Polysomnographic Phenotype of Positional Obstructive Sleep Apnea

Article information

J Rhinol. 2024;31(3):168-175

Publication date (electronic) : 2024 November 30

doi : https://doi.org/10.18787/jr.2024.00039

Jae-Seon Park, MD¹, Young Bok Kim, MD, PhD², Il Seok Park, MD, PhD², Sun A Han, MD², Sung Hun Kang, PhD¹, Kyung Chul Lee, MD, PhD¹, Seok Jin Hong, MD, PhD^,1

¹Department of Otorhinolaryngology-Head and Neck Surgery, Kangbuk Samsung Hospital, Sungkyunkwan University School of Medicine, Seoul, Republic of Korea

²Department of Otorhinolaryngology-Head and Neck Surgery, Dongtan Sacred Heart Hospital, Hallym University College of Medicine, Seoul, Republic of Korea

Address for correspondence: Seok Jin Hong, MD, PhD, Department of Otorhinolaryngology- Head and Neck Surgery, Kangbuk Samsung Hospital, Sungkyunkwan University School of Medicine, 29 Saemunan-ro, Jongno-gu, Seoul 03181, Republic of Korea Tel: +82-2-2001-2264, Fax: +82-2-2001-2273, E-mail: enthsj@hanmail.net

Received 2024 November 7; Revised 2024 November 21; Accepted 2024 November 22.

Abstract

INTRODUCTION

Obstructive sleep apnea (OSA) is a disorder characterized by recurrent episodes of complete or partial upper airway obstruction, involving the nasal cavity, pharynx, base of the tongue, and larynx, during sleep [1]. This condition can lead to symptoms such as snoring, daytime sleepiness, impaired concentration, memory impairment, and fatigue, and has been associated with serious complications, including systemic hypertension, cardiovascular disease, and metabolic disorders [2-5].

Factors contributing to OSA include the craniofacial structure, soft tissue hypertrophy, sleeping posture, age, male sex, nasal obstruction, and pharyngeal fat deposition [6,7]. Both anatomical factors, such as bony structure and soft tissue abnormalities, and physiological factors, such as reduced muscle tone, influence the severity of OSA [6,7]. Sleep position plays a significant role, particularly the supine position, which can cause the tongue to shift posteriorly, resulting in mucosal edema and more frequent apneic episodes [8]. Laboratory assessments of sleep posture may overestimate or underestimate OSA severity compared to typical home sleeping positions, thus emphasizing the importance of clinical phenotyping [8].

Patients with positional OSA (POSA) are defined as those whose supine respiratory disturbance index (RDI) or apneahypopnea index (AHI) is at least twice as high as their lateral RDI or AHI. Conversely, non-positional patients have a supine RDI or AHI that is less than twice their lateral RDI or AHI [9,10].

This study investigated the positional characteristics of patients with OSA and analyzed the impact of supine versus lateral sleeping positions on OSA symptoms.

METHODS

This retrospective study analyzed patients who visited Hallym University Dongtan Sacred Heart Hospital from January 1, 2018, to December 31, 2021, with complaints of snoring or sleep apnea. All participants underwent type 1 PSG to assess sleep-related parameters. We analyzed data from medical records and test results to evaluate clinical and PSG characteristics. Patients with severe musculoskeletal, craniofacial, or psychiatric disorders; those under 13 years of age; or those with a total sleep time (TST) of less than 4 hours were excluded. The study protocol received approval from the Institutional Review Board of Hallym University Dongtan Sacred Heart Hospital (IRB No. 2021-09-002).

Demographic and clinical data collected from the patients’ medical records included age, sex, height, weight, and body mass index (BMI). Polysomnographic data, such as TST, sleep efficiency, AHI, RDI, REM sleep percentage, arousal index (AI), nadir peripheral oxygen saturation (SpO2), oxygen desaturation index (ODI), and time spent in supine and lateral positions, were also recorded. Additionally, positional AHI values for supine and lateral positions were documented. Patients were classified into three OSA severity groups: mild (5≤AHI<15), moderate (15≤AHI<30), and severe (AHI≥30). POSA was defined as a supine AHI that was at least twice as high as the lateral AHI, with an overall AHI of 5 or greater.

