Introduction and Definitions
Sleep-disordered breathing is a very common clinical problem. Pathologic changes in breathing during sleep may take the form of discrete episodes of absent (apnea) or reduced (hypopnea) breathing or of more sustained reductions in breathing during sleep compared with wakefulness (hypoventilation). The most common form of sleep-disordered breathing results from closure of the upper airway during sleep and is called obstructive sleep apnea (OSA). Apneas may also be due to transient loss of respiratory drive output from the central respiratory controller (central apnea). “Mixed apneas” are individual events that begin as central apneas but become obstructive. Mixed apnea should not be confused with the combination of pure obstructive and pure central events in the same patient during sleep. The latter situation is referred to as “complex sleep apnea.” Complex sleep apnea is used specifically when there is either a combination of obstructive and central events on the diagnostic sleep study or when central events emerge during titration of continuous positive airway pressure (CPAP) for obstructive sleep apnea. Sleep-associated hypoventilation is defined as a sustained reduction in breathing associated with an increase in P co 2 or sustained decrease in arterial oxygen saturation (S o 2 ) of less than 90% during sleep without discrete apneas or hypopneas.
Fundamental to identifying sleep-disordered breathing is a description of breathing events that happen during sleep ( Fig. 88-1 ). In adults an apnea is defined as a cessation of breathing that lasts longer than 10 seconds. Apneas are identified as being obstructive if there are sustained respiratory efforts during the event, usually identified by use of thoracic and abdominal respiratory inductance plethysmography bands, which demonstrate paradoxical thoracoabdominal motion. In contrast, during central apneas there is absence of respiratory effort during the event. Hypopneas are events associated with a reduction rather than complete cessation of airflow. These discrete reductions in ventilation may be associated with a fall in oxygen saturation or may terminate in association with a brief arousal from sleep ( microarousal ). Although the definition of apneas is well standardized, several different criteria have been used to identify hypopneas, based on the extent of flow reduction, degree of oxygen desaturation, and the presence of microarousal. For example, the American Academy of Sleep Medicine (AASM) currently recognizes two alternate definitions of hypopnea ( Table 88-1 ). Another type of event is termed a respiratory effort–related arousal (RERA) . These are episodes characterized by mild upper airway narrowing during sleep, with increased respiratory effort required to maintain a slightly reduced level of airflow not large enough to be scored as hypopnea. When the increased inspiratory effort required to maintain ventilation is associated with a microarousal, a RERA is scored ( Fig. 88-2 ). Traditionally detection of RERAs has required use of an esophageal pressure catheter to measure respiratory effort. However, with the advent of nasal pressure cannula for measurement of airflow, subtle reductions in flow, accompanied by airflow limitation identified as flattening of the inspiratory airflow signal, can now be used to score RERAs. Changes in pulse transit time (the time from the onset of the electrocardiographic QRS complex to the pulse wave in the finger) also accurately reflect increasing effort and arousal. The scoring criteria for respiratory events recommended by the AASM are shown in Table 88-1 .
|
The standard metric for assessment of OSA severity is the apnea-hypopnea index (AHI), which is calculated as the number of apneas and hypopneas during sleep divided by total sleep time. OSA severity is graded as follows: normal (no OSA; AHI < 5 episodes/hr), mild sleep apnea (AHI ≥ 5 and < 15 episodes per hour), moderate sleep apnea (AHI ≥ 15 and < 30 episodes per hour), and severe sleep apnea (AHI ≥ 30 episodes per hour). Current AASM criteria for polysomnographic (PSG) analysis also define the respiratory disturbance index as the number of apneas, hypopneas, and RERAs per hour of sleep.
Several factors can influence the AHI value derived from PSG recordings. For example, the technology used to assess airflow has evolved from reliance on oronasal thermistor technology to addition of the nasal pressure signal, which is more sensitive for detection of mild reductions in airflow. The current prevalence estimates for OSA are derived from studies that used thermistor technology only. Addition of nasal pressure analysis would likely increase detection of respiratory events and thus OSA prevalence. Differences in hypopnea definitions can also lead to sizeable differences in measurement of OSA severity. In one analysis performed by the Sleep Heart Health Study investigators, varying the hypopnea scoring criteria (from no requirement for associated desaturation or arousal to a severe desaturation requirement) led to as much as a 10-fold difference in mean AHI value. This may therefore affect both the severity estimate and the presence versus absence of OSA based on AHI cutoff criteria. Thus it is essential in interpreting clinical reports of PSG studies or published research data that the criteria for respiratory event scoring be stated clearly and taken into account in interpreting the findings.
