Etiology

In general terms, OSA results from an anatomically or functionally narrowed upper airway; this typically involves some combination of decreased upper airway patency (upper airway obstruction and/or decreased upper airway diameter), increased upper airway collapsibility (reduced pharyngeal muscle tone), and decreased drive to breathe in the face of reduced upper airway patency (reduced central ventilatory drive) (Table 17-4). Upper airway obstruction varies in degree and level (i.e., nose, nasopharynx/oropharynx, hypopharynx) and is most commonly due to adenotonsillar hypertrophy, although tonsillar size does not necessarily correlate with degree of obstruction, especially in older children. Other causes of airway obstruction include allergies associated with chronic rhinitis/nasal obstruction; craniofacial abnormalities, including hypoplasia/displacement of the maxilla and mandible; gastroesophageal reflux with resulting pharyngeal reactive edema; nasal septal deviation; and velopharyngeal flap cleft palate repair. Reduced upper airway tone may result from neuromuscular diseases, including hypotonic cerebral palsy and muscular dystrophies, or hypothyroidism. Reduced central ventilatory drive may be present in some children with Arnold-Chiari malformation and meningomyelocele. In other situations, the etiology is mixed; individuals with Down syndrome, by virtue of their facial anatomy, hypotonia, macroglossia, and central adiposity, as well as the increased incidence of hypothyroidism, are at particularly high risk for OSA, with some estimates of as great as 70% prevalence.

Although many children with OSA are of normal weight, an increasingly large percentage are overweight or obese, and many of these children are school-aged and younger. There is a significant correlation between weight and SDB (habitual snoring, OSA, central apneas). While adenotonsillar hypertrophy also plays an important etiologic role in overweight/obese children with OSA, mechanical factors related to an increase in the amount of adipose tissue in the throat (pharyngeal fat pads), neck (increased neck circumference), and chest wall and abdomen can create increased upper airway resistance, worsen gas exchange, and increased work of breathing, particularly in the supine position and during REM sleep. There may be a component of blunted central ventilatory drive in response to hypoxia/hypercapnia and hypoventilation as well, particularly in children with morbid or syndrome-based (Prader-Willi) obesity. Overweight and obese children and adolescents are at a particularly high risk for metabolic and cardiovascular complications of SDB, such as insulin resistance and systemic hypertension; morbidly obese children may also be at increased risk for postoperative complications following adenotonsillectomy.

Epidemiology

Overall prevalence of parent-reported snoring in the pediatric population is about 8%; “always” snoring is reported in 1.5-6%, and “often” snoring in 3-15%. When defined by parent-reported symptoms, the prevalence of OSA is 4-11%. The prevalence of pediatric OSA as documented by overnight sleep studies utilizing ventilatory monitoring procedures (e.g., in-lab PSG, home studies) is 1-4% overall, with a reported range of 0.1-13%. Prevalence is also affected by the demographic characteristics, such as age (increased prevalence between 2 and 8 yr), gender (more common in boys, especially after puberty), race/ethnicity (increased prevalence in African-American and Asian children), and family history of OSA.

Pathogenesis

The upregulation of inflammatory pathways, as indicated by an increase in peripheral markers of inflammation such as C-reactive protein (CRP), appear to be linked to metabolic dysfunction (e.g., insulin resistance, dyslipidemia) in both obese and non-obese children with OSA. Both systemic inflammation and arousal-mediated increases in sympathetic autonomic nervous system activity with altered vasomotor tone may be key contributors to increased cardiovascular risk in both adults and children with OSA. Mechanical stress on the upper airway induced by chronic snoring may also result in both local mucosal inflammation of adenotonsillar tissues and subsequent upregulation of inflammatory molecules, most notably leukotrienes. Another potential mechanism that may mediate cardiovascular sequelae in both adults and children with OSA is altered endothelial function.

Although yet to be fully elucidated, one of the primary mechanisms by which OSA is believed to exert negative influences on cognitive function appears to involve repeated episodic arousals from sleep leading to sleep fragmentation and resulting sleepiness. An equally important role may be intermittent hypoxia that leads directly to systemic inflammatory vascular changes in the brain. Levels of inflammatory markers such as CRP and cytokine IL-6 are elevated in children with OSA and are also associated with cognitive dysfunction.

