Obstructive Sleep Apnea

Elise A. Maher

Lawrence J. Epstein

Obstructive sleep apnea (OSA) is a repetitive pattern of upper airway obstruction that occurs during sleep and is associated with sleep fragmentation and/or oxygen desaturation. The combination of sleep fragmentation, oxygen desaturation, and the other pathophysiologic changes causes daytime impairment, particularly excessive daytime sleepiness (EDS), and cardiovascular disease. The disorder was first clinically described as recently as (1). There is a gradient of airway closure in OSA ranging from mild and partial obstruction to a complete and sustained blockage of the upper airway. An event with complete obstruction is called an “apnea,” whereas the one with partial obstruction is called a “hypopnea.” A milder respiratory disturbance called a respiratory effort-related arousal (RERA) fragments sleep with even minimal increases in upper airway resistance. The clinical syndrome has variously been called OSA, the obstructive sleep apnea syndrome, and the obstructive sleep apnea/hypopnea syndrome (OSAHS). The current terminology recommended by the International Classification of Sleep Disorders, 3rd edition (2), and used in this chapter, is OSA. Partial obstruction or hypopnea is frequently seen in the syndrome and so the term OSAHS is used interchangeably with OSA.

Mild obstruction of the upper airway causes turbulent airflow that leads the tissues to vibrate and creates snoring, and sometimes choking, sounds. The sound levels resulting from snoring range from soft to decibel levels loud enough to be considered unsafe in the workplace (>80 dB). Obstructive events can last from 10 seconds to as long as 2 minutes, after which a brief central nervous system arousal from sleep restores the patency (opening or unblocking) of the airway. Severity is defined by the number of breathing disruptions per hour, or apnea-hypopnea index (AHI). Some authors use the term “respiratory disturbance index” (RDI) interchangeably with AHI; however, the American Academy of Sleep Medicine (AASM) Manual for the Scoring of Sleep and Associated Events defines RDI as the number of apneas + hypopneas + RERAs per hour of sleep. An AHI below 5 is considered normal in adults. Mild OSA is in the range of 5 to 14, moderate 15 to 30, and severe more than 30. The number of times that a patient’s oxygen level drops below the minimum normal saturation of 88% to 90% is another indicator of severity, with greater stress on the cardiovascular system and increased risk of cardiovascular disorders related to the more frequent and larger desaturation.

SLEEP AND BREATHING

Most body systems slow down during sleep. The breathing rate slows, and tidal volume is reduced as breaths become shallower, resulting in mild hypoventilation (2 to 8 mm Hg increase in PaCO₂). At the same time, the heart rate slows down, and oxygenation levels dip a little below waking levels. Muscles relax and the airway narrows, leading to an increase in upper airway resistance in the pharynx (the soft tissue pathway starting from the nasal and oral airways that leads down to the larynx; Fig. 12-1). The anatomic and neuromuscular systems
that keep the airway open for normal breathing during wakefulness are not as active during sleep and can fail to compensate for the changes that occur in sleep.

Figure 12-1 This is a sagittal section through the pharyngeal airway of a normal human. The opening to the nose is just above the maxilla. The other structures of the airway include the mandible, hard palate, soft palate, and uvula. The body of the tongue is the genioglossus muscle. Pharyngeal collapse in apneic patients occurs somewhere between the choana and the epiglottis, usually behind the uvula and the soft palate, behind the tongue, or some combination of these sites. In humans, the hyoid bone is not firmly attached to bony structures, thus increasing dependency on the upper airway muscles to keep the airway open. (Modified with permission from Moore, K. L., Dalley, A. F., & Agur, A. M. R. (2013). Clinically oriented anatomy [7th ed., Figure 7.83]. Baltimore, MD: Lippincott Williams & Wilkins/Wolters Kluwer.)

The human airway is anatomically at greater risk than other animals. The hyoid bone that anchors the pharyngeal muscles is not rigidly attached in humans as it is in other mammals, which allows us wide-ranging vocalizations needed for speech. However, this feature makes the upper airway more collapsible when the many muscles responsible for dilating and maintaining the airway are not fully functioning during sleep. The areas of obstruction differ considerably, depending on body position, craniofacial structure, and weight distribution. It has been shown that obese patients tend to have collapse of the velopharynx (soft palate and side-walls of the pharynx), whereas nonobese patients with a recessed chin showed collapse in the oropharynx and velopharynx (3).

