Pathophysiology of upper airway obstruction in obstructive sleep apnea in adults

INTRODUCTION — Obstructive sleep apnea (OSA) is a fairly common disorder with significant adverse health consequences [1-4]. OSA is characterized by recurrent obstruction of the pharyngeal airway during sleep, with resultant hypoxia and sleep fragmentation. The pathogenesis of OSA is due to the interaction between unfavorable anatomic upper airway (UA) susceptibility and sleep-related changes in UA function [5]. However, the mechanisms linking sleep-related physiologic changes to UA obstruction in some individuals are not fully understood. In addition, the majority of studies investigating UA obstruction during sleep have been conducted during nonrapid eye movement (NREM) sleep, given the difficulty in achieving rapid eye movement (REM) during invasive studies in the laboratory environment.

This topic will review the effects of sleep on respiratory mechanics, the determinants of UA patency, and the pathophysiology of UA obstruction during sleep. The pathophysiology of OSA in children and the clinical features, diagnosis, and treatment of OSA in children and adults are reviewed separately. (See "Mechanisms and predisposing factors for sleep-related breathing disorders in children" and "Clinical presentation and diagnosis of obstructive sleep apnea in adults" and "Obstructive sleep apnea: Overview of management in adults" and "Management of obstructive sleep apnea in children".)

EFFECT OF SLEEP ON RESPIRATORY MECHANICS — Sleep is accompanied by multiple physiologic changes relevant to ventilation and respiration (algorithm 1). Sleep is viewed as a quiet resting period, judging by the limited movement, decreased responsiveness, and the passive appearance of a sleeping individual. Sleep is associated with a decreased metabolic rate, loss of the wakefulness drive to breathe [5], and a subsequent decrease in ventilatory motor output to respiratory muscles, including upper airway (UA) muscle. Furthermore, the loss of the wakefulness drive to breathe renders respiration during sleep critically dependent on the level of chemoreceptor and mechanoreceptor stimuli [6], and hence susceptible to central apnea and to upper airway obstruction. The pathogenesis of central apnea is discussed separately. (See "Central sleep apnea: Pathogenesis".)

Upper airway mechanics

Decreased muscle activity — Reduced UA muscle activity during sleep is a physiologic phenomenon of little consequence in healthy individuals, but it may promote UA narrowing in susceptible individuals.

Decreased ventilatory motor output is associated with decreased UA muscle activity, particularly in muscles that display tonic activity (independent of the phase of respiration). For example, the tensor palatini demonstrates immediate decrease in activity with sleep onset, with associated reduction in inspiratory flow [7,8]. Studies investigating respiratory muscle activity at sleep onset demonstrate that the activity of respiratory pump muscles and UA-dilating muscles changes less when the dominant electroencephalogram (EEG) waveform is theta (light sleep) versus alpha (wakefulness) [9].

Changes in caliber and compliance — The sleep state is associated with decreased pharyngeal caliber, increased UA resistance [10], and increased UA compliance [11]. In other words, the lumen of the UA is smaller and the airway walls are more deformable during sleep. Thus, sleep renders the UA more susceptible to closure in the presence of a collapsing transmural pressure.

UA narrowing appears to be a universal finding, resulting in increased turbulent flow, which may explain why breathing during sleep is audible even in healthy individuals. A substantial increase in UA resistance leads to increased flow turbulence, inspiratory flow limitation, and fluttering of the soft palate and UA soft tissue [12,13]. Snoring is the acoustic corollary of inspiratory flow limitation, which manifests as a plateau in inspiratory flow despite persistent downstream driving pressure.

Pressure-flow relationships differ between normal, non-flow-limited breathing (NIFL) versus high-resistance and flow-limited breathing (figure 1). Breathing during wakefulness is non-flow-limited. Likewise, breathing during sleep in non-snorers is also non-flow-limited. Snoring and flow limitation suggest increased propensity for UA collapse during sleep. In extreme cases of UA narrowing, complete closure may occur, leading to obstructive sleep apnea (OSA).

Loss of load compensation — Sleep-related UA narrowing and increased resistance to flow represent an added internal load on the respiratory system. The ability of the ventilatory control system to respond to added loads is critical for the preservation of alveolar ventilation.

Breathing through high-resistance tubing (like a straw) during wakefulness is associated with an immediate increase in ventilatory effort to maintain alveolar ventilation and PaCO₂. During sleep, loads are not perceived; thus, ventilation decreases and PaCO₂ increases [14,15]. The teleological reason for lack of immediate load perception is uncertain but may indicate that sleep reigns supreme, a biological need that is protected through reliance on chemical and mechanoreceptor feedback. Conversely, impaired resistive load compensation has been noted in healthy children of patients with OSA. Thus, the ability to compensate for increased loads may be an inherited trait that contributes to preservation of upper airway patency during sleep [16].

