Physiology Of Sleep Disordered

Otolaryngol Clin N Am 40 (2007) 691–711

Physiology of Sleep Disordered Breathing B. Tucker Woodson, MD, FACSa,*, Rose Franco, MDb a

Division of Sleep Medicine, Department of Otolaryngology and Communication Sciences, Medical College of Wisconsin at Froedtert West, 9200 West Wisconsin Avenue, Milwaukee, WI 53226, USA b Division of Pulmonary and Critical Care Medicine and Division of Sleep Medicine, Medical College of Wisconsin, Milwaukee, WI 53226, USA

Obstructive sleep apnea (OSA) is a common disorder resulting from collapse of the pharyngeal airway during sleep. The cause and mechanisms of this collapse are multifactorial, but ultimately, it results from the interdependence of structurally vulnerable upper airway anatomy interacting with physiologic mechanisms of ventilatory instability and reduced or absent dilator muscle control in the sleep state. Our understanding of the pathophysiology of OSA has evolved over the last several decades. Historically, two basic schools of thought existed to describe the genesis of airway collapse: ‘‘active’’ versus ‘‘passive’’ mechanisms. The active theory proposed by Weitzman and colleagues [1] in 1978 resulted from their observation of spasmodic closure of the lateral pharyngeal walls and closure of the velopharynx timed at the end of expiration in animal models. This sphincteric closure ‘‘apparently by active muscle contraction’’ was maintained for the duration of inspiration, and was followed by airway opening occurring following arousal from sleep. Because electromyographic studies of pharyngeal constrictors in humans fail to demonstrate expiratory muscle activity during airway collapse, the concept of active muscular contraction causing airway closure in OSA has been abandoned in favor of other explanatory mechanisms [2]. An alternative to the active theory is based on passive mechanisms that do not require active neuromuscular contraction of pharyngeal muscle to close the airway. Airway obstruction instead results from the interaction of loss of dilating activity of pharyngeal muscles during sleep, as well as the mass of the tongue and other tissues, and their interplay with inspiratory

* Corresponding author. E-mail address: [email protected] (B.T. Woodson). 0030-6665/07/$ - see front matter Ó 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.otc.2007.04.002

oto.theclinics.com

692

WOODSON & FRANCO

intraluminal pressures [3]. As early as 1978, Harper and Sauerland [4] observed a loss of both phasic and tonic electromyographic tongue activity during non-rapid eye movement (NREM) sleep in OSA subjects. This loss of activity was not seen in snoring subjects (is this then a compensatory mechanism in snoring that protects them from obstruction?). Both OSA and snoring subjects contrasted with normal nonsnoring individuals. Normal subjects did not have significant electromyographic activity in the genioglossus muscle during wakefulness or NREM sleep. Their findings suggested that muscle activity was important in maintaining stability in OSA subjects, but that loss of muscle activity alone was inadequate to create airway closure. Additional structural features such as a ‘‘large tongue’’ contributed to collapse, which when combined with negative inspiratory airway pressures, obstructed the airway. One could thus conclude that if the subject has a structurally compromised airway anatomy, he will suffer from OSA; however, this is not always the case. There appears to be a synergistic impact of the anatomy and the ventilatory control stability. Jordan and colleagues have provided evidence to explain the development of OSA through ventilatory instability in the OSA population with a loop gain model has been described [5]. These concepts of structural compromise interacting with nervous system response to ventilatory disturbance form the current working model of OSA physiology. The human upper airway has a complex task. It has the constant challenge of maintaining ventilation, while simultaneously allowing for alimentation, phonation, and speech. In other mammals, the larynx is positioned near the skull base, creating a physiologically separate respiratory and alimentary pathway [6]. Anatomically, the tongue is an oral structure. The intimate relationship to the skull base results in a highly stable airway that is independent of muscle tone. In adult humans, however, the larynx resides in the neck, separated from the bony enclosure of the craniofacial skeleton (Fig. 1). The tongue is oral and pharyngeal, and is more critical in supporting ventilation. This results in a longer pharyngeal airway, which further increases with age, male gender, and OSA. The laryngeal descent process (klinoraphy), while facilitating speech development, predisposes the soft tissue supra-glottic pharyngeal airway to obstruction and requires compensatory mechanisms to maintain stability [7,8]. In such a setting, the addition of otherwise nonpathologic structural or physiologic changes may lead to upper airway collapse during sleep. Multiple mechanisms are intricately involved to maintain upper airway stability. Anatomical structure, neuromuscular tone, ventilatory control mechanisms, level of consciousness (sleep and arousal effects), upper airway reflexes, peripheral nervous system mechanisms, craniofacial and soft tissue structure, body position, vascular tone, surface tension forces, lung volume effects, and expiratory collapse may all contribute to upper airway stability.

PHYSIOLOGY OF SLEEP DISORDERED BREATHING

693

Fig. 1. Differences in upper airway length between infants (left), non-apneic adults (middle), and OSA (right) are shown. In infants, the tongue composes a shorter segment of the pharyngeal airway, and the larynx may reside at the level of the second cervical vertebra. In nonapneic adults, the larynx may reside at the fourth cervical vertebra, and both airway length and tongue area are greater in obstructive sleep apnea.

