Obstructive Sleep Apnea in Children: Accurate Diagnosis, Effective Treatment

ABSTRACT: Obstructive sleep apnea (OSA) has a high prevalence in the pediatric population and is associated with significant morbidity, both physical and in the realms of development, cognition, behavior, and school performance. Certain subgroups of children are more susceptible to the development of OSA. Although adenotonsillectomy—the most common treatment for childhood OSA—improves symptoms in the majority of patients, a number of children respond less favorably. Positive airway pressure (PAP) is the second most common treatment for childhood OSA. Success in achieving adherence to PAP often requires a stepwise, multidisciplinary approach with close follow-up to address and quickly resolve any problems that arise.

Obstructive sleep apnea (OSA) in children and adolescents is much more prevalent than most practitioners realize: it is found in 1% to 4% of children.¹ Moreover, the incidence of pediatric OSA appears to be on the upswing, probably because of the notable increase in childhood and adolescent obesity over the past decade.² The peak incidence of OSA is seen in children between 3 and 6 years of age,³ around the time that the tonsils and the adenoids reach their maximum size relative to the dimensions of the upper airway. In young children, OSA is equally common in boys and girls, but following the onset of puberty, it is more common in boys.

In this article, I discuss the effects of OSA in children and provide concrete guidance on diagnosing the disorder and ensuring that treatment is as effective as possible.

THE SPECTRUM OF PEDIATRIC OSA

The muscles of the upper airway relax during sleep, leading to a narrowing of its caliber. In addition, the negative pressure generated during inspiration causes the soft tissue of the upper airway to collapse inward. Varying degrees of obstruction can result. With mild obstruction, the only manifestation may be snoring. As the degree of obstruction increases, so does the work of breathing. This tends to result in an increased number of arousals. Arousals in the setting of airflow that remains greater than 50% of baseline constitute what is known as upper airway resistance syndrome. Reductions of airflow to less than 50% of baseline for 2 breath cycles, accompanied by arousal and/or desaturation, are called "hypopneas"; and reductions to less than 10% of baseline are termed "apneas."

Children with obstructed breathing during sleep are less likely to have arousals than are adults. It is not unusual for children—especially younger children—to present instead with obstructive hypoventilation, in which persistent obstruction and decreased airflow lead to carbon dioxide retention and hypoxemia that can persist for minutes without causing an arousal (which might prompt a change in body position, leading to resolution of the obstruction).

CAUSES OF OSA IN CHILDREN

Pediatric OSA is caused by the interaction of many different factors that together result in a critical degree of narrowing and ultimate collapse of the upper airway. Among these are anatomical factors, conditions that result in inflammation of the soft tissue of the upper airway, and reduced baseline central muscle tone of various etiologies⁴ (Table 1). OSA is also typically worse in the supine position and during rapid eye movement (REM) sleep, when almost all muscle tone except for that of the eye muscles and diaphragm is lost. The large number of possible contributory factors explains why the approach to the treatment of OSA varies from one child to the next.

OSA in infants is usually caused by craniofacial abnormalities, low muscle tone, and/or altered soft tissue size.⁵ The 2 most common causes of OSA in older children are adenotonsillar hypertrophy⁶ and obesity.⁷ MRI studies of children with OSA have shown that most obstructions are in the area of the adenoids (which reach their maximum size at age 6 years) and the soft palate; these studies show that the size of the soft palate in children between the ages of 3 and 7 years who have OSA is increased by 30% compared with soft palates in children who do not have OSA.⁸ The increase in the size of the soft palate represents edema, which is probably secondary to vibratory trauma.

Obesity increases the likelihood of childhood OSA by a factor of 1.3- to 9-fold^7,9; it does this by increasing the size of the fat pads in the neck, causing fatty muscle infiltration and producing changes in chest wall mechanics. This can lead to a decrease in the functional residual capacity of the lungs, which, in turn, increases sensitivity to relatively small changes in respiratory patterns.

MORBIDITY ASSOCIATED WITH OSA IN CHILDREN

The following effects of OSA are seen in patients of all ages:

Researchers are discovering that in children, OSA has additional deleterious neurocognitive, developmental, and behavioral consequences. Connections between OSA and reduced verbal IQ,¹⁵ decreased executive function,¹⁶ lower Bailey developmental scores,¹⁷ poor school performance,¹⁸ attention-deficit/hyperactivity disorder,¹⁹ and failure to thrive²⁰ have become widely recognized. Many of the studies cited have described improvement in the parameters examined following successful treatment of OSA—an improvement not seen in untreated children. It is still unclear what the precise mechanisms are that cause the neurocognitive and behavioral deficits; however, there is evidence that changes in cerebral blood flow²¹ and the presence of inflammation²² play significant roles.

DIAGNOSIS OF CHILDHOOD OSA

It is always a good idea to ask parents whether their child snores. If parents report that their child snores—or volunteer information about other signs and symptoms listed in Table 2—a workup for OSA is indicated. The first step in the workup is to take a detailed history. Questions to ask the child and parent are listed in Table 2.