The consequences of decreased load perception during sleep are mild in normal humans, as increased PaCO₂ restores ventilation to near-normal levels. In contrast, patients with abnormal respiratory mechanics, such as those with chronic obstructive pulmonary disease (COPD), may experience worsening of respiration and gas exchange as a result of impaired load compensation. (See "Mechanisms, causes, and effects of hypercapnia", section on 'Chronic obstructive pulmonary disease'.)

Thoracic cage dynamics — The relative contributions of the rib cage and abdominal muscles to tidal volume change during sleep. During non-rapid eye movement (NREM) sleep, the ratio of the rib cage to abdominal muscle contribution increases compared with wakefulness. In contrast, the ratio is lower during rapid eye movement (REM) sleep, when there is loss of intercostal muscle activity.

It is noteworthy that most of the studies on sleep effect have been conducted during NREM sleep, since REM is difficult to achieve under instrumented conditions. REM sleep is associated with muscle atonia affecting many UA dilators and intercostal muscles, while sparing the diaphragm. During REM, minute ventilation decreases even more and the respiratory rate becomes more irregular, particularly during phasic REM sleep [5,17].

Increased arterial carbon dioxide — Relative to wakefulness, sleep is associated with decreased ventilation and increased PaCO₂ (figure 2). During sleep, PaCO₂ rises by 4 to 5 mmHg. This physiologic hypercapnia is due to a combination of increased UA resistance and decreased ventilatory motor output [18].

Implications — The consequences of physiologic sleep-related changes in respiratory mechanics depend on individual host factors:

●Individuals with favorable UA anatomy and lung mechanics are able to sustain rhythmic breathing, normal gas exchange, and stable sleep, albeit at a higher PaCO₂.

●Individuals with intermediate susceptibility to collapse develop snoring and inspiratory flow limitation, while maintaining stable sleep and breathing.

●Those with highly compromised UAs develop complete UA obstruction.

Similarly, changes in the relative contributions of the rib cage and abdominal muscles to ventilation are inconsequential in healthy individuals with normal lung mechanics but may lead to worsening ventilation–perfusion mismatch and hypoxia in obese individuals and those with pulmonary disease. The ensuing increase in ventilation may contribute to unstable breathing via activation of peripheral chemoreceptors.

DETERMINANTS OF UPPER AIRWAY PATENCY — The human upper airway (UA) is a multipurpose conduit serving respiration, deglutition, and vocalization. The pharyngeal airway lacks structural bony or cartilaginous support; therefore, it is a deformable tube susceptible to collapse if sufficient transmural pressure is applied across a compliant pharyngeal wall. The determinants of pharyngeal compliance or intrinsic "stiffness" are not fully understood, given the multitude of structures that constitute the pharyngeal wall.

Determinants of UA patency can be broadly divided into two categories: structural factors and neuromuscular factors. Structural determinants include craniofacial structure; surrounding soft tissue, including adipose tissue; vascular structures; and mucosal factors. Neuromuscular factors include ventilatory motor output, UA muscle activity, and the thoracic-upper airway link via caudal traction. (See 'Effect of lung volume' below.)

Structural factors

Craniofacial structure — Craniofacial structure is a critical determinant of UA patency [19,20]. The UA is enclosed in a bony box composed of the mandible, maxilla, skull base, and cervical spine. A small bony enclosure, or a craniofacial constraint, such as retrognathia, may result in tissue "crowding," increased pressure in the tissues surrounding the UA, and increased collapsing transmural pressure, with resultant inferior displacement of the hyoid bone.

Support for the role of bony structures in the propensity for UA obstruction comes from observational and experimental studies. In a case-control study that used three-dimensional magnetic resonance imaging (MRI), increased mandibular length was associated with decreased risk for obstructive sleep apnea (OSA) in men but not in women [21]. In addition, differences in position of the hyoid bone between patients with OSA and controls was largely determined by tongue volume, suggesting that inferior displacement of the hyoid bone in patients with OSA is due to relative tongue volume and increased surrounding pressure. Likewise, another study found that the salutary effect of mandibular-advancement oral appliances may be explained by decreased surrounding tissue pressure [22].

In addition to these studies, there is epidemiologic evidence that differences in craniofacial indices may contribute to racial and ethnic differences in the prevalence of sleep apnea [23-25] and may interact with obesity to promote UA obstruction during sleep.

