Epidemiology

Obstructive sleep apnea-hypopnea (OSAH) is an increasingly important disease with numerous clinically relevant consequences, including neurocognitive and cardiovascular sequelae.^1–3 The prevalence of this disease varies, depending on the definitions (of hypopnea) used. Young and colleagues⁴,⁵ showed that 4% of men and 2% of women in a middle-aged North American population had symptoms of OSAH and an apnea-hypopnea index exceeding 5. However, 24% of men aged 30 to 60 and 9% of women had an abnormal apnea-hypopnea index but without excessive sleepiness, which had been used to define the former statistics. Cardiovascular risk assessments, however, have shown a dose-response relationship between the apnea-hypopnea index and various sequelae; thus, the definition and epidemiology are still evolving (Young, Peppard, Gottlieb 2002).

Pathophysiology

Considerable progress has been made in understanding the genesis of obstructive events. The upper airway is anatomically small, and augmented pharyngeal dilator muscle activation maintains airway patency while the patient is awake but not while asleep, when an increase in upper airway resistance is found. Snoring, an important marker of increased upper airway resistance, is in part genetically determined,⁶ which perhaps reflects anatomical contributions such as a degree of retrognathia or overbite. Racial differences may be explained by this (apnea is more frequent among African Americans).⁷ Airway muscle tone insufficient for the airway size may allow intraluminal negative pressure to collapse the pharyngeal “tube.” Additional anatomical factors include enlarged tonsils or adenoids, vascular perfusion, the posture of the individual (supine versus lateral) and, of importance, fat accumulated in the pads in the lateral pharyngeal wall (Fig. 16-1).⁸

During wakefulness, augmented pharyngeal dilator muscle activity maintains airway potency. At sleep onset and/or during rapid-eye-movement sleep (with active inhibition of muscles), this reflex activity is diminished, and if airway anatomy is abnormal, the airway is compromised, which leads to hypopneas and/or apneas. As a result, hypoxia and hypercapnia occur; ventilation is stimulated, often with arousal from sleep; and airway patency is reestablished. With the return to sleep, the cycle is repeated. It is possible, then, to conceive of a continuum of disordered breathing from snoring alone to an inability to breathe and sleep at the same time. Additional risk factors for OSAH are obesity, male gender, and increasing age. Of patients with OSAH, 70% are obese (pharyngeal size is diminished); sleep laboratories report a fivefold or sixfold increased risk of OSAH in men in comparison with women; and the prevalence increases with age.⁹ An evolving literature¹⁰ also suggests an important concept of snoring-induced traumas causing sensory and/or motor neuronal damage, as well as actual damage to the muscle (Boyd, Petrof, and Hamid, 2004).

Clinical Manifestations/Sequelae

OSAH should be suspected in patients who snore intrusively and who are obese (body mass index > 30) and/or in whom apneas have been witnessed. However, more subtle manifestations can occur (e.g., in the 30% who are not obese); therefore, questioning with regard to daytime sleepiness and sleep quality is mandatory.

Poor sleep quality and daytime sleepiness are largely the results of sleep fragmentation by repetitive arousals. The neurocognitive sequelae of recurrent arousals also include reduced performance in neuropsychological tests, lengthened reaction times, altered quality of life, and an increased risk of vehicular accidents and work-related accidents.¹,⁸ A causal relationship to all is supported by the response to treatment with continuous positive airway pressure (CPAP), which improves these sequelae.^11–13 Because of its practical importance, more should be said about sleep apnea and driving.

Human error is a major determinant in automobile accidents; inattention, improper lookout, and other perceptual and cognitive errors account for up to 40% of cases. Progressive daytime sleepiness can enhance inattention and thereby increase the risk of accidents in such patients. OSAH is an important cause of daytime sleepiness, along with cognitive impairment, and consequently contributes to the problem of drowsy driving.