Clinical Manifestations

The clinical manifestations of OSA may be divided into sleep-related and daytime symptoms. The most common nocturnal manifestations of OSA in children and adolescents are loud, frequent, and disruptive snoring, breathing pauses, choking or gasping arousals, restless sleep, and nocturnal diaphoresis. Many children who snore do not have OSA, but very few children with OSA do not snore. Most children, like adults, tend to have more frequent and more severe obstructive events in REM sleep and when sleeping in the supine position. Children with OSA may adopt unusual sleeping positions, keeping their necks hyperextended in order to maintain airway patency. Frequent arousals associated with obstruction may result in nocturnal awakenings, but are more likely to cause fragmented sleep.

Daytime symptoms of OSA include mouth breathing and dry mouth, chronic nasal congestion/rhinorrhea, hyponasal speech, morning headaches, difficulty swallowing, and poor appetite. Children with OSA may have secondary enuresis, most likely as a result of the disruption of the normal nocturnal pattern of antidiuretic hormone secretion. Partial arousal parasomnias (sleepwalking and sleep terrors) may occur more frequently in children with OSA, related to the frequent associated arousals and an increased percentage of delta sleep, or SWS.

One of the most important but frequently overlooked sequelae of OSA in children is the effect on mood, behavior, learning, and academic functioning. The neurobehavioral consequences of OSA in children include daytime sleepiness with drowsiness, difficulty in morning waking, and unplanned napping or dozing off during activities, although evidence of frank hypersomnolence tends to be less common in children compared to adults with OSA. Mood changes include increased irritability, mood instability and emotional dysregulation, low frustration tolerance, and depression/anxiety. Behavioral issues include both “internalizing” (i.e., increased somatic complaints and social withdrawal) and “externalizing” behaviors, including aggression, impulsivity, hyperactivity, oppositional behavior, and conduct problems. There is a substantial overlap between the clinical impairments associated with OSA and the diagnostic criteria for ADHD, including inattention, poor concentration, and distractibility (Chapter 30). There also appears to be a selective impact of OSA specifically on “executive functions,” which include cognitive flexibility, task initiation, self-monitoring, planning, organization, and self-regulation of affect and arousal; executive function deficits are also a hallmark of ADHD.

Studies that have looked at changes in behavior and neuropsychologic functioning in children following treatment (usually adenotonsillectomy) for OSA have largely documented significant improvement in outcomes, in both the short and long term, of OSA syndrome post-treatment, including daytime sleepiness, mood, behavior, academics, and quality of life. Many studies have failed to find a dose-dependent relationship between OSA in children and specific neurobehavioral/neurocognitive deficits, suggesting that other factors may influence neurocognitive outcomes, including individual genetic susceptibility, environmental influences such as passive smoking exposure, and co-morbid conditions, such as obesity, shortened sleep duration, and the presence of other sleep disorders.

Diagnosis

The American Academy of Pediatrics clinical practice guidelines provide excellent information for the evaluation and management of uncomplicated childhood OSA (Table 17-5). There are no physical examination findings that are truly pathognomonic for OSA, and most healthy children with OSA appear normal; certain physical examination findings may suggest OSA. Growth parameters may be abnormal (obesity or, less commonly, failure to thrive), and there may be evidence of chronic nasal obstruction (hyponasal speech, mouth breathing, septal deviation, “adenoidal facies”), as well as signs of atopic disease (i.e., “allergic shiners”). Oropharyngeal examination may reveal enlarged tonsils, excess soft tissue in the posterior pharynx, and a narrowed posterior pharyngeal space. Any abnormalities of the facial structure, such as retrognathia and/or micrognathia, midfacial hypoplasia, best appreciated by inspection of the lateral facial profile, increase the likelihood of OSA and should be noted. In very severe cases, there may be evidence of pulmonary hypertension, right-sided heart failure, and cor pulmonale; systemic hypertension, unlike in adults, is relatively uncommon.

Because no combination of clinical history and physical findings can accurately predict which children with snoring have OSA, the gold standard for diagnosing OSA remains an overnight polysomnogram (PSG).

