Positional obstructive sleep apnea in an obese pediatric population

Abstract

Study Objectives:

Positional obstructive sleep apnea (POSA) is a phenotype of obstructive sleep apnea (OSA) where sleep-related obstructive events occur predominantly in the supine position. Limited knowledge exists regarding the presence of POSA in children with obesity. The study objective was to determine the prevalence of POSA while identifying factors associated with POSA in children with obesity.

Methods:

This was a cross-sectional study of children with obesity, aged 8 to 18 years, with a diagnostic polysomnogram (PSG) between 2012 to 2019, who were referred for the evaluation of sleep-related breathing. POSA was defined as an overall obstructive apnea-hypopnea index (OAHI) ≥5 events/h and a supine OAHI to nonsupine OAHI ratio of ≥2. Patient demographics, anthropometrics, and PSG data were recorded.

Results:

Of the 112 children with obesity with a diagnostic PSG, 43 (38%) had OSA. Among those with OSA, 25 of 43 (58%) had POSA (mean age: 14.6 ± 2.3 years; mean body mass index: 37.7 ± 7.6 kg/m²; 68% male) and 18 of 43 (42%) had non-POSA (mean age: 13.9 ± 2.8 years; mean body mass index: 37.9 ± 7.2 kg/m²; 78% male). Among those with POSA, 13 of 25 (52%) had mild OSA, 7 of 25 (28%) had moderate OSA, and 5 of 25 (20%) had severe OSA. No significant differences were found in age, sex, and anthropometric measures between POSA and non-POSA groups. Time spent in supine and nonsupine sleep did not differ significantly between groups.

Conclusions:

In children with obesity and OSA, POSA occurs frequently. Identifying POSA allows for potential targeted positional therapy for children with obesity.

Citation:

Selvadurai S, Voutsas G, Massicotte C, et al. Positional obstructive sleep apnea in an obese pediatric population. J Clin Sleep Med. 2020;16(8):1295–1301.

BRIEF SUMMARY

Current Knowledge/Study Rationale: Strong evidence suggests severity of obstructive sleep apnea (OSA) can worsen when sleeping supine, known as positional OSA (POSA). While POSA is shown to be prevalent among adults, there are limited data on the presence of POSA in children with obesity. Our study assessed the prevalence of POSA in a cohort of children with obesity and coexisting OSA.

Study Impact: We found that POSA occurs frequently in children with obesity, especially among those with milder forms of OSA. These data support that a personalized treatment such as positional devices should be considered when treating OSA and may be beneficial with children who are nonadherent to continuous positive airway pressure therapy.

INTRODUCTION

Obstructive sleep apnea (OSA) is a common sleep-related breathing disorder in children. OSA is characterized by snoring, recurrent partial and/or complete obstruction of the upper airway, and intermittent nocturnal oxyhemoglobin desaturations and sleep disruption.¹ While OSA affects 1–4% of otherwise healthy children,² the prevalence of OSA among children and adolescents with obesity is as high as 60%.³ Untreated OSA is associated with adverse outcomes, including increased cardiovascular and metabolic risk^4,5 as well as neurocognitive and behavioral deficits.⁶

The first-line treatment for OSA in children with obesity is typically an adenotonsillectomy in addition to lifestyle modifications such as weight-loss strategies.⁷ For extreme cases of comorbid obesity, bariatric surgery may also be considered for treating OSA.⁸ However, among the majority of children with obesity, an adenotonsillectomy has not been shown to be as effective, with 50% of children experiencing persistent OSA.⁹ Weight-loss strategies have also been shown to be difficult to achieve and maintain,⁷ and patients with obesity and OSA are particularly vulnerable to surgical complications, including difficulties with intubation, extubation, and pain management.¹⁰ As such, continuous positive airway pressure (CPAP) therapy is primarily prescribed for the treatment of OSA.¹¹ While CPAP has been shown to be effective in treating OSA, adherence is a significant challenge among the pediatric population with concurrent obesity,¹² whereby more than 50% of adolescents are unable to tolerate CPAP due to discomfort and lack of support and perceived benefit.¹³ As a result, many adolescents with obesity and OSA remain untreated. Novel alternative treatments are urgently needed to improve the health of children with obesity and OSA to prevent long-term morbidity.