Not all patients with OSA will have symptoms related to the respiratory disturbance. The term sleep apnea syndrome is used to refer to the concurrence of OSA with symptoms referable to the respiratory disturbance, such as excessive daytime sleepiness (EDS). The latter can be identified by subjective questionnaires. A commonly used questionnaire is the Epworth Sleepiness Scale, in which the respondent indicates the likelihood of dozing (scale from 0 to 3) in eight common circumstances associated with sleepiness ( Table 88-2 ). A score of 11 or higher (out of 24) is commonly accepted as indicating excessive sleepiness. The International Classification of Sleep Disorders , third edition (ICSD-3) identifies a diagnosis of OSA in adults based on a AHI of 5 or higher with symptoms, or AHI of 15 or higher regardless of symptoms ( Table 88-3 ).
| Situation | Chance of Dozing |
|---|---|
| Sitting and reading | |
| Watching TV | |
| Sitting inactive in a public place (such as a theater or meeting) | |
| Riding as a passenger in a car for an hour without a break | |
| Lying down in the afternoon when circumstances permit | |
| Sitting and talking to someone | |
| Sitting quietly after a lunch without alcohol | |
| In a car, while stopped for a few minutes in traffic | |
| Score = total (normal < 11) | |
|
* Sleep recording may be either laboratory polysomnography (PSG) or out of laboratory portable testing which usually does record signals to allow sleep staging. Respiratory event numbers are expressed per hour of sleep for PSG and per hour of recording for portable monitors. Portable monitors tend to underestimate event indices as compared with PSG as patients usually are not asleep for an entire recording period. Hypopneas associated with arousals and RERAs cannot be scored from portable monitors without sleep staging capability as arousals cannot be identified.
Another term used in the spectrum of obstructive sleep-disordered breathing is the upper airway resistance syndrome . Whereas this entity is now subsumed under “Obstructive Sleep Apnea” in ICSD-3, the term was originally used by Guilleminault and colleagues to describe patients with increased upper airway resistance during sleep, recurrent arousals, and daytime symptoms, but in whom conventional criteria for thermistor-based identification of hypopneas were not met ( eFig. 88-1 ). The definition and identification of this syndrome has evolved in the ensuing years, and it is now commonly characterized as a sleep-disordered breathing syndrome in which PSG shows more than 50% of events are RERAs. A recent review of the upper airway resistance syndrome is available.
Pathogenesis of OSA
The pathogenesis of OSA involves a complex interaction of factors, including altered upper airway anatomy tissue characteristics and neuromuscular function, sleep-related decrements in upper airway dilator muscle activity, attenuated protective dilator reflexes, and altered ventilatory and arousal responses to chemical and other respiratory stimuli ( Fig. 88-3 ). Different factors may predominate in different individuals, yielding different OSA phenotypes. The upper airway in OSA closes only during sleep, indicating that sleep-dependent changes in output to the dilator muscles of the upper airway are clearly a fundamental mechanism. Airway collapse ensues when dilator muscle activity and compensatory reflexes are no longer sufficient to maintain patency of the compromised airway. The key elements of sleep-related influences on respiratory motor output are discussed in Chapter 85 and will not be covered here. Rather this section will focus on structural and functional changes in the upper airway that predispose to airway closure. Several comprehensive reviews of OSA pathogenesis are available.
Upper Airway Size
A variety of imaging methods have shown that upper airway dimensions are reduced in OSA. The reduction in airway caliber is predominantly in the lateral rather than anteroposterior dimension. Airway size may be compromised by alterations in bony structures such as a small retrognathic mandible or by increased volume of soft tissues (tongue, lateral walls). Although cross-sectional dimensions are reduced, airway length may be increased in OSA patients, with increased length predisposing to collapse. Upper airway caliber is also influenced by lung volume due to tracheal traction, so that in obese subjects, marked reductions in lung volume during recumbency can lead to reduced upper airway patency.
Upper Airway Collapsibility
Due to a lack of rigid supporting structures, much of the human upper airway is a collapsible tube. Pressure-flow relationships in the airway have been modeled using the Starling resistor model, in which collapsibility is expressed as the critical closing pressure (P crit ). P crit is measured during sleep by changing intraluminal pressure (e.g., lowering CPAP level in an OSA patient or applying negative pressure in a normal subject) and assessing flow reductions. A more negative P crit denotes a less collapsible airway. Studies have shown a continuum of upper airway collapsibility in subjects with normal breathing (P crit < −10 cm H 2 O), nonapneic snoring (P crit range, −10 to −5 cm H 2 O), obstructive hypopnea (P crit range, −5 to 0 cm H 2 O), and obstructive apnea (P crit > 0 cm H 2 O). These values are for “passive” P crit , which reflects the passive mechanical properties of the airway. Techniques are also available to measure “active” P crit , which reflects active neuromuscular compensation for reduced intraluminal pressure. Active P crit measurements demonstrate a significant impairment in neuromuscular compensation in OSA patients.
Neuromuscular Factors
Several factors may contribute to impaired neuromuscular upper airway function in OSA in addition to the attenuation of protective reflexes during sleep discussed in Chapter 85 . Upper airway muscle function may be impaired in OSA, although this remains controversial. In response to the loading of the upper airway dilators, the muscles appear to adapt, and whereas contractility appears preserved in most patients, the muscles are more fatigable. In some patients with severe OSA, muscles may be injured, leading to reduced contractility, as in the English bulldog model of OSA. There is also evidence for upper airway neuropathy in OSA, with evidence for sensory/afferent impairment and efferent neuropathy in the form of muscle denervation, both of which could impair neuromuscular compensatory responses.
Upper Airway Inflammation
There is increased inflammation in upper airway tissues in OSA, which may contribute to OSA pathogenesis. Tissue trauma from snoring, oxidative stress, acid-pepsin reflux, smoking, and alcohol could all lead to inflammation. Increased inflammation in turn may produce edema and airway narrowing, lead to changes in soft tissue composition (e.g., increased collagen deposition) and mechanics, adversely affect muscle contractility, and contribute to upper airway afferent and efferent neuropathy. Few studies to date have assessed the effects of anti-inflammatory therapy on OSA, though inhaled nasal steroids have been shown to improve mild OSA.