The physical examination includes assessment of body mass index, blood pressure, and central muscle tone. Examine the nose for septal deviation; nasal polyps; and size, shape, and patency of the nares. Examine the midface for retrusion/hypoplasia and the mouth for evidence of crossbite, overjet, overbite, or underbite. Note the size of the tonsils and the Mallampati classification (Figure), the symmetry of the soft palate, and any evidence of inflammation of the soft palate and uvula. Check the mandible for retrognathia, and measure the neck circumference. Note whether acanthosis nigricans is evident, and assess for scoliosis. Identify any signs of right-sided heart failure.

Although lateral neck films are sometimes used to assess for adenoidal hypertrophy, remember that a lateral neck film that is obtained in an awake child in the upright position may underrepresent dynamic changes that occur in the upper airway while the child is in supine REM sleep.²³ Also keep in mind that in contrast to the tonsils, adenoids can grow back after removal. Consider the possibility of adenoidal regrowth when treating a child with symptoms of OSA who has had his or her adenoids removed; a second adenoidectomy, with or without tonsillectomy, may be sufficient to treat the OSA.

The current gold standard for determining whether a child has OSA is attended polysomnography (PSG). In PSG, defined physiological parameters (electroencephalogram, ECG, electromyogram, eye movements, pulse oximetry, end-tidal carbon dioxide levels, air flow, respiratory effort) are monitored while the child sleeps in a controlled environment under the supervision of a sleep technologist.²⁴

Figure – The Mallampati classification is a rough estimate of the size of the tongue relative to the size of the oral cavity.

Whether every child with symptoms of sleep disordered breathing needs to undergo a diagnostic PSG before initiation of treatment (adenotonsillectomy in most cases) is still an open question.^25-27 PSG is very useful in determining the severity of OSA and in guiding perioperative care. However, distinctions between abnormal and normal results are still not well defined. In fact, one study found that the behavior of children referred for adenotonsillectomy because of clinical symptoms of OSA improved irrespective of the findings on PSG¹⁹; these results suggest that relying solely on PSG to decide whether to proceed to adenotonsillectomy in otherwise healthy children who snore and who have enlarged adenoids and tonsils may lead to undertreatment of OSA. Because of the lack of clear PSG criteria for the diagnosis of OSA, many clinicians rely on the clinical history and physical examination. A recent survey revealed that fewer than 10% of pediatric otolaryngologists ordered PSG for children with suspected OSA before proceeding to adenotonsillectomy, and over 60% stated that they would perform adenotonsillectomy in a child with clinical symptoms of OSA regardless of the findings on PSG.²⁸

TREATMENT OF OSA IN CHILDREN

Pharmacotherapy and positional therapy for mild OSA. In children with mild OSA, both nasal corticosteroids and montelukast(Drug information on montelukast) have been shown to reduce symptoms and the degree of obstruction, although they are not considered to be as effective as adenotonsillectomy.^29,30 Elimination of other secondary causes of soft tissue inflammation (such as exposure to environmental tobacco smoke) and treatment of allergies and gastroesophageal reflux disease can also significantly reduce the degree of obstruction. Positional treatments can be effective as well. For example, if obstruction is seen only while the child is in the supine position, a wedge—or a T-shirt with a tennis ball sewn into the pocket and worn backwards—can be used to keep the child out of that position. Similarly, in some children, especially those with central hypotonia and a restrictive lung disease, sleeping with the head of the bed elevated is often the only intervention required to control obstruction.

Adenotonsillectomy. In most children with OSA, however, the firstline treatment is adenotonsillectomy. This is curative in 79% to 92% of children,³¹ although some series have demonstrated complete resolution of obstruction in smaller numbers. Adenotonsillectomy is less likely to be successful in older and obese children^32,33 and in children who have underlying chromosomal or craniofacial abnormalities.

When a decision is made to treat OSA surgically, both the adenoids and the tonsils should be removed, barring a compelling reason not to do so.³⁴ Even though the tonsils may appear small on inspection, differences in muscle tone between wakefulness and sleep, especially REM sleep, may result in otherwise small-appearing tonsils having a significant impact on the caliber of the upper airway during sleep.

Following adenotonsillectomy, some children may require extubation to continuous positive airway pressure (CPAP) and/or admission to the ICU until postoperative swelling has receded and the upper airway is stable. In 2002, the American Academy of Pediatrics published guidelines recommending postoperative hospital admission and close monitoring of children at high risk for perioperative complications, including those with severe OSA; children younger than 3 years; and children with cardiac complications of OSA, failure to thrive, obesity, prematurity, recent respiratory infection, craniofacial abnormalities, or neuromuscular disorders.³⁵ In children identified before surgery as having severe OSA, and in those with other underlying risk factors, such as obesity or craniofacial, chromosomal, or muscle tone abnormalities, consider obtaining a follow-up PSG about 8 weeks after surgery to check for full resolution.