Statistical analyses were conducted using SPSS version 24.0 (IBM Corp., Armonk, NY, USA). Categorical variables were analyzed using the chi-square test. The t-test was utilized to compare PSG parameters and positional data between OSA and non-OSA groups, as well as between the POSA and non-POSA groups. Analysis of variance was used to compare different OSA severity groups. Statistical significance was established at p<0.05, with two-tailed p-values.

RESULTS

Our study included a total of 556 participants, with 478 men (86.0%) and 78 women (14.0%). The average age was 44.8±12.0 years, and the average BMI was 26.8±3.8 kg/m². Participants had an average TST of 349.6±32.0 minutes and a sleep efficiency of 87.9%±10.2%. The average AHI was 28.0±23.4/h. Regarding sleep positions, participants spent 61.7%±26.2% of their sleep time in the supine position and 37.9%±25.8% in the lateral decubitus position (Fig. 1). The average supine AHI was 37.9±27.3/h, compared to a lateral AHI of 12.1±20.8/h. The non-OSA group had a significantly higher proportion of females (30.4% vs. 12.2%; p<0.001). The average age was significantly higher in the OSA group (45.9±11.7 years) than in the non-OSA group (35.5±10.1 years) (p<0.001). The average BMI was also significantly higher in the OSA group (27.0±3.8 kg/m²) than in the non-OSA group (25.2±3.8 kg/m²) (p=0.001). Additionally, the OSA group had a significantly lower percentage of REM sleep (12.9%±5.2% vs. 14.6%± 4.7%; p=0.017) and a significantly higher AI (34.0±20.1/h vs. 12.6±5.9/h; p<0.001) than the non-OSA group. No significant differences were observed in the distribution of sleep positions between the OSA and non-OSA groups (Table 1).

Fig. 1.

Proportion of sleep time spent in different positions.

Table 1.

Comparison of polysomnography and positional data among all participants, the OSA group, and the non-OSA group

Among the 500 participants with OSA, 317 (63.4%) were classified as having POSA and 183 as non-POSA. Women were significantly more prevalent in the POSA group, comprising 15.1% compared to 7.1% in the non-positional group (p=0.008). The POSA group also had a lower BMI, averaging 26.3±3.4 kg/m² versus 28.1±4.1 kg/m² in the non-positional group (p<0.001), and a longer average TST of 353.7±31.2 minutes compared to 340.1±30.5 minutes (p<0.001). The average AHI was significantly lower in the POSA group at 23.3±15.3/h, compared to 43.9±27.9/h in the non-positional group (p<0.001). Additionally, the positional group had a higher average percentage of REM sleep (13.4%±4.9% vs. 12.0%±5.7%; p=0.005) and a higher nadir SpO2 (82.8%±6.6% vs. 77.1%±9.8%; p<0.001). Patients with POSA also exhibited significantly lower AI values (27.4±13.3/h vs. 45.4±24.4/h; p<0.001) and ODI values (19.3±15.7/h vs. 38.3±27.9/h; p<0.001).

Furthermore, the POSA group spent less time in the supine position (56.0%±22.9 vs. 72.5%±28.5; p<0.001) and more time in the lateral decubitus position (43.8%±22.7% vs. 27.2%±28.2%; p<0.001). The supine AHI was significantly lower in the POSA group (38.6±23.3/h vs. 47.1±29.9/h; p=0.001), as was the lateral AHI (5.9±8.2/h vs. 26.4±30.0/h; p<0.001) (Table 2).

Table 2.