Fluid Shift
Studies by Bradley and colleagues have provided evidence that spontaneous shifts in fluid volume from the legs to the neck during recumbency or sleep may play a role in the pathophysiology of OSA. Cephalad displacement of fluid from the legs in normal subjects using shock pants increases neck circumference and pharyngeal resistance, as well as upper airway collapsibility. The extent of spontaneous overnight fluid shift correlates with AHI in nonobese OSA subjects. In OSA patients with venous insufficiency, the use of compression stockings to prevent daytime fluid accumulation in the legs was associated with a significant reduction in nocturnal fluid shift and in AHI. Consideration should therefore be given to this approach in managing OSA among older sedentary patients with venous insufficiency.
Clinical Factors Predisposing to OSA
Obesity
There are strong links between obesity and OSA. Approximately 58% of patients with moderate to severe OSA are obese. Obesity may reduce upper airway caliber because of adipose tissue deposition as well as through a lung volume-dependent effect. Obesity and OSA are both associated with oxidative stress and systemic inflammation, and the two conditions may interact to potentiate each other. Weight change is associated with a change in AHI. In the Wisconsin cohort over 8 years’ follow-up, a 10% increase in body weight was associated with a 32% increase in AHI and a sixfold increase in risk for developing an AHI of 15 or higher. A 10% weight loss was associated with a 26% reduction in AHI. Weight-related changes in AHI are more marked in men than women. Despite the strong links between OSA and obesity, many obese individuals do not have OSA, and up to one third of OSA patients are not obese.
Upper Airway Anatomic Abnormalities
Abnormalities that may predispose to OSA include marked craniofacial disproportion as is seen in Pierre Robin syndrome, benign tonsillar hypertrophy, oropharyngeal malignancies, macroglossia, and acromegaly. Nasal obstruction may contribute to OSA by increasing negative inspiratory driving pressure to maintain airflow, which may contribute to dynamic oropharyngeal collapse. However, surgical correction of mechanical nasal obstruction alone typically has minimal effects on OSA severity.
Gravity/Body Position
The frequency of OSA events may increase in the supine position compared with the lateral decubitus position because of the effects of gravity on upper airway size and shape. Positional OSA is commonly defined as a supine AHI at least double that in the lateral position. The duration of obstructive events and extent of associated oxygen desaturation may also worsen in the supine position. Although prevalence estimates vary, positional OSA has been identified in 49.5% of mild (AHI 5 to 15), 19.4% of moderate (AHI 15 to 30), and 6.5% of severe (AHI ≥ 30) OSA patients.
Genetic Factors
Studies in diverse populations have demonstrated familial propensity for OSA. The relative risk for OSA in a first-degree relative of an affected patient is approximately 2.0. Pedigree and twin studies indicate that the heritability of the AHI is approximately 35% to 40%. Thus inquiring about a family history of snoring and other symptoms should be part of the routine evaluation of patients for OSA. Obesity, which is a major OSA risk factor, also has a genetic basis, and it is estimated that approximately 40% of the genetic variance in AHI is shared with pathways that mediate obesity. The remainder of heritability likely resides in genes controlling craniofacial structure, ventilatory control, sleep-wake patterns, and inflammation. Various techniques have been used to identify specific genetic pathways that mediate OSA and its complications. Although progress has been made, there are few findings with direct clinical impact. Several reviews of genetic studies in OSA are available.
Endocrine Disturbances
OSA has been reported to be present in 25% to 35% of patients with untreated hypothyroidism. Predisposition to OSA may be due to increased mucopolysaccharide and protein deposition in upper airway tissues. Altered central respiratory control or neuropathic changes in upper airway muscles may also play a role. Thyroid replacement therapy improves OSA in many patients, though there may be residual sleep-disordered breathing that requires continued standard treatment. Although a clinical history suggesting hypothyroidism should be sought during evaluation of OSA patients, routine biochemical screening of all potential OSA patients is not warranted.
Sleep-disordered breathing is found in approximately 70% of patients with acromegaly. Although OSA predominates, central sleep apnea may also be observed. Upper airway dimensions are reduced due to both soft tissue (glycosaminoglycan and collagen deposition, edema) and bony changes. Upper airway myopathy may also play a role. Correction of the endocrine disturbances in acromegaly results in variable improvements in OSA; continued treatment with CPAP is often required.
OSA is found in up to 70% of women with polycystic ovary syndrome. Potential mechanisms include hormonal changes (relative androgen excess) and increased central adiposity. Polycystic ovary syndrome is associated with a high rate of metabolic dysfunction. OSA may worsen metabolic function, and OSA treatment may improve metabolic parameters in these patients. Clinicians managing patients with these endocrine disturbances should have a low threshold for requesting sleep studies if symptoms suggestive of OSA are present.
Smoking
Cigarette smoking has been linked to snoring and OSA in cross-sectional epidemiologic studies. Although the Sleep Heart Health Study investigators found that smoking was less prevalent in OSA than non-OSA patients, analysis of a subset of the Wisconsin cohort study found a significant positive dose-response relationship between cigarette smoking and OSA severity. Possible mechanisms include worsening of upper airway inflammation and sleep-disruptive effects of nicotine producing respiratory instability.
