gms | German Medical Science

There are many interactions between sleep and breathing. Nevertheless, there are only minor changes in blood gases in healthy subjects. However, in patients with pre-existing lung disease, for example, sleep-related changes of the respiratory system may lead to marked impairment of gas exchange. The symptoms of the “classical” obstructive and restrictive pulmonary diseases worsen during sleep compared with wakefulness [1]. Furthermore, there are typical illnesses or symptoms, which preferably occur during sleep.

This paper deals with the pathophysiology of these sleep-related breathing disorders.

We shall distinguish between

1.

Central hypoventilation

2.

Central apneas and hypopneas and periodic breathing

3.

Upper airway obstruction.

Most of the sleep-related breathing disorders are based on or maintained by disorders of the respiratory control systems in the central nervous system.

The so-called respiratory neurons within the medulla oblongata and the pons guarantee the regular inspiration and expiration. Neurophysiologic research resulted in the identification of neurons with respiration-related activity at widespread sites within the brainstem. Therefore, there is no single neuronal breathing center, but a network of connected neurons, which works as a rhythm generator for the timing of in- and expiration, as well as a pattern generator with highly complex efferent activity to the primary and accessory respiratory muscles and the upper airways including the nose, pharynx, larynx and the intrathoracic airways. There is as precise timing of muscle activation during inspiration, starting at the nostrils, then upper and lower airways. Milliseconds later the respiratory muscles generate the intrathoracic sub pressure necessary to inhale air into the lungs. During expiration, airway resistance increases due to the reduction of the activity of airway dilator muscles.

Single cell recordings in the brainstem and pons identified several types of respiratory-related neurons, which are active during well-defined phases of the respiratory cycle, e.g. early inspiratory differed from ramp-inspiratory and end-inspiratory neurons [2]. Inspiratory neurons are located bilaterally within the ventral respiratory group (VRG) next to the nucleus ambiguous and more rostrally as dorsal ventilatory group (DRG) close to the nucleus tractus solitarius (NTS, Figure 1 [Fig. 1]). Furthermore, there are inspiratory neurons at the spinal level of C1 and C2. Expiratory neurons are found within the pons (pontine respiratory group, PRG) and the caudal brainstem. Respiratory neurons can hardly be identified morphologically, because they are very similar to the surrounding cells of the reticular formation.

The lung inflation reflex (Hering-Breuer-reflex) is part of the neuronal control of breathing. Stretch receptors within the lung measure the stretching of the tissue during the respiratory cycle. During the inspiration lung inflation and afferent vagal activity increase, during the expiratory phase, afferent activity decreases. Lung inflation causes a reflex inhibition of further inspiration and fosters the next expiration. On the other hand, the lung deflation reflex endorses the next inspiration during deep expiration. These reflexes lead to an optimization of the work of breathing and gas exchange, particularly since breathing slowly, but deeply markedly increases the work of breathing due to the higher elastic forces that have to be overcome. On the other hand, rapid shallow breathing increases the ineffective dead space ventilation.

Recently, several genes have been identified, which control the breathing frequency during spontaneous ventilation [3]. During wakefulness, there was a moderate correlation between respiratory frequencies and the genotype, which markedly increased during sleep. These possible candidate genes code for the adenosine and 5-HT-receptors, which are also involved in the respiratory network and respiratory afferents of various origin.

The phasic activity of the respiratory network is dependent upon a non-rhythmic, tonic activation by afferents from the surrounding reticular formation of the brainstem. This tonic activation is the neuronal substrate of the so-called respiratory drives. Here, we distinguish between drives with feedback and those without feedback (Figure 2 [Fig. 2]). A suppression of the tonic activity below a critical threshold causes respiratory arrest.

Non-feedback respiratory drives change the ventilation independent from the actual blood gases and may therefore cause deviations of blood gases in case of hyper- or hypoventilation. Almost all sensory afferents influence the reticular activity and thus the respiratory activity by either stimulation or inhibition. Besides these direct influences by the surrounding reticular formation other non-feedback respiratory drives have been identified: During wakefulness respiration is not only controlled by the autonomic nervous system, but also by the behaviour, such as vocalization, breath holding or hyperventilation as well as non-voluntary acts as coughing or sneezing. Additional respiratory drives arise from the serotonergic and norepinephrine-containing neurons, which are involved in the regulation of the cortical activity and arousal level during different states of wakefulness and sleep.

