Consequences of Obstructive Sleep Apnea in Children

Obstructive sleep apnea syndrome (OSAS) has various negative health and behavioral consequences in the pediatric population. As shown in adults, there are metabolic derangements such as obesity, insulin sensitivity, dyslipidemia, and metabolic syndrome, as well as cardiovascular derangements like hypertension, chronic inflammation, endothelial dysfunction, ventricular size/function abnormalities, and even elevated pulmonary arterial pressures, that can be seen in children with OSAS. The first two sections will discuss the metabolic and cardiovascular consequences on OSAS in children. The last section summarizes selected studies and reviews on the behavioral, neurocognitive and academic consequences of OSAS in children.

Introduction

Obstructive sleep apnea syndrome (OSAS) has various negative health and behavioral consequences in the pediatric population. As shown in adults, there are both metabolic and cardiovascular derangements that can be seen in children with OSAS. The first two sections will discuss the metabolic and cardiovascular consequences on OSAS in children. The last section summarizes selected studies and reviews on the behavioral, neurocognitive and academic consequences of OSAS in children.

The prevalence of childhood obesity is estimated at 17%, including approximately 8% of 2–5-year olds and 20% of 12–19 year olds.¹ Obesity is associated with cardiovascular disease, dyslipidemia, nonalcoholic fatty liver disease, insulin resistance, type 2 diabetes, and OSAS.² Additionally, the compilation of obesity, dyslipidemia, hypertension, and impaired glucose tolerance/insulin resistance is recognized as the metabolic syndrome (MetS). MetS is seen in up to 10% of overweight/obese adolescents.³ Patients with MetS are at increased risk of heart disease, type 2 diabetes, and inflammatory disorders.4, 5 Studies have demonstrated a connection between OSAS and obesity, dyslipidemia, insulin sensitivity, and the metabolic syndrome. This section will address these relationships.

There is a strong connection with obesity and the development of OSAS, leading to a prevalence of OSAS in the obese pediatric population of over 50%.⁶ This is due to a combination of several factors. Obese children have increased subcutaneous fat in the neck and fatty infiltration of upper airway structures like the tongue, narrowing the upper airway. Lymphoid-tissue hypertrophy including tonsils and adenoids also contribute to airway narrowing.⁷ Another factor includes decreased lung volume and oxygen reserve secondary to fat deposition around viscera and in abdominal and thoracic walls. This is exacerbated while lying supine. Reduced effectiveness of leptin, or leptin resistance, secondary to obesity contributes to a weakened ventilatory response. Leptin acts as a respiratory stimulant in collaboration with both central and peripheral chemoreceptors.⁸

OSAS may cause or worsen obesity as well. Sleep fragmentation and decreased quantity of sleep results in daytime sleepiness, which can limit physical activity. It may also alter metabolic hormones including leptin and ghrelin. Leptin is produced by adipocytes and is a satiety-producing hormone whose effects are mediated in the arcuate nucleus of the hypothalamus.⁹ Ghrelin is produced by cells in the stomach and is an appetite stimulating hormone. Both are believed to play a role in body weight regulation.¹⁰ With sleep deprivation, there is a significant decrease in levels of leptin and an increase in levels of ghrelin.¹¹ It has also been shown that with sleep restriction, there is an increase in hunger and appetite, especially for carbohydrate-rich foods.¹²

Most studies demonstrate an independent correlation between OSAS and fasting insulin. A study by Bhushan et al.¹³ evaluated young obese children, aged 2–12 years and found significantly higher fasting insulin, blood glucose, and homeostasis model assessment (HOMA) in patients with OSAS compared to those with no OSAS; these indicators were also worsened with greater severity of OSAS. This relationship is strengthened by data showing treatment of OSAS is associated with improvement in insulin levels, lipid homeostasis, and HOMA in the absence of body mass index (BMI) changes.¹⁴ However, there are few studies that argue an opposing viewpoint. One such view is that there is no correlation between OSAS and serum insulin, serum glucose, and HOMA, as was displayed by Kaditis et al.⁹ in a study composed of 110 non-obese children. Another view is that obesity, rather than OSAS, is the main determinant of insulin resistance.¹⁵

The relationship between OSAS and obesity, insulin resistance, dyslipidemia, and hypertension, as described previously, belies the role OSAS can have in developing the metabolic syndrome. The mechanisms of metabolic disturbance due to OSAS includes increased production of serum cortisol and sympathetic activity,¹⁶ secondary to activation of the stress response system including the hypothalamic–pituitary–adrenocortical (HPA) axis, and formation of reactive oxygen species. The potential long term negative impact of these health issues stresses the need for therapy. While typical OSAS therapy of adenotonsillectomy (AT) is not as effective in obese children,¹⁷ AT has been associated with normalization of salivary cortisol.¹⁸ As well, weight loss can improve both OSAS and metabolic function, given the role obesity plays in OSAS and metabolic function.