Sleep-related vehicular accidents are not only more common than is generally realized (Maycock found that 29% of 4600 respondents in a U.K. survey admitted to having felt close to falling asleep at the wheel in the previous year, and 18% had accidents in the previous 3 years) but are also more liable to result in death or serious injury as a result of the relatively high speed of the vehicles on impact. The financial and human costs can be considerable. The determination that sleeping at the wheel is the cause of an accident is based on the following:

The absence of skid marks.

The fact that for 7 seconds, the driver could have clearly seen at the point of runoff or the object hit (which implies prolonged inattention rather than momentary distraction).

Other causes such as mechanical failure are eliminated.

It is appreciated that drivers who are able to respond after these accidents seldom acknowledge having fallen asleep.

A strong association between sleep apnea and the risk of traffic accidents is now well documented. A Spanish study revealed that 102 drivers received emergency treatment after vehicular accidents and were more likely by a factor of 6 to have OSAH. Results of a French study suggested that approximately one half of drivers involved in sleep-related vehicular accidents have sleep disorders and that 31% have clear indications of OSAH. In addition, patients with OSAH in many other studies have been shown to have an increased rate of accidents. It is important to stress, however, that although patients with OSAH as a group are at increased risk, not all patients are at the same risk; results of the largest study to date suggested that increased automobile accidents may be restricted to patients with more severe apnea [age > 40], although sleep-related vehicular accidents are recognized to be multifactorial in origin.

Driver performance can be measured by simulators of varying degrees of sophistication, and some patients with OSAH perform as poorly as subjects intoxicated with alcohol. Beneficial effects of treatment, including CPAP and surgery, have also been shown with these simulators. The U.K. Driver and Vehicle Licensing Agency (DVLA) has a guide for medical practitioners in which it is pointed out that it is the duty of the license holder to notify the DVLA of any medical condition that may affect safe driving. There are some circumstances in which the license holder cannot, or will not, do this. Under these circumstances, the General Medical Council has issued clear guidelines:

The cardiovascular sequelae are best considered as immediate and delayed. The immediate response to the obstructed breathing is an increase in negative intrapleural pressure with increased venous return (and increased output of atrial natriuretic peptide and resulting nocturia) and reduced cardiac output (due to the increased afterload).

At the same time, the associated hypoxemia promotes sympathetic activation and circulatory vasoconstriction. With the return of airflow, the augmented preload leads to increases in stroke volume and in systemic blood pressure. This occurs repeatedly, and the normal nocturnal fall in blood pressure may be lost. A delayed effect on diurnal blood pressure may then follow. Indeed, it is appreciated that as much as one third of “essential” hypertension is associated with OSAH.¹⁴,¹⁵ A causal relationship is, again, supported by the response to treatment (with CPAP)¹⁶ (Pepperell, 2002). The combination of immediate and delayed hemodynamic effects in OSAH have been associated with increased risk of myocardial infarction and congestive heart failure, and there is evidence of a link between these and stroke.¹⁷ Additional links have been demonstrated with insulin resistance.¹⁸ The combination of obesity, insulin resistance (with or without diabetes), hypertension, and cardiovascular disease is typical of the metabolic syndrome, or syndrome X. Because these may all be associated with OSAH also, OSAH should obviously be considered as well; some authorities refer to it as syndrome Z.¹⁹

Finally, consideration might be given to the role of chronic hypercapnia in the setting of OSAH. Obesity is complicated in 10% of patients with OSAH by CO₂ retention (in part caused by the increased load on the respiratory system), but OSAH alone may produce this through repeated bouts of CO₂ retention at night, compensatory bicarbonate retention, and a daytime metabolic alkalosis that necessitates compensation. Evidence for this comes, again, from response to treatment with CPAP. One clinical variant of this is the pickwickian syndrome, so named after the “fat boy,” Joe, in Charles Dickens’ “Posthumous Papers of the Pickwick Club” who was, like the patient reported, obese, a snorer, sleepy, and in heart failure (dropsy). All these patients also have CO₂ retention and sleep apnea, usually obstructive.

