Because voluntary and automatic control of respiration are separate, automatic control is sometimes disrupted without loss of voluntary control. The clinical condition that results has been called Ondine’s curse. In German legend, Ondine was a water nymph who had an unfaithful mortal lover. The king of the water nymphs punished the lover by casting a curse on him that took away all his automatic functions. In this state, he could stay alive only by staying awake and remembering to breathe. He eventually fell asleep from sheer exhaustion, and his respiration stopped. Patients with this intriguing condition generally have bulbar poliomyelitis or disease processes that compress the medulla.


Inhibition of respiration and closure of the glottis during vomiting, swallowing, and sneezing not only prevent the aspiration of food or vomitus into the trachea but, in the case of vomiting, fix the chest so that contraction of the abdominal muscles increases the intra-abdominal pressure. Similar glottic closure and inhibition of respiration occur during voluntary and involuntary straining.

Hiccup is a spasmodic contraction of the diaphragm and other inspiratory muscles that produces an inspiration during which the glottis suddenly closes. The glottic closure is responsible for the characteristic sensation and sound. Hiccups occur in the fetus in utero as well as throughout extrauterine life. Their function is unknown. Most attacks of hiccups are usually of short duration, and they often respond to breath holding or other measures that increase arterial PCO2. Intractable hiccups, which can be debilitating, sometimes respond to dopamine antagonists and perhaps to some centrally acting analgesic compounds.

Yawning is a peculiar “infectious” respiratory act whose physiologic basis and significance are uncertain. Like hiccuping, it occurs in utero, and it occurs in fish and tortoises as well as mammals. The view that it is needed to increase O2 intake has been discredited. Underventilated alveoli have a tendency to collapse, and it has been suggested that the deep inspiration and stretching them open prevents the development of atelectasis. However, in actual experiments, no atelectasis-preventing effect of yawning could be demonstrated. Yawning increases venous return to the heart, which may benefit the circulation. It has been suggested that yawning is a nonverbal signal used for communication between monkeys in a group, and one could argue that on a different level, the same thing is true in humans.


Afferent fibers from the baroreceptors in the carotid sinuses, aortic arch, atria, and ventricles relay to the respiratory neurons, as well as the vasomotor and cardioinhibitory neurons in the medulla. Impulses in them inhibit respiration, but the inhibitory effect is slight and of little physiologic importance. The hyperventilation in shock is due to chemoreceptor stimulation caused by acidosis and hypoxia secondary to local stagnation of blood flow, and is not baroreceptor-mediated. The activity of inspiratory neurons affects blood pressure and heart rate, and activity in the vasomotor and cardiac areas in the medulla may have minor effects on respiration.


Respiration is less rigorously controlled during sleep than in the waking state, and brief periods of apnea occur in normal sleeping adults. Changes in the ventilatory response to hypoxia vary. If the PCO2 falls during the waking state, various stimuli from proprioceptors and the environment maintain respiration, but during sleep, these stimuli are decreased and a decrease in PCO2 can cause apnea. During rapid eye movement (REM) sleep, breathing is irregular and the CO2 response is highly variable.



In asphyxia produced by occlusion of the airway, acute hypercapnia and hypoxia develop together. Stimulation of respiration is pronounced, with violent respiratory efforts. Blood pressure and heart rate rise sharply, catecholamine secretion is increased, and blood pH drops. Eventually the respiratory efforts cease, the blood pressure falls, and the heart slows. Asphyxiated animals can still be revived at this point by artificial respiration, although they are prone to ventricular fibrillation, probably because of the combination of hypoxic myocardial damage and high circulating catecholamine levels. If artificial respiration is not started, cardiac arrest occurs in 4–5 min.


Drowning is asphyxia caused by immersion, usually in water. In about 10% of drownings, the first gasp of water after the losing struggle not to breathe triggers laryngospasm, and death results from asphyxia without any water in the lungs. In the remaining cases, the glottic muscles eventually relax and fluid enters the lungs. Fresh water is rapidly absorbed, diluting the plasma and causing intravascular hemolysis. Ocean water is markedly hypertonic and draws fluid from the vascular system into the lungs, decreasing plasma volume. The immediate goal in the treatment of drowning is, of course, resuscitation, but long-term treatment must also take into account the circulatory effects of the water in the lungs.