Studies of adults have shown that OSAS is a risk factor for dyslipidemia¹⁹; however, it must be noted that most adult patients with OSAS are obese, making it difficult to determine the degree of which OSAS contributes to dyslipidemia versus obesity׳s known effect. In the pediatric population, there are studies that demonstrate evidence for and against this linkage. Tauman et al. reviewed fasting serum glucose, insulin, and lipids after polysomnographic evaluation in 116 snoring children and 19 control subjects. Of the total 135 children, 70 were obese.¹⁵ They found no significant correlation between severity of OSAS and dyslipidemia but positive correlations between insulin/glucose ratio and relative BMI. A recent study by Bhushan et al.¹³ had similar findings among 76 obese children demonstrating no significant difference in total cholesterol, triglycerides, or high- and low-density lipoprotein levels in patients with or without OSAS. Another study of 62 obese and non-obese children with OSAS demonstrated significant improvement in lipid profiles, improvement of apolipoprotein B serum levels, and C-reactive protein in both non-obese and obese children following treatment of OSAS via AT.¹⁴ Based on these results, it is presumable that OSAS does play a role in dyslipidemia and systemic inflammation, which is independent of obesity.

During sleep the sympathetic nervous system activity decreases, resulting in lowered heart rate, blood pressure, and stroke volume during non-rapid eye movement (NREM). Sympathetic nervous system output increases during rapid eye movement (REM) sleep yielding a rise in heart rate and blood pressure. These changes are altered during sleep disordered breathing (SDB).

Adult studies have established a link between OSAS and cardiovascular disease.²⁰ Impacts on the heart includes an effect on blood pressure, arrhythmogenics, ventricular structure/function, endothelial function, and cardiac autonomic activity. These factors, alone and especially combined, can lead to significant cardiovascular morbidity. The link between OSAS and cardiovascular disease in children has become more evident over time.

In the adult population, OSAS is associated with increased sympathetic activity.²¹ The repetitive activation of the sympathetic nervous system occurs via central and peripheral chemoreceptor activation secondary to frequent arousals and intermittent hypoxia and hypercapnia. Amongst other outcomes, this leads to elevated blood pressure. This process can also be seen in children and was illustrated in a study of 96 children by O׳Driscoll et al.²² They demonstrated increased noradrenaline and adrenaline in patients with OSAS and found that levels of overnight urinary noradrenaline and adrenaline were independently related to AHI in children, representing increased sympathetic nervous system activity in pediatric OSAS.

Angiotensin II, a potent vasoconstrictor that has an important role in the renin–angiotensin–aldosterone system (RAS), also plays a role in blood pressure regulation through its control on extracellular fluid volume and vascular resistance. Angiotensin II levels are increased in adults with OSAS compared to healthy controls.²³ This system has also been shown to impact ambulatory blood pressure in the pediatric population.

Hypertension is a known consequence of OSAS in adult patients. It has become increasingly clear that children with OSAS are at higher risk for developing elevated blood pressure as well. Of 105 elementary-school-aged children with OSAS, measured blood pressure while awake and asleep was 10–15 mmHg greater than 36 non-snoring control patients.²⁴ This finding was independent of OSAS severity. Leung et al. recruited 96 children, among them 41 were obese, and classified them into two groups, a high apnea–hypopnea (AHI) index group (AHI>5/h) and a low-AHI group (AHI <or =5/h). They recorded 24-h systolic and diastolic blood pressures and found that the high-AHI group children had a significantly higher systolic blood pressure while awake and higher systolic blood pressure and diastolic blood pressure while asleep.²⁵ They also found that obese children in the high-AHI group had a significantly higher prevalence of hypertension than their obese counterparts in the low-AHI group. To further strengthen this association, significant reductions in 24-h ambulatory blood pressures were noted after AT.

Chronic inflammation is plays a significant role in patients with obstructive sleep apnea. Patients with OSAS have elevated levels of interleukin-6, TNF-alpha, and C-reactive.²⁶ In addition, inflammatory markers such as TNF-alpha significantly improve after treatment of OSAS.²⁷ A potential negative consequence of these inflammatory mediators is vascular injury.