The typical reclining sleeping posture causes a gravitational effect on the soft tissue of the upper airway, promoting collapse. This positional effect is at its worst when patients are sleeping on their back (supine). Patients sometimes unknowingly self-treat their OSA by preferring to sleep propped up with extra pillows or in a recliner. In rapid eye movement (REM) sleep, the upper airway can collapse further when dilator muscle activation, reduced the greatest amount compared with waking levels, is chemically blocked.

There are two general types of breathing disturbances in sleep: central and obstructive. Central breathing events are caused by a malfunction in the respiratory control system, leading the person to not take a breath, whereas obstructive events are caused by a physical block of the airway that prevents airflow despite efforts to breathe. Central and obstructive apneas represent a complete lack of airflow, obstructive hypopneas (and less commonly described central hypopneas) represent a partial obstruction to airflow, and both of these can cause fragmented sleep and oxygen desaturation. On the mildest end of the spectrum is the RERA. The limitation of airflow from mild obstruction requires more muscle effort to breathe, like having to suck harder through a narrow straw. This increased effort can cause a brief arousal from sleep or RERA. Taken altogether, these patterns of abnormal breathing—snoring, RERAs, hypopneas, and apneas—are called sleep-disordered breathing (SDB).

EPIDEMIOLOGY

It is difficult to accurately determine the prevalence of OSA in the population because of the various ways that breathing obstructions in sleep are defined and measured, how individuals are screened, and how they are tested. A widely referenced study found that the occurrence of more than five obstructive events per hour during sleep is high, reported at 24% in men and 9% in women aged 30 to 60 years. The same study found the prevalence of OSA, defined as more than five obstructive events per hour and symptoms of sleepiness, was 2% in women and 4% in men aged 30 to 60 years (4). This seminal study was published in 1993 and likely underestimates the prevalence of OSA because of the rapid increase in obesity in the general population since that time. More recent studies suggest prevalence rates of 5% in women and 14% in men, with higher rates associated with older age and higher weight (5).

RISK FACTORS

There are multiple and often intertwined risk factors that predispose to OSA. Longitudinal studies have found that the most significant independent risk factor for OSA is being overweight (5, 6, 7). There is a 4-fold risk for OSA associated with a body mass index (BMI) increase of one standard deviation (this amount of change in BMI is approximately the difference between normal and overweight or between overweight and obese) (1).
Even a 10% weight reduction can lower the AHI by an average of 26% (6). For patients with mild OSA, AHI may be objectively reduced to normal levels after significant weight loss. In a 2009 study that had patients follow a very low-calorie diet, the AHI was brought below 5 when patients lost an average of 10.6% of initial weight (8).

Gender plays a large role in the relative expression of OSA symptoms and health consequences, with males having a reported 2:1 or 3:1 prevalence over females. Hormones and fat distribution both predispose males toward a more collapsible airway. Interestingly, the ratio of males to females who present to the sleep center for testing is closer to 8:1. Males may tend to be identified more readily by primary care doctors as being candidates for OSA or may be prompted more often by female partners to seek treatment (9). The gender difference in prevalence starts to equalize postmenopause.

There is a linear relationship between age and risk for OSA, which starts in middle age and plateaus at about age 65 (10). In women, there is a marked increase in OSA after menopause when the protective effects of progesterone on the respiratory system diminish (11). Although the incidence of OSA is higher in old age than in middle age, at 70% for men and 56% for women, the symptoms and risk factors are different (12). The AHI is not as strongly correlated with obesity (10), EDS, cognitive dysfunction, or hypertension (13, 14) in older individuals. Snoring, one of the main complaints in middle age, subsides in old age. This could be because of hearing loss or because a central component to respiration emerges (15).