In the example shown in Figure 16-2, a sleep-onset central apnea is followed by a hypopnea associated with efforts to breathe, registered by abdominal and thoracic impedance plethysmography. The hypopnea in this instance, however, resulted from ineffectual diaphragm contraction, evidenced by the paradoxical inward movement of the abdomen, presumably caused by the excessive abdominal load.

Figure 16-2 A composite of: abdominal and thoracic respiband (respiratory inductance plethysmography) set-up; overnight oximetry demonstrating repetitive desaturations typical of obstructive apnoeas; a 30 sec epoch of the polysomnogram with electroocculogram and electroencephalogram showing sleep onset and a mixed apnoea with initial absence of effort (sleep onset induced) followed by paradoxical movement of the abdomen and thoracic bands indicating effort.

Diagnosis

The initial evaluation of the patient with suspected OSAH is based on identification of disease markers by history and, to a lesser extent, physical examination. The physician may then choose to proceed with a relatively simple and inexpensive investigation such as overnight oximetry to confirm the presence of sleep-disordered breathing before other investigations such as nocturnal polysomnography.

Treatment

When the primary problem is an abnormally small upper airway, treatment is aimed at rectifying this. Cures may be potentially effected with surgical removal of obstructing tonsils and adenoids, substantial weight loss, and, in rare cases, prevention of supine sleep. There currently exist no drugs that increase further upper airway dilator activity.

Medical treatment then relies on devices to produce dilatation of the pharyngeal airway: nasal CPAP, which acts as a pneumatic splint, and oral appliances, which advance the mandible and hence the tongue.

Nasal Continuous Positive Airway Pressure

This remains the treatment of choice with substantial grade A evidence summarized by a Cochrane Review. Important immediate benefits are seen in sleepiness, cognition, quality of life, accidents, and blood pressure reduction, and reduced cardiovascular events are anticipated. However, compliance with CPAP is imperfect; approximately 20% of patients do not adhere to the CPAP regimen. Remedial, if imperfect, strategies include heated humidification, use of nasal decongestants, and intensive follow-up (cognitive behavioral therapy). Bilevel positive airway pressure is no better tolerated (Fig. 16-3).

Figure 16-3 Continuous positive airway pressure (CPAP) mask.

Etiology and Pathogenesis

CSA is a heterogeneous disease entity. In general, patients can be classified into two broad groups based on wakefulness levels of arterial carbon dioxide tension (PaCO₂) and their ventilatory response to carbon dioxide. The first group consists of patients who tend to hypoventilate, have high levels of PaCO₂ in the absence of intrinsic lung disease, and have a blunted ventilatory response to carbon dioxide. They tend to have recurrent episodes of respiratory failure. Patients within this group often have a clinical picture that merges into the spectrum of primary alveolar hypoventilation. Many of them are obese and have features of the obesity-hypoventilation syndrome. At the other end of the spectrum of CSA is the second group of patients, who either ventilate normally or hyperventilate and have normal or low wakefulness PaCO₂ levels and a normal or exaggerated ventilatory response to carbon dioxide. These patients often present with clinical features typical of sleep apnea. Many of them have Cheyne-Stokes respiration (CSR). Even though the clinical and physiological differences between the two groups are marked, the two groups may have similar nocturnal apneic events and sleep architecture.

CSR is characterized by alternating periods of hyperventilation and hypoventilation or apnea. It was first described by Hippocrates, but the classic descriptions were made by John Cheyne and William Stokes in the 19th century. The etiology and pathogenesis of CSR have been argued since the description by John Cheyne of a patient with both enlarged cerebral ventricles and heart disease who had periodic breathing. Autopsy studies have shown that most subjects with CSR have structural abnormalities of the brain. The disease is, however, seen frequently in patients with cardiovascular disease, particularly heart failure. Therefore, theories suggesting both neurological and cardiovascular mechanisms in the pathogenesis of the disease have been postulated. Proposed neurological abnormalities include cyclical medullary depression or medullary hyperexcitability alternating with periods of depression. Described cardiovascular abnormalities involve a circulatory delay related to heart failure.

