Pediatric obstructive sleep apnea (OSA) is characterized by partial or complete obstruction of the upper airway during normal sleep, disrupting normal sleep and breathing [1]. With a prevalence of 1%-5%, pediatric OSA is recognized as a relatively common condition that can have potentially serious consequences for children’s growth. Recent studies suggest that delayed diagnosis of pediatric OSA may lead to several complications, including neurobehavioral developmental abnormalities, cardiopulmonary dysfunction, and insulin resistance, which may predispose the child to adult OSA [2,3].

OSA is increasingly recognized as a complex and heterogeneous disorder [4]. Recent studies have shown that the heterogeneity of this disorder spans multiple domains, including symptoms [5], physiologic causes [6], comorbidity [7] and treatment outcomes [8,9]. Despite this recognition, the diagnosis, evaluation, and treatment of OSA still rely primarily on a single metric, the apnea-hypopnea index (AHI) which makes understanding and resolving this complex condition difficult [1]. To address these issues, a “phenotyping” approach has been proposed that classifies OSA into smaller, more homogeneous categories [10].

Phenotyping, based on clinical, pathophysiological, cellular, or molecular characteristics, can provide important information about treatment response, as recent studies suggest that patients with similar pathophysiological characteristics may respond differently to the same treatment [11]. This is particularly relevant in pediatric OSA, which is characterized by a variety of phenotypes. Differentiating phenotypes in each patient may lead to more targeted diagnostic approaches and help identify subgroups more likely to respond to specific treatments, playing a critical role in the development of personalized medicine strategies for pediatric OSA patients.

In adult patients with OSA, it is well-known that altered sleep architecture results in inadequate recovery after sleep, regardless of sleep duration [12,13]. Furthermore, recent studies have shown that these changes in sleep architecture vary according to patient phenotype, leading to differences in treatment outcomes [14]. However, research regarding this in children is very limited. While there have been studies on sleep structure and a few phenotypes in children with OSA, the results have been inconsistent [13,15-17]. It is well known that although pediatric OSA shares many similarities with adult OSA in terms of upper airway obstruction during sleep, its causes, clinical manifestations, and treatments differ from those of adult OSA. Therefore, the aim of this review is to explore the phenotypes of pediatric patients with OSA, describe changes in sleep structure according to phenotype across studies, and seek consistency among findings.

Phenotypes of pediatric OSA can be broadly categorized into 3 types: common, adult and congenital phenotypes. The common phenotype is characterized by enlarged tonsils and/or adenoids, a long face, a narrow palate, or mild malocclusion. The peak incidence of pediatric OSA occurs between 3 and 6 years of age, coinciding with the typical period of tonsil and adenoid enlargement. Tonsil and/or adenoid hypertrophy can cause partial or complete obstruction of the nasal and oropharyngeal airways during sleep, leading to airway compromise and respiratory distress in children [2,18,19]. The adult phenotype in OSA children is often characterized by obesity, a short neck, and midfacial hypoplasia. Although most children with OSA have enlarged tonsils and/or adenoids or obesity, certain subgroups with underlying conditions are also at increased risk. These children often have congenital phenotypes that manifest OSA symptoms from infancy and may have a history of prematurity. They may also be associated with conditions such as Down syndrome, neuromuscular disorders, and craniofacial anomalies commonly associated with the Pierre Robin sequence. These conditions may be associated with micrognathia or retrognathia, which increases the risk of upper airway obstruction during sleep [19,20]. However, these phenotypes are not always distinctly separated; it is common for pediatric phenotypes to present with varying degrees of adenoid and tonsil hypertrophy, ranging from mild to severe [21].

Polysomnography (PSG) remains the gold standard for diagnosing pediatric OSA and provides an objective assessment of the severity of the condition [19,22,23]. Although PSG is the definitive diagnostic test for sleep problem, it has inherent limitations, including cost, time, and challenges that are particularly daunting for children [19]. PSG is usually performed in a hospital setting and often requires changes to a child’s usual bedtime routine, which can alter sleep quality [24]. Despite these challenges, PSG is highly recommended, especially in children who are obese, have congenital syndromes, neuromuscular disorders, or craniofacial anomalies. Phenotyping pediatric OSA with PSG can confirm the presence of specific features associated with different phenotypes, providing important information about the severity of the disease and its underlying mechanisms. Since the phenotype of pediatric OSA can also be determined by PSG, which can provide important information about the severity of the disease and the underlying mechanisms of the disease, it is worthwhile analyzing PSG results for some of the characteristic phenotypes of pediatric sleep apnea mentioned above and consider their implications.

PSG in adult patients with OSA shows that sleep cycles are fragmented, with an increase in the proportion of N1 sleep stages and a decrease in the proportion of deep sleep stages. This results in a lack of restorative sleep, and frequent arousals after respiratory events that disrupt sleep, and lead to poor sleep quality. As a result, patients may show excessive daytime sleepiness and fatigue [12,13,25,26].

As noted above, there is a lack of evidence that pediatric OSA leads to changes in sleep architecture as commonly observed in adults. Research evaluating the relationship between pediatric OSA phenotypes and sleep architecture is limited, and most of these studies have focused on children with typical phenotypes.

In the study of pediatric OSA, Goh, et al. [27] reported that children with a common phenotype of OSA had normal sleep architecture, this study is limited by its small sample size of 20 patients. Similarly, Scholle and Zwacka [15] described those children with the common phenotype of OSA had an increased arousal index compared to controls, but generally maintained a normal sleep structure, a finding that contrasts with the sleep structure changes observed in adults with OSA. However, this study was also limited by its small sample size, with only 12 children in the experimental group and 20 in the control group.