There is evidence that severity of OSA is associated with body position and is particularly worse when supine.¹⁴ This particular phenotype of OSA is known as positional OSA (POSA). Among adults, the prevalence of POSA is estimated to be 55%.¹⁵ Currently, there are limited studies evaluating POSA in children. Furthermore, there remains ambiguity in defining POSA in the literature given the lack of a universally accepted definition, presenting a major barrier to the development of clear management guidelines for POSA.¹⁶ Most commonly, POSA is defined by an obstructive apnea-hypopnea index (OAHI) in the supine position that is 2 or more times greater than the OAHI in the nonsupine position.¹⁷

Several factors may influence the effect of body position on OSA severity, including age, obesity, and history of an adenotonsillectomy.^18,19 For example, the prevalence of POSA is higher in nonobese adults than in those who are obese, as well as in adults with mild or moderate OSA rather than severe OSA.²⁰ In children, obstructive events have been found to be more common during supine sleep as well as rapid eye movement (REM) sleep.²¹ Sleeping in a supine position can result in reduced craniofacial volume,²² reduced lung volume,²³ and an inability of airway dilator muscles to compensate for airway collapsibility during an obstruction.²⁴ Therefore, individuals with obesity may be more susceptible to upper airway obstruction during supine sleep due to increased fat deposits found within the pharyngeal space, which could contribute to a reduced upper airway.²⁵

Currently, there is a lack of knowledge regarding the prevalence of POSA in children with obesity, in whom a personalized type of treatment may be considered as an alternative to CPAP. The objective of this study was to determine the prevalence of POSA and to identify factors that are associated with POSA in an obese pediatric population. By assessing the prevalence as well as identifying factors associated with POSA in children and adolescents with obesity, who are known to have low rates of CPAP adherence, potential targeted therapy can be offered to mitigate related adverse outcomes.

METHODS

This was a cross-sectional study of children with obesity, aged 8–18 years, who were prospectively enrolled between May 2012 and February 2019. Following enrollment, participants had a diagnostic polysomnogram (PSG). Children with obesity were referred to the Hospital for Sick Children and reviewed by the sleep physician for the evaluation of sleep-related breathing. Children were excluded if they had a known chronic illness or underlying comorbidity associated with high risk for OSA (ie, Down syndrome) or obesity related to other diseases and syndromes (eg, Prader-Willi syndrome). Patient demographics and anthropometrics (age, sex, height, weight, body mass index); symptoms including snoring, nasal congestion, and presence of tonsillar and/or adenoid hypertrophy; and PSG data were recorded systematically. Waist and hip circumferences were measured according to the National Institutes of Health and World Health Organization guidelines, respectively.²⁶ Neck circumference was measured according to a previously published protocol specific to pediatric populations.²⁷ Obesity was defined as a body mass index more than the 95th percentile for age and sex.²⁸

Overnight PSG was performed using XLTEK data acquisition and analysis systems (Natus Medical, San Carlos, CA). PSG measurements included electroencephalograms, electrooculograms, and submental and bilateral anterior tibialis electromyograms. Chest wall and abdominal movements were measured using chest wall and abdominal belts. Other respiratory measurements included nasal air pressure transducer, oronasal thermal sensor, oxygen saturation, transcutaneous carbon dioxide, and end-tidal carbon dioxide monitoring. Sleep architecture was assessed using standard techniques.²⁹ Respiratory events were scored according to the American Academy of Sleep Medicine guidelines²⁹ by a registered, certified PSG technician and were reviewed and interpreted by 1 of 3 experienced pediatric sleep physicians.

OSA severity was graded according to the OAHI—the total number of obstructive apneas, mixed apneas, and obstructive hypopneas per hour during sleep. For children aged 13 years and younger, PSG data were scored based on pediatric guidelines: mild OSA was defined as OAHI ≥1.5 to <5 events/h, moderate OSA as OAHI ≥5 to <10 events/h, and severe OSA as OAHI ≥10 events/h. For those older than 13 years of age, PSG data were scored using adult guidelines: mild OSA was defined as OAHI ≥5 to <15 events/h, moderate OSA as OAHI ≥15 to <30 events/h, and severe OSA as OAHI ≥30 events/h.^30,31 The central apnea-hypopnea index was defined as the number of central apneas and central hypopneas per hour during sleep. Central sleep apnea was defined as a central apnea-hypopnea index ≥5 events/h.^32,33 The supine OAHI was defined as the total number obstructive apneas, mixed apneas, and obstructive hypopnea events while in a supine position divided by the total supine sleep time. Similarly, the nonsupine OAHI was defined as the total count of total number obstructive apneas, mixed apneas, and obstructive hypopnea events in a nonsupine position divided by the total nonsupine sleep time. POSA was defined as an overall OAHI ≥5 events/h and a supine OAHI to nonsupine OAHI ratio of ≥2.¹⁷ In addition, a minimum of 20 minutes spent sleeping in a supine and nonsupine position was required for each participant.³⁴