Positive airway pressure (PAP). This is the second most common treatment for OSA in children. The pressurized air is generated by a compressor and is delivered through an air hose via an interface: a nasal mask with or without a chinstrap, a full-face mask, or nasal pillows. The amount of pressure required is determined during a titration PSG in which PAP is administered at increasing increments until no further obstruction is seen. Either CPAP or bi-level PAP (BiPAP) can be used. CPAP is used more often, but in children in whom high pressure is required, BiPAP may prove to be more comfortable. PAP is used either until testing shows it is no longer needed or until further surgical or orthodontic interventions can be tried.³⁶ Although PAP is very effective, its success depends greatly on regular use. For a description of an approach to introducing children to PAP and monitoring its use afterward that may increase the odds of success, see the Box.

Less common treatment options. Additional surgical interventions, such as midface advancement, mandibular distraction and advancement,³⁷ tongue reduction, and uvulopalatopharyngoplasty,³⁸ are also used to treat OSA, albeit rarely. These interventions are usually attempted when adenotonsillectomy and CPAP have not succeeded in resolving the obstruction or in children in whom there is a clear indication for a surgical or orthodontic procedure, eg, a high-arched and narrow palate or retromicrognathia; in cases such as these, CPAP is used as a bridge. Although rarely performed for this purpose, tracheostomy can also be used to bypass the upper airway, thereby eliminating obstruction.

Although dental appliances are used to treat mild to moderate OSA in adults, their use is not common in children because of the potential for alteration of the bite and for temporomandibular joint disease.³⁹ Orthodontic interventions, such as maxillary expansion, have been shown to be very successful, both as standalone procedures and in conjunction with adenotonsillectomy, in treating OSA in children.^40,41

Children's Hospital Boston: One Model of How Pediatric PAP Is Initiated and Monitored

At Children's Hospital Boston, children in need of positive airway pressure (PAP) meet with one of the sleep technologists in advance of the titration so that they can be fitted with a comfortable and suitable mask. The mask-fitting sessions generally last for 20 to 30 minutes, during which the children are engaged, with the parents present, and the placement and removal of the interface is introduced in a playful and non-threatening way.

The majority of children are able to proceed immediately to a titration study after meeting with the technologist. Others, however, require a period of acclimatization and desensitization at home. In these cases, the families are given the mask to take home and are encouraged to work on getting the child to adapt to the mask at his or her own pace, first by having the child wear it while playing, then while watching television or being read to, and finally by putting it on before falling asleep. A number of parents have reported that they incorporated the headgear for the mask into a game (eg, a mask with mesh-like headgear was referred to as a "Spiderman costume" by one of the children). The habituation process is highly specific, depends on many factors, and can range in duration from a few days to weeks. The families are asked to contact the sleep lab once they feel their child is comfortable wearing the mask; the titration study is then scheduled.

First follow-up visit: a time to review treatment details. Following the titration study, the PAP is ordered with downloadable compliance and efficacy capacity so that adherence can be monitored longitudinally. The child and parent(s) are asked to return in 1 to 2 weeks for follow-up and to bring the mask and machine with them. The machine is checked to make sure it is programmed to the correct settings, and the mask is applied to make sure that the fit is appropriate—not loose enough to cause an air leak, not so tight as to cause skin abrasion or breakdown. The use of the heated humidifier is reviewed, and the family is taught that its use results in diminished nasal stuffiness. Other issues typically discussed at the first follow-up visit include the following:

•	Ways to avoid water condensation in the air hose.
•	Possible need for adjuvant medications such as nasal corticosteroids, antihistamines, or montelukast(Drug information on montelukast).
•	How frequently to expect to need to replace the mask and tubing.
•	Cleaning instructions.

Monitoring for compliance, effectiveness, and accommodation to growth and change. The downloadable compliance data are monitored at varying intervals. At Children's Hospital Boston, the downloadable compliance data are reviewed 2 weeks, 6 weeks, and 3 months after initiation and, if all is going well, every 3 months thereafter. This helps not only to alert providers to any compliance problems patients may be having that they are not reporting (or that the parents may not even be aware of) but also to meet insurance requirements of documented use.

Because children, especially younger children, are constantly growing, leading to changes in airway caliber and compliance, muscle tone, and body mass index, it is important to restudy those receiving PAP therapy on a regular basis. There have been reports of midface hypoplasia developing in infants with longterm PAP delivered via mask⁴²; thus, it is important to watch for this and to discontinue PAP once it is no longer necessary, especially in very young children. Conversely, children who gain a significant amount of weight or who experience a recurrence of symptoms such as snoring, excessive daytime sleepiness, or behavioral disturbances despite documented use of PAP may require higher pressures.

At Children's Hospital Boston, most children younger than 2 years who are receiving PAP are restudied every 3 to 6 months to determine whether PAP is still necessary and whether the settings are appropriate. Children older than 2 years are restudied every 1 to 2 years (unless there are clinical reasons to do so more often).