Soft tissue structures — The UA lumen is surrounded by the soft tissue of the neck, including connective, adipose, vascular, and lymphatic tissue. Consequently, factors that increase surrounding tissue pressure tend to promote UA narrowing.

Several soft tissue factors are associated with higher risk of OSA, including increased tongue size, increased size of lateral pharyngeal walls, and increased total soft tissue volume [26]. Increased soft tissue may be a heritable trait, as evidenced by the familial aggregation of UA soft tissue structure in normal individuals and those with OSA, independent of body mass index (BMI) and neck circumference [27]. Increased adipose tissue in the UA or the tongue secondary to obesity may also increase collapsing tissue pressure [28].

Enlarged tonsils can increase the susceptibility to UA obstruction by encroaching on the pharyngeal lumen. Enlarged tonsils is a recognized risk factor for OSA, especially in children [29-31]. (See "Mechanisms and predisposing factors for sleep-related breathing disorders in children", section on 'Enlarged tonsils and adenoids'.)

Vascular factors

Rostral fluid displacement — Increased vascular volume in the neck (ie, rostral fluid shifts) may promote UA obstruction by increasing surrounding tissue volume and pressure. Experimentally, vasoconstriction and vasodilatation have been shown to cause a decrease and increase in UA resistance, respectively [32,33]. Changes in chemical stimuli, such as hypoxia or hypercapnia, may additionally alter vascular tone and adversely affect UA patency, especially in patients with unfavorable UA anatomy.

Several small observational studies have consistently demonstrated that, during recumbent sleep, an increase in rostral volume may contribute to the severity of OSA, particularly in patients with high volume states (eg, congestive heart failure, end-stage kidney disease, and refractory hypertension) and that reducing lower extremity fluid volume (eg, with compression stockings, diuresis) may attenuate this process [34-40]. As examples:

●In one study of patients with uncontrolled hypertension and OSA, aggressive diuretic therapy was associated with modest reductions in the apnea-hypopnea index (AHI; 49 versus 58 events per hour), neck circumference (0.7 versus 1.2 cm) and leg fluid volume (308 versus 418 mL) [35].

●Similarly, another study of non-obese sedentary men with OSA reported that wearing compression stockings while awake led to measurable reductions in AHI (23 versus 31 events per hour), together with a 40 percent reduction in leg volume and neck circumference [40].

●In a study of 17 men with non-severe OSA or no OSA, the infusion of similar amounts of saline in older men, as compared with younger men, caused a greater increase in neck circumference and AHI (2 versus 32); however, substantial methodologic flaws prohibit accurate interpretation of these results [37].

The role of rostral fluid shifts in intraoperative fluid management for patients with OSA is discussed separately. (See "Intraoperative management of adults with obstructive sleep apnea", section on 'Intravenous fluid management'.)

Neuromuscular factors

Upper airway muscle activity — Upper airway muscles support multiple critical functions, including respiration, deglutition, and phonation. The majority of upper airway muscles demonstrate activity that is independent of the phase of respiration (tonic activity), whereas some UA muscles demonstrate electrical activity during one part of the respiratory cycle. (See 'Effect of sleep on respiratory mechanics' above.)

For example, the tensor palatini demonstrate tonic electrical activity, which decreases with sleep onset [7,9,41]. It is thought that the tensor palatini stiffens the UA and decreases pharyngeal collapsibility. By contrast, the genioglossus demonstrates inspiratory activity (above the tonic level); the genioglossus is classified as a pharyngeal dilator that is activated before the thoracic pump muscle to prepare the airway for inspiratory flow.

Decreased activity — The loss of the wakefulness drive to breathe is associated with a reduction in the electromyography (EMG) activity of the respiratory pump muscles and UA muscles [9]. Available evidence indicates that a variety of UA muscles have reduced tonic or phasic activity during non-rapid eye movement (NREM), thereby promoting UA narrowing and increased UA resistance.

Decreased responsiveness — During wakefulness, application of negative pressure to the UA is associated with a robust reflex increase in UA muscle activity [42,43]. Sleep is associated with significant attenuation on the negative pressure reflex response.

Mechanical corollary — The changes in UA muscle activity and responsiveness described above are based on electrical activity alone. It is unclear, however, if these changes are accompanied by mechanical consequence.

Determining the relative contribution of decreased UA muscle activity to sleep-related narrowing is difficult because of the many influences on UA muscle activity, such as changes in flow and magnitude of negative pressure. Similarly, the mechanical corollary of decreased UA muscle responsiveness is also unclear; it may indicate an attenuation of UA muscles to preserve pharyngeal patency during physiologic challenges.