Data were analyzed with SPSS version 24.0 (IBM Corporation, Armonk, NY). Descriptive statistics, including frequencies of and percentages for categorical variables and means (±SDs) for continuous variables, were computed for all demographics and PSG parameters. Categorical variables were analyzed using the χ² test. Continuous variables were analyzed using either independent-samples t tests or Mann-Whitney U test according to data distribution. A P value ≤.05 determined significance.

RESULTS

A total of 112 children with obesity were recruited and underwent a PSG. Of the 112 children included in the study, 43 of 112 (38%) were diagnosed with OSA and 69 of 112 (62%) did not have OSA. Among those with OSA, 25 of 43 (58%) were defined as having POSA and 18 of 43 (42%) had non-POSA (Table 1). The mean (±SD) age of those with POSA and without POSA was 14.6 ± 2.3 and 13.9 ± 2.8 years, respectively. In the POSA group, 22 (88%) were aged 12 years and older. Seventeen (68%) with POSA and 14 (78%) without POSA were male. There were no significant differences between those with and without POSA with respect to age, sex, height, weight, body mass index, neck, waist and hip circumferences, snoring, or adenotonsillar hypertrophy. In the POSA and non-POSA groups, 8 of 25 (32%) and 6 of 18 (33%) had a previous adenotonsillectomy, respectively. Among those with POSA, 13 of 25 (52%) had mild OSA, 7 of 25 (28%) had moderate OSA, and 5 of 25 (20%) had severe OSA (Figure 1).

Table 1.

Baseline characteristics for children with and without POSA.

	Non-POSA (n = 18)	POSA (n = 25)	P
Age, years	13.9 ± 2.8	14.6 ± 2.3	.41
Male sex, n (%)	14 (78)	17 (68)	.48
Height, cm	162.5 ± 14.7	167.5 ± 11.2	.23
Weight, kg	102.6 ± 32.4	106.7 ± 27.2	.67
BMI z score	2.6 ± 0.3	2.4 ± 0.3	.26
Neck circumference, cm	39.0 ± 6.9	38.9 ± 6.7	.98
Waist circumference, cm	110.3 ± 16.0	112.9 ± 18.5	.64
Hip circumference, cm	117.5 ± 15.9	120.5 ± 15.9	.56
Waist to height ratio	0.68 ± 0.07	0.67 ± 0.09	.80
Neck to waist ratio	0.35 ± 0.04	0.35 ± 0.03	.75
Neck to height ratio	0.24 ± 0.03	0.23 ± 0.03	.44
Snoring, n (%)	15 (83)	23 (92)	.38
Tonsillar hypertrophy, n (%)	3 (17)	10 (40)	.22
Adenoid hypertrophy, n (%)	11 (61)	13 (52)	.09
Previous adenotonsillectomy, n (%)	6 (33)	8 (32)	.93

Data are presented as mean ± SD unless otherwise indicated. P values were calculated using χ² analysis for categorical variables and independent-samples t tests adjusted for unequal variance for continuous variables. BMI = body mass index; POSA = positional obstructive sleep apnea.

Figure 1. Relationship between obstructive sleep apnea severity and positional sleep apnea.

POSA = positional obstructive sleep apnea.