Alcohol, Drugs
Alcohol is known to worsen snoring and OSA. This may be due to direct effects on upper airway motor activity or to a deepening of sleep and impairment of arousal responses. Other drugs with similar effects include muscle relaxants, sedative-hypnotics, and opioids, although the latter may also produce central and/or complex sleep apnea. Pharmacologic effects on upper airway motor control have been reviewed.
Epidemiology of OSA
Prevalence
The first major community-based assessment of OSA prevalence was the Wisconsin Sleep Cohort Study, which reported the prevalence of OSA syndrome (AHI ≥ 5 events per hour with daytime sleepiness) as 4% in men and 2% in women. The prevalence of AHI of 15 or higher regardless of symptoms was 9% in men and 4% in women. Although other prevalence estimates have varied somewhat based on ethnicity, recruitment methods, technology used to record airflow, and definitions of hypopnea, other studies performed in the United States, Australia, Asia, and Spain have yielded similar values. However, it is important to note that, since the original estimates published in the early to mid-1990s, obesity has increased markedly in the United States and other Western countries. Given the links between OSA and obesity, the earlier figures likely now underestimate the true prevalence. The Wisconsin Cohort investigators published updated estimates of OSA prevalence with OSA syndrome (AHI ≥ 5 events per hour with Epworth sleepiness score ≥ 11) estimated at 14% in men and 5% in women 30 to 70 years of age. OSA is therefore highly prevalent in the general population.
Gender Differences
Epidemiologic studies have consistently demonstrated that OSA is two to three times more prevalent among men than women. Several factors may account for this male predominance, including differences in body fat distribution, upper airway anatomy (length, cross-sectional area) and collapsibility, and a protective effect of female sex hormones. The latter is supported by the observation that OSA is approximately three times more prevalent among postmenopausal than premenopausal women. However, the mechanisms by which sex hormones affect OSA remain unclear. Epidemiologic studies suggest that hormone replacement therapy may have a protective effect, but hormone replacement appears to have little impact on OSA once present.
Ethnicity
Most of the data on OSA prevalence are derived from predominantly white populations, though several studies have addressed OSA prevalence in African American, Hispanic, and Asian populations. In African Americans, overall OSA prevalence is similar to that in whites, although prevalence is comparatively higher among African Americans younger than 25 or older than 65 years of age. Initial reports suggested that symptoms of OSA were more prevalent among Hispanics than whites, but a more recent study found similar rates of OSA. Several studies have shown that OSA prevalence in Asians is comparable to whites, despite a lower prevalence of obesity among Asians. Although obesity remains an important risk factor for OSA among Asians, craniofacial structure may play a more prominent role in OSA pathogenesis. Studies on ethnicity and OSA may be confounded by socioeconomic and residential factors. Poor socioeconomic status has been linked to OSA risk among children. Living in poor neighborhoods is associated with shorter sleep time and increased sleep disruption; air quality may also be worse, which could contribute to worsening of OSA through increased airway inflammation. CPAP compliance appears to be lower in African American patients, which may be related to socioeconomic status and shorter sleep duration.
Aging
The prevalence of OSA increases with age through midlife. In the update from the Wisconsin Sleep Cohort, the estimated prevalence of OSA (AHI ≥ 5 with Epworth sleepiness score ≥ 11) among men was 12% for age 30 to 49 and 18% for age 50 to 70 and among women 3% for age 30 to 49 and 8% for age 50 to 70. Studies evaluating OSA prevalence in individuals older than 65 years have reported a high incidence of OSA (>50%). In one sample of community-dwelling men and women 65 years or older, 81% had an AHI of 5 or higher, and 62% had an AHI of 10 or higher. In another study the odds ratio (OR) for having an AHI of 10 or higher was 6.6 (95% confidence interval [CI], 2.6 to 16.7) for men 65 to 100 years, compared with 20 to 44 years. The comparable OR for women for an AHI of 15 or higher was 6.8 (95% CI, 0.8 to 25.9). Although it is clear that elderly patients with symptomatic OSA benefit from treatment, controversy exists with respect to the adverse health consequences of OSA in older adults. Randomized controlled trials evaluating effects of OSA treatment in older adults are required to advance knowledge in this area.
Pregnancy
A growing body of data indicates that OSA may develop or worsen over the course of pregnancy. Although there are no large-scale prospective epidemiologic PSG studies in pregnant women, based on symptoms of OSA (e.g., snoring) and witnessed apneas, together with limited PSG studies, OSA may be present in up to 20% of pregnant women by the third trimester. This is clinically relevant in that maternal OSA appears to contribute to gestational hypertension and preeclampsia, gestational diabetes, and possibly low infant birth weight.
Diagnosis of OSA
Questionnaires/Prediction Equations
Several questionnaires such as the Berlin and Stop-BANG questionnaires have been developed to grade OSA risk. Clinical prediction models that incorporate symptoms and anthropometric measurements (e.g., body mass index, neck circumference) have been developed. These models tend to be relatively sensitive (76% to 96%) but not very specific (13% to 54%) when compared with PSG. Thus, although questionnaires or prediction models are useful for screening or estimating pretest probability, objective sleep recording is required to establish a diagnosis of OSA ( Fig. 88-4A ).