Changes during rapid eye movement sleep — Rapid eye movement (REM) sleep is associated with muscle atonia affecting antigravity muscles, particularly during periods of phasic rapid eye movements. Muscle atonia affects phasic UA-dilating muscles and the intercostal muscles, but not the diaphragm.

These changes are particularly relevant in patients with neuromuscular disease, diaphragmatic dysfunction (such as patients with chronic obstructive pulmonary disease with hyperinflation [44]), or gas exchange defects (such as patients with interstitial lung disease). In these settings, patients develop hypoventilation and worsening of gas exchange during REM sleep hypoventilation. This is a form of sleep-disordered breathing that is different from OSA, although it may be difficult to distinguish without polysomnography.

It is of note that pharyngeal compliance is not increased during REM sleep [10,45], despite the attenuated UA-dilating muscle activity. In fact, the retropalatal airway is less compliant during REM sleep relative to NREM sleep. This finding points to the significance of additional, non-neuromuscular factors in regulating UA patency.

Effect of lung volume — Changes in lung volume during the respiratory cycle are paralleled by changes in UA caliber [46]. Independent of UA-dilating muscle activity, there is an inspiratory increase and an expiratory decrease in UA luminal size [47,48]. In fact, pharyngeal cross-sectional area reaches a nadir at end expiration, especially in patients with sleep apnea [49].

The underlying mechanism is the anatomic connection between the intrathoracic and extrathoracic airways. Accordingly, inspiratory activity displaces the carina and trachea caudally and stretches the connective tissue linking the trachea to the UA. Thus, increased UA caliber during inspiration is due to caudal traction on the UA that is proportional to inspiratory thoracic activity and independent of UA-dilating muscle activity [50].

From a mechanical standpoint, caudal traction transmits subatmospheric pressure through the trachea and ventrolateral cervical structures to the soft tissues surrounding the UA. In this way, caudal traction promotes UA patency by increasing transmural pressure and/or stiffening the pharyngeal wall [50]. Furthermore, increased lung volumes during sleep are associated with decreased UA collapsibility, perhaps by increasing the longitudinal tension of the pharyngeal airway [51,52]. Accordingly, inspiratory thoracic activity exerts a salutary effect on UA patency via caudal traction, with subsequent dilatation and stiffening of the UA. Decreased lung volume in obesity may contribute to pharyngeal narrowing in obese individuals though this mechanism. (See "Epidemiology and pathogenesis of obesity hypoventilation syndrome", section on 'Pathogenesis'.)

PHARYNGEAL OBSTRUCTION DURING SLEEP — The occurrence of upper airway (UA) obstruction during sleep reflects an interplay between the removal of the wakefulness drive (which helps to maintain airway patency) and an individual susceptibility to collapse. Although individual risk factors are known, the precise pathophysiologic pathways leading to UA obstruction in patients with obstructive sleep apnea (OSA) remain elusive.

As discussed above, the loss of wakefulness drive to breathe results in decreased UA neuromuscular activity and responsiveness, leading to decreased UA caliber, increased UA resistance, and increased pharyngeal compliance (see 'Upper airway mechanics' above). The response to these physiologic changes depends on the underlying susceptibility to pharyngeal collapse, which is determined by baseline UA caliber, surrounding tissue pressure, craniofacial structure, and the intrinsic properties of the UA.

Upper airway mechanics — Sleep-related changes in pharyngeal mechanics are inconsequential in individuals with favorable UA anatomy, manifesting only by slight physiologic increase in partial pressure of carbon dioxide (PaCO₂). Snoring occurs in individuals with moderate susceptibility to collapse, leading to fluttering of the soft palate due to turbulent flow and inspiratory flow limitation. In extreme cases of UA narrowing, complete closure may occur, leading to OSA.

The UA is a deformable tube, prone to collapse if subjected to a collapsing transmural pressure, either by intraluminal subatmospheric (negative) pressure during inspiration and/or by extraluminal surrounding tissue pressure during either phase of respiration. Although UA obstruction during sleep has been attributed to collapsing intraluminal subatmospheric pressure effectively sucking the hypotonic pharyngeal airway shut [53], conclusive evidence implicating negative intraluminal pressure alone is lacking.

