Childhood obstructive sleep apnoea

Obstructive sleep apnoea is characterised by oxygen desaturation and reduced oro-nasal air flow despite preserved thoracic and abdominal respiratory effort.¹ It occurs in 1-2% of children and is more common in prematurely born infants and in black and Hispanic children.² As our knowledge of this condition has grown, so has concern about it among parents and clinicians. How is it best diagnosed and managed?

Habitual snoring, breathing through the mouth, periods of observed apnoea, restless sleep, urinary incontinence, inattentiveness, daytime hyperactivity, mood swings, and failure to thrive are the most common clinical manifestations of childhood obstructive sleep apnoea. How loudly a child snores is not correlated to the presence or severity of sleep disordered breathing. The most common predisposing factors are adenotonsillar hypertrophy, neuromuscular disorders, and craniofacial anomalies associated with maxillary hypoplasia, retrognathia, or macroglossia. The local release of proinflammatory cytokines such as C reactive protein, tumour necrosis factor α, and interleukin 6 might also play a part in exacerbating mucosal swelling and airway narrowing.³

The neuropsychological sequelae of classic childhood obstructive sleep apnoea have now been firmly established. O'Brien et al recently described 35 children with obstructive sleep apnoea (mean age 6.7 years) and 35 closely matched controls.⁴ The children with sleep apnoea had notable deficits in attention span, executive function, phonological processing, visual attention, and general conceptual ability compared with the controls. The deficit in phonological processing is worrying since this serves as a basic building block in the development of reading skills.

Nocturnal polysomnographic observations in paediatric obstructive sleep apnoea were first made by Guilleminault et al in 1975.⁵ Polysomnography consists of the simultaneous recording of cardiorespiratory, electromyographical, and electroencephalographic variables. The threshold of oxygen desaturation that should be used for scoring respiratory events during polysomnography remains unresolved. Federal guidelines in the United States (Medicare) stipulate a 4% oxygen drop from the baseline during respiratory events, whereas the Cleveland heart health study² applied a 3% desaturation threshold. These disparities are not trivial and can lead to inconsistencies from one sleep laboratory to another in diagnosing obstructive sleep apnoea. Standardisation of sleep monitoring techniques and the universal application of validated criteria for diagnosing childhood obstructive sleep apnoea remain a priority.

Though nocturnal polysomnography is the accepted gold standard for diagnosing childhood obstructive sleep apnoea, it is expensive and not widely available. The small number of sleep centres with trained paediatric staff coupled with increasing public demand for childhood sleep evaluations have strained resources. Alternative methods of diagnosing childhood obstructive sleep apnoea therefore need consideration.

Goldstein et al have shown that a comprehensive clinical assessment can diagnose childhood obstructive sleep apnoea with 48% accuracy.⁶ The assessment consists of a validated quality of life survey, review of an audiotape for snoring and “snort arousals” (sensitivity 88%, specificity 52%, positive predictive value 62%), evaluating the lateral neck radiograph for adenoidal hypertrophy, voice recording assessment for hypernasality, and an electrocardiogram. Though 48% accuracy is far from optimum, this assessment is nevertheless a useful tool for the general practitioner.

Another substitute for nocturnal polysomnography is physical examination combined with the inexpensive and easily available technique of nocturnal pulse oximetry. Brouillette et al from the Montreal Children's Hospital have shown that when the clinical history suggests obstructive sleep apnoea, an abnormal overnight oximetry trend graph has a 97% positive predictive value.⁷ A normal overnight oximetry study does not, however, exclude mild or moderate obstructive sleep apnoea as only 93/210 (44%) of their patients with sleep apnoea proved by polysomnography showed abnormal results on oximetry.

Though adenotonsillectomy is widely used for treating childhood sleep apnoea, its benefit has not been definitively established through more evidence based research. About a fifth of patients below the age of 36 months develop dangerous postoperative airway oedema after adenotonsillectomy, so all children in this age group should be closely monitored for cardiorespiratory compromise for at least 24 hours. Positive pressure airway breathing is helpful in those with obstructive sleep apnoea that persists despite adenotonsillectomy. Weight reduction measures intuitively make sense, but again their value has not been established through randomised controlled trials. Corticosteroids such as intranasal fluticasone seem to reduce the size of adenoids and the severity of symptoms slightly⁸ and serve as an adjunctive treatment.

Increased interest in childhood obstructive sleep apnoea should prompt the randomised trials that are now necessary to assess the best ways of managing the condition.