Structural considerations: anatomy abnormalities Structure significantly predisposes to development, and is a fundamental abnormality of OSA [9,10]. Structure effects of airway size, compliance, and shape are all critical to flow. Although the pharynx is the core, no single morphologic abnormality exists (Fig. 2) that explains all OSA. The three major structural determinates are obesity, soft tissue, and skeletal morphology [11,12]. In addition, structures associated with sleep disordered breathing may also be impacted by the additional factors of ethnicity, gender, obesity, and age [13,14]. In children, adenotonsillar hypertrophy predominates as a cause. This condition is not unique to children, and other craniofacial and soft tissue abnormalities also are present in childhood [5,15,16]. In fact, these other abnormalities may be the main predisposing factors leading to obstruction, and adenotonsillar hypertrophy may simply weight the scales further toward OSA. In adults, no single structural abnormality has been identified as causative, and often multiple anatomic features are associated with OSA. Obesity increases the risk and severity of OSA and the severity of hypoxemia during sleep. Obesity and OSA are each independently associated with one another, with each contributing to the other’s severity. Fat distribution around the neck and airway has long been postulated to compromise the airway; however, obesity effects on metabolism, ventilation, and lung volume might be more important than volume and mass effects to the upper airway [17–19]. Leptin and various inflammatory cytokines, which are either produced or influenced by adiposity, increase CO2 response, and may contribute to

694

WOODSON & FRANCO

Fig. 2. Common lateral cephalometric radiograph abnormalities associated with OSA are depicted: (A) An inferior placed hyoid, (B) increased length and width of the soft palate, (C) increased size of the tongue and increased apposition of the tongue and palate with compromised oro-palatal airway, and (D) decreased projection of the maxilla (PV-A) and mandible (PV-B). A, subspinale; B, supramentale, H, hyoid; PV, porion vertical.

central ventilatory sensitivity, and in turn result in further increased risk for obstructive breathing, as discussed below.

Craniofacial characteristics Most studies assessing structure lack appropriate control groups, with groups differing not only in OSA but often also age, weight, ethnicity, and gender. ‘‘Normal’’ may be a subjective definition; however, few would argue that the skeletal and cartilage framework, which supports the soft tissues, which ultimately determine airway characteristics, are not critical. Framework abnormalities vary, and a constellation of abnormalities is consistently observed. Compared with nonsnoring controls, craniofacial variables associated with OSA include: increased distance of the hyoid bone

PHYSIOLOGY OF SLEEP DISORDERED BREATHING

695

from the mandibular plane, a decreased mandibular and maxillary projection, a downward and posterior rotation of the mandibular and maxillary growth, increased vertical length of the face, increased vertical length of the posterior airway, and increased cervical angulation. In the Wisconsin Cohort of Sleep Disordered Breathing study, two thirds or more of the variability of the apnea hypopnea index was explained by facial structure and obesity [11]. In nonobese subjects, the major contributor was facial structure. Abnormalities in maxillary position and width explained much of OSA in all patient groups. From that study, it may be postulated that OSA is primarily a disorder of maxillary, not mandibular, development. An inferiorly based hyoid position has also been consistently associated with sleep disordered breathing. This displacement may signify abnormalities in airway length or increased tongue size.

Nose Nasal obstruction contributes to the presence and severity of OSA [20]. Nasal blockage might [21–23] Reduce nasal afferent reflexes that help to maintain muscular tone of the upper airway Augment the tendency for mouth opening, which destabilizes the lower pharyngeal airway (by posterior rotation, vertical opening, and inferior displacement of the hyoid) Reduce humidification, increase mucus viscosity, and increase surface tension forces Increase upstream airway resistance, predisposing to downstream airway collapse. A multitude of pathologies cause nasal obstruction, and warrant appropriate evaluation.

Pharyngeal soft tissues In adults, no single soft tissue structure contributes to OSA. The relative contribution of soft tissue size differs among individuals, and particularly among ethnic groups [24]. The size as well as position of the tongue is an important consideration in OSA. In the supine position, the tongue projects posteriorly and is counteracted by the tone of the genioglossal muscle. MRI volumetric studies have identified tongue size as a major predictor of OSA. Patients who have severe OSA, obesity, and marked lateral wall collapse demonstrated sagittally oriented airways that were speculated to result in unfavorable muscular mechanics for reopening the airway [25]. Circular tubes are more efficient than those that are flat [26].

696

WOODSON & FRANCO

The soft palate is both longer and wider in OSA individuals. The tongue is larger, the oral to palatal airspace is smaller, and the posterior airspace behind both the tongue and palate is narrower. In some, the epiglottis is posteriorly placed. Cross-sectional shape of the airway in apneics tends to be more elliptical rather than more circular. The elliptical shape increases the surface area of the airway and frictional resistance compared with a more circular conduit. This leads to a biomechanically weaker structure that is more easily collapsed at less negative pressures. Additionally, when the airway is oriented with its long axis in an anterior posterior direction (in the midsagittal plane), contraction of major airway dilators (such as the genioglossus muscle) is less effective.

Body/jaw position/gravity Body position alters airway size and collapsibility. Airway size decreases following movements from sitting to supine, as well as lateral decubitus to supine [27]. Changes are greater in OSA. Changes in tissue mass, lung volume, tracheal tug, and vascular volume from nonsupine to supine positioning may contribute to airway collapsibility, because they are independent (see discussion below). Gravity affects the lower pharynx and retro-epiglottic airway more than other segments during parabolic flight [27]. Gravity has minimal effect on the position and airway of the upper pharynx and upper tongue base in nonapneic individuals. In apneic subjects, gravity likely contributes to obstruction at the lower pharynx. It may be speculated that because gravity alters the airway by affecting massdas mass increases, gravity effects will increase.

Structural considerations Balance of forces A potentially large number of anatomic and physiologic processes must be integrated into a model of upper airway obstruction during sleep. One model that allows at least some integration is the concept of ‘‘balance of forces.’’ The balance of forces model allows an accurate description of how multiple variables alter upper airway size (Fig. 3). Airway size is determined by both dilating and collapsing forces. Dilating forces include upper airway muscle tone, mechanical force of the airway wall structure, and positive intraluminal airway pressure. Collapsing forces include tissue mass, surface adhesive forces, and negative intraluminal pressures. The resulting difference in these forces is the distending force, which acts on the wall of the upper airway. When the distending force increases, the airway size increases; when it decreases, the airway size decreases.