PSG parameters were compared between those with and without POSA (Table 2). The median (interquartile range) overall OAHI for the POSA and non-POSA groups were 11.3 (7.6–21.6) and 12.0 (9.0–26.7) events/h, respectively. No significant difference was found across groups for overall OAHI. Similarly, OAHI values during non-REM (NREM) and REM sleep were compared between both groups. For the POSA and non-POSA groups, median OAHI values during NREM sleep were 7.9 (5.5–18.5) and 10.8 (6.9–18.8) events/h, respectively. Median OAHI values during REM sleep were 21.8 (6.1–32.1) events/h for the POSA group and 23.6 (11.8–45.5) events/h for the non-POSA group. No significant differences were found for OAHI during NREM sleep and REM sleep between the groups. Further, for the POSA group, the median supine and nonsupine OAHI values were 22.0 (11.2–40.7) and 3.9 (2.3–8.5) events/h, respectively. For the non-POSA group, the median supine and nonsupine OAHI values were 11.5 (6.3–25.4) and 12.1 (8.8–32.3) events/h, respectively. Compared with the non-POSA group, the supine OAHI was significantly higher while the nonsupine OAHI was significantly lower in the POSA group. Of those in the non-POSA and POSA groups, 2 of 18 (11%) and 15 of 25 (60%) had a nonsupine OAHI below the threshold requiring CPAP treatment (eg, ≥5 events/h), respectively. Further, 5 of 25 (20%) in the POSA group had exclusively supine OSA as defined by a nonsupine OAHI below 2 events/h. Specifically, during REM sleep, supine OAHI and nonsupine OAHI did not significantly differ between both groups. However, compared with the non-POSA group, the POSA group showed significantly higher supine OAHI (20.4 [9.3–37.3] vs 6.1 [.0–14.2] events/h) and lower nonsupine OAHI (2.4 [.0–6.0] vs 10.5 [2.4–24.5] events/h) during NREM sleep. Among those with POSA, no significant differences were found between supine REM OAHI and supine NREM OAHI. The time spent sleeping in a supine and nonsupine position was also compared between the POSA and non-POSA groups. While the POSA group spent more time in a nonsupine position (177.5 [92.8–251.8] minutes) compared with the non-POSA group (144.5 [73.5–208.9] minutes), the difference was not statistically significant. No significant differences were also found with the time spent in supine position between the non-POSA and POSA groups (237.5 [110.3–279.3] vs 155.5 [67.5–252.3] minutes). Similarly, nonsignificant results were observed with time spent sleeping in a supine and nonsupine position during NREM or REM sleep between both groups. Additionally, mean oxygen saturation was significantly lower in the non-POSA group (96.6 [96.0–97.3] vs 97.5 [96.6–98.1]).

Table 2.

Polysomnography data for children with and without POSA.

	Non-POSA (n = 18)	POSA (n = 25)	P
Total sleep time, minutes	367.3 (325.8–429.9)	346.5 (311.5–388.3)	.07
Sleep-onset latency, minutes	24.1 (6.4–49.5)	15.0 (5.6–47.5)	.35
REM latency, minutes	117.3 (71.9–239.5)	132.0 (95.8–219.2)	.40
Sleep efficiency, %	85.5 (72.4–92.4)	77.7 (72.4–89.1)	.40
Wake after sleep onset, minutes	34.0 (15.5–64.9)	45.0 (21.0–97.0)	.20
Arousal index, events/h	15.6 (9.6–19.5)	15.5 (10.3–22.8)	.66
REM sleep, %	18.1 (11.7–26.1)	16.6 (10.8–20.0)	.25
NREM-1 sleep, %	10.8 (5.4–12.7)	8.2 (4.4–13.1)	.42
NREM-2 sleep, %	46.3 (42.1–52.6)	51.6 (44.2–55.2)	.20
NREM-3 sleep, %	22.0 (17.7–33.5)	23.4 (18.3–29.5)	.98
Obstructive AHI NREM, events/h	10.8 (6.9–18.8)	7.9 (5.5–18.5)	.65
Obstructive AHI REM, events/h	23.6 (11.8–45.5)	21.8 (6.1–32.1)	.17
Obstructive AHI, events/h	12.0 (9.0–26.7)	11.3 (7.6–21.6)	.28
Supine AHI, events/h	11.5 (6.3–25.4)	22.0 (11.2–40.7)	.02a
Nonsupine AHI, events/h	12.1 (8.8–32.3)	3.9 (2.3–8.5)	<.001a
NREM supine AHI, events/h	6.1 (0.0–14.2)	20.4 (9.3–37.3)	.01a
NREM nonsupine AHI, events/h	10.5 (2.4–24.5)	2.4 (0.0–6.0)	.02a
REM supine AHI, events/h	24.0 (2.9–60.5)	23.8 (0.0–44.8)	.27
REM nonsupine AHI, events/h	3.2 (0.0–24.2)	4.1 (0.0–9.0)	.81
Total time spent in supine, minutes	237.5 (110.3–279.3)	155.5 (67.5–252.3)	.25
Total time spent in nonsupine, minutes	144.5 (73.5–208.9)	177.5 (92.8–251.8)	.48
Total time spent in supine during NREM sleep, minutes	199.5 (2.5–238.0)	95.5 (51.0–227.0)	.42
Total time spent in nonsupine during NREM sleep, minutes	108.0 (38.8–140.3)	117.0 (55.0–216.0)	.55
Total time spent in supine during REM sleep, minutes	29.3 (1.1–43.4)	23.5 (2.5–44.5)	.84
Total time spent in nonsupine during REM sleep, minutes	25.0 (1.1–89.0)	23.5 (3.5–61.5)	.70
Mean SaO2, %	96.6 (96.0–97.3)	97.5 (96.6–98.1)	.02a
Nadir SaO2, %	86.7 (78.8–89.8)	87.0 (83.5–90.6)	.68
Desaturation index, events/h	11.1 (5.4–20.1)	7.0 (4.0–12.2)	.17