In the study by Walter, et al. [16], they found that duration of sleep stages, total sleep time (TST), and sleep period time (SPT) and sleep latency were relatively well maintained in patients with common types of pediatric OSA. However, their sleep efficiency was reduced, and notably, the proportion of rapid eye movement (REM) sleep in SPT was significantly lower. Similarly, Tauman, et al. [17] found in their study that children with OSA had a significantly increased proportion of slow-wave sleep compared to controls, but a decrease in REM sleep.

Recent research by Sun, et al. [28] presents contrasting findings in school-aged children with OSA, reporting a significant decrease in N3 stage and an increase in N1 stage compared to controls. In addition, research by Zhu, et al. [29] suggests that preadolescent school-aged children with primary snoring and a common OSA phenotype generally have a normal sleep structure. However, in adolescents, there is an increase in N1 sleep stage and wake after sleep onset.

According to a study by Durdik, et al. [30], in children with the common phenotype of pediatric OSA, the distribution of sleep cycles and stages may remain intact, with no significant differences in SPT, TST, arousals during sleep, REM sleep, N2 stage sleep, and sleep efficiency compared to controls. Similar to the findings of Sun, et al. [28], Durdik, et al. [30] observed that children with OSA had a significantly shorter duration of N3 stage sleep and a significantly longer duration of N1 stage sleep. Table 1 shows the summary of these studies on the characteristics and Polysomnographic findings in children with Common phenotype of OSA.

In pediatric OSA, the adult phenotype is particularly associated with obesity-related and positional sleep apnea. OSA is known to have a high prevalence in obese children, affecting up to 60% of children in the obesity categories [31]. This phenotype is more common in older children and is characterized by a higher body mass index (BMI). Children with obesity-related pediatric OSA typically show frequent arousals during sleep on PSG, resulting in fragmented sleep and severe nocturnal hypoxemia [32,33]. Prolonged nocturnal hypoxemia can be associated with neurocognitive dysfunction and cardiovascular metabolic complications that may persist into adulthood, underscoring the importance of timely diagnosis and intervention [34].

Positional sleep apnea is one of the dominant phenotypes in pediatric patients with the adult phenotype OSA. Positional OSA is generally diagnosed when the AHI is at least twice as high in the supine position as in the non-supine position [35]. While approximately 60% of adult OSA patients have a positional phenotype, the prevalence in pediatric patients is lower, ranging from 19% to 48%, indicating a lower incidence compared to adults [36-38]. Most studies suggest that the positional OSA phenotype tends to be less severe in terms of the overall degree of sleep apnea and is more likely to preserve sleep architecture compared to other phenotypes [39,40]. According to Wang, et al. [41], children with the positional OSA phenotype had higher AHI scores than those without, especially when lying in the supine position. In addition, these children spent significantly more time in non-supine positions during sleep, which was associated with fewer respiratory events in these positions. Table 2 shows the summary of these studies on the characteristics and Polysomnographic findings in children with Common phenotype of OSA.

Congenital phenotypes of pediatric OSA are closely associated with craniofacial anomalies that predispose children to a higher risk of airway obstruction [42]. Here’s an overview based on various studies regarding congenital phenotypes in pediatric OSA. MacLean, et al. [43] reported that infants with non-syndromic cleft palate had a high mean AHI of 7.6 events per hour during sleep, linking structural anomalies such as cleft palate and midface hypoplasia to an increased risk of airway obstruction. Because conditions such as Pierre Robin sequence and Treacher Collins syndrome are often associated with a small lower jaw and tongue displacement that blocks the airway, the incidence of severe OSA is high in infants with these conditions [44]. Anderson, et al. [45] found an extremely high mean AHI of 33.5 events per hour in infants with Pierre Robin sequence. Children with Down syndrome often have midface hypoplasia, macroglossia and low muscle tone, which contribute to airway obstruction. A study by Rosen [46] found that 31%-100% of children with Down syndrome had gas exchange abnormalities on sleep PSG. Children with achondroplasia, characterized by midface hypoplasia and upper airway narrowing, have a high incidence of both obstructive and central sleep apnea [47]. More than half of sleep studies in these children show abnormalities, with hypoxemia being the most common finding [48]. Table 3 shows the summary of these studies on the characteristics and polysomnographic findings in children with common phenotype of OSA.

In summary, the reviewed studies highlight that pediatric OSA phenotypes are more distinctly categorized than in adults, with significant differences in symptoms, diagnosis, and treatment strategies between phenotypes. It can also be assumed that in most phenotypes of pediatric OSA patients, sleep architecture is altered on PSG, which can negatively affect sleep quality in children, as it does in adults, and can lead to various complications. However, as noted above, large-scale studies examining the relationship between pediatric OSA phenotypes and sleep architecture remain scarce. Further clinical research is essential to validate the proposed hypotheses and to better understand the nuanced interplay between phenotypes and sleep disorders in children. Also, this study was conducted on pediatric OSA phenotypes. It would be beneficial to extend this research to include comparisons between adult and pediatric phenotypes, as there may be significant differences in the phenotypes and management strategies between these age groups. Finally, these findings highlight the importance of early detection and accurate management of all pediatric OSA patients, particularly by using diagnostic tools such as PSG to identify each phenotype early and by attempting to personalize care for each patient by considering other polysomnographic variables along with apnea-hypopnea levels to improve outcomes.