Laboratory Polysomnography
The gold standard test for OSA has long been considered to be in-laboratory, technologist-attended complete overnight PSG, referred to as “type 1 sleep testing.” Standards for the performance and analysis of full PSG have been established by the AASM. Type 1 recordings include monitoring of electroencephalography, electro-oculography, chin electromyography, electrocardiography, oronasal airflow and snoring, pulse oximetry, thoracic and abdominal movement, body position, and tibialis anterior electromyography for scoring periodic leg movements (see Fig. 88-1 ). Infrared video monitoring is also used to record complex behaviors and movements to diagnose parasomnias. Sleep-wake state is scored using electroencephalography, electrooculography, and electromyography signals. Respiratory events are scored based on airflow, thoracoabdominal motion, and oximetry signals using the criteria described in Table 88-1 . PSG records should be scored manually by trained technologists according to standard criteria and summary data on sleep, respiratory, and other associated events generated in tabular and graphic summary format.
Unattended Sleep Studies
Although complete in-laboratory PSG is the gold standard for OSA diagnosis, there is limited access to this resource in many locations. There are now “out-of-center” or portable sleep testing options available that are adequate for diagnosis of OSA in many instances. In addition to type 1 sleep testing (in-laboratory PSG) there are three other types of sleep testing defined by the AASM based on number of channels recorded. Type 2 PSG testing, which involves complete PSG performed in an unattended, nonlaboratory setting, has been used in population-based studies and has clinical utility when complete PSG recording is desirable but the patient cannot come to the sleep laboratory (e.g., intensive care unit or disabled patients). Type 3 monitors acquire respiratory airflow (nasal pressure and/or oronasal thermistor), respiratory effort, oximetry, and often snoring and body position ( eFig. 88-2 ). Type 4 monitors record oximetry and sometimes one other signal such as airflow, thus yielding less information than type 3 monitors. Type 3 devices in particular are being increasingly used for the diagnosis and management of OSA. Recent guidelines recommend use of portable monitoring for evaluation of moderate- to high-pretest probability patients. Positive test results in such patients will rule in a diagnosis of OSA, whereas negative test results do not exclude more subtle forms of disease, and complete PSG should then be performed. Current guidelines do not recommend the use of portable monitoring for OSA diagnosis in the presence of major medical comorbidities or for diagnosis of sleep hypoventilation or central sleep apnea. However, for patients with moderate to severe OSA, data show that the use of portable monitors in ambulatory-based clinical management algorithms results in treatment adherence and clinical outcomes similar to conventional type 1–based approaches (see “ Disease Management Strategies for OSA ” section later).
Clinical Presentation of OSA
Symptoms, Signs
The classic clinical presentation of OSA is a history of heavy habitual snoring and EDS. However, OSA is a heterogenous condition and may present with a variety of manifestations, and some patients with even moderate to severe OSA have few if any symptoms. OSA symptoms typically develop over many years in association with aging, weight gain, or onset of menopause and may be underappreciated by the patient. A comprehensive sleep history should be obtained in evaluating patients to assess not only features of OSA but the impact of sleep habits and other potential sleep disorders on clinical symptoms.
Patients with OSA should be questioned about having symptoms at night or during the day ( Table 88-4 ). When possible, the spouse or partner should be questioned as well. During sleep most OSA patients manifest loud, disruptive snoring, which may lead the spouse to sleep in a separate room. Snoring is not a specific symptom for OSA and has a high prevalence in the general population. An important finding is a report by the bed partner of witnessed apneas. Although the frequency or duration of events reported may not be accurate, a description of episodes during which breathing stops, followed by loud gasping or snoring is highly suggestive of OSA. Patients themselves are often unaware of apneic episodes, and it is surprisingly uncommon for patients to report nocturnal choking or shortness of breath upon awakening. The differential diagnosis for nocturnal choking includes paroxysmal nocturnal dyspnea related to heart failure and Cheyne-Stokes respiration, nocturnal asthma, laryngospasm (idiopathic or due to acid-pepsin reflux), orthopnea due to diaphragm dysfunction, or insular cortical seizures. Most patients with OSA do not complain of insomnia, which when present often appears to be a separate process because OSA treatment does not generally reduce complaints of insomnia. However, OSA may rarely present with prominent insomnia symptoms that respond to OSA treatment. Other nocturnal symptoms of OSA include restless sleep, nocturia, enuresis (in severe cases), diaphoresis, and reduced libido and impotence. These symptoms also respond to OSA treatment, suggesting a causal link. Patients may complain of dry mouth upon awakening and a feeling of awakening unrefreshed. Morning headache may be present, which may be associated with nocturnal hypercapnia due to concomitant obesity hypoventilation.
| NOCTURNAL |
|
| DAYTIME |
|
The cardinal daytime symptom of OSA is EDS, defined as a propensity to fall asleep in unwanted situations during normal waking hours. EDS may develop insidiously and be underrecognized by the patient, with many patients describing symptoms of fatigue or lack of energy rather than sleepiness per se. Patients undergoing evaluation for OSA should be asked whether they have experienced drowsiness while driving or have fallen asleep at the wheel, because OSA increases the risk for crashes or near misses. Other symptoms of OSA include cognitive impairment such as difficulty with concentration or memory loss, mood disturbances such as irritability or depression, and impaired quality of life (discussed in “ Neurocognitive Complications of OSA ” section later).