Data are presented as medians (25th–75th percentile) unless otherwise indicated. P values were calculated using Mann-Whitney U independent-samples tests adjusted for unequal variance. AHI = apnea-hypopnea index; NREM = non–rapid eye movement; POSA = positional obstructive sleep apnea; REM = rapid eye movement; SaO₂ = oxygen saturation. ^aP ≤ .05 represents a statistically significant difference between both groups.

DISCUSSION

The principal finding of our study was that 58% of obese children with no other known significant comorbidities had POSA. This is similar to previous reports of adult populations where 55% of adults with OSA were diagnosed with POSA.¹⁵ Few studies have assessed the prevalence of POSA among children with different clinical phenotypes. POSA has been reported to be less prevalent among nonobese children and those with medical comorbidities such as Down syndrome, ranging from 9% to 31%.^19,35 More recently, a study by Verhelst and colleagues³⁶ assessed POSA among a large heterogenous cohort of otherwise healthy children and children without medical comorbidities, where 19% of the population had POSA. Their study population also included a group of children with obesity, where 31% had POSA. Although the prevalence of POSA was lower compared with that in our study, only 27% of their entire study population were obese. Furthermore, a majority of their population were nonsyndromic, nonobese children who were younger, ranging between 6 and 12 years. In our study, POSA was more common among older children, where 88% of the POSA group was aged 12 years and older. Previous findings have shown that supine OAHI can be higher among older children, thus increasing the risk of POSA with age³⁷ and resulting in a higher prevalence among obese youth.

Within our study cohort, we found that POSA was common among those with less severe OSA. A similar trend has been reported with adult populations, where POSA is more prevalent among those diagnosed with milder OSA.²⁰ Current treatment for milder forms of OSA include intranasal corticosteroids and montelukast, which have been found to show only modest improvement in reducing OAHI among 50–60% of younger children.^38,39 Considering strategies targeted toward treating POSA may be favorable among those with milder OSA, especially for those who prefer less intrusive and nonpharmacologic treatment. This is important as these children may be at particular risk of worsened OSA symptoms if not properly treated.³⁹

Similarly, treating POSA may be effective among those with more severe OSA, where we found POSA among 28% and 20% of children with moderate and severe OSA, respectively. Moderate–severe OSA tends to be treated with an adenotonsillectomy, lifestyle modifications including exercise and dietary restrictions,⁷ and bariatric surgery for those who are morbidly obese.⁸ However, among children with obesity, the risk of persistent OSA is high following an adenotonsillectomy while weight loss through diet and exercise is poorly maintained.⁹ Bariatric surgery has been shown to significantly improve OSA severity but some children can still experience residual OSA following surgery.^8,40 Moreover, surgical interventions are considered with caution in children with obesity and OSA as they are particularly vulnerable to surgical complications such as respiratory depression and additional cardiopulmonary problems.¹⁰ Therefore, CPAP therapy is generally prescribed for the treatment of moderate–severe OSA; however, more than 50% of youth with obesity poorly tolerate CPAP therapy, resulting in suboptimal adherence rates ranging between 3 and 4 hours of CPAP usage per night.⁴¹ Since children with obesity poorly tolerate CPAP therapy, they may be at risk of worsened OSA symptoms as well as neurocognitive, behavioral, cardiovascular, and metabolic consequences.⁴² As such, evaluating the efficacy of personalized alternative therapies could be beneficial. For example, positional therapy, where various devices or appliances are used to prevent sleeping in a supine position,¹⁶ may be a more effective, noninvasive treatment strategy given the increased prevalence of POSA in this population. The use of positional therapy has been assessed in adults with POSA and has shown to be a simple, effective, and inexpensive treatment option.²⁷ Of importance, adults have demonstrated increased short- and long-term adherence to positional therapy compared with CPAP therapy.^43,44