The acute effects of voluntary hyperventilation demonstrate the interaction of the chemical mechanisms regulating respiration. When a normal individual hyperventilates for 2–3 min, then stops and permits respiration to continue without exerting any voluntary control over it, a period of apnea occurs. This is followed by a few shallow breaths and then by another period of apnea, followed again by a few breaths (periodic breathing). The cycles may last for some time before normal breathing is resumed (Figure 36–12). The apnea apparently is due to a lack of CO2 because it does not occur following hyperventilation with gas mixtures containing 5% CO2. During the apnea, the alveolar PO2 falls and the PCO2 rises. Breathing resumes because of hypoxic stimulation of the carotid and aortic chemoreceptors before the CO2 level has returned to normal. A few breaths eliminate the hypoxic stimulus, and breathing stops until the alveolar PO2 falls again. Gradually, however, the PCO2 returns to normal, and normal breathing resumes. Changes in breathing patterns can be symptomatic of disease (Clinical Box 36–3).


FIGURE 36–12 Changes in breathing and composition of alveolar air after forced hyperventilation for 2 min. Bars in bottom indicate breathing, whereas blank spaces are indicative of apnea.


Periodic Breathing in Disease

Cheyne–Stokes Respiration

Periodic breathing occurs in various disease states and is often called Cheyne–Stokes respiration. It is seen most commonly in patients with congestive heart failure and uremia, but it occurs also in patients with brain disease and during sleep in some normal individuals. Some of the patients with Cheyne–Stokes respiration have increased sensitivity to CO2. The increased response is apparently due to disruption of neural pathways that normally inhibit respiration. In these individuals, CO2 causes relative hyperventilation, lowering the arterial PCO2. During the resultant apnea, the arterial PCO2 again rises to normal, but the respiratory mechanism again overresponds to CO2. Breathing ceases, and the cycle repeats.

Another cause of periodic breathing in patients with cardiac disease is prolongation of the lung-to-brain circulation time, so that it takes longer for changes in arterial gas tensions to affect the respiratory area in the medulla. When individuals with a slower circulation hyperventilate, they lower the PCO2 of the blood in their lungs, but it takes longer than normal for the blood with a low PCO2 to reach the brain. During this time, the PCO2 in the pulmonary capillary blood continues to be lowered, and when this blood reaches the brain, the low PCO2 inhibits the respiratory area, producing apnea. In other words, the respiratory control system oscillates because the negative feedback loop from lungs to brain is abnormally long.

Sleep Apnea

Episodes of apnea during sleep can be central in origin (ie, due to failure of discharge in the nerves producing respiration) or they can be due to airway obstruction (obstructive sleep apnea). Apnea can occur at any age and is produced when the pharyngeal muscles relax during sleep. In some cases, failure of the genioglossus muscles to contract during inspiration contributes to the blockage. The genioglossus muscles pull the tongue forward, and without (or with weak) contraction the tongue can obstruct the airway. After several increasingly strong respiratory efforts, the patient wakes up, takes a few normal breaths, and falls back to sleep. Apneic episodes are most common during REM sleep, when the muscles are most hypotonic. The symptoms are loud snoring, morning headaches, fatigue, and daytime sleepiness. When severe and prolonged, the condition can lead to hypertension and its complications. Frequent apneas can lead to numerous brief awakenings during sleep and to sleepiness during waking hours. With this in mind, it is not surprising to find that the incidence of motor vehicle accidents in sleep apnea patients is seven times greater than it is in the general driving population.


Treatment of sleep apnea is dependent on the patient and on the cause (if known). Treatments range from mild to moderate interventions to surgery. Interventions including positional therapy, dental appliances that rearrange the architecture of the airway, avoidance of muscle relaxants (eg, alcohol) or drugs that reduce respiratory drive, or continuous positive airway pressure. Because sleep apnea is increased in overweight or obese individuals, weight loss can also be effective.