Recently orexinergic neurons have been identified in the hypothalamus, which constitute a widespread network to nuclei of the sleep-wakefulness regulation. These neurons drive respiration especially during wakefulness and may be the neuronal substrate of the so-called “wakefulness drives” of respiration [4].

The specific chemical afferents during hypoxia, hypercapnia, and acidosis play an important role in the regulation of breathing. These afferents form closed feedback loops, which lead to a stabilization of oxygen and carbon dioxide partial pressures and the pH in the arterial blood.

The sensors of this feedback loop are located in the peripheral chemoreceptor sites in the carotid and aortic glomera, which respond to hypoxia, hypercapnia, and metabolic acidosis by increasing their firing rate. The afferents from the peripheral chemoreceptors are integrated into the respiratory network via the solitary tract nucleus (NTS) within the dorsal brainstem. The peripheral chemoreceptors are the only sensor systems of the body, which directly drive ventilation during hypoxemia. Their impact on breathing is fast, but they tend to adapt. They contribute 15 – 40 % of the whole respiratory drive during rest [5]. Denervation causes an immediate drop of minute ventilation and an increase in arterial PCO₂ by 5 – 10 Torr. Ventilation, however, recovers, until the PaCO₂ reaches normocapnic values again. The hypoxic respiratory sensitivity, however, remains blunted.

Even after experimental elimination of the peripheral chemoreceptors there is still a highly sensitive and very effective CO₂- and pH-dependent respiratory drive, in which the ventral medullary surface of the brainstem plays an important role [6]. Minimal changes of the PCO₂ or the pH in these superficial layers, in the cerebrospinal fluid or in the arterial blood cause marked changes in ventilation. Experimental elimination of these areas leads to a loss of CO₂-sensitivity of the respiratory system, followed by a destabilization of blood gas homeostasis dependent on the level of vigilance: Hyperventilation can be observed during active wakefulness, severe hypoventilation during sleep [7]. In animal experiments several brain areas have been identified, where neurons increase their firing rate during an increase of the local PCO₂ or a decrease of the local pH dependent upon the state of sleep and wakefulness, which coincided with an acceleration or deepening of respiration [8], [9]. This central chemical drive of ventilation, called ‘central chemosensitivity’, is characterized by a much longer time constant than the peripheral chemoreceptors in the range of minutes, contributes 60 – 85 % of the respiratory drive at rest and shows only little adaptation.

Sleep affects the tonic activation of the respiratory network in different ways: The reduction of sensory input reduces the reticular activity of the brainstem. Single neuron recordings showed that neurons with mostly phasic activity are less affected than those neurons with more tonic activity. Some of these nerve cells are silenced during certain sleep states of NREM sleep, and can experimentally be activated by application of excitatory neurotransmitters, which induces rhythmic activity. This phenomenon demonstrates that a “sub-threshold” respiratory rhythm remains in these cells [10].

During REM sleep, respiration becomes irregular. There are short apneas and short periods of hyperpnea. The mean respiratory rate increases slightly, the mean tidal volume decreases compared with NREM sleep. One reason for these changes is the loss of the muscle tone during REM sleep, which not only affects the motor system, but also the respiratory muscles except the diaphragm. The rapid, shallow breathing can be interpreted as compensatory mechanism. A breath-by-breath analysis demonstrated that breathing was transiently suppressed during bursts of rapid eye movements of phasic REM sleep [11]. The short-term activation of breathing during REM sleep originates from higher CNS regions and persists even after vagotomy, which eliminates the mechanical feedback. It is concluded that these non-feedback respiratory drives during REM sleep are of endogenous nature. Attempts to correlate the irregularity of breathing during REM sleep with dream contents have failed so far.

Some neurophysiologic features of the respiratory system are in favour of destabilization, especially during sleep onset and during changes of sleep states (Figure 3 [Fig. 3]) [12]. During wakefulness, there is a short-term potentiation within the respiratory network, which stabilizes the respiratory rhythm: Short-term increases in ventilation, e.g. sighs or augmented breaths induced by unspecific respiratory drives are maintained for several breaths. During sleep, however, this mechanism is suppressed. A transient hyperventilation may therefore be followed by an apnea [13]. During hyperpnea the PCO₂ may drop below the so-called apnea threshold, which results in the appearance of central apneas [14]. The ventilatory responsiveness below eupnea is considered as a determinant of ventilatory stability in sleep [15]. After cessation of breathing, the PCO₂ has to rise above eucapnia, until the respiratory rhythm is re-established. This control system inertia favours central apneas and periodic breathing during sleep [16]. After sighs and augmented breaths the lung inflation reflex can suppress the next inspiration and may account for the instability of breathing during sleep [17].

