Chapter 11: Central and Peripheral Neural Controls of Respiration

As demonstrated in earlier chapters, ventilation is simple in concept but complex in execution. The brain controls the basic pattern of breathing, integrating multiple influences within lower motor neurons of the brainstem and spinal cord to drive pharyngeal, laryngeal, diaphragmatic, intercostal, and other respiratory muscles. Recall that V̇_E (L/min) = V_T · f. The central nervous system regulates respiration by controlling the rhythm and pattern of its output to respiratory muscles, adjusting f, V_T or both, depending on overall ventilatory needs for a greater or lesser V̇_E (Fig. 11.1).

The frequency of respiration, or its rhythm, is intrinsic to the brainstem. All vertebrates that use tidal oscillations for the exchange of O₂ and CO₂ in their lungs have such movements generated in their medulla oblongata. Indeed surgically isolated brainstem preparations, lacking afferent inputs from chemoreceptors and mechanoreceptors, still produce rhythmic outputs along the same cranial nerves as during normal respiration. One critical rhythm generator within a small area of the medulla and rostral to the obex is called the pre-Bötzinger complex. Bilateral lesions in this portion of the medulla induce complete respiratory arrest in humans. Whether this rhythmic discharge initiates within individual pacemaker cells, or from a network of such cells, remains a matter of debate. At this point, it is important to understand that a rhythm is constantly generated by the medulla, a rhythm that is modifiable by afferent input from sensory receptors.

The central pattern generator, or the brainstem output controlling all muscles involved in respiration, is much more complex. This pattern generator algebraically sums all the afferent inputs to produce well-coordinated activations of the diaphragm, intercostal muscles, and abdominal muscles, and if needed, the accessory muscles of respiration. Like respiratory rhythm, the goal of such pattern generation is maintenance of normal P_aO₂, P_aco₂, and pH_a. Extreme examples of modulating both the rhythm and pattern of respiration are seen during rigorous exercise and ascents to high altitude. The pattern of respiration is also modulated by events like coughing, speech, sleep, vomiting, micturition, and defecation, particularly as the latter may mimic a Valsalva maneuver. Although some of these events are episodic or relatively infrequent, they can affect the normal respiratory pattern in dramatic ways.

• Eupnea: is normal, quiet breathing at rest. Individuals are usually unaware of it.

• Tachypnea or polypnea: is an increase in f without an increase in V_T. Tachypnea is not a normal stress response, unless hyperthermia or other factor has induced panting.

• Hyperpnea or hyperventilation: denotes an increase in pulmonary ventilation involving both V_T and f, but without the subjectively stressful sensations of dyspnea (Fig. 11.2).

• Dyspnea: is the sensation of inadequate or stressful respiration, with exaggerated awareness of one's need for increased respiratory effort. Dyspnea implies labored breathing, often involving accessory respiratory muscles. Many stimuli induce dyspnea.

• Cheyne-Stokes respiration: is the most common form of abnormal breathing, with weak respiratory efforts that decrease to an apnea and then increase to hyperpnea. The "crescendo-decrescendo" oscillations of Cheyne-Stokes are most often caused by hypoxemia and are a frequently encountered symptom (see Chap. 25) (Fig. 11.2).

• Apneusis: is an abnormally patterned breathing with prolonged inspirations that alternate with short expiratory movements (Fig. 11.2). Apneustic breathing is commonly noted after lesions in the pontine pneumotaxic center discussed below.

• Ataxis or ataxic respiration: is an abnormal pattern with completely irregular breathing and increasing periods of apnea. As the pattern deteriorates, it may merge with agonal respiration. It is caused by damage to the medulla oblongata by stroke or trauma.

• Apnea: is the absence of breathing. As generally used, apnea implies that the cessation is temporary. A prolonged apnea for any cause is considered respiratory arrest.

The Spinal Cord

Spinal cord neurons are organized into ventral horn (motor), dorsal horn (sensory), and lateral horn (autonomic) regions. Those in the ventral horn are most critical to respiration, since they include lower motor neurons innervating somatic striated muscles (Fig. 11.3). Major somatic striated muscles involved in respiration include:

• Diaphragm: its innervating motoneurons exit at C3-C5 as the phrenic nerve.

• Intercostal muscles: innervated by motoneurons within the thoracic ventral horn with axons that exit the spinal cord and distribute via the intercostal nerves.