Exercise provides a physiological example to explore many of the control systems discussed above. Of course, many cardiovascular and respiratory mechanisms must operate in an integrated fashion if the O2 needs of the active tissue are to be met and the extra CO2 and heat removed from the body during exercise. Circulatory changes increase muscle blood flow while maintaining adequate circulation in the rest of the body. In addition, there is an increase in the extraction of O2 from the blood in exercising muscles and an increase in ventilation. This provides extra O2, eliminates some of the heat, and excretes extra CO2. A focus on regulation of ventilation and tissue O2 is presented below, as many other aspects of regulation have been presented in previous chapters.


During exercise, the amount of O2 entering the blood in the lungs is increased because the amount of O2 added to each unit of blood and the pulmonary blood flow per minute are increased. The PO2 of blood flowing into the pulmonary capillaries falls from 40 to 25 mm Hg or less, so that the alveolar–capillary PO2 gradient is increased and more O2 enters the blood. Blood flow per minute is increased from 5.5 L/min to as much as 20–35 L/min. The total amount of O2 entering the blood therefore increases from 250 mL/min at rest to values as high as 4000 mL/min. The amount of CO2 removed from each unit of blood is increased, and CO2 excretion increases from 200 mL/min to as much as 8000 mL/min. The increase in O2 uptake is proportional to work load, up to a maximum. Above this maximum, O2 consumption levels off and the blood lactate level continues to rise (Figure 36–13). The lactate comes from muscles in which aerobic resynthesis of energy stores cannot keep pace with their utilization, and an oxygen debt is being incurred.


FIGURE 36–13 Relation between work load, blood lactate level, and O2 uptake. I–VI, increasing work loads produced by increasing the speed and grade of a treadmill on which the subjects worked. (Reproduced with permission from Mitchell JH, Blomqvist G: Maximal oxygen uptake. N Engl J Med 1971;284:1018.)

Ventilation increases abruptly with the onset of exercise, which is followed after a brief pause by a further, more gradual increase (Figure 36–14). With moderate exercise, the increase is due mostly to an increase in the depth of respiration; this is accompanied by an increase in the respiratory rate when the exercise is more strenuous. Ventilation abruptly decreases when exercise ceases, which is followed after a brief pause by a more gradual decline to preexercise values. The abrupt increase at the start of exercise is presumably due to psychic stimuli and afferent impulses from proprioceptors in muscles, tendons, and joints. The more gradual increase is presumably humoral, even though arterial pH, PCO2, and PO2 remain constant during moderate exercise. The increase in ventilation is proportional to the increase in O2 consumption, but the mechanisms responsible for the stimulation of respiration are still the subject of much debate. The increase in body temperature may play a role. Exercise increases the plasma K+ level, and this increase may stimulate the peripheral chemoreceptors. In addition, it may be that the sensitivity of the neurons controlling the response to CO2 is increased or that the respiratory fluctuations in arterial PCO2 increase so that, even though the mean arterial PCO2 does not rise, it is CO2 that is responsible for the increase in ventilation. O2also seems to play some role, despite the lack of a decrease in arterial PO2, since during the performance of a given amount of work, the increase in ventilation while breathing 100% O2 is 10–20% less than the increase while breathing air. Thus, it currently appears that a number of different factors combine to produce the increase in ventilation seen during moderate exercise.


FIGURE 36–14 Diagrammatic representation of changes in ventilation during exercise. See text for details.

When exercise becomes more vigorous, buffering of the increased amounts of lactic acid that are produced liberates more CO2, and this further increases ventilation. The response to graded exercise is shown in Figure 36–15.With increased production of acid, the increases in ventilation and CO2 production remain proportional, so alveolar and arterial CO2 change relatively little (isocapnic buffering). Because of the hyperventilation, alveolar PO2increases. With further accumulation of lactic acid, the increase in ventilation outstrips CO2 production and alveolar PCO2 falls, as does arterial PCO2. The decline in arterial PCO2 provides respiratory compensation for the metabolic acidosis produced by the additional lactic acid. The additional increase in ventilation produced by the acidosis is dependent on the carotid bodies and does not occur if they are removed.


