Mo J, Lee CH, Rhee C, Yoon I, Kim J. Positional Dependency in Asian Patients With Obstructive Sleep Apnea and Its Implication for Hypertension. Arch Otolaryngol Head Neck Surg. 2011;137(8):786-790. doi:10.1001/archoto.2011.122
Author Affiliations: Department of Otorhinolaryngology, Dankook University College of Medicine, Cheonan, South Korea (Dr Mo); Departments of Otorhinolaryngology (Drs Lee, Rhee, and Kim) and Psychiatry (Dr Yoon), Seoul National University College of Medicine, Seoul, South Korea; Departments of Otorhinolaryngology (Drs Lee, Rhee, and Kim) and Psychiatry (Dr Yoon), National University Bundang Hospital, Seongnam, South Korea; and Sensory Organ Research Institute, Seoul National University Medical Research Center, Seoul (Drs Lee, Rhee, and Kim).
Objectives To investigate the relationship of obstructive sleep apnea (OSA) with positional dependency and to identify its clinical implication in an Asian population.
Design Retrospective analysis.
Setting Academic tertiary referral center.
Patients A total of 1170 adults (1003 men and 167 women; mean [SD] age, 50.8 [12.9] years) with OSA were included from February 1, 2004, through October 31, 2008.
Intervention and Main Outcome Measures All patients underwent full-night polysomnography. The anthropometric or polysomnographic variables between the patients with positional OSA (PPs) and those with nonpositional OSA (NPPs) were characterized, and multivariate analysis was performed to find the determining factors of positional dependency. The prevalence of hypertension was also investigated.
Results Nearly 75% of the patients (874 [74.7%]) had positional dependency. Positional dependency was present in 87.0% of the patients with mild OSA (apnea hypopnea index [AHI], ≥5 but <20), in 84.2% of those with moderate OSA (20 ≤ AHI < 40), and in 43.1% of those with severe OSA (AHI ≥ 40). The prevalence of PPs was 46.4% among severely obese patients (body mass index [BMI], ≥30, calculated as weight in kilograms divided by height in meters squared) and 82.7% among the nonobese patients (BMI < 25) and 74.6% among obese patients (25 ≤ BMI <30). Multivariate analysis showed that the AHI was the most dominant variable that determined positional dependency, followed by the BMI. In the PP group, the percentages of deep sleep and rapid eye movement sleep were significantly greater compared with those in the NPP group. The Epworth Sleepiness Scale score was lower in the PP group. The prevalence of hypertension was 34.4% and 49.7% in the PP and NPP groups, respectively.
Conclusions This study demonstrates that the prevalence of PPs among Asians is almost three-fourths of the patients and that the AHI is the most dominant factor for determining positional dependency, followed by BMI. The PP group had lower BMI, a lower AHI, longer deep sleep, longer rapid eye movement sleep, and less daytime sleepiness than did the NPPs. The prevalence of hypertension was also affected by positional dependency.
The severity of obstructive sleep apnea (OSA) can be affected by many factors, such as obesity, craniofacial characteristics, sex, age, and sleep position. Among those, sleep