The arousal thresholds and the thresholds of the hypoxic and hypercapnic ventilatory drives become less sensitive during sleep. Obviously, the organism makes a compromise on the guarantee of gas exchange and an undisturbed sleep. Lowest arousability is found during REM sleep. Here severe hypoxia or hypercapnia is necessary to induce a change in sleep states or awakening.

The ventilatory drive by hypoxia, hypercapnia, or metabolic acidosis changes dependent upon the states of sleep and wakefulness. Nevertheless, these control circuits stay in operation during sleep and prevent larger deviations of the blood gases. In general the hypoxic ventilatory drive is suppressed during sleep to minimal values in REM sleep [18]. This is true for men and for women. Women, however, already have a lower hypoxic drive during wakefulness. Therefore, the sleep-related inhibition is significant only during REM sleep [19]. The hypoxic ventilatory response during NREM sleep does not differ between men and women [20].

The hypercapnic ventilatory response markedly varies during sleep. Rebreathing tests during well defined sleep states demonstrated a reduction of the CO₂ response to 50% in NREM sleep and to 30% in REM sleep compared with the level during wakefulness before sleep onset [21]. Furthermore there is a circadian rhythm of the CO₂ sensitivity independent of sleep with minimal hypercapnic responses in the early morning hours [22], [23]. In combination of these two findings, a marked reduction of CO₂ sensitivity might be expected during REM sleep in the early morning. This high variability during sleep has been proven by repetitive CO₂ challenges throughout the whole night [24]. In this experiment in ten healthy subjects, the mean hypercapnic sensitivity ranged from a maximum of 1.45 ± 0.35 L/min/Torr to a minimum of 0.31 ± 0.13 L/min/Torr. The lowest values occurred during the second half of the night between 3:00 and 6:30 a.m. During REM sleep, the tidal volume response was markedly reduced. The respiratory rate slightly increased, but did not compensate for the reduction of the tidal volume [25]. Despite the reduced chemical drives of ventilation the average minute ventilation at rest was slightly increased during REM sleep and the mean PCO₂ was slightly lower than during NREM sleep. This indicates that additional non-feedback drives stimulate breathing during REM sleep independent of the blood gas situation. The CO₂ response curves were flattened but shifted to the left. Consequently, the minute ventilation at normal PCO₂ is higher. The reduced hypercapnic drive becomes apparent only at higher PCO₂ values (Figure 4 [Fig. 4]).

Sleep deprivation suppresses the chemical drives of respiration. After one night of sleep deprivation in healthy men the hypoxic ventilatory response was reduced by 29% and the hypercapnic ventilatory response by 24% [26]. Studies in patients suffering from obstructive sleep apnea showed that the repetitive hypoxic-hypercapnic events with arousals change the chemical drives of ventilation. In contrast to healthy subjects, the patients showed a 30% increase in chemical drives after one night without therapy. This phenomenon further destabilizes the respiratory rhythm.

The diameter of the upper airways is a result of the interplay of anatomical, functional-mechanical and neuro-muscular components. Some of the anatomical factors can be measured by cephalometry, e.g. the position of the upper and lower jaw and the palate, the size of the tongue and the thickness of the connecting and fatty tissues in the pharynx. The tension of the pharynx belongs to the mechanical factors. It is dependent from the caudal traction of the trachea and decreases with lower functional residual capacity, which increases the collapsibility of the upper airways. In addition, more than 20 muscles below the pharyngeal mucosa stiffen and widen the upper airways [27]. Mainly tonically active muscles as the tensor palatini can be distinguished from mainly phasically active muscles as the genioglossus. The respiratory network controls this phasic activity. During inspiration, the muscle activity continuously increases and thus widens and stiffens the airway to prevent a collapse due the negative intraluminal pressure. In addition to the respiratory innervation there are local mechanoreceptor reflexes, which can experimentally be triggered by short negative pressure pulses within the pharynx and which result in an increased tension of specific pharyngeal muscles [28]. The palatopharyngeous [29] and the genioglossus increase their basic tone during negative pressure pulses. Superficial anaesthesia, however, suppresses this effect markedly [30]. CPAP also suppresses the genioglossus and tensor palatini activities [31]. During wakefulness upper airway muscle activity markedly increases, when the respiratory drive is enhanced, e.g. by hypercapnia [32]. This effect, however, is blunted during sleep.