FIGURE 36–15 Physiologic responses to work rate during exercise. Changes in alveolar PCO2, alveolar PO, ventilation image consumption production image consumption image arterial image, and arterial pH with graded increases in work by an adult male on a bicycle ergometer. Resp comp, respiratory compensation. STPD, standard temperature (0°C) and pressure (760 mm Hg), dry. Dashed lines emphasize deviation from linear response. See text for additional details. (Reproduced with permission from Wasserman K: Breathing during exercise. NEJM 1978 Apr 6;298(14):780–785.)

The respiratory rate after exercise does not reach basal levels until the O2 debt is repaid. This may take as long as 90 min. The stimulus to ventilation after exercise is not the arterial PCO2, which is normal or low, or the arterial PO2, which is normal or high, but the elevated arterial H+ concentration due to the lactic acidemia. The magnitude of the O2 debt is the amount by which O2 consumption exceeds basal consumption from the end of exertion until the O2consumption has returned to preexercise basal levels. During repayment of the O2 debt, the O2 concentration in muscle myoglobin rises slightly. ATP and phosphorylcreatine are resynthesized, and lactic acid is removed. Eighty per cent of the lactic acid is converted to glycogen and 20% is metabolized to CO2 and H2O.


Maximum O2 uptake during exercise is limited by the maximum rate at which O2 is transported to the mitochondria in the exercising muscle. However, this limitation is not normally due to deficient O2 uptake in the lungs, and hemoglobin in arterial blood is saturated even during the most severe exercise.

During exercise, the contracting muscles use more O2, and the tissue PO2 and the PO2 in venous blood from exercising muscle fall nearly to zero. More O2 diffuses from the blood, the blood PO2 of the blood in the muscles drops, and more O2 is removed from hemoglobin. Because the capillary bed of contracting muscle is dilated and many previously closed capillaries are open, the mean distance from the blood to the tissue cells is greatly decreased; this facilitates the movement of O2 from blood to cells. The oxygen–hemoglobin dissociation curve is steep in the PO2 range below 60 mm Hg, and a relatively large amount of O2 is supplied for each drop of 1 mm Hg in PO2 (see Figure 35–2). Additional O2 is supplied because, as a result of the accumulation of CO2 and the rise in temperature in active tissues—and perhaps because of a rise in red blood cell 2,3-diphosphoglycerate (2,3-DPG)—the dissociation curve shifts to the right. The net effect is a threefold increase in O2 extraction from each unit of blood (see Figure 35–3). Because this increase is accompanied by a 30-fold or greater increase in blood flow, it permits the metabolic rate of muscle to rise as much as 100-fold during exercise.


What determines the maximum amount of exercise that can be performed by an individual? Obviously, exercise tolerance has a time as well as an intensity dimension. For example, a fit young man can produce a power output on a bicycle of about 700 W for 1 min, 300 W for 5 min, and 200 W for 40 min. It used to be argued that the limiting factors in exercise performance were the rate at which O2 could be delivered to the tissues or the rate at which O2 could enter the body in the lungs. These factors play a role, but it is clear that other factors also contribute and that exercise stops when the sensation of fatigue progresses to the sensation of exhaustion. Fatigue is produced in part by bombardment of the brain by neural impulses from muscles, and the decline in blood pH produced by lactic acidosis also makes one feel tired, as does the rise in body temperature, dyspnea, and, perhaps, the uncomfortable sensations produced by activation of the J receptors in the lungs.


image Breathing is under both voluntary control (located in the cerebral cortex) and automatic control (driven by pacemaker cells in the medulla). There is a reciprocal innervation to expiratory and inspiratory muscles in that motor neurons supplying expiratory muscles are inactive when motor neurons supplying inspiratory muscles are active, and vice versa.

image The pre-Bötzinger complex on either side of the medulla contains synaptically coupled pacemaker cells that allow for rhythmic generation of breathing. The spontaneous activity of these neurons can be altered by neurons in the pneumotaxic center, although the full regulatory function of these neurons on normal breathing is not understood.

image Breathing patterns are sensitive to chemicals in the blood through activation of respiratory chemoreceptors. There are chemoreceptors in the carotid and aortic bodies and in collections of cells in the medulla. These chemoreceptors respond to changes in PO2 and PCO2 as well as H+ to regulate breathing.

image Receptors in the airway are additionally innervated by slowly adapting and rapidly adapting myelinated vagal fibers. Slowly adapting receptors can be activated by lung inflation. Rapidly adapting receptors, or irritant receptors, can be activated by chemicals such as histamine and result in cough or even hyperpnea.

image Receptors in the airway are also innervated by unmyelinated vagal fibers (C fibers) that are typically found next to pulmonary vessels. They are stimulated by hyperinflation (or exogenous substances including capsaicin) and lead to the pulmonary chemoreflex. The physiologic role for this response is not fully understood.