Comparison of polysomnography and positional data between the positional and non-positional OSA groups

Among all participants, 56 (10.1%) were classified as nonOSA, 166 (29.8%) as mild OSA, 125 (22.5%) as moderate OSA, and 209 (37.6%) as severe OSA (Fig. 2). Significant differences were observed across all parameters among the mild, moderate, and severe OSA groups, with the exception of sleep efficiency. The proportion of females decreased with increasing severity: 19.9% in mild, 12.0% in moderate, and 6.2% in severe OSA groups (p<0.001). BMI showed a significant increase across the groups, with averages of 25.6±3.2 kg/m² in mild, 26.5±3.6 kg/m² in moderate, and 28.3±3.9 kg/m² in severe OSA (p<0.001). TST decreased with severity, from 357.8±31.1 min in mild to 353.5±32.2 min in moderate, and 338.6±28.7 min in severe OSA (p<0.001). AHI was 9.5±2.9/h in mild, 21.3±4.4/h in moderate, and 53.4±17.9/h in severe OSA (p<0.001). REM sleep percentage also decreased with severity, from 14.3%±5.1% in mild to 13.2%±5.0% in moderate, and 11.5%±5.1% in severe OSA (p<0.001). Regarding sleep position, the supine percentage was 60.2% in mild, 57.3% in moderate, and 66.3% in severe OSA (p=0.006), while the lateral decubitus position percentage was 39.4% in mild, 42.3% in moderate, and 33.5% in severe OSA (p=0.007). POSA was most common in mild cases (76.5%) and declined with severity (72.8% in moderate and 47.4% in severe cases; p<0.001) (Table 3).

Fig. 2.

Distribution of participants across AHI severity groups. AHI, apnea–hypopnea index.

Table 3.

Comparison of polysomnography and positional data by OSA severity

Table 4 presents a comparative analysis of the POSA and non-POSA groups across mild, moderate, and severe categories. The POSA group showed a significantly higher proportion of time spent in the lateral decubitus position and a significantly lower lateral AHI compared to the non-positional group at all severity levels. In mild POSA, the supine AHI was significantly higher at 18.9±11.9/h compared to 9.5±3.7/h in the non-positional group (p<0.001). Similarly, in moderate POSA, the supine AHI was higher (40.0±14.6/h vs. 22.0±6.7/h, p<0.001). However, in severe POSA, the supine AHI was significantly lower (62.6±16.9/h vs. 68.1±18.1/h, p=0.025). Among mild OSA patients, the average age was significantly higher in the POSA group (45.6±11.7 years) compared to the nonPOSA group (38.7±12.3 years, p=0.002), with no notable differences in other variables. In the moderate POSA group, there were no significant differences in variables other than the positional data. For severe POSA, TST (346.9±28.1 min vs. 331.2±27.3 min, p<0.001), REM sleep percentage (12.8%±5.1% vs. 10.4%±4.8%, p=0.001), and nadir SpO2 (78.2%±6.3% vs. 72.3%±9.2%, p<0.001) were significantly higher. Conversely, BMI (27.2±3.4 vs. 29.2±4.0, p<0.001), AHI (42.5±10.3/h vs. 63.3±17.6/h, p<0.001), AI (43.1±9.1/h vs. 61.7±16.7/h, p<0.001), and ODI (37.5±12.8/h vs. 56.5±20.6/h, p<0.001) were lower.

Table 4.

Comparison of polysomnography and positional data according to OSA severity and the presence of positional OSA

DISCUSSION

OSA is a complex disease, and enhanced phenotyping could enable personalized treatments that improve patient symptoms [11]. Phenotyping of OSA, based on pathophysiological factors such as anatomical predisposition, arousal threshold, loop gain, and upper airway muscle responsiveness, has been well-developed [11,12]. Sleep position might represent another phenotype of OSA, which could lead to tailored approaches for patients based on their clinical characteristics.

The morbidity rate of POSA in Korea has been reported to approach 76.8% [13], and in 1984, Cartwright [14] reported a rate of 58.3%. A recent study from the Multi-Ethnic Study of Atherosclerosis found that 63.9% of patients were supine-positional [15]. In this study, 317 out of 500 patients (63.4%) were diagnosed with POSA, aligning with previous reports from Korea. The BMI and AHI were significantly higher in the non-POSA group. Conversely, patients with POSA exhibited significantly lower ODI and AI values than those without POSA.