An overnight PSG is a technician-supervised, monitored study that documents physiologic variables during sleep; sleep staging, arousal measurement, cardiovascular parameters, and body movements (electroencepholography, electrooculography, chin and leg electromyography, ECG, body position sensors, and video recording), and a combination of breathing monitors (oronasal thermal sensor and nasal air pressure transducer for airflow), chest/abdominal monitors (e.g., inductance plethysmography for respiratory effort, pulse oximeter for O₂ saturation, end-tidal or transcutaneous CO₂ for hypercarbia, snore microphone). The polysomnographic parameter most commonly used in evaluating for sleep disordered breathing is the apnea/hypopnea index (AHI), which indicates the number of apneic and hypopneic events per hr of sleep. It should be noted that currently there are no universally accepted polysomnographic normal reference values and parameters for diagnosing OSA in children, and it is still unclear which parameters best predict morbidity. Normal preschool and early school-aged children may have a total AHI of less than 1.5, and this is the most widely used cutoff value for OSA in children 12 yr and below; in adolescents, the adult cutoff of an AHI ≥5 is generally used. In cases in which the AHI is between 1 and 5 obstructive events per hour, clinical judgment regarding risk factors for SDB, evidence of daytime sequelae, and the technical quality of the overnight sleep study should determine further management.

Treatment

There are presently no universally accepted guidelines regarding the indications for treatment of pediatric SDB (i.e., including primary snoring and OSA). Current recommendations largely emphasize weighing what is known about the potential cardiovascular, metabolic, and neurocognitive sequelae of SDB in children in combination with the individual health care professional’s clinical judgment. The decision of whether and how to treat OSA specifically in children is contingent on a number of parameters, including severity (nocturnal symptoms, daytime sequelae, sleep study results), duration of disease, and individual patient variables such as age, co-morbid conditions, and underlying etiologic factors. In the case of moderate to severe disease (AHI >10), the decision to treat is usually straightforward, and most pediatric sleep experts recommend that any child with an apnea index >5 should be treated.

In the majority of cases of pediatric OSA, adenotonsillectomy is the first-line treatment in any child with significant adenotonsillar hypertrophy, even in the presence of additional risk factors such as obesity. Adenotonsillectomy in uncomplicated cases generally (70-90% of children) results in complete resolution of symptoms; regrowth of adenoidal tissue after surgical removal occurs in some cases. Groups considered high-risk include young children (<3 yr old), as well as those with severe OSA documented on polysomnography, significant clinical sequelae of OSA (e.g., failure to thrive), or associated medical conditions, such as craniofacial syndromes, morbid obesity, and hypotonia. All patients should be re-evaluated postoperatively to determine whether additional evaluation and/or treatment are required. If there are significant residual risk factors (e.g., obesity) or continued symptoms of OSA, a follow-up sleep study at least 6 wk post-adenotonsillectomy may be indicated.

Additional treatment measures that may be appropriate include weight loss, positional therapy (attaching a firm object, such as a tennis ball, to the back of a sleep garment to prevent the child from sleeping in the supine position), and aggressive treatment of additional risk factors when present, such as asthma, seasonal allergies, and gastroesophageal reflux; there is some evidence that intranasal corticosteroids and leukotriene inhibitors may be helpful in mild OSA. Other surgical procedures, such as uvulopharyngopalatoplasty, and maxillofacial surgery (mandibular distraction osteogenesis and maxillomandibular advancement), are seldom performed in children but may be indicated in selected cases. Oral appliances, such as mandibular advancing devices and tongue retainers, are typically considered for adolescents in whom facial bone growth is largely complete.

Continuous or bilevel positive airway pressure (nasal CPAP or BiPAP) is the most common treatment for OSA in adults and can be used successfully in children and adolescents. CPAP delivers humidified, warmed air through an interface (mask, nasal pillows) that, under pressure, effectively “splints” the upper airway open. Optimal pressure settings (that abolish or significantly reduce respiratory events without increasing arousals or central apneas) are determined in the sleep lab during a full night CPAP titration. Efficacy studies at the current pressure and retitrations should be conducted periodically with long-term use (every 6 mo in young children and at least yearly or with significant weight changes in older children and adolescents). CPAP may be recommended if removing the adenoids and tonsils is not indicated, if there is residual disease following adenotonsillectomy, or if there are major risk factors that are not amenable to treatment with surgery (obesity, hypotonia).