PHYSIOLOGY OF SLEEP DISORDERED BREATHING

697

Fig. 3. The principal of the balances of forces is diagramed. Transmural forces (Ptm) on the upper airway are depicted. As Ptm increases, the airway enlarges; as Ptm decreases, the airway collapses. Ptm is described by the basic equation Ptm ¼ Pout  Pin. Ptm. It may also be described as the difference between tissue forces (Ptissue) and luminal (Pluminal) or airway forces (ie, Ptm ¼ Ptissue  Pluminal) (see text for details).

The distending force of the upper airway is the transmural pressure (Ptm) of the airway. The equation Ptm ¼ Pout  Pin defines transmural pressure; Pout represents the sum of the dilating the upper airway forces, and Pin represents the sum of the collapsing forces. Another more clinically relevant means to conceive of the forces that act on the upper airway is by considering the skeletal airway structure as a constant and describing the dynamic forces as being either tissue pressures or luminal pressures (Ptm ¼ Ptissue  Pluminal). Tissue pressure includes the forces from tissue mass, tissue elastance, surface tension, and neuromuscular dilating and collapsing forces. Luminal pressures include the segmental airway pressure (Pairway) and pressures relating to airflow (Pflow). Airway pressures may be dilating (ie, such as positive pressures during expiration or with the application of external positive pressuredcontinuous positive airway pressure [CPAP]) or collapsing (negative inspiratory pressure). Although seemingly esoteric, such a model (Ptm ¼ Pluminal  Ptissue) provides a means of quantifying and describing upper airway collapse. Studies have been able to replicate a syndrome identical to OSAS in non-OSA subjects by applying negative intraluminal airway pressures to the upper airway during sleep. The airway pressure required to collapse the pharyngeal airway has been described by the critical closing pressure (Pcrit) [28–30]. When Ptm ¼ Pcrit, airway collapse occurs. Change in area/change in pressure (dA/dP) of the upper airway represents the tendency of the upper airway to collapse, and can be calculated allowing measurement of the intrinsic collapsibility of the upper airway. These measures, however, require controlling for both airflow velocity and muscular tone. Airflow velocity effects on luminal pressures are described by Bernoulli’s equation, and if velocity is zero, flow effects on the luminal

698

WOODSON & FRANCO

pressure are eliminated (Pluminal ¼ Pairway þ 0). When neuromuscular tone is held constant, then tissue forces are constant (Ptissue¼ k [constant]). In this situation, measured airway pressure represents the distending or transmural pressure of the upper airway (Ptm ¼ Pairway  k), and combined with measures of upper airway size, allows for calculation of airway compliance (dA/dP) independent of physiologic influences. Furthermore, airway pressure can be measured and manipulated (such as with nasal CPAP) to assess changes in airway size and compliance. Starling resistor In contrast to a structurally patent airway, where the pressure difference across the tube’s wall is of minimal importance and flow is determined by driving pressure, in sleep disordered breathing the tube wall characteristic becomes the major determinate of airway cross sectional area and air flow. The upper airway is a collapsible conduit with its flow described by the Starling resistor concept (Fig. 4). The Starling Resistor concept builds upon Poiseuille’s law, which describes flow in noncollapsible tubes. Poiseuille’s law states: V ¼ P1  P2/R (V ¼ flow, P1 ¼ pressure upstream, P2 ¼ pressure downstream, P1  P2¼ driving pressure). The resistance component R is determined by length of the tube (L), fluid viscosity (h), and the radius R of the tube (R ¼ resistance ¼ 8hL/pr4). Changes in resistance relate inversely to changes in the area because viscosity and length are usually

Fig. 4. Characteristics of a Starling resistor are shown as a model basin with two attached rigid tubes spanned by a collapsible segment (A). Behavior of an ideal Starling resistor is depicted for differing conditions of upstream pressure (Pus) in B–D. The pressure difference (ie, transmural pressure ¼ Pin  Pout ¼ Ptm) across the airway determines airway size. In (B), fluid fills the basin and the pressure outside the tube (Pout) is greater than pressure inside (Pout O Pin), the tube collapses and no flow occurs. In (C), upstream pressure is increased. When dilating pressures are greater than collapsing pressures (Pin O Pout), the tube is patent and flow occurs. In (D) flow increases with increased positive upstream pressures and unchanged downstream pressures. The driving pressure (downstream  upstream pressure) does not determine flow. Flow is determined by upstream pressure.

PHYSIOLOGY OF SLEEP DISORDERED BREATHING

699

constant. For a rigid tube, flow is determined by the driving pressure across the tube (P1  P2), Ohlm’s Law (V ¼ I/R). This is in contrast to a collapsible tube, as is the case in the upper airway during sleep, where flow may be independent of driving pressure. Resistance changes, and is determined by airway size, which in turn fluctuates with airway luminal pressures, airflow, and all the other surrounding forces acting on the airway (such as described by the balance of forces). In the most basic form of collapsible tube ‘‘a simple collapsible tube’’ (defined as a tube having a wall without intrinsic structural forces), variations in airway size are determined primarily by the upstream and downstream airway pressures, flow, and the transmural pressures of the upper airway. A simple collapsible tube is often modeled with a soft, thin-walled rubber drain placed between two rigid tubes. Three possible conditions of flow across this tube exist: unimpeded flow, flutter, and obstruction. These roughly correlate with the three basic clinical patterns of normal breathing, snoring, and obstruction. In the human upper airway, the supraglottic pharynx is the collapsible tube with a transmural pressure (Ptm); the downstream pressure (Pds) is negative inspiratory pressure (trachea), and upstream pressure (Pus) is ambient pressure (nose) (Pcrit). During wakefulness, low negative inspiratory intraluminal pressures (ie, 5 cm H2O) combined with a large upper airway area (a positive transmural pressure) result in unimpeded flow. During sleep, although the balance of forces changes in ‘‘normals’’ without snoring or apnea, a structurally larger and stable upper airway remains patent. That is, the Pcrit remains lower than the Ptm, preventing collapse. Because transmural pressure remains greater than both downstream and upstream pressures (Ptm O Pus O Pds), the airway behaves more or less as a rigid tube; collapse does not occur and airway size and resistance are not altered. In contrast, apneic subjects have a structurally smaller and more unstable upper airway, resulting in transmural pressures during sleep becoming lower than both downstream and upstream pressures (Pus O Pds O Ptm). With transmural pressures more negative than the closing pressure of the airway during inspiration, airway size is zero. No flow occurs, regardless of the driving pressure across the tube. An intermediate condition occurs in snorers. In snorers, upstream (ambient) pressures are higher than transmural pressures, but downstream pressures (Pus O Ptm O Pds) are less. This results in flutter caused by a choke point of the airway exposed to alternating negative transmural pressure (ie, Pds) and positive transmural pressure (ie, Pus). When the airway is open, it is exposed to downstream pressure, which acts to collapse the choke point. When the choke point is closed, it suddenly is no longer exposed to negative downstream pressures; instead, the segment is exposed to more positive upstream pressures. These pressures dilate and open the airway. An open airway is now exposed to negative downstream pressures and the airway collapses. Repeating of this cycle creates snoring at the choke point.