• Abdominal muscles: their motoneurons have axons that track within the lower thoracic and upper lumbar cord regions.

• Accessory muscles: include all muscles that elevate and splay the ribs, notably the levator costalis, scalene, transverse thoracic, and sternocleidomastoid.

The Medulla

In addition to its role as the principal area of sensorimotor integration for respiration, the medulla contains lower motor neurons with fibers that exit via cranial nerves to innervate striated muscles in the head and neck (Fig 11.4). Motor neurons innervating the tongue muscles are found in the hypoglossal nucleus, while those innervating laryngeal, pharyngeal and facial muscles are found in the ventrolateral medulla. All of these muscles receive rhythmic central nervous system (CNS) motor input during every breath. Major cranial striated muscles that are involved during normal respiration include:

• Laryngeal muscles: both laryngeal abductors and adductors are innervated by motoneurons in the nucleus ambiguus, whose axons travel with the vagus nerve.

• Pharyngeal muscles: also receive motor input from the nucleus ambiguus by neurons whose axons exit via the glossopharyngeal and vagus nerves.

• Facial muscles: most notably the m. nasalis, by motoneurons within the facial motor nucleus whose axons exit via the facial nerve.

• Tongue muscles: principally the genioglossus muscle by motoneurons in the hypoglossal nucleus whose axons exit via the hypoglossal nerve.

Within the medulla are two recognized groups of neurons involved in the integration and coordination of breathing, whose functions continue to be intensively investigated. The first of these, the ventral respiratory column, is a longitudinal array of respiratory-related neurons that fire synchronously with each phrenic nerve discharge. These neurons are found in the ventrolateral reticular formation of the medulla, generally just ventral to the nucleus ambiguus. A subject's basic respiratory rhythm persists if only these brainstem neurons are intact, albeit poorly controlled. Four main components of the ventral respiratory column have been identified (Fig. 11.5):

• Caudal ventral respiratory group (cVRG): an expiratory area running from the spino-medullary junction to the obex;

• Rostral ventral respiratory group (rVRG): the area of mixed inspiratory and expiratory respiratory neurons just rostral to the obex;

• Pre-Bötzinger Complex: an area of neurons that is rostral to the rVRG and considered of central importance for rhythm generation;

• Bötzinger Complex: an expiratory area just caudal to the facial nucleus.

In addition to the ventral respiratory column, a second brainstem area of importance to respiratory control contains the dorsal respiratory group (DRG), being inspiratory neurons in the ventrolateral part of the nucleus tractus solitarii (NTS) (Fig 11.5).

The Pons

Situated rostral to the VRG and DRG neurons of the brainstem, the pons contains the parabrachial nucleus and the Kölliker-Fuse area, a neurophil surrounding the brachium conjunctivum (Fig. 11.6). These two regions contain neurons considered important as the main "off-switch" for spontaneous inspiration, and have been called the pontine pneumotaxic center because lesions here result in apneustic respiration.

Pontine and medullary respiratory neurons are quite interconnected, making it difficult to assign unambiguous functions to any particular group of neurons. Investigations of respiratory control usually are conducted while eliminating variables that are known to affect the rhythm or pattern of respiration. This usually means maintaining constant P_ao₂ or P_aco₂ levels in animals that are anesthetized, vagotomized, and sometimes spinalized. Despite such limitations, it is clear that afferent inputs from higher brain areas, as well as from chemoreceptors and mechanoreceptors, converge on this network of brainstem respiratory neurons (Fig. 11.5). There they produce a pattern of muscle contractions and thus ventilation that are appropriate for metabolic needs.

Respiratory rhythm generation is intrinsic to the medulla and proceeds even without additional sensory input. However, central respiratory neurons are modulated by afferent inputs affecting the depth, rate, and pattern of respiration. Chemo-receptors respond to changes in the composition of blood or other fluids around them. Major groups of chemoreceptors are located in the peripheral and central nervous systems.