One vascular abnormality seen in association with OSAS is endothelial dysfunction, which is characterized by reduced vasodilation, enhanced vasoconstriction and chronic inflammatory and prothrombotic activity, contributing to the development of atherosclerosis and cardiovascular disorders. The link between OSAS and endothelial dysfunction may be a consequence of the episodes of intermittent hypoxia seen in OSAS, leading to oxidative stress on blood vessels and further elevating markers of inflammation. Studies have demonstrated that endothelin-1, a marker of endothelial dysfunction, is elevated in OSAS and then reduced after treatment of OSAS²⁸ and that flow-mediated dilatation of the brachial artery, an indirect measure of endothelial dysfunction, is abnormal in OSAS but improves after treatment of OSAS. However, it must be noted that the relative contribution of OSAS versus obesity to vascular damage is not entirely known. In a cohort of 31 obese adolescents, Koren et al.²⁹ examined OSAS׳s independent contribution to macrovascular disease risk and found that the primary predictor of measures of carotid structure changes, such as maximal carotid intima-media thickness, and arterial stiffness was the degree of obesity as opposed to the presence of OSAS. It is postulated that this degree of vascular damage by OSAS may take years to manifest and therefore, is uncommonly seen in pediatric populations.

Ventricular size, and therefore function, may also be altered in OSAS. Specifically, left ventricular hypertrophy with greater relative wall thickness and left ventricular mass has been seen in pediatric OSAS.³⁰ There are studies implicating impairment of right ventricular function as well in patients with adenotonsillar hypertrophy.³¹ The mechanism of left ventricular hypertrophy remains unclear. Possible causes of cardiac remodeling are RAS activation secondary to hypoxia, hypertension, or chronic sympathetic activation. With resolution of OSAS in non-obese children, there are improvements in left ventricular diastolic function and right ventricular myocardial performance index. However, as with other cardiovascular morbidities, the contribution of obesity to these findings is not insignificant, as obese patients without OSAS can have a higher left ventricular mass index and reduced diastolic and systolic function compared to lean controls.³²

Finally, there are few case reports and small-sample-sized studies that have demonstrated elevated pulmonary arterial pressures in OSAS. This is believed to be due to recurrent pulmonary vasoconstriction as a result of episodic hypoxemia in children with OSAS, leading to pulmonary artery hypertension.

There is growing empirical evidence to suggest that pediatric OSAS is associated with a number of behavioral consequences during childhood. Recent empirical studies, meta-analyses, and comprehensive reviews have linked untreated OSAS symptoms in children to both externalizing33, 34, 35, 36, 37, 38, 39, 40 and internalizing41, 42, 43 behavior problems, as well as to deficits in adaptive,⁴⁴ neurocognitive, and academic functioning.45, 46, 47, 48 Externalizing problems refer to difficulties with behavior regulation, compliance, aggression, impulsivity, and hyperactivity, whereas internalizing problems refer to difficulties with anxiety, depressed mood, and emotion regulation.⁴⁹ Neurocognitive deficits refer to impairments in various neurologic and cognitive abilities, such as attention and other executive functions, visual-spatial skills, and working memory, although untreated OSAS symptoms are most robustly linked to diminished attention and executive functioning skills.45, 46, 48 The Table provides a summary of recent studies and meta-analytic work on the behavioral, neurocognitive, and academic outcomes of pediatric OSAS and as well as SDB. It should be noted that this table and article do not include an exhaustive listing of studies, but a selection of recent work in this regard. Additionally, given that subclinical nighttime breathing difficulties have been associated with diminished behavioral, neurocognitive, and academic functioning,34, 35, 37, 41, 42, 44, 46 studies examining developmental outcomes along the continuum of SDB (i.e., from primary snoring to severe OSAS) are included in summary.