700

WOODSON & FRANCO

Static and dynamic collapse When conceptualizing the upper airway, it is often very easy to oversimplify the many complex interactions. The size of the upper airway is not static. Dynamic collapse occurs during inspiration when negative inspiratory pressure and airflow factors predominate. Major dynamic forces include phasic neuromuscular tone and airflow [7,31]. Other dynamic components may include surface tension forces that increase collapse and impede opening, vascular compliance (congestion) that may alter upper airway size, and segmental interactions of the upper airway during inspiration and expiration [32,33]. Aberrations in inspiratory flow result in hypoxemia, arousal, and the morbidity of OSA [34]. The timing of collapse of the upper airway and increased resistance occurs not only during inspiration, but also in expiration in OSA [22,35,36]. The contribution of expiratory collapse is underemphasized, yet critical, with maximum airway collapse occurring at end expiration, and precedes dynamic obstruction (Fig. 5). Expiratory collapse is essentially a static process. The main static forces contributing include structure (craniofacial and soft tissue), tonic muscle tone, which is diminished by the low lung volumes of end expiration, and passive declining pressures in the airway lumen. Expiratory obstruction and flow limitation are common in adults who snore and in OSA patients, but not normals [37,38]. Expiratory obstruction increases the work of breathing because of the need to overcome the effects of positive airway pressure (auto-positive end expiratory pressure [PEEP]) [39]. How this alters airway collapse during sleep is unknown. More directly, both nonapneic and apneic individuals demonstrate that positive expiratory intraluminal pressure dilates the airway during both wakefulness and sleep [40,41].

Fig. 5. Pattern of collapse as a function of the ventilatory cycle in both obstructive sleep apnea syndrome (solid line) and normals (dashed line) is shown. (1) Phasic activation at the onset of inspiration slightly increases upper airway size. (2) Early expiration is associated with an increase in airway size in OSA. (3) Airway collapse begins in later expiration. (4) The smallest airway size is at end expiration. Note normals have little airway size variability during the ventilatory cycle.

PHYSIOLOGY OF SLEEP DISORDERED BREATHING

701

Patients who have OSA with tongue base levels of obstruction loose the dilating effects of positive expiratory pressures resulting in greater retropalatal airway collapse (Fig. 6) [42]. Decreased retropalatal size during expiration further predisposes the upper airway to obstruction on subsequent inspiratory breaths. Airway size is ventilatory cycle dependent. Changes are greatest in OSA and minimal in normals. In OSA, cross-sectional area increases during early expiration and narrows during the last half of expiration. The magnitude of the expiratory collapse increases and is progressively worse in the several breaths before apnea (Fig. 7) [43]. Ultimately, this progressive narrowing either completely obstructs the airway or results in a critical size, where the combination of negative inspiratory pressure, Bernoulli forces, and surface adhesive forces combine to create airway closure during the following inspiration. In OSA, only the inspiratory breath immediately preceding apnea demonstrates abnormal resistance and collapse.

Neuromuscular tone and state of consciousness Neuromuscular tone of the airway contributes to the balance of forces and the patency of the airway. In the awake state, the increased tone dilates and stiffens the airway walls. The genioglossus muscle is considered prototypical, but others also contribute (such as the geniohyoid, ala nasi, tensor

Fig. 6. Behavior of multi-element model of the pharynx is depicted for expiration. Without lower pharyngeal obstruction (left), positive pressure contributes to stability of both lower and upper pharyngeal segments. With obstruction lower pharyngeal obstruction (right), positive expiratory pressure (þ) is not transmitted to upper pharyngeal segments. Cross-sectional retropalatal airway size collapses more on obstructed (solid line) than non-obstructed (dotted line) breaths (lower portion of graph).

702

WOODSON & FRANCO

Fig. 7. Progressive expiratory collapse of the upper airway preceding apnea is shown TI, inspiratory time as gray; TE, expiratory time as white. Note that flow and pressure (upper two panels) remain constant until apnea at end of tracing. Cross-sectional airway size at end expiration (arrows) progressively decreases in the three breaths (B-3 to B-1) preceding apnea. Airway size during each inspiration is stable until the actual apnea.

and levator palatini, stylopharyngeus, and styloglossus). Muscle tone is determined by multiple factors, including       

Voluntary activity Postural tone Drive from central respiratory neurons Consciousness level (wake, NREM, and REM sleep) Ventilation (hypercarbia and hypoxia) Lung volume changes Upper airway mechanoreceptors