Etiology

Partial arousal parasomnias, which include sleepwalking, sleep terrors, and confusional arousals are more common in preschool and school-aged children because of the relatively higher percentage of SWS in younger children. Any factor that is associated with an increase in the relative percentage of SWS (certain medications, previous sleep deprivation) may increase the frequency of events in a predisposed child. There appears to be a genetic predisposition for both sleepwalking and night terrors. In contrast, nightmares, which are much more common than the partial arousal parasomnias but are often confused with them, are concentrated in the last third of the night, when REM sleep is most prominent. Partial arousal parasomnias may also be difficult to distinguish from nocturnal seizures. Table 17-6 summarizes similarities and differences among these nocturnal arousal events.

Epidemiology

Many children (15-40%) sleepwalk on at least one occasion; the prevalence of children who regularly sleepwalk is approximately 17%, and 3-4% have frequent episodes. Sleepwalking may persist into adulthood, with the prevalence in adults of about 4%. The prevalence is approximately 10 times greater in children with a family history of sleepwalking. Approximately 1-6% of children experience sleep terrors, primarily during the preschool and elementary school years, and the age of onset is usually between 4 and 12 yr. Because of the common genetic predisposition, the prevalence of sleep terrors in children who sleepwalk is about 10%. Although sleep terrors can occur at any age from infancy through adulthood, most individuals outgrow sleep terrors by adolescence. Confusional arousals commonly co-occur with sleepwalking and sleep terrors; prevalence rates have been estimated to be upwards of 15% in children ages 3-13 yr.

Clinical Manifestations

The partial arousal parasomnias have several features in common. Because they typically occur at the transition out of “deep” or SWS, partial arousal parasomnias have clinical features of both the awake (ambulation, vocalizations) and the sleeping (high arousal threshold, unresponsiveness to the environment) states; there is usually amnesia for the events. The typical timing of partial arousal parasomnias during the first few hours of sleep is related to the predominance of SWS in the first third of the night; the duration is typically a few minutes (sleep terrors) to an hour (confusional arousals). Sleep terrors are sudden in onset and characteristically involve a high degree of autonomic arousal (i.e., tachycardia, dilated pupils), while confusional arousals typically arise more gradually from sleep, may involve thrashing around but usually not displacement from bed, and are often accompanied by slow mentation on arousal from sleep (“sleep inertia”). Sleepwalking may be associated with safety concerns (e.g., falling out of windows, wandering outside). Avoidance of, or increased agitation with, comforting by parents or attempts at awakening are also common features of all partial arousal parasomnias.

Treatment

Management of partial arousal parasomnias involves some combination of parental education and reassurance, good sleep hygiene, and avoidance of exacerbating factors such as sleep deprivation and caffeine. Particularly in the case of sleepwalking, it is important to institute safety precautions such as use of gates in doorways and at the top of staircases, locking of outside doors and windows, and installation of parent notification systems such as bedroom door alarms. Scheduled awakenings, a behavioral intervention that involve having the parent wake the child approximately 15 to 30 min before the time of night that the first parasomnia episode is most likely to be successful in situations in which partial arousal episodes occur on a nightly basis. Pharmacotherapy is rarely necessary, but may be indicated in cases of frequent or severe episodes, high risk of injury, violent behavior, or serious disruption to the family; the primary pharmacologic agents used are potent SWS suppressants, primarily benzodiazepines and tricyclic antidepressants.

Etiology

“Early-onset” RLS (i.e., onset of symptoms before 35-40 yr of age), often termed “primary” RLS, appears to have a particularly strong genetic component. Variants of the PTPRD gene are associated with RLS. Low serum iron levels in both adults and children may be an important etiologic factor for the presence and severity of both RLS symptoms and PLMs. As a marker of decreased iron stores, serum ferritin levels in both children and adults with RLS are frequently low. The underlying mechanism that has been postulated is related to the role of iron as a cofactor of tyrosine hydroxylase in a rate-limiting step of the synthesis of dopamine; in turn, dopaminergic dysfunction has been implicated as playing a key role particularly in the genesis of the sensory component of RLS, as well as in PLMD. Certain medical conditions, including diabetes mellitus, end-stage renal disease, cancer, rheumatoid arthritis, hypothyroidism, and pregnancy, may also be associated with RLS/PLMD, as are specific medications (i.e., antihistamines such as diphenhydramine, antidepressants, and H-2 blockers such as cimetidine) and substances (e.g., caffeine).