It is now known that both neurological and cardiovascular factors contribute to the pathogenesis of the disease and that if the relationship between these cardiac and neurological components is altered, the stability of the respiratory control system is lost. Such disturbance in the control system may arise by prolongation of the circulation time or by the system’s becoming more dependent on the arterial partial pressure of oxygen rather than carbon dioxide.

Patients with heart failure and CSR often hyperventilate. They have lower PaCO₂ both when awake and during sleep than do control subjects with heart failure but no CSR or CSA. Circulatory delay is a well-known feature of heart failure, and animal models suggest that circulatory delay can indeed lead to periodic breathing. Whether the changes observed in animal models of CSR are also valid in humans with heart failure remains controversial, because the magnitude of circulatory delay necessary to produce CSR in animals is rarely if ever seen in humans. Nonetheless, a strong correlation has been noted between circulation time and CSR-CSA cycle length in humans.

Arousal and apnea termination are associated with the hyperventilation stage of CSR. Termination of apnea in patients with periodic breathing appears to be related largely to chemoreceptor input, which is in contrast to the proposed mechanism for apnea termination in OSAH, in which mechanoreceptor input from the lungs is believed to be of primary importance. Arousals disrupt sleep and are associated with lack of slow-wave sleep but surprisingly little daytime hypersomnolence (Cormican, Williams 2005).

Prevalence

The exact prevalence of CSA remains unclear. It does, however, appear to be particularly high in patients with neurological disease, including structural brainstem and cerebrovascular disease, as well as in patients with cardiac dysfunction.

Several studies have shown that significant left ventricular impairment is associated with sleep-disordered breathing, CSA, and CSR. Sleep-disordered breathing has been reported to occur in up to 50% of patients with stable congestive heart failure, and left ventricular systolic dysfunction may be an independent risk factor for sleep apnea in these patients.²² One small study showed that approximately 40% of patients on a heart transplantation waiting list had periodic breathing and CSA. In another study, patients with left ventricular impairment caused by ischemic heart disease were found to have cyclical oxygen desaturations with a frequency 10 times higher than those observed in healthy controls.²³ Similar results suggesting a very high prevalence of CSR and central apnea have been reported in patients with dilated cardiomyopathy.

Morbidity and Mortality

In one relatively small study designed to determine the effect of CSA with or without CSR on morbidity and mortality, CSR was found in 60% of patients with CSA. Patients with severe CSR had more central apneas, more but shorter desaturations, and more awakenings and spent more time awake during the night. Heart failure was associated with CSR. Even though patients with severe CSR were at almost twice the risk of dying than were those with no apnea, CSR was not found to be an independent risk factor for increased mortality risk. On the basis of the limited information available, it appears that even though CSA is associated with significant morbidity, the prognosis in patients with CSA is largely dependent on the underlying disease and not on the presence of sleep apnea per se.

Diagnosis

A history of cardiac dysfunction, cerebrovascular disease, or structural brainstem disease should alert the clinician to the possibility of CSA. Obese patients may have central apnea as part of the obesity-hypoventilation syndrome. Daytime hypercapnia in the absence of structural lung disease may be an indicator of worsening hypoventilation during sleep. Most patients with CSA, however, have normal or exaggerated ventilation during wakefulness. Most patients with cardiac failure belong to this group. The bed partner may note apneic episodes in the absence of respiratory effort.

Screening tests such as overnight oximetry and some more sophisticated techniques are often useful, but their sensitivity and specificity in patients with CSA have not been established. The diagnosis relies ultimately on demonstration of repetitive apnea or hypopnea in the absence of respiratory effort during polysomnography.

Treatment

In patients with CSA, treatment of the underlying cause, whenever possible, is of paramount importance. In patients with heart failure, angiotensin-converting enzyme inhibition with captopril increases the proportion of sleep spent in slow-wave and rapid-eye-movement sleep. Apneic episodes, arousals, and episodes of desaturation are also reduced. In patients with advanced heart failure, CSR has been cured by heart transplantation.