Mechanoreceptors in the larynx with superior laryngeal nerve afferent activity and increased hypoglossal output are activated with exposure to negative pressure in the airway [43]. Independent of this mechanism, ventilatory related muscle tone is influenced by central respiratory drive. This drive is affected by sleep state (wake, NREM, and REM sleep), chemical control (hypercarbia and hypoxia), and upper airway mechanoreceptors. Central respiratory neurons in the medulla innervate the diaphragm and upper airway muscles via the phrenic, vagus, glossopharyngeal, and hypoglossal nerves. Activation is nonuniform and activity is hierarchical. Ventilatory activity

PHYSIOLOGY OF SLEEP DISORDERED BREATHING

703

is highest in the diaphragm, and is reduced or absent in most upper airway muscles (inversely proportional to postural activation), unless ventilatory drive is increased. For this reason, the effects of sleep/wake state on upper airway muscles are not uniform [44]. Respiratory generators in the medulla activate the genioglossal in anticipation of activation of the diaphragms independent of the mechanoreceptors input [45]. Postural upper airway muscle tonic muscle activity decreases progressively with depth of sleep. A portion of this activity is maintained during NREM (but not REM) sleep (Fig. 8). In OSA patients, muscle tone is increased during wake compared with nonapneic normals. This likely compensates for a structurally smaller and more collapsible upper airway. This augmented tonic and muscle tone is reduced during sleep (Fig. 9). Thus upper airway phasic muscles activity is linked to the ventilatory cycle, and is activated during inspiration. It results from both inherent central respiratory neuron activity and reflexes mediated via peripheral nervous system mechanoreceptors. Central mechanisms preactivate upper airway muscles during inspiration. The resulting upper airway muscle activity stabilizes the upper airway before the potentially detrimental negative inspiratory forces generated by the diaphragm [46]. This activity is independent of upper airway mechanoreceptors. Upper airway preactivation, but not diaphragm activity, is suppressed by various sedative medications, including alcohol and benzodaizepam medications. This is clinically relevant in the setting of habitual alcohol or sedative use. With the loss of preactivation, the upper airway may be unable to compensate to negative airway pressure and other collapsing forces, and may worsen upper airway obstruction. Upper airway mechanoreceptors primarily located at the level of the epiglottis react to negative airway pressure and drive phasic upper airway muscle activity [47–49]. Both OSA and non-OSA individuals have the potential to

Fig. 8. A sleep-related decrease in tonic muscle tone (squares) and its associated increase in upper airway resistance (circles) with differing levels of non-REM sleep are shown. Conceptually, phasic upper airway reflexes (triangle) occur acutely at sleep onset and are reduced in stage 2; increased reflex activity in stages 3 and 4 stabilizes ventilation despite high airway resistance.

704

WOODSON & FRANCO

Fig. 9. Changes in muscle tone in patients with OSA (upper) compared with normal subjects (lower). Sleep onset (arrow) is associated with decrease in tonic muscle tone in both groups. Loss of wakefulness results in loss of phasic muscle tone in OSA (not depicted). Increased negative inspiratory pressure and arousals associated with apneas result in augmented phasic inspiratory EMG activity in OSA (seen in upper graph).

demonstrate this reflex during wakefulness; however, in the normal airway, the reflex is not active at rest and requires augmented breathing (such as exercise). In OSA patients, this reflex primarily mediates increased upper airway muscle activity that is present during wakefulness. This reflex is also state-dependent, and in OSA patients is acutely lost following the transition from wake to NREM sleep [50]. Decreased phasic upper airway muscle tone worsens airway obstruction.

Vascular effects on structure Blood volume changes in the head and neck may affect upper airway size. In human subjects, pharyngeal upper airway size may be altered by changes in leg elevation [34]. This effect is likely mediated through changes in central venous pressure. The location and exact cause of this change in size or its relationship to OSA are not known; however, indirect evidence suggests that abnormalities in blood vessels or blood volume contribute. The soft tissue of the lateral pharyngeal walls is a major abnormality in OSA [51]. Anatomically, they are both enlarged and abnormally compliant in OSA. The reason is not known, yet potential abnormalities may involve muscle, blood vessels, and fat. Given that CPAP’s structural effectiveness is mediated primarily by dilation of the lateral walls at low CPAP pressures (5–15 cm H2O), venous blood volume likely contributes significantly to these observed changes. Whether vascular abnormalities contribute to the development of OSA awaits further study [52].

PHYSIOLOGY OF SLEEP DISORDERED BREATHING

705

Surface tension Other mechanisms contributing to airway narrowing may include tissue surface adhesive forces [53]. Increased surface tissue adhesion worsens airway collapse, and lubricants decrease airway collapse. Upper airway pressure flow characteristics during NREM sleep demonstrate hysteresis. For a given pressure flow parameter, flow is decreased in late inspiration compared with early inspiration. Similar airflow findings have been observed in the lower airway, and have been explained using collapse from surface adhesive forces. Lung volume Changes in lung volume significantly alter pharyngeal upper airway size. This ‘‘lung volume dependence’’ of pharyngeal airway size occurs during wakefulness and sleep [52]. Increased lung volumes increase pharyngeal size, and decreased lung volumes contribute to pharyngeal collapse. Although when initially observed, reflex activation of upper airway dilator muscles was speculated, subsequent studies have demonstrated that changes are a mechanical effect of tracheal and thoracic traction. Thoracic traction, commonly referred to as ‘‘tracheal tug,’’ markedly influences pharyngeal size and patency, and is mediated through the mediastinum, intrathoracic pressures, and the trachea. Changes are independent of neuromuscular activity or upper airway muscle support. Passive tracheal traction likely alters pharyngeal collapsibility by increasing longitudinal tension and stability on the pharyngeal wall (Fig. 10) [35].

Ventilatory control mechanisms Awakeness, REM, and NREM sleep differ in arousal mechanisms and ventilatory sensitivity to hypoxia and hypercapnia [45]. Differences in these

Fig. 10. The effects of tracheal tug and longitudinal tension on the airway are demonstrated. Inferior displacement of the trachea or hyoid bone increases longitudinal tension of the pharyngeal airway and decreases collapsibility.