Race is a risk factor for OSA although not as widely researched as weight, gender, age, or other factors. African Americans younger than 25 and older than 65 have an elevated risk over African American males between 25 and 65 or males of other races (16, 17). Asians have similar rates of OSA as those in Western cultures. However, BMI rates are lower, and craniofacial features may be the determining factor. Chinese males, for example, have a shorter cranial base, maxilla (upper jaw), and mandible (lower jaw), causing more crowding of the airway (18).

Craniofacial features can have a large impact on the airway, particularly during sleep when muscle relaxation, body position, and gravity can compromise the airway patency of susceptible individuals. There are two types of features that impact potential upper airway obstruction: soft tissue and bone structure. Soft tissue from large turbinates, a deviated septum, occluded sinuses, polyps, a large tongue or uvula, enlarged tonsils or adenoids, or an elongated soft palate can crowd the oronasal airway and cause obstruction and reduced airflow. A large neck size (>17 inches in men and >16 inches in women) is one of the most predictive features of OSA (2). The larger the volume of tissue surrounding the airway, whether fatty or muscular, the more prone it is to partially or fully collapse.

A receding (retrognathic) or small (micrognathic) chin or narrow/high arched palate can be problematic and cause OSA in nonobese patients. Congenital conditions that impact facial bones and soft tissue of the upper airway are associated with increased risk for OSA that can be especially difficult to treat. Features of Down syndrome include a large tongue (macroglossia), small chin, narrow airway, and generalized hypotonia. In a 2009 study, 94% of Down syndrome patients were found to have at least mild OSA and 69% were in the severe range (19). Pierre Robin, Marfan syndrome, and achondroplasia are some of the other genetic disorders that result in decreased airway size and predispose for OSA.

Familial incidence of OSA increases the risk to an individual for OSA. There is a direct increase in risk for OSA with each diagnosed family member (20). This may be because of a combination of genetic and environmental factors. For instance, obesity is an independent risk factor and can be affected by familial lifestyle and dietary factors.

Smoking and sedative use (including alcohol), particularly near bedtime, are additional predisposing factors. Smoking may affect the function of upper airway tissue because of inflammation and toughening from prolonged smoke exposure. One study showed that active smokers are three times more likely to have OSA than nonsmokers (21). Sedatives act as muscle relaxants and also reduce reactivity in the respiratory control system. Sedatives can induce apneas and snoring in people without OSA and worsen respiratory events in symptomatic patients (22, 23, 24).

Other clinical conditions that affect airway size or airway muscle function are risk factors for OSA, including polycystic ovary syndrome, hypothyroidism, and pregnancy. Although hormones can have a protective effect on the airway in sleep, gestational weight gain and pressure on the thoracic cavity can cause or increase snoring and apneas. Apneas during pregnancy that are associated with hypoxemia can be associated with low fetal birth weight and lower Apgar scores (25).

CARDIOVASCULAR DISEASE

Cardiac disease and progression frequently coexist with OSA. Patients with OSA are at increased risk for hypertension, arrhythmias, myocardial infarction (MI), coronary artery disease, pulmonary hypertension, stroke, and sudden cardiac death. Cardiac arrhythmias are more common in patients with OSA as a result of sympathetic nervous system activation (which causes
vasoconstriction and increased blood pressure), hypoxia (low oxygen levels), hypercapnia (elevated carbon dioxide levels), and increased fluctuations in intrathoracic pressure. The increase in negative intrathoracic pressure during an obstructive event negatively affects left ventricular function and decreases stroke volume and cardiac output. Many patients with OSA develop a brady-tachy heart rate pattern in which bradycardia occurs during the apneic phase and tachycardia occurs when breathing returns (26).

Of note, the Sleep Heart Health Study (SHHS) showed that hypopneas accompanied by an oxyhemoglobin desaturation of more than or equal to 4% were associated with prevalent cardiovascular disease independently of confounding covariates (27). In contrast, no association was observed between cardiovascular disease and hypopneas associated with milder desaturation or arousals.

Hypertension

Analysis of data from more than 6,000 adults participating in the SHHS showed that those in the upper quartile of AHI (≥11.0 events per hour) had 42% greater odds of cardiovascular disease than those in the lowest quartile (AHI <1.3 events per hour) (28). Patients with hypertension have an increased incidence of OSA (30%) and 50% of patients with OSA have hypertension (29, 30). Hypertension is strongly associated with OSA, even when mild (31).