For all questions, select the single best answer unless otherwise directed.

1. The main respiratory control neurons

A. send out regular bursts of impulses to expiratory muscles during quiet respiration.

B. are unaffected by stimulation of pain receptors.

C. are located in the pons.

D. send out regular bursts of impulses to inspiratory muscles during quiet respiration.

E. are unaffected by impulses from the cerebral cortex.

2. Intravenous lactic acid increases ventilation. The receptors responsible for this effect are located in the

A. medulla oblongata.

B. carotid bodies.

C. lung parenchyma.

D. aortic baroreceptors.

E. trachea and large bronchi.

3. Spontaneous respiration ceases after

A. transection of the brain stem above the pons.

B. transection of the brain stem at the caudal end of the medulla.

C. bilateral vagotomy.

D. bilateral vagotomy combined with transection of the brain stem at the superior border of the pons.

E. transection of the spinal cord at the level of the first thoracic segment.

4. The following physiologic events that occur in vivo are listed in random order: (1) decreased CSF pH; (2) increased arterial PCO2(3) increased CSF PCO2(4) stimulation of medullary chemoreceptors; (5) increased alveolar PCO2.
    What is the usual sequence in which they occur when they affect respiration?

A. 1, 2, 3, 4, 5

B. 4, 1, 3, 2, 5

C. 3, 4, 5, 1, 2

D. 5, 2, 3, 1, 4

E. 5, 3, 2, 4, 1

5. The following events that occur in the carotid bodies when they are exposed to hypoxia are listed in random order: (1) depolarization of type I glomus cells; (2) excitation of afferent nerve endings; (3) reduced conductance of hypoxia-sensitive K+ channels in type I glomus cells; (4) Ca2+ entry into type I glomus cells; (5) decreased K+ efflux.
    What is the usual sequence in which they occur on exposure to hypoxia?

A. 1, 3, 4, 5, 2

B. 1, 4, 2, 5, 3

C. 3, 4, 5, 1, 2

D. 3, 1, 4, 5, 2

E. 3, 5, 1, 4, 2

6. Injection of a drug that stimulates the carotid bodies would be expected to cause

A. a decrease in the pH of arterial blood.

B. a decrease in the PCO2 of arterial blood.

C. an increase in the image concentration of arterial blood.

D. an increase in urinary Na+ excretion.

E. an increase in plasma Cl.

7. Variations in which of the following components of blood or CSF do not affect respiration?

A. Arterial image concentration

B. Arterial H+ concentration

C. Arterial Na+ concentration

D. CSF CO2 concentration

E. CSF H+ concentration


Barnes PJ: Chronic obstructive pulmonary disease. N Engl J Med 2000;343:269.

Crystal RG, West JB (editors): The Lung: Scientific Foundations, 2nd ed. Lippincott-Raven, 1997.

Fishman AP, Elias JA, Fishman JA, et al. (editors): Fishman’s Pulmonary Diseases and Disorders, 4th ed. McGraw-Hill, 2008.

Jones NL, Killian KJ: Exercise limitation in health and disease. N Engl J Med 2000;343:632.

Laffey JG, Kavanagh BP: Hypocapnia. N Engl J Med 2002;347:43.

Putnam RW, Dean JB, Ballantyne D (editors): Central chemosensitivity. Respir Physiol 2001;129:1.

Rekling JC, Feldman JL: Pre-Bötzinger complex and pacemaker neurons: Hypothesized site and kernel for respiratory rhythm generation. Annu Rev Physiol 1998;60:385.

West, JB: Respiratory Physiology: The Essentials, 8th ed. Wolers Kluwer/Lippincott Williams & Wilkins, 2008.