706

WOODSON & FRANCO

may be present between those who have OSA and normals. Central ventilatory control is destabilized by changes in level of consciousness, the cyclic alternating potential (CAP), and by hypoxia and hypercapnia. Hypercapneic drive is the most tightly controlled response, and is the primary determinant for ventilatory control. This drive decreases from wake to NREM and REM sleep, and may result in hypoventilation during sleep if uncompensated. In the normal subject, hypoxic ventilatory drive also decreases from wake to NREM, and is lowest in REM sleep. Ventilation during NREM sleep is primarily mediated by carbon dioxide and chemical control, through both a central effect on the medullary centers and indirectly through the carotid bodies. Both classic arousals and oscillations in the cyclic alternating potential (which alter arousal threshold sensitivity) affect ventilation and ventilatory drive, because of the central nervous system (CNS) effects on the responsiveness of the afferent receptors and the central controllers, and this further affects airway stability (Fig. 11) in an interdependent fashion, along with changes in airway tone and flow. Arousal acutely activates and stabilizes the upper airway muscles in sleep. Cyclic arousals and ventilatory changes may worsen airway collapse. Brief awakenings rapidly shift central CO2 sensitivity lower to the levels of wakefulness. This increases ventilation and quickly reduces CO2. Rapid resumption to sleep (or in the case of CAP, to more stable NREM sleep), results in CO2 levels below ‘‘stable NREM’s’’ critical hypocapneic ventilatory threshold. A ventilatory ‘‘overshoot’’ occurs, and ventilatory drive decreases. This systematic regulation through feedback loops is best explained using the engineering theory of loop gain. This concept describes the overall gain of a system controlled by feedback loops. The ventilatory loop gain system and its interplay with the airway anatomy, both passive and dynamic can

Fig. 11. The interaction of small upper airways, increased resistance (with associated increase in carbon dioxide) and the effects of sleep on muscle tone are diagramed. Changes in sleep state and the associated changes in central ventilatory control augment normal physiologic reflexes, which result in decreased central respiratory drive and worsening of upper airway obstruction (see text for details).

PHYSIOLOGY OF SLEEP DISORDERED BREATHING

707

be measured. The primary variable in the respiratory control is chemoresponsiveness to hyercapnia and hypoxia, and it is termed controller gain. The higher the controller gain, the brisker the responsive to hypercapnia. The effectiveness of the level of ventilation to eliminate CO2 is termed plant gain, and variables causing a high plant gain are things that affect the alveolar volume changes and alveolar to arterial CO2 ratio, such as a low functional residual capacity, low dead space ratio, low metabolic rate, low cardiac output, and a high PaCO2. Both types of gain are important in ventilatory stability. High loop gain resulting in ventilatory overshoots may contribute to central apneas, central hypopneas, periodic breathing, and oscillating decreases and increases in upper airway muscles activity. To propagate the repeating pattern of obstruction, arousal, ventilation, and stability, the system must have a phase delay between the central controllers of ventilation and the sensing portion of the system, as in the carotid bodies and medulla. The longer the phase delay, the more likely the subject is to fall into a waxing and waning cycle of periodic breathing. In structurally vulnerable upper airways, periodic breathing destabilizes the upper airway because of the nonuniformity of neuromuscular drive to lower and upper airway muscles. Because the genioglossus and other upper airway muscle are less tightly linked to central motor neurons than the diaphragm, a decrease in drive that has little effect on the diaphragm function may eliminate motor activity to upper airway muscles, and cause decreased upper airway size, increased airway resistance, and overt obstruction.

Peripheral nervous system effects Interactions of the peripheral nervous system contribute to OSA via altering arousal threshold, compensatory reflexes as discussed in the loop gain model, and muscle tone affecting both upper and lower airway volume and function. It is ultimately the increased work of breathing and mechanoreceptor stimulation associated with obstructive breathing and not asphyxia or hypoxia that results in brainstem, spinal cord, or cortical arousal. Both the arousal, the mechanical obstruction, and the hypoxia contribute to autonomic activation (blood pressure, heart rate) and sleep fragmentation [34]. Mechanoreceptor-mediated reflexes appear to be critical in compensation for OSA but not normal subjects. Negative inspiratory pressure reflexes stimulation of upper airway muscles is reduced or eliminated by both application of nasal CPAP or topical pharyngeal anesthesia. In OSA patients during wakefulness, CPAP or topical upper airway anesthesia markedly decreases genioglossus EMG activity, whereas normal subjects have no change [54,55]. Damage to these afferent mechanisms may worsen OSA. In fact, abnormalities such as pharyngeal nerve damage and decreases in pharyngeal tactile sensitivity are observed in OSA. Imunohistopathologic muscle fiber-type changes observed in OSA may be caused by muscle denervation and

708

WOODSON & FRANCO

reinervation [42,56–58]. Muscle fiber hypertrophy and degenerative changes resulting from denervation and reinervation and eccentric contraction during stretch may worsen OSA by affecting muscle elastance, strength of contraction, and force of dilation. Changes to a greater percentage of Type II muscle fiber types have been observed, as well as muscle hypertrophy, which if progressive, would contribute to impingement on airway size [59]. Vestibular afferents may also affect sleep disordered breathing in both children and adults. Vestibular system abnormalities affecting the autonomic nervous system have been demonstrated. The vestibular system has also been shown to mediate changes in upper airway muscle tone with positional changes. (Is there also an impact on lung volume with vestibular inputs?)