With sleep onset, the upper airway resistance significantly increases as compared with wakefulness. This happens rapidly and coincides with the occurrence of theta activity in the sleep EEG [33]. The basal dilator tone decreases [34] and the phasic activity is reduced and less sensitive to enhancements of the respiratory drive (Figure 5 [Fig. 5]) [35]. This effect favours the occurrence of upper airway obstructions. After short periods of hyperpnea, the inspiratory muscles are activated more strongly and before the upper airway muscles. Thus, they generate an inspiratory negative pressure, while the upper airways are more collapsible. The muscle atonia during REM sleep also affects the pharyngeal muscles, which further increases the collapsibility of the upper airways. The horizontal position during sleep favours the obstruction of the upper airways. Even the habitual snoring improves in an elevated upper body position [36].

With sleep onset, the influence of non-feedback respiratory drives is markedly reduced. The so-called wakefulness drives and stimulation by sensory afferents cease. The chemical feedback drives of respiration are reduced dependent upon the sleep state, the arousal thresholds increase. The muscle tone decreases. Thus the airway resistance, especially of the upper airways, increases. In addition, the respiratory muscles are affected by the reduction of muscle tone, which is most pronounced during REM sleep. During REM sleep, however, endogenous drives increase ventilation independent of the blood gas situation. In contrast, during NREM sleep the maintenance of the respiratory rhythm is dependent upon the functioning of the metabolic feedback drives of respiration, e.g. the hypoxic and the hypercapnic ventilatory drives. In spite of these challenges, the ventilation in healthy subjects only slightly changes during sleep, which is in favour of an adequate compensation of the respiratory control system. Spirometric measurements of the ventilation during sleep onset in healthy subjects demonstrated a 12% reduction in minute ventilation from wakefulness to light sleep and a 16% reduction from wakefulness to deep sleep (Szczyrba and Schäfer, unpublished). This reduction was due to a slight decrease in respiratory rate and a decrease in tidal volume. The PCO₂ slightly rose by 2.7 and 4.1 Torr, respectively.

In case of an exogenous or an endogenous disturbance of the control system of breathing, breathing disorders will usually appear during sleep at first. During wakefulness, compensatory mechanisms are more effective. In the following chapters, the pathophysiology of three sleep-related breathing disorders will be represented: Hypoventilation, central sleep apnea and periodic breathing, and sleep-related obstructions of the upper airways.

Hypoventilation is characterized by an increased arterial PCO₂. The ratio of CO₂ production and CO₂ elimination is shifted due to a disturbed respiratory pump. This disturbance may be caused by peripheral mechanisms, as the “classical” obstructive and resistive lung diseases, anatomical factors due to inborn errors involving the skull and neuromuscular or muscular disorders. In cases are summarized as “secondary” hypoventilation. The symptoms improve, if the underlying disease is efficiently treated. Very often, the first symptoms of respiratory insufficiency occur during sleep, while the blood gas values during wakefulness remain in the normal range.

“Central” hypoventilation is due to a reduced respiratory drive. During wakefulness, the patients can voluntarily increase their ventilation. During sleep, however, they tend to marked hypercapnia and hypoxia. Children suffering from a congenital, central hypoventilation syndrome (CCHS) have extensively been examined [7]. During wakefulness, their blood gases are very instable, they tend to hyperventilate, but in boring situations, they might also hypoventilate. Dyspnea is almost unknown. Therefore, they have to be cautious about diving. The results with respect to the arousal reactions on hypoxia and hypercapnia are contradictory. During sleep onset the tidal volume is reduced, the respiratory rate changes only slightly. Sometimes there are central apneas. Hypoxia and hypercapnia develop rapidly (Figure 6 [Fig. 6]). Without treatment PCO₂, values above 130 Torr can be observed, despite eucapnic values during wakefulness. During spontaneous breathing as well as during mechanical ventilation a strong dependence from the sleep states can be observed: The hypoventilation is most pronounced during deep sleep, during REM sleep hypoventilation may be less severe. This points to the differences in the respiratory drives during REM sleep, where endogenous stimuli can drive ventilation in addition to the metabolic, chemical drives, which play the main role during NREM sleep (Figure 7 [Fig. 7]). During mechanical ventilation in NREM sleep, breathing is often completely passive, while during REM sleep spontaneous activity can be observed, which may cause hyperventilation.