Measurement of Sleepiness
Questionnaires
The most commonly used questionnaire to quantify daytime sleepiness is the Epworth sleep questionnaire described earlier (see Table 88-2 ). Values of 11 or higher indicate excessive sleepiness. The Stanford Sleepiness Scale asks subjects to rate their degree of sleepiness at a single moment in time, such as just prior to a multiple sleep latency test. Choices range from “feeling active and vital, alert wide awake” to “almost in reverie, sleep onset soon.” This scale has been used more commonly in research, and normative clinical values have not been established, limiting its applicability.
Objective Sleepiness Measurements
Objective measurements can be made in the sleep laboratory during a series of scheduled daytime nap sessions to measure either physiologic sleepiness (multiple sleep latency test) or the ability to remain awake (maintenance of wakefulness test). A more accessible behavioral test that does not require PSG recording (Osler Test) has been found to approximate findings on the maintenance of wakefulness test.
Differential Diagnosis of Excessive Sleepiness
Although EDS is a common symptom of OSA, it is important to consider other potential causes of sleepiness ( Table 88-5 ). The commonest cause of EDS is insufficient sleep duration. This may be due to poor sleep or social habits, the demands of work and family life, unfavorable sleeping environment, or other factors. It is therefore essential to obtain details of usual bedtime, time to fall asleep, number and duration of nocturnal awakenings, usual waking time, and whether wakening is spontaneous or requires one or more alarms. The regularity and timing of work and sleep schedules should also be ascertained. Short sleep duration during the week, with recovery sleep time on weekends is often a sign of insufficient sleep. When habitually present for 3 months or longer, this is termed “insufficient sleep syndrome.” Management involves altering the sleep schedule on a consistent basis to provide adequate nocturnal sleep.
|
Central Disorders of Hypersomnolence
Other medical sleep disorders may also produce EDS. These include central disorders of hypersomnolence, the most common of which is narcolepsy. Narcolepsy typically presents in the second or third decade of life and is characterized by a tetrad of symptoms, including daytime sleepiness, cataplexy, hypnagogic or hypnopompic hallucinations (hallucinations that take place in the transition between wakefulness and sleep), and sleep paralysis. Narcolepsy is a disorder of rapid eye movement (REM or stage R) with an “unpackaging” of REM phenomena so that aspects of REM sleep become apparent during wakefulness. Cataplexy is the sudden loss of muscle tone, as takes place in REM sleep, which is precipitated by emotion (laughter, anger, surprise) during wakefulness. Cataplexy is highly specific to narcolepsy and is a virtually pathognomonic symptom, although narcolepsy can be seen without cataplexy. The atonia of cataplexy can be intense, leading to collapse, and can mimic syncope, though there is no true loss of consciousness. Cataplexy is associated with objective transient areflexia. Hypnagogic or hypnopompic hallucinations are vivid dreams typically with a strong visual component that are experienced either at sleep onset or when awakening from sleep. Individuals with sleep paralysis wake up from sleep unable to move, because they are emerging from REM sleep and the REM-associated atonia has not been switched off. Although particularly common in narcolepsy, sleep paralysis may also be experienced by otherwise healthy individuals.
Excessive sleepiness in narcolepsy is commonly treated with stimulant medication. Modafinil (Provigil) is the drug of choice. When ineffective, alternate stimulants are methylphenidate (Ritalin) or amphetamine preparations. Improvements in EDS can be associated with improvement in other symptoms such as cataplexy and sleep paralysis, though specific medication may be required. Cataplexy is treated with antidepressants (tricyclic, selective serotonin-noradrenalin reuptake and monoamine oxidase inhibitor classes). Sodium oxybate (Xyrem) is approved for treatment of cataplexy and may also improve nocturnal sleep quality and EDS, although its use is limited by cost and potential side effects.
Periodic Hypersomnia
Klein-Levin syndrome is a rare cause of EDS that usually presents in adolescence and is characterized by intermittent episodes of intense hypersomnia with normal intervening sleep and alertness. The episodes may last for days, and patients may sleep for up to 20 hr/day. Episodes are associated with behavioral abnormalities, including binge eating and hypersexuality. Treatment is usually supportive during the episodes, with a limited role for stimulants; lithium may be effective in some cases.
Idiopathic Hypersomnia
This disorder is a diagnosis of exclusion and is characterized by persistent marked excessive sleepiness despite documented adequate sleep duration and hygiene, frequent daytime napping, subjectively uninterrupted nocturnal sleep, and normal nocturnal PSG recording, without symptoms of another sleep disorder or medical cause for hypersomnolence. Although most patients have prolonged nocturnal sleep, some may not. The ICSD-3 does not distinguish between these two variants. Management is with stimulant medication similar to narcolepsy, although symptoms may be more difficult to control.
Sleep-Related Movement Disorders
Movement disorders during sleep may cause sleep disruption and EDS. The most common is the restless legs syndrome (RLS), which is characterized by an urge to move the legs usually accompanied by uncomfortable or unpleasant sensations in the legs that (1) gets worse with periods of physical inactivity or rest, (2) is partially or totally relieved by movement such as walking or stretching, and (3) is worse or is felt only in the evening or night. The timing of symptom onset is such that the patient has difficulty falling asleep until the uncomfortable sensations and urge to move subside, which may be only after 2 to 3 hours after bedtime. There are several factors that can precipitate or exacerbate RLS, the most common of which is iron deficiency. Serum iron and ferritin levels should be checked in all RLS patients. Caffeine, stimulants, and other medications may exacerbate RLS. RLS is common in renal dialysis and congestive heart failure patients and in patients with neuropathic disorders and may emerge or worsen during pregnancy. Familial RLS has been described. Genetic studies have identified several loci linked to RLS, though to date these have not led to major insights into etiology, prevention, or treatment.