A number of studies have linked OSAS or SDB to externalizing problems, and specifically to symptoms of attention-deficit/hyperactivity disorder (ADHD).34, 35, 36, 44, 46 For instance, longitudinal work by Perfect et al.⁴⁴ has shown that compared to youth with no history of SDB (defined as a respiratory disturbance index of≥1 per hour and an associated oxygen desaturation of≥3%), those with SDB that persisted over the 5-year study period were more significantly likely to have parent-reported behavior problems consistent with ADHD, such as broad externalizing and hyperactivity concerns. Experimental research has additionally demonstrated that OSAS treatment via AT35, 36, 38 or positive airway pressure (PAP)⁵⁰ can significantly reduce these concerns. In a prospective, non-randomized study that compared youth with OSAS referred for AT to controls scheduled for non-AT surgery, those with OSAS were found to have poorer attention functioning and more psychiatrist-diagnosed ADHD and disruptive behavior problems compared to controls.35, 36 At post-surgery follow-up, however, 50% of AT patients with baseline ADHD no longer met diagnostic criteria, and there were no differences in rates of externalizing diagnoses across the AT and non-AT surgery groups.³⁶ More recently, Marcus et al.³⁸ randomized youth with OSAS to receive AT or to a watchful waiting and supportive care condition, and found that children who underwent AT showed significantly greater reductions from pre-surgery baseline to 7-month follow-up in parent- and teacher-reported ADHD and other externalizing problems, among other gains in areas like parent-reported executive functions and parent- and child-reported quality of life.⁵¹

Research findings have been less consistent in documenting associations between OSAS or SDB and internalizing outcomes,⁴⁶ although meta-analytic data show a medium effect size for the link between pediatric OSAS and depressive symptoms, as well as for improvement in depressive symptoms from pre to post-AT.⁴³ Similarly, there is less consistency in the extant literature for the impact of OSAS and SDB on specific neurocognitive and academic outcomes.45, 46, 48 Several meta-analytic and literature reviews have concluded that there are mixed findings on the extent to which OSAS can impact broad intellectual capacity (IQ), visual and motor functioning, working memory, and performance on academic achievement tests.33, 46 However, across studies there is strong evidence that untreated OSAS and SDB are associated with diminished attention and broad executive functioning,45, 46, 47, 48 and some evidence for an association with lower child grade point average.⁴¹ For example, a meta-analysis of 25 pediatric OSAS studies found that youth with untreated OSAS had substantial deficits in attention vigilance and executive functioning compared to both healthy controls and available normative data.⁴⁵ Although studies do not reliably show that OSAS and SDB are associated with poor performance on academic achievement tests,⁴⁶ some studies have demonstrated that youth with OSAS and SDB are more likely to have teacher-reported learning problems, lower self-reported grades,⁴¹ and greater rates of academic functioning impairments than expected for non-SDB youth.⁴⁷

Importantly, variation in the extent to which OSAS and SDB are associated with child behavioral, neurocognitive, and academic impairments may be due to the use of report-based (i.e., parent questionnaire) versus performance-based (i.e., behavioral tasks) measures to assess these skills.45, 46 A critical direction for future research is to examine the consistency in parent-, teacher-, and self-reported child impairments and in behavioral tasks or more objective observations of these impairments in applied settings.⁴⁶ Several other directions for future research include additional longitudinal studies of OSAS and SDB, as well as research on the mechanisms linking OSAS and SDB to poor child behavioral outcomes. Recently, a set of cross-sectional studies conducted by Bourke et al.⁴⁷ and Jackman et al.³⁹ revealed that the emergence of specific SDB-related child impairments may vary according to child age. Whereas children aged 3–5 years across a continuum of SDB evidenced behavioral, but not neurocognitive deficits, children aged 7–12 years on the SDB continuum did show diminished neurocognitive skills, suggesting that early childhood may be a particularly important period for SDB and OSAS treatment to prevent subsequent neurocognitive impairments. However, experimental and longitudinal studies that examine the emergence and management of broad SDB- and OSAS-related child difficulties across development are needed.

Although there is little available mechanism research on pediatric SDB or OSAS and behavioral consequences, some studies have begun to examine this topic. Beebe and Gozal46, 52 have proposed a causal model to explain the negative behavioral outcomes for children and adults with untreated OSAS, based on both human and animal research related to the consequences of sleep disruption and intermittent hypoxia and hypercarbia. These OSAS characteristics are thought to impact individuals׳ cellular and chemical processes, which can then disrupt prefrontal cortical functioning. The authors have suggested that these resultant prefrontal cortex difficulties impact the executive functioning system, which, in turn, leads to poorer daytime functioning in behavioral, neurocognitive, and academic domains.

Other studies have begun to examine OSAS symptoms as mediating factors in linking risk factors for OSAS such as child obesity with poor neurobehavioral outcomes. For instance, Xanthopoulos et al.⁴⁰ recently compared behavioral and neurocognitive functioning among obese adolescents with OSAS, obese controls, and lean controls, and found evidence that the level of OSAS, indexed via the apnea–hypopnea index, significantly mediated the effect of body mass on neurobehavioral outcomes. Future studies that replicate and extend these findings to other pediatric populations and child behavioral outcomes are needed.