Functional magnetic resonance imaging (fMRI) has been used to identify brain regions, which are involved in the CO₂-dependent respiratory drive [37], [38].

The diagnostic criteria of hypoventilation contain clinical symptoms of chronic hypoxia, such as cor pulmonale, pulmonary hypertension, excessive daytime sleepiness, which not explained by other factors, erythrocytosis, and an increase of the arterial PCO₂ from rest during wakefulness to sleep by more than 10 Torr and/or desaturations, which are not due to apneas or hypopneas [39].

Central sleep apneas may be due to a blunted respiratory drive and thus appear during hypercapnia. On the other hand, they can be the consequence of an increased respiratory drive. In this case, they appear during eu- or hypocapnia [40]. As shown in chapter 2, the respiratory system has features, which favour the emergence of central apneas and periodic breathing during sleep. In healthy subjects, these instabilities are small and rare. There are neither larger blood gas changes, nor arousal reactions or sleep disturbances. Enhanced respiratory drives with only small a CO₂ reserve until the apnea threshold [15] is reached, may unmask the inherent instability and may induce long-lasting periodic breathing with repetitive central apneas, hyperventilatory phases and accompanying arousal reactions. The sleep is strongly disturbed and followed by excessive daytime sleepiness. The factors inducing periodic breathing in humans [41] and the interaction between arousal and chemical drive in sleep-disordered breathing [42] have been mathematically modelled. An enhanced hypercapnic and/or hypoxic respiratory drive and a delayed feedback of the blood gas changes to the respiratory network can induce periodic breathing with central apneas. Patients with idiopathic central sleep apneas showed an exaggerated hypercapnic ventilatory response and a hyperventilation with hypocapnia [43].

Very often periodic breathing in the form of Cheyne-Stokes respiration (CSR) is observed in patients suffering from heart failure [44]. CSR is characterized by a regular waxing and weaning of the tidal volume between central apnea and hyperpnea (Figure 8 [Fig. 8]). Often this happens during hypocapnia. The respiratory drive is exaggerated, mainly by non-feedback drives which originate from lung afferents, which are stimulated by the lung congestion due to the heart failure [45]. Similar facts may be responsible for the central apneas, which occur in patients with primary pulmonary hypertension [46].

Interestingly the upper airways may completely collapse during central apneas and thus lead to an obstruction of the upper airways without an additional negative transluminal pressure [47]. Strong, transient reductions of the respiratory drive also induce an inhibition of dilator muscle activity in the pharynx and may activate the constrictor muscles, which causes an active obstruction during the central apnea.

The diagnostic criteria comprise excessive daytime sleepiness and/or frequent arousals during sleep and five or more central apneas or hypopneas per hour sleep. The PCO₂ during wakefulness should not exceed 45 Torr [39].

Narrowing of the upper airways during sleep occurs in different degrees. It begins with primary snoring and a negligible flow limitation during inspiration without disturbance of the ventilation, the gas exchange, and the sleep quality. The elevated upper airway resistance is completely compensated by an increased work of breathing. A more severe form is the obstructive snoring or “upper airway resistance syndrome” (UARS), with proven increase of the work of breathing and frequent arousals, which result in a non-restorative sleep. Obstructions during sleep, which impede the ventilation so strongly that the flow is reduced by at least 50% during 10 seconds or longer and/or that the oxygen saturation drops by at least 4% or more, are called obstructive hypopneas. A complete collapse of the upper airways with an interruption of breathing during inspiration for at least 10 seconds is called obstructive apnea. Usually arousal reactions in the sleep EEG accompany these obstructive events. These arousals may not lead to wakefulness, but they considerably disturb the sleep structure (Figure 9 [Fig. 9]). The suggested clinical criteria for the obstructive sleep apnea/hypopnea syndrome (OSAHS) are excessive daytime sleepiness, and/or two or more of the following symptoms: choking or gasping during sleep, recurrent awakenings form sleep, unrefreshing sleep, daytime fatigue, impaired concentration. Overnight monitoring demonstrates five or more obstructive events (apneas or hypopneas) per hour during sleep [39].