The SHHS showed a linear relationship between increasing AHI and incidence of hypertension. Prevalence rates of hypertension in increasing AHI categories were 43% (<1.5 per hour), 53% (1.5 to 4.9 per hour), 59% (5 to 14.9 per hour), 62% (15 to 29.9 per hour), and 67% (≥30 per hour) (32). It is likely that hypertension itself increases the risks of other types of cardiac disease.

Arrhythmias

Arrhythmias are more common in patients with OSA, even in the absence of structural or electrical heart conduction problems. The higher incidence of abnormal rhythms is most likely caused by a combination of hypoxemia, the vagal response to obstructed breathing, increased sympathetic activation, acute increases in blood pressure, and rapid changes in intrathoracic pressure.

Cardiac variability is commonly seen in OSA patients with repetitive alternations between bradycardia and tachycardia in synchrony with obstructive events. Bradycardia is caused by the vagal response to apnea and is followed by an overshoot to tachycardia during the hyperpneic (recovery) phase of breathing. Decreased heart rate variability in response to repetitive apneas is a marker for subsequent development of hypertension because of excessive sympathetic activity and potential loss of vagal control (33).

Other bradycardic patterns, including second- and third-degree atrioventricular (AV) blocks and sinus pauses, are frequently seen in OSA patients. In the absence of any electrophysiologic abnormalities of the sinus or AV node, heart block was found in 10% of patients with OSA. The blocks included AV block type II (Mobitz), third-degree block, and sinus pauses and occurred more often in REM sleep and during 4% or greater oxygen desaturations (34). Sinus pauses of up to 12 seconds can be seen in REM sleep and during long apneas with desaturation.

Ventricular ectopy is also frequently seen in patients with OSA and increases in frequency along with increasing AHI. Individuals with SDB had three times the odds of nonsustained ventricular tachycardia and almost twice the odds of complex ventricular ectopy (nonsustained ventricular tachycardia with bigeminy, trigeminy, or quadrigeminy) (35). In a study of 400 OSA patients conducted by Guilleminault et al. (1), premature ventricular contractions were noted in 193 (48%) of subjects. The common arrhythmias seen in patients with OSA without known cardiac history often occur during sleep (35, 36). Positive airway pressure (PAP) therapy has been shown to decrease ventricular ectopy in patients with heart failure and OSA (37).

There is a 4-fold increase in atrial fibrillation in patients with OSA (35). There is also a higher recurrence of atrial fibrillation after cardioversion in patients with untreated SDB and atrial fibrillation than in patients without a known SDB diagnosis (38).

Congestive Heart Failure

Congestive heart failure (CHF) is strongly associated with Cheyne-Stokes respiration, a type of central sleep apnea (CSA). However, many patients with CSA also have an obstructive component, and there is increased risk of heart failure with OSA as well. In a sample of 81 male patients with CHF, 40% had CSA and 11% had OSA (39). Patients with OSA were heavier, on average, than those with CSA. There is a high frequency of atrial fibrillation and ventricular arrhythmias in patients with both OSA and CHF. The SHHS reported that OSA is associated with a relative odds ratio of 2.38 for heart failure, independent of other known risk factors (28, 40).

Coronary Artery Disease

There is increased risk for MI in patients with OSA, reported to be about four times greater than in those without OSA (41, 42, 43). Additionally, OSA patients may be at higher risk for subsequent cardiovascular events after an MI, possibly because of hypoxemia (44).

Stroke

There is a high incidence of OSA in stroke patients (45, 46). In the SHHS, there was a positive relationship between the history of stroke and the AHI (28). Prospective studies have shown an increased risk of stroke in patients with OSA, with an increased hazard ratio of 2.5 in those with OSA (47).

Sudden Cardiac Death

Although sudden cardiac death tends to occur between 6 and 11 a.m. in the general population, there is 2.5 times the risk of sudden cardiac death between 12 and 6 a.m. for patients with known OSA (48). Increased nocturnal arrhythmias in patients with OSA could be a contributing factor to nocturnal sudden cardiac death (35).

Only gold members can continue reading. Log In or Register to continue