Summary Upper airway competence involves complex interactions between anatomy and physiology. For most, OSA is an abnormality of a structurally small and abnormally collapsible upper airway interacting with normal physiologic mechanisms. The mechanisms individually or in combination may or may not be pathologic, yet together airway instability, ventilatory control instability, and the state of sleep conspire to create cyclic obstructive breathing. Understanding the pathophysiology provides insight into why and how to treat the disorder. Expanding on simplistic models of airway collapse may lead to potential novel new medical and surgical treatments for OSA. References [1] Weitzman ED, Pollak CP, Borowiecki BB, et al. The hypersomnia-sleep apnea syndrome: site and mechanism of upper airway obstruction. In: Guilleminault C, Dements WC, editors. Sleep apnea syndromes, Kroc Foundation Series. vol. 11. New York: Alan R Liss Inc.; 1978. p. 235–48. [2] Guilleminault C, Hill M, Simmons FB, et al. Passive constriction of the upper airway during central apneas: fiberoptic and EMG investigations. Respir Physiol 1997;108:11–22. [3] Remmers JE, deGroot WJ, Sauerland EK, et al. Pathogenesis of upper airway occlusion during sleep. J Appl Physiol 1978;44:931–8. [4] Sauerland EK, Harper RM. The human tongue during sleep: electromyographic activity of the genioglossus muscle. Exp Neurol 1976;51:160–70. [5] Jordan A, White DP, et al. Respiratory control stability and upper airway collapsibility in men and women with obstructive sleep apnea. J Appl Physiol 2005;99:2020–7. [6] Lieberman DE, Mc Carthy RC. The ontogeny of cranial base angulation in humans and chimpanzees and its implications for reconstructing pharyngeal dimensions. J Hum Evol 1999;36:487–517. [7] Malhotra A, Pillar G, Fogel RB, et al. Pharyngeal pressure and flow effects on genioglossus activation in normal subjects. Am J Respir Crit Care Med 2002;165:71–7.

PHYSIOLOGY OF SLEEP DISORDERED BREATHING

709

[8] Pae EK, Lowe AA, Fleetham JA. A role of pharyngeal length in obstructive sleep apnea patients. Am J Orthod Dentofacial Orthop 1997;111:12–7. [9] Galvin JR, Rooholamini SA, Standford W. Obstructive sleep apnea: diagnosis with ultrafast CT. Radiology 1989;171:775–8. [10] Isono S, Remmers J, Tanaka A, et al. Anatomy of the pharynx in patients with obstructive sleep apnea and in normal subjects. J Appl Physiol 1997;82:1319–26. [11] Dempsy JA, Skatrud JB, Jacques AJ, et al. Anatomical determinates of sleep disordered breathing across the spectrum of clinical and non-clinical subjects. Chest 2002;122: 840–51. [12] Do K, Ferreyra H, Healy J, et al. Does tongue size differ between patients with and without sleep-disordered breathing?’’. Laryngoscope 2000;110:1552–5. [13] Lyberg T, Krogstad O, Djupesland G. Cephalometric analysis in patients with obstructive sleep apnea syndrome: skeletal morphology. J Laryngol Otol 1989;103:287–92. [14] Will MJ, Ester MS, Ramirez SG, et al. Comparison of cephalometric analysis with ethnicity in obstructive sleep apnea syndrome. Sleep 1995;18:873–5. [15] Aren R, McDonough JM, Costarino AT, et al. Magnetic resonance imaging of the upper airway structure of children with obstructive sleep apnea syndrome. Am J Respir Crit Care Med 2001;164:698–703. [16] Zucconi M, Caprioglio A, Calori G, et al. Craniofacial modifications in children with habitual snoring and obstructive sleep apnoea: a case control study. Eur Respir J 1999;13: 411–7. [17] Bradley T, Brown I, Grossman R, et al. Pharyngeal size in snorers, nonsnorers, and patients with obstructive sleep apnea. N Engl J Med 1986;315:1327–31. [18] Kryger M, Felipe L, Holder D, et al. The sleep deprivation syndrome of the obese patientda problem of periodic nocturnal upper airway obstruction. Am J Med 1974;56: 531–9. [19] Mortimore I, Marshall I, Wraith P, et al. Neck and total body fat deposition in non-obese and obese patients with obstructive sleep apnea compared with that in control subjects. Am J Respir Crit Care Med 1998;157:280–3. [20] Young T, Finn L, Kim H. Nasal obstruction as a risk factor for sleep-disordered breathing. Allergy Clin Immunol 1997;99:757–62. [21] Meurice J, Marc I, Carrier G, et al. Effects of mouth opening on upper airway collapsibility in normal sleeping subjects. Am J Respir Crit Care Med 1996;153:255–9. [22] Schwab RJ, Gefter WB, Hoffman EA, et al. Dynamic upper airway imaging during respiration in normal subjects and patients with sleep disordered breathing. Am Rev Respir Dis 1993;148:1385–400. [23] Schwab RJ, Parirstein M, Pierson R, et al. Identification of upper airway anatomic risk factors for obstructive sleep apnea with volumetric MRI. Am J Respir Crit Care Med 2003; 167(9):1176–80. [24] Redline S, Tishler PV, Hans MG, et al. Differences in sleep disordered breathing in African Americans and Caucasians. Am J Respir Crit Care Med 1997;155:186–92. [25] Rodenstein DO, Dooms G, Thomas Y, et al. Pharyngeal shape and dimensions in healthy subjects, snorers, and patients with obstructive sleep apnoea. Thorax 1990;45:723–7. [26] Leiter JC. Upper airway shape. Is it important in the pathogenesis of obstructive sleep apnea? Am J Respir Crit Care Med 1996;153:894–8. [27] Beaumont M, Fodil R, Isabey D, et al. Gravity effects on upper airway area and lung volumes during parabolic flight. J Appl Physiol 1998;84(5):1639–45. [28] Schwartz AR, Smith PL, Wise RA, et al. Induction of upper aiway occlusion in sleeping individuals with subatmospheric nasal pressure. J Appl Physiol 1988;64:535–42. [29] Fogel R, Malhotra A, Pillar G, et al. Genioglossal activation in patients with obstructive sleep apnea versus control subjectsdmechanisms of muscle control. Am J Respir Crit Care Med 2001;164:2025–30.