Treatment of RLS first involves correction of potential precipitating factors, in particular iron replacement if stores are low. A variety of medications may be used to treat RLS, including dopamine agonists such as pramipexole (Mirapex) and ropinirole (Requip), as well as anticonvulsants such as gabapentin and pregabalin.
Circadian Rhythm Sleep-Wake Disorders
This set of clinical disorders is characterized by disturbances of the internal “biologic clock” so that sleep is usually normal, but takes place at abnormal times. The most common disturbance is delayed sleep-wake phase disorder. Affected individuals are unable to fall asleep until the early morning hours (e.g., 3 am or later), sleep normally once asleep, and awaken late in the morning (e.g., 11 am ). If able to follow their “natural” sleep schedule, there are few daytime symptoms. However, when work or school imposes a different schedule, sleep time is restricted and daytime sleepiness ensues. Management of delayed sleep-wake phase disorder includes interventions to shift the biologic clock, such as administration of melatonin in the evening and exposure to bright light in the morning. Although changes in sleep schedule can be achieved, the effects may not be dramatic, and patients tend to revert easily to a delayed schedule if circadian measures are not rigorously maintained. Some patients choose occupations or work schedules that are compatible with their delayed phase.
Other circadian disorders include advanced sleep-wake phase disorder, in which sleep begins in the early evening, with early morning awakening. Treatment includes sleep schedule modifications and bright light exposure in the evening. Jet lag disorder results from circadian misalignment because of travel across time zones with subsequent sleep disruption and daytime symptoms. Management includes sleep scheduling, appropriately timed light therapy, and possible use of melatonin. Shift work disorder is characterized by more than 1 month of excessive sleepiness during scheduled work time, and insomnia during scheduled sleep times in the context of nonconventional and/or rotating work schedules. Management includes optimizing the sleep environment and sleep schedule, preshift napping, bright light exposure during night shifts, and avoidance of bright light during the return trip home in the morning. Evidence supports the use of melatonin and judicious use of hypnotics to promote sleep, and modafinil and caffeine to promote wakefulness during work.
Other Conditions and Medications
Many medical conditions may be associated with sleep disruption and/or result in excessive sleepiness. These include nocturnal respiratory disease with nocturnal cough or dyspnea, gastroesophageal reflux, nocturnal urinary frequency, chronic renal failure, various infectious diseases, and chronic pain syndromes. Psychiatric disorders such as depression may present with excessive sleepiness. Numerous medications can affect nocturnal sleep quality and/or contribute to excessive sleepiness. Thus medication history and potential side effects should be carefully assessed.
Neurocognitive Complications of OSA
Pathophysiology
A majority of apnea and hypopnea events terminate in association with arousal (see Fig. 88-1 ). Thus recurrent respiratory events over the course of the night lead to marked disruption of sleep continuity or sleep fragmentation. OSA is also typically associated with a reduction in the duration of deeper, more “restorative” sleep, including stage N3 (slow-wave) and stage R (rapid eye movement) sleep. These characteristic changes in sleep architecture are one factor leading to EDS and other neurocognitive sequelae of OSA. Application of nasal CPAP to treat OSA restores sleep continuity and may be associated acutely with a rebound increase in N3 and stage R sleep, which patients perceive as improved sleep quality.
Although respiratory-related sleep fragmentation alone (i.e., obstructive events without associated oxygen desaturation) can produce EDS, in studies of moderate to severe OSA, sleepiness and other neurocognitive sequelae correlate more closely with OSA-associated hypoxemia than with measures of sleep disruption. Studies in mice of cycling intermittent hypoxia mimicking OSA showed that severe hypoxia can produce excessive sleepiness, which persists up to 6 months following return to normoxia. Veasey and colleagues identified neuronal injury in specific wake-promoting areas (monoaminergic neurons in locus coeruleus, periaqueductal gray) in these mice and have shown that injury is mediated by nicotinamide adenine dinucleotide phosphate , reduced form (NADPH) oxidase–dependent oxidative injury. Nair and colleagues demonstrated that intermittent hypoxia in mice also induces deficits in learning and memory, with the hippocampus being a particular site of neuronal injury. Neuroimaging studies in OSA patients have demonstrated changes in the hippocampal region and in other areas that subserve cognitive functions known to be impaired in OSA. It is therefore likely the hypoxia-reoxygenation associated with human OSA may produce sleepiness and other neurocognitive changes through similar pathways to those identified in experimental animals. However, debate continues concerning the relative roles of sleep fragmentation and hypoxic-mediated injury in producing OSA-related neurocognitive deficits.
Excessive sleepiness in OSA patients may also be influenced by factors other than those directly related to OSA itself. In addition, several studies have demonstrated that sleep schedule, obesity per se, and depression may all contribute to symptoms of sleepiness in OSA patients.