In our study, the prevalence of POSA varied according to the severity of OSA: 76.5% in mild cases, 72.8% in moderate cases, and only 47.4% in severe cases exhibited positional dependency. This suggests that positional dependency is more common in mild-to-moderate OSA, making positional therapy a viable treatment option for these patients, as demonstrated in our research. Conversely, the absence of positional dependency occurs relatively frequently in patients with severe OSA, where sleeping on the lateral side proved ineffective. These findings indicate that the effectiveness of positional therapy in improving sleep apnea is not universal and must be customized for each patient. Therefore, it is crucial to confirm POSA through PSG before recommending positional therapy.

Our study revealed that women are more frequently diagnosed with mild OSA and POSA compared to men. Men possess a lower laryngeal position, a longer upper airway, and a larger cross-sectional area, factors that contribute to greater airway collapsibility and a higher susceptibility to OSA [16]. Additionally, previous studies that compared drug-induced sleep endoscopy findings based on POSA status indicated that both POSA and non-POSA groups experienced worsening velum anteroposterior and tongue obstruction in the supine position. However, the non-POSA group also showed increased velum lateral and oropharyngeal lateral obstruction in the lateral position [17]. These findings suggest that anatomical differences may result in lower airway collapsibility in women, potentially explaining their lower OSA severity and higher prevalence of POSA. Further research is needed to investigate this hypothesis directly.

In our study, patients with severe POSA exhibited milder values across most PSG parameters (TST, AHI, RDI, REM sleep time, AI, nadir SpO2, ODI, supine AHI, and lateral AHI) compared to those without POSA. Additionally, both weight and BMI were lower in patients with severe POSA. However, when comparing mild and moderate POSA cases to non-POSA cases at similar AHI levels, no significant differences were observed in most PSG parameters. Therefore, the differences noted in PSG parameters between the severe POSA and nonPOSA groups may primarily be attributed to the lower mean AHI in the severe POSA group, suggesting a generally lower average disease severity. This disparity in severity, rather than other factors, likely explains the significant differences in TST, AI, ODI, nadir SpO2, and other PSG metrics.

The supine AHI was found to be higher in the mild and moderate POSA groups compared to the non-POSA group. This indicates that relying solely on the total AHI to assess severity in patients with mild and moderate POSA might result in an underestimation of their condition. Additionally, the reduced proportion of REM sleep observed in patients with mild POSA compared to those without POSA could partially explain this underestimation. Moreover, the total AHI can vary with changes in the proportion of sleep spent in different positions. These findings suggest that positional therapy could be a viable treatment option for patients with mild or moderate POSA.

Mandibular advancement devices (MADs) are an effective alternative to positive airway pressure therapy for OSA, particularly in patients with POSA [18]. This may be attributed to the cephalometric characteristics of patients with POSA, such as larger posterior airway space, shorter soft palate, and wider lateral airway space [18]. Furthermore, an autotitrating MAD (AMAD) has been recently developed. This device automatically adjusts the length of mandibular advancement based on the patient’s sleep position, and its clinical efficacy and safety have been confirmed [19]. In our study, patients with POSA spent 43.8% of their sleep time in the lateral position, during which the mean AHI was 5.9, significantly reducing the need for mandibular advancement. This finding suggests that AMAD could be a valuable treatment option for patients with POSA, potentially minimizing side effects.

This study had several limitations. First, we were unable to conduct a comparative analysis of subjective symptoms, anatomical findings, and cephalometric parameters based on the POSA status, which underscores the need for further research in this area. Additionally, due to subgrouping, the mild and moderate non-POSA groups each had sample sizes of approximately 30 patients. This small sample size may have limited the statistical significance of the comparisons between variables; thus, future studies should include larger patient cohorts. Finally, relying on a single PSG may not accurately reflect a patient’s typical positional profile, emphasizing the need for future studies to incorporate multiple PSG sessions to derive more comprehensive insights.

In conclusion, this study found that patients with POSA typically displayed less severe symptoms and more favorable PSG profiles compared to those without POSA. POSA was primarily seen in mild to moderate cases, while non-positional dependency occurred more commonly in severe cases. Identifying POSA through PSG offers crucial insights that can guide and enhance treatment approaches.

Article information Continued