Epidemiology

Previous studies have found prevalence rates of RLS in the pediatric population ranging from 1-6%; the percent of 8-17 yr olds meeting criteria for “definite” RLS is approximately 2%. Prevalence rates of PLMs greater than 5 per hour in clinical populations of children referred for sleep studies range from 5-27%; in survey studies of PLM symptoms, rates are 8-12%. Several studies in referral populations have found that PLMs occur in as much as one fourth of children diagnosed with ADHD.

Clinical Manifestations

In addition to the sensory component and the urge to move the legs, most RLS episodes begin or are exacerbated by rest or inactivity, such as lying in bed to fall asleep or riding in a car for prolonged periods. A unique feature of RLS is that the timing of symptoms also appears to have a circadian component, in that they often peak in the evening hours. Some children may complain of “growing pains,” although this is considered a nonspecific feature. Because RLS symptoms are usually worse in the evening, bedtime struggles and difficulty falling asleep are 2 of the most common presenting complaints. In contrast to patients with RLS, individuals with PLMs are usually unaware of these movements; these movements may result in arousals during sleep and consequent significant sleep disruption. Parents of children with RLS/PLMD may complain that their child is a restless sleeper, moves around or even falls out of bed during the night.

Treatment

The decision of whether and how to treat RLS depends on the level of severity (intensity, frequency, and periodicity) of sensory symptoms, the degree of interference with sleep, and the impact of daytime sequelae in a particular child or adolescent. With PLMs, for an index (PLMs per hr) less than 5, usually no treatment is recommended; for an index over 5, the decision to specifically treat PLMs should be based on the presence or absence of nocturnal symptoms (restless or nonrestorative sleep) and daytime clinical sequelae. A reasonable initial approach would be to promote good sleep hygiene (including restricting caffeine) and instituting iron supplements in children if serum ferritin levels are low (<50); the recommended dose of ferrous sulfate is typically in the range of 4-6 mg/kg/day for a duration of 3-6 mo. Medications that increase dopamine levels in the CNS, such as ropinirole and pramipexole, have been found to be effective in relieving RLS/PLMD symptoms in adults; data in children are extremely limited.

Sleep-related rhythmic movements, including head banging, body rocking, and head rolling, are characterized by repetitive, stereotyped, and rhythmic movements or behaviors that involve large muscle groups. These behaviors typically occur with the transition at sleep at bedtime, but also at nap times and following nighttime arousals. Children typically engage in these behaviors as a means of soothing themselves to (or back to) sleep; they are much more common in the 1st yr of life and usually disappear by 4 yr of age. In most instances, rhythmic movement behaviors are benign, because sleep is not significantly disrupted as a result of these movements and associated significant injury is rare. These behaviors typically occur in normally developing children, and in the vast majority of cases their presence does not indicate that there is some underlying neurological or psychological problem. Usually, the most important aspect in management of sleep-related rhythmic movements is reassurance to the family that this behavior is normal, common, benign, and self-limited.

Etiology

There is a specific deficit in the hypothalamic orexin/hypocretin neurotransmitter system in the genesis of narcolepsy with cataplexy. The underlying pathogenesis of narcolepsy involves selective loss of cells that secrete hypocretin/orexin in the lateral hypothalamus; it has been postulated that autoimmune mechanisms, possibly triggered by viral infections, in combination with a genetic predisposition and environmental factors, may be involved. Human leukocyte antigen testing also shows a strong association with narcolepsy; however, the vast majority of individuals with this antigen do not have narcolepsy. Although the majority of cases of narcolepsy are considered idiopathic, “secondary” narcolepsy with cataplexy may also result from CNS insults.

Epidemiology

The prevalence of narcolepsy is reported to be between 3 and 16 per 10,000, with the prevalence of narcolepsy with cataplexy approximately 0.2-0.5/10,000. The risk of developing narcolepsy with cataplexy in a first-degree relative of a narcoleptic patient is estimated at 1-2%; this represents an increase of 10- to 40-fold compared to the general population.

Clinical Manifestations and Diagnosis

The typical onset of symptoms of narcolepsy is in adolescence and early adulthood, although symptoms may initially present in school-aged and even younger children. The early manifestations of narcolepsy are often ignored, misinterpreted, or misdiagnosed as other medical, neurologic, and psychiatric conditions, and the appropriate diagnosis is frequently delayed for a number of years.