Peripheral chemoreceptors are located in parenchymal lobules termed the carotid bodies above the bifurcations of the common carotid arteries (Fig. 11.7), and the aortic bodies located along the superior aspect of the aortic arch. Although the lobules are organized similarly at each location, the carotid bodies send afferent impulses via the carotid sinus nerve branch of cranial nerve IX, while the aortic bodies send signals via cranial nerve X to the NTS in the CNS. In general terms, these peripheral chemoreceptors respond quickly to decreasing P_ao₂ and pH_a, and increasing P_aco₂, with discharge rates alterable during a single respiratory cycle. Importantly, they cause all increased ventilation in response to arterial hypoxemia, although their effect is not appreciable until P_ao₂ declines to ~40 mm Hg. Thus, their role in regulating eupneic breathing is small. It is also thought that their response to increased P_aco₂ is less important than that of central chemoreceptors.

The major populations of central chemoreceptors are located near the ventral surface of the medulla, many near levels of the exiting hypoglossal nerve (Fig. 11.8). They are bathed in brain extracellular fluid (ECF) that rapidly equilibrates with gaseous CO₂ diffusing from blood vessels into cerebrospinal fluid (CSF). This local rise in Pco₂ acidifies CSF and thus stimulates the chemoreceptors, even though [H⁺] and [HCO₃^–] do not readily cross the blood-brain barrier. In this manner, a reduction in pH_a stimulates ventilation centrally while an increased pH_a is inhibitory (Fig. 11.8).

Systemically the body seldom experiences isolated hypoxia, hypercapnia, or acidemia. Moreover, the ventilatory response to a set of variables is greater than the response to each component, due to integration of these modulatory influences within the CNS.

There are several important types of sensory receptors associated with afferent fibers in the vagus nerve that respond to either mechanical or chemical stimulation of the tracheobronchial tree, including its respiratory mucosa (Table 11.1).

Important subtypes of receptors within the lower airways and/or the lung parenchyma include three main groups:

• Slowly adapting pulmonary stretch receptors (SARs): Myelinated SARs are situated among airway smooth muscle cells and send afferent fibers centrally via the vagus nerve. They discharge in response to distension of the lung, and show little adaptation over time to sustained lung inflation. Activation of SARs reduces respiratory frequency by increasing the expiratory time interval, the so-called Hering-Breuer reflex. These SARs are inactive in awake subjects unless V_T is large (>1 L), but they may be prominent in anesthetized animal preparations.

• Rapidly adapting pulmonary receptors (RARs), also termed by some authors as irritant receptors: Myelinated RARs lie between airway epithelial cells and also send afferent fibers centrally via cranial nerve X (vagus). Some RARs are mechanoreceptors that fire during hyperinflation or forced deflation of the lungs. Most are chemoreceptive RARs activated by exogenous agents (eg, noxious gases, cigarette smoke, and cold air) and endogenous chemicals (eg, histamine, prostaglandins, and serotonin). Their reflex effects include bronchoconstriction, hyperpnea, and cough. Indeed, a presumptive cough receptor has been discovered recently. It is not clear whether RARs function during eupneic respiration. However, their responsiveness to inflammatory mediators and other endogenous chemicals indicate they may be important in disease (Chap. 10).

• Tracheobronchial C-fibers: Unmyelinated C-fibers comprise >75% of afferent fibers emerging from the lungs. Their nerve endings generally function as chemoreceptors that are sensitive to histamine, prostaglandins, bradykinin, and serotonin, as well as cigarette smoke, noxious gases, and capsaicin. Thus they probably play a prominent role in asthma, allergic bronchoconstriction, pulmonary vascular congestion, and pulmonary embolism. Although C-fibers are relatively inactive during eupneic breathing, they respond to lung hyperinflation and can induce reflex apnea (followed by tachypnea), bronchoconstriction, hypotension, bradycardia, and mucus secretion.

• Receptors of the upper airways: There are unencapsulated (free) nerve endings arrayed within the respiratory mucosa from the nasal cavity to the larynx (Chap. 2). Their afferent fibers travel via the trigeminal, glossopharyngeal, and vagus nerves and respond to chemical and mechanical stimuli. Induced respiratory responses caused by their stimulation include sneezing, coughing, bronchoconstriction, and apnea, depending on stimulus nature and the region within the upper respiratory tract being stimulated. Such sensory innervations of the larynx via the superior laryngeal nerve are of considerable clinical importance, since these receptors monitor airflow, laryngeal temperature, and other mechanical and chemical stimuli (Fig. 11.9).