The pharynx resembles a starling resistor in form of a collapsible tube, which is fixed between the non-collapsible nose and larynx. Its condition is dependent upon the ratio of the intraluminal pressure and the external tissue pressure, the driving pressure difference upstream versus downstream, and the compliance of the tissue. In healthy subjects, the upper airways remain open even during forced inspiration. The so-called critical closing pressure Pcrit is negative at about -14 cm H₂O (Figure 10 [Fig. 10]), as proven by application of negative pressure via a nose mask [48]. Snorers demonstrated a less negative Pcrit. At a Pcrit of -5 cm H₂O, the pharynx tends to collapse during inspiration. In patients with obstructive hyperpneas without obstructive apneas, the Pcrit is slightly below the atmospheric pressure. In patients with obstructive sleep apnea, the Pcrit during sleep is found at positive values, which calls for the application of nasal continuous positive airway pressure (nCPAP) as the adequate treatment. Otherwise, the upper airways collapse during inspiration. The individual level of the Pcrit is dependent upon anatomical, mechanical, and neuromuscular factors. The research of the pathophysiological mechanisms of the upper airway obstructions during sleep has revealed several obvious as well as subtile differences between healthy subjects and patients with OSAS (Figure 11 [Fig. 11]).

OSA patients often suffer from a metabolic disorder, which leads to weight gain or prevents from weight loss, possibly by a reduced Insulin sensitivity, reduced blood levels of anabolic hormones, and a disturbed central serotonergic tone [49]. This leads to a typical male distribution of the adipose tissue with an increased neck circumference. The cross-section of the pharynx typically changes due to the deposit of fats in the lateral pharyngeal walls [50], [51]. In addition the diameter markedly differs between in- and expiration. The crosswise oval cross-section in healthy subjects changes to a longwise oval cross-section in OSA patients. Moreover, the obesity reduces the functional residual capacity and therefore the caudal traction of the trachea, which increases the pharyngeal compliance [52]. Opening of the mouth and mouth breathing increase the pharyngeal airway resistance by narrowing the posterior airway space [53]. A handicapped nasal breathing and the narrowing of the pharynx release the Venturi- or Bernoulli effect, which increases the collapsibility due to the higher airflow and the lower intraluminal pressure.

During wakefulness OSA patients demonstrate a higher activity of their pharyngeal dilator muscles, both, in the tonic and the average and the maximum phasic activity [54], which may be necessary to compensate for the increased airway resistance. The comparison of nasal breathing and breathing through a tracheal cannula in tracheotomized OSA patients demonstrates that local mechanisms triggered by the negative intraluminal pressures play an important role [55]. Healthy subjects and OSA patients significantly differed in the sensitivity of the pharyngeal mucosa. OSA patients have a significantly reduced 2-point-dicrimination sensibility and a reduced vibration sensibility. The latter improves after six month of CPAP therapy [56]. The loss of tactile sensitivity of the pharynx may be causally connected to the incomplete compensation of the pharyngeal obstruction. fMRI studies during inspiratory loading in OSA patients revealed significantly lower activations in the primary sensory thalamus and the sensory cortex compared with healthy subjects [57].

Studies on the connection of the central respiratory drive and the critical occlusion pressure Pcrit revealed a significant correlation only in those patients with a Pcrit around 0 cm H₂O [58]. There is evidence that the chronic intermittent hypoxia induces a vicious circle, which leads to a conversion of the pharyngeal muscle fibres, which, again, increases the collapsibility: Experimental studies with intermittent hypoxia demonstrated after only 10 hours that the genioglossus suffered a conversion of myosin heavy chain type 2A to type 2B, which are markedly more fatigable. This effect lasted for at least 30 hours [59]. Besides the effect of sleep deprivation on the respiratory drive and the activity of the pharyngeal muscles, and the reduced mechanical sensitivity of the pharyngeal mucosa, the repetitive hypoxia also leads to and perpetuates an increased collapsibility of the upper airways in man [60] (Figure 12 [Fig. 12]).

Sleep-related breathing disorders mainly arise as exaggerations of sleep-related physiologic features of the respiratory system. Involved are the direct neuronal effects on the respiratory network, the muscle atonia during REM sleep, the blunting of the arousal reactions and the hypoxic and hypercapnic chemoreflexes, as well as the reduced tonic and phasic activities of the pharyngeal muscles. In combination with an anatomical disposition given by the obesity, the obstructive sleep apnea syndrome is maintained by just several vicious cycles.