The most prominent clinical manifestation of narcolepsy is profound daytime sleepiness, characterized by both an increased baseline level of daytime drowsiness and by the repeated occurrence of sudden and unpredictable sleep episodes. These “sleep attacks” are often described as “irresistible” in that the child or adolescent is unable to stay awake despite considerable effort, and they occur even in the context of normally stimulating activities (e.g., during meals, in the middle of a conversation). Very brief (several seconds) sleep attacks may also occur in which the individual may “stare off,” appear unresponsive, or continue to engage in an ongoing activity (automatic behavior). Cateplexy is considered pathognomonic for narcolepsy. Cataplexy is rarely the first symptom of narcolepsy, but it often develops within the 1st year of the onset of EDS. It is described as an abrupt, bilateral, partial or complete loss of muscle tone, classically triggered by an intense positive emotion (e.g., laughter, surprise). The cataplectic attacks are typically brief (seconds to minutes), and fully reversible, with complete recovery of normal tone when the episode ends. Hypnogogic/hypnopompic hallucinations involve vivid visual, auditory, and sometimes tactile sensory experiences occurring during transitions between sleep and wakefulness, primarily at sleep onset (hypnogogic) and sleep offset (hypnopompic). Sleep paralysis is the inability to move or speak for a few seconds or minutes at sleep onset or offset, and often accompanies the hallucinations. Other symptoms associated with narcolepsy include disrupted nocturnal sleep, inattention, and behavioral and mood issues.

Overnight polysomnography followed by a multiple sleep latency test (MSLT) are strongly recommended components of the evaluation of a patient with profound unexplained daytime sleepiness or suspected narcolepsy. The purpose of the overnight PSG is to evaluate for primary sleep disorders, such as OSA that may cause EDS. The MSLT involves a series of 5 opportunities to nap (20 min long), during which narcoleptics demonstrate a pathologically shortened sleep onset latency as well as periods of REM sleep occurring immediately after sleep onset.

Treatment

An individualized narcolepsy treatment plan usually involves education, good sleep hygiene, behavioral changes, and medication. Scheduled naps may be helpful. Medications such as psychostimulants and modafinil are often prescribed to control the EDS. The goal should be to allow the fullest possible return of normal functioning in school, at home, and in social situations. Medications such as tricyclic antidepressants and serotonin reuptake inhibitors may also be used to control the REM-associated phenomena, such as cataplexy, hypnogogic hallucinations, and sleep paralysis.

Etiology

Individuals with DSPD often start out as night owls; that is, they have an underlying predisposition or circadian preference for staying up late at night and sleeping late in the morning, especially on weekends, holidays, and summer vacations. The underlying pathophysiology of DSPD is still unknown, although some authors have theorized that it involves an intrinsic abnormality in the circadian oscillators that govern the timing of the sleep period.

Epidemiology

Studies indicate that DSPD affects approximately 7-16% of adolescents.

Clinical Manifestations

The most common clinical presentation is sleep initiation insomnia when the individual attempts to fall asleep at a “socially acceptable” desired bedtime, accompanied by extreme difficulty getting up in the morning even for desired activities, and daytime sleepiness. Sleep maintenance is generally not problematic, and no sleep onset insomnia is experienced if bedtime coincides with the preferred sleep onset time (e.g., on weekends, school vacations). School tardiness and frequent absenteeism are often present.

Treatment

The goal in the treatment of DSPD is basically 2-fold: first, shifting the sleep-wake schedule to an earlier time, and second, maintaining the new schedule. Gradual advancement of bedtime in the evening and rise time in the morning typically involves shifting bedtime/wake time earlier by 15-30 min increments; more significant phase delays (difference between current sleep onset and desired bedtime) may require “chronotherapy,” which involves delaying bedtime and wake time by 2-3 hr daily to every other day. Exposure to light in the morning (either natural light or a “light box”) and avoidance of evening light exposure are often beneficial. Exogenous oral melatonin supplementation may also be used; larger doses (i.e., 5 mg) are typically given at bedtime, but some studies have suggested that physiologic doses of oral melatonin (0.3-0.5 mg) administered in the afternoon or early evening (i.e., 5-7 hr before the habitual sleep onset time) seem to be most effective in advancing the sleep phase.