Excessive Sleepiness, Reduced Performance, and Traffic /Workplace Complications
EDS is the most common symptom of OSA and can have profound adverse effects on quality of life, social relationships, and professional safety and performance. Patients with untreated OSA and daytime sleepiness are at significantly increased risk for motor vehicle crashes. Two meta-analyses have identified a twofold to threefold increase in risk for vehicular crashes among patients with untreated OSA (e.g., OR, 2.4; 95% CI, 1.2 to 4.9 ). A recent meta-analysis found that treatment with positive airway pressure for severe OSA decreased the risk ratio for motor vehicle crashes to 0.28 (95% CI, 0.22 to 0.35), translating into a reduction of between 65% and 78% in crash rate with OSA treatment.
There are important public safety and medicolegal implications of OSA for the practicing physician. The American Thoracic Society has recently published an updated practice guideline on OSA, sleepiness, and crash risk. Although the prediction of crash risk in individual OSA patients is imprecise, it is important to identify high-risk patients with symptoms of excessive sleepiness and a previous crash or near miss due to sleepiness. Such individuals should be warned of the dangers of driving before treatment, undergo diagnostic sleep testing within 1 month, and start CPAP treatment following a positive diagnostic test. Clinicians should also familiarize themselves thoroughly with legal requirements in their jurisdiction for reporting of OSA and driving restrictions.
One area of particular concern is the increased rate of motor vehicle crashes among commercial drivers. Several studies have shown a high prevalence of OSA among commercial drivers. Although no study to date has demonstrated that OSA leads to increased accident rates specifically among commercial drivers, extension from the general OSA driving literature, combined with the long distances driven commercially raises substantial concern. Screening for OSA is recommended as part of the general medical evaluation for commercial drivers. However, such screening relies heavily on self-reporting of sleepiness and OSA symptoms by drivers, who are known to underreport due to concerns about work restrictions. One study suggests that drivers may be more likely to report symptoms if they can do so anonymously online. A multisociety statement includes recommendations concerning OSA screening of commercial drivers and assessment of fitness for duty and return to work following OSA treatment.
Cognitive Impairment
OSA is associated with cognitive impairment. OSA produces impairments in attention and vigilance, visuospatial construction abilities, verbal episodic and visuospatial memory, and subdomains of executive function. Treatment of OSA can improve global cognitive function, attention/vigilance, verbal and visual memory, and executive function. Most studies have evaluated treatment effects after only a few months, however, so that long-term effects are unknown. The Apnea Positive Pressure Long-term Efficacy Study (APPLES) assessed domains of cognitive function in OSA at 2 and 6 months of effective versus sham CPAP treatment. Improvement was observed in a measure of executive and frontal lobe function at 2 months but was not sustained at 6 months. However, baseline function was high in study participants; further studies on long-term treatment in OSA patients with more marked baseline cognitive impairment are required.
Depression
Depressive symptoms are common in OSA patients and are more prevalent in women. Although epidemiologic studies indicate that the presence of untreated OSA is a risk factor for developing depression, the interaction between depression and OSA remains poorly understood. Depression may worsen symptoms of sleepiness and fatigue in OSA. Although responses are not uniform, some studies report improvement in depression scores with OSA treatment. The prevalence of OSA among patients with clinical depression is not known, but symptoms of OSA should be sought in patients with depression, given the potential beneficial effects of OSA treatment.
Other Complications
OSA can lead to erectile dysfunction, which has been linked to nitric oxide and NADPH-related pathways in animal models. Treatment of OSA may improve erectile function.
The effects of OSA can have a substantial negative impact on patients’ overall quality of life. Disease-specific indices have been developed to assess the impact of OSA on quality of life, including the Functional Outcomes of Sleep Questionnaire, Calgary Sleep Apnea Quality of Life Index, and the Quebec Sleep Questionnaire. Studies using these measures indicate that OSA-related impairment can be substantial, and that OSA treatment can significantly improve quality of life.
Cardiometabolic Complications of OSA
Pathophysiology
Obstructive apneas and hypopneas trigger a cascade of acute hemodynamic, autonomic, biochemical, inflammatory, and metabolic effects that can produce both acute and long-term changes in cardiovascular function ( Fig. 88-5 ). Autonomic tone is altered during apneas, with acute sympathetic-mediated increases in blood pressure and heart rate observed at airway reopening ( Fig. 88-6 ). Blood vessel shear stress is also increased due to hypertension and hemodynamic surges. The negative intrathoracic pressures generated during obstructed inspiratory efforts increase left ventricular transmural pressure and therefore afterload, a stimulus for ventricular hypertrophy and an excessive load for a compromised ventricle. The increased sympathetic neural activity during apneas may persist into wakefulness, contributing to sustained hypertension. OSA-associated hypoxia-reoxygenation has analogous biologic effects to ischemia-reperfusion, leading to generation of reactive oxygen species. This oxidative stress activates nuclear transcription factors, including hypoxia inducible factor-1α and nuclear factor-κB, which activate a diversity of proinflammatory pathways. Inflammatory mediators in turn may adversely affect endothelial cell function and promote atherogenesis, have antifibrinolytic/prothrombotic effects that could contribute to acute vascular events, and, in concert with increased sympathetic activity, may increase insulin resistance, further contributing to cardiovascular risk. Animal studies indicate that intermittent hypoxia can also adversely affect lipid metabolism, further contributing to the proatherogenic effects of OSA. There is growing evidence that, on the basis of these pathophysiologic mechanisms, OSA has important clinical effects on cardiovascular morbidity and mortality.