• Receptors in muscle: Dynamic or static contractions of limb muscles induce many autonomic effects including hyperpnea, although P_aco₂, P_ao₂, and pH_a change little during moderate exercise. Moreover, experimental data indicate that neither the carotid chemoreceptors nor respiratory muscle afferent fibers provide the primary stimulus for such hypernea. Rather, most evidence suggests that the small myelinated and unmyelinated fibers of sensory nerves (Groups III and IV) that innervate striated muscle initiate this response. It also has been proposed that a central command mechanism may initiate hypernea during exercise. This topic is addressed further in Chap. 13.

Sleep is a complex diurnal rhythm having a different time scale that is superimposed on the normal respiratory rhythm. Although sleep consists of various stages, it is often simplified to be either rapid eye movement (REM) sleep or non-REM sleep. Respiration slows and chemoreception is blunted during sleep, such that prolonged apnea and marked CO₂ retention may occur. From a cardiovascular perspective, sleep also induces bradycardia and mild hypotension. These autonomic effects are especially evident during REM sleep, during which there is atonia or hypotonia of skeletal muscles. This hypotonia increases upper airway resistance by relaxing pharyngeal muscles. It is thought that snoring, as well as obstructive sleep apneas (OSA) is the result of this atonia (Chap. 25). Perhaps the most important effect of sleep on respiration is loss of what some researchers refer to as the wakefulness stimulus, an overriding controller or integrator of all other respiratory inputs and functions.

Eliminating suprapontine inputs to brainstem respiratory centers has only minor effects on resting respiration, suggesting that such areas modulate but do not control basal respiratory patterns and rhythms. Suprapontine areas are active in volitionally altered respiration associated with speech, singing, and similar activities (Fig. 11.10). Thus, cortical areas are capable of affecting the pattern generator and of activating different muscle groups of respiration. Although little is known currently about the anatomy and physiology of cerebral cortical areas that modulate respiration, new imaging methods using functional MRI and/or PET scans are beginning to identify important regions.

Respiratory rhythm is usually regular and continues throughout life without conscious effort. Indeed, consciousness of one's breathing often signifies the presence of extreme environmental conditions or pathological states. When respirations become arrhythmic, the term periodic breathing is applied, while apnea refers to an absence of breathing that is usually temporary. Apneas are obstructive when caused by upper-airway blockage despite efferent impulses to breathe, or central when without respiratory movements, or may be of mixed type. However other apneas exist, including obstructive expiratory apnea, obstructive hypoventilation, and central hypoventilation syndrome ("Ondine's curse"); these will be discussed in Chap. 25.

The term dyspnea was introduced earlier to describe a patient's sensation of inadequate or stressful respiration, or an exaggerated consciousness of a need for increased respiratory effort. Clinically, dyspnea usually implies labored respiration and the use of some or all accessory respiratory muscles, as in many patients with acute lung injury and the acute respiratory distress syndrome (Chap. 28). Many factors induce dyspnea, but it is important to remember that it is a sensation that may not have an underlying pathophysiological explanation in every case.

Periodic Breathing

Periodic breathing is characterized as clusters of breaths separated by intervals of apnea or near-apnea. A clinically useful definition of such periodic breathing might be when at least three respiratory pauses, each lasting 3 seconds or longer, are separated by periods of normal breathing, each lasting less than 20 seconds. Cheyne-Stokes breathing as commonly seen in congestive heart failure is an example of such periodic breathing (Fig. 11.11). Periodic breathing usually occurs during sleep and can occur in healthy individuals, and the apnea there is usually of central rather than obstructive origin.

Periodic breathing is a normal respiratory pattern in premature infants during active REM-sleep and non-REM sleep. It persists in infants, but decreases as a percentage of the population as children mature. Among term infants, periodic breathing usually is confined to REM sleep. Persistence of periodic breathing during longer portions of sleep may be abnormal, and reflect immaturity or an abnormality of brainstem respiratory control (Chap. 39). Sleep apnea occurs in 2%-3% of children, 3%-7% of middle-aged adults, and 10%-15% of otherwise healthy adults who are > 65 years old. Periodic breathing is generally considered to be a consequence of chemoreceptor function. When P_aco₂ or P_Aco₂ is far below normal eupneic values, or when P_ao₂ or P_Ao₂ is far above, then at least transient apnea will almost certainly ensue. Thus, there are apneic thresholds for both hypocapnia and hyperoxia that may differ strikingly among healthy individuals. Whether their periodic breathing is more, the result of activation of central versus peripheral chemoreceptors remains unresolved to date.