710

WOODSON & FRANCO

[30] Jokic R, Klimaszewski A, Mink J, et al. Surface tension forces in sleep apnea: the role of a soft tissue lubricantda randomized double-blind, placebo-controlled trial. Am J Respir Crit Care Med 1998;157:1522–5. [31] Shepard J, Pevernagie D, Stanson A, et al. Effects of changes in central venous pressure on upper airway size in patients with obstructive sleep apnea. Am J Respir Crit Care Med 1996; 153:250–4. [32] Stanescu D, Kostinavev S, Sonna A, et al. Expiratory flow limitation during sleep in heavy snorers and obstructive sleep apnea patients. Eur Respir J 1996;9:2116–21. [33] Kimoff RJ, Cheong TH, Olha AE, et al. Mechanisms of apnea termination in obstructive sleep apnea: role of chemoreceptor and mechanoreceptor stimuli. Am J Respir Crit Care Med 1994;149:707–14. [34] Tuck S, Remmers J. Mechanical properties of the passive pharynx in Vietnamese pot-bellied pigs. II. Dynamics. J Appl Physiol 2002;92:2236–44. [35] Van de Graaff W. Thoracic traction on the trachea: mechanisms and magnitude. J Appl Physiol 1991;70(3):1328–36. [36] Lofasa F, Lorino AM, Fodil R, et al. Heavy snoring with upper airway resistance syndrome may induce positive end-expiratory pressure. J Appl Physiol 1998;85:860–6. [37] Badr SM, Dawak A, Skatrud JB, et al. Effect of induced hypocapnic hypopnea on upper airway patency in humans during NREM sleep. Respir Physiol 1997;110:33–45. [38] Rowley JA, Sannders CS, Zahn BR, et al. Effect of REM sleep on retroglossal cross-sectional area and compliance in normal subjects. J Appl Physiol 2001;91(1):239–48. [39] Woodson BT. Expiratory pharyngeal airway obstruction during sleep: a multiple element model. Laryngoscope 2003;113:1450–9. [40] Morrell MJ, Arabi Y, Zahn B, et al. Progressive retropalatal narrowing preceding obstructive apnea. Am J Respir Crit Care Med 1998;158:1974–81. [41] Series F. Upper airway muscles awake and asleep. Sleep Med Rev 2002;6:195–212. [42] Van Lunteren E. Muscles of the pharynx: structural and contractile properties. Ear Nose Throat J 1993;72:27–9. [43] Horner R. Impact of brainstem sleep mechanisms on pharyngeal motor control. Respir Physiol 2000;119:113–21. [44] Strohl KP, Hensley MJ, Hallet M, et al. Activation of upper airway muscles before onset of inspiration in normal humans. J Appl Physiol 1983;53:87–98. [45] Malhotra A, Huang Y, Gogel R, et al. The male predisposition to pharyngeal collapse: importance of airway length. Am J Respir Crit Care Med 2002;166:13888–95. [46] Martin S, Marshall I, Douglas N. The effect of posture on airway caliber with the sleepapnea/hypopnea syndrome. Am J Respir Crit Care Med 1995;152:721–4. [47] Mezzanote WS, Tangle DJ, White DP. Influence of sleep onset on upper-airway muscle activity in apnea patients versus normal controls. Am Rev Respir Crit Care Med 1996;153: 1880–7. [48] Mezzanote WS, Tangle DJ, White DP. Waking genioglossal electromyogram in sleep apnea patients versus normala controls (a neuromuscular compensation mechanism). J Clin Invest 1992;89:1571–9. [49] Schwab RJ, Gupta KB, Gefter WB, et al. Upper airway and soft tissue anatomy in normal subjects and patients with sleep-disordered breathing. significance of the lateral pharyngeal walls. Am J Respir Crit Care Med 1995;152:1673–89. [50] Kuna ST, Deepak GB, Ryckman C. Effect of nasal airway positive pressure on upper airway size and configuration. Am Rev Respir Dis 1988;138:969–75. [51] Van der Touw T, Crawford ABH, Wheatley JR. Effects of a synthetic lung surfactant on pharyngeal patency in awake human subjects. J Appl Physiol 1997;79:78–85. [52] Dempsey J, Smith C, Harms C, et al. Sleep and breathing state of the art review: sleepinduced breathing instability. Sleep 1996;19(3):236–47. [53] Deegan PC, Nolan P, Carey M, et al. Effects of positive airway pressure on upper airway dilator muscle activity and ventilatory timing. J Appl Physiol 1996;81(1):470–9.

PHYSIOLOGY OF SLEEP DISORDERED BREATHING

711

[54] Liistro G, Stanescu D, Veriter C, et al. Upper airway anesthesia induces airflow limitation in awake humans. Am Rev Respir Dis 1992;146:581–5. [55] Friberg D. Heavy snorer’s disease: a progressive local neuropathy. Acta Otolaryngol 1999; 119(8):925–33. [56] Petrof BJ, Pack AI, Kelly AM, et al. Pharyngeal myopathy of loaded upper airway in dogs with sleep apnea. J Appl Physiol 1994;76(4):1746–52. [57] Woodson BT, Garancis JC, Toohill RJ. Histopathologic changes in snoring and obstructive sleep apnea syndrome. Laryngoscope 1991;1010:1318–22. [58] Se´rie`s F, Cote C, St. Pierre S. Dysfunctional mechanical coupling of upper airway tissues in sleep apnea syndrome. Am J Respir Crit Care Med 1999;159:1551–5. [59] Series F, Marc I. Influence of lung volume dependence of upper airway resistance during continuous negative airway pressure. J Appl Physiol 1994;77:840–4.