Ventilatory motor output — Changes in ventilatory drive may also influence UA patency. Oscillation of ventilatory motor output during periodic breathing is associated with pharyngeal narrowing or obstruction at the nadir of ventilatory motor output, when collapsing pressures are feeble, especially in individuals with a high propensity for UA collapse [54,55]. In fact, UA narrowing or occlusion occurs during a spontaneous or induced hypocapnic central apnea, supporting the notion that negative pressure is not required for the development of UA obstruction [56]. Similarly, in a study examining the mechanics of the pharynx in anesthetized, paralyzed healthy individuals and those with OSA, the pharynx was patent at atmospheric pressure in normal individuals and closed in patients with OSA (in the absence of negative pressure) [57]. Decreased ventilatory output can also be caused by a physiologic event such as swallowing, which is associated with transient inhibition of phrenic motor neurons.

Surrounding tissue — This is a significant determinant of UA patency during sleep. Examples of collapsing extraluminal pressure include passive gravitational forces, high tissue pressure in humans with mandibular deficiency, large tongue size, fat deposits in the UA, or pharyngeal wall edema secondary to rostral fluid shifts in the recumbent position. (See 'Rostral fluid displacement' above.)

Role of expiratory narrowing — Accumulating evidence implicates expiratory narrowing as a possible mechanism of the initial narrowing. For example, an obstructive apnea is often preceded by expiratory narrowing of the UA, as evidenced by increased expiratory resistance [58] or progressive expiratory narrowing [59,60]. Likewise, induced hypocapnic hypopnea is associated with expiratory pharyngeal narrowing and increased pharyngeal expiratory UA compliance [61]. Therefore, UA obstruction may occur at either inspiration or expiration. Individuals with a high surrounding tissue pressure may be particularly susceptible to expiratory pharyngeal narrowing under conditions of low ventilatory motor output and reduced expiratory driving pressure.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Sleep-related breathing disorders in adults".)

SUMMARY

●Definition – Obstructive sleep apnea (OSA) is characterized by recurrent obstruction of the pharyngeal airway during sleep, with resultant hypoxemia and sleep fragmentation. The pathogenesis of OSA, while not completely understood, is likely due to the interaction between unfavorable anatomic upper airway (UA) susceptibility to upper airway collapse, and sleep-related changes in UA function. (See 'Introduction' above.)

●Reduced UA muscle activity – Upper airway mechanics play a role in the pathogenesis of OSA (algorithm 1) (see 'Upper airway mechanics' above):

•Reduced UA muscle activity during sleep is a physiologic phenomenon of little consequence in healthy individuals, but it may promote UA narrowing in susceptible individuals. (See 'Decreased muscle activity' above.)

•Changes in UA caliber and compliance and loss of load compensation during sleep are additional factors that may promote obstruction. (See 'Changes in caliber and compliance' above.)

•Despite multiple physiologic sleep-related changes in respiratory mechanics and a rise in partial pressure of carbon dioxide (PaCO₂), not all individuals develop complete UA obstruction. Individuals with favorable UA anatomy and lung mechanics are able to sustain rhythmic breathing and normal gas exchange, whereas those with highly compromised upper airways may develop complete obstruction. (See 'Implications' above.)

●Structural, vascular, and neuromuscular factors – Determinants of UA patency can be broadly divided into structural, vascular, and neuromuscular factors.

•Structural determinants include craniofacial structure; surrounding soft tissue, including adipose tissue; vascular structures; and mucosal factors. A craniofacial constraint, such as retrognathia, can limit the potential space shared by soft tissue and the airway lumen. Craniofacial features are heritable traits and may be influenced by racial/ethnic factors. (See 'Structural factors' above.)

•Vascular factors include rostral fluid shifts that occur during recumbent sleep. Increased soft tissue surrounding the airway could be due to increased adipose tissue, enlarged tonsils, or increased vascular volume. (See 'Soft tissue structures' above and 'Rostral fluid displacement' above.)

•Neuromuscular factors include ventilatory motor output, UA muscle activity, and the thoracic-upper airway link via caudal traction. (See 'Neuromuscular factors' above.)

•Obesity is a major risk factor for OSA through multiple mechanisms, including decreased lung volume, increased soft tissue volume, and potential impairment of the mechanical output of UA muscles.

●Pharyngeal obstruction during sleep – Although individual risk factors are known, the precise pathophysiologic pathways leading to UA obstruction in patients with OSA are not well understood. Important triggers of pharyngeal obstruction during sleep in susceptible individuals may include changes in upper airway mechanics during sleep, changes in ventilatory motor output, surrounding tissue pressure, and expiratory narrowing. (See 'Pharyngeal obstruction during sleep' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Kathe G Henke, PhD, who contributed to earlier versions of this topic review.