As expected, we found a significant difference in supine OAHI and nonsupine OAHI among children with obesity. Similar trends were observed by Verhelst and colleagues³⁶ among children with and without POSA. However, the difference in supine OAHI in their study was smaller between groups, which could be attributed to the heterogeneity of their study cohort. Previous research has also shown that the presence of obesity in children with OSA can increase the effect of sleeping in a supine position on respiratory disturbance, resulting in a higher supine OAHI.¹⁸ A possible pathophysiological mechanism associated with POSA could involve significant increases in loop gain accompanied by decreases in functional residual capacity while sleeping in a supine position, resulting in overall ventilatory instability.⁴⁵ Therefore, patients with the POSA phenotype demonstrate a greater propensity for more frequent respiratory events. Obesity further increases the risk of airway obstruction among those with POSA due to the presence of increased fat deposits along the pharyngeal wall, leading to a narrowed airway.²⁵ Such findings could explain the increased prevalence of POSA observed within an obese pediatric population as compared with children who are healthy or diagnosed with other comorbidities. As such, implementing therapies targeted toward preventing supine sleep such as positional therapy may be beneficial in alleviating airway obstruction and reducing the severity of OSA among those who are obese. Additionally, because OSA tends to worsen during REM sleep, POSA may be a REM-specific phenomenon, where supine OAHI is greater during REM sleep as compared with NREM sleep.³⁶ In our study, between the POSA and non-POSA groups, no significant differences were found in supine and nonsupine OAHI values during REM sleep as well as the time spent sleeping in a supine and nonsupine position during REM sleep, which may suggest that POSA is not specifically related to REM sleep in children with obesity. Such findings are similar to those found in adults, where POSA does not worsen during REM sleep.

There have been a limited number of pediatric studies that have identified risk factors for POSA. For example, a retrospective review of Asian adults with POSA by Chung and colleagues⁴⁶ found that OAHI severity was significantly associated with increasing age and percentage of time spent snoring during sleep. Other studies with adults have found an opposing relationship, where POSA was associated with younger age among Western populations.^47,48 Additional studies involving Asian adult populations have found no association between age and POSA.^49,50 Such a discrepancy in findings may be explained by the influence of racial differences on the relationship between age and POSA.⁴⁶ In our study, we did not identify any significant risk factors for POSA in a population of obese children including age and body mass index, but further research may be beneficial in determining whether racial differences could influence the occurrence and severity of POSA.

Our study has several limitations. First, although participants were prospectively enrolled, the study population consisted of children with obesity who were primarily referred for a history of snoring and screening of sleep-related breathing disorders; thus, this study may represent a highly biased population and may not represent the general population of children with obesity. Second, the sample size of our study groups was small, resulting in limited subgroup analyses. Third, our analyses were not adjusted for multiple comparisons in order to avoid false-negative results. Our study also did not assess airway anatomy or the presence of pharyngeal overcrowding among those with POSA. Thus, further larger-scale studies are required to confirm our findings. Additionally, our PSG data reflect sleep during a single night outside of the home environment. As such, we were unable to consider the night-to-night variability in OAHI and sleeping position in patients with OSA. Similarly, sleeping while attached to PSG equipment may predispose an individual to sleep in a particular position. As a result, the severity of POSA may either be under- or overestimated. Nonetheless, our study adds to the literature by determining a high prevalence of POSA in children with obesity, especially among those with milder forms of OSA.

CONCLUSION

Among children with obesity and OSA, POSA was found in 58% of the cohort. Given the high prevalence of POSA, a personalized approach for alternative therapies should be considered. This is particularly important in youth with obesity as these individuals are more likely to become adults with coexisting OSA and obesity, who are at high risk of developing related comorbidities. The use of positional devices has been investigated in adults with OSA, where such devices have been proven to be an effective, cost-efficient, and noninvasive treatment option for POSA. Future studies are required to identify the clinical phenotypes of youth with obesity and POSA who will be adherent to positional therapy and who are most likely to benefit from positional therapy.

DISCLOSURE STATEMENT

All authors have reviewed and approved this manuscript. Work for this study was performed at the Hospital for Sick Children, Toronto, Canada. The authors report no conflicts of interest.

ABBREVIATIONS

CPAP

continuous positive airway pressure

NREM

non–rapid eye movement

OAHI

obstructive apnea-hypopnea index

OSA

obstructive sleep apnea

POSA

positional obstructive sleep apnea

PSG

polysomnogram

REM

rapid eye movement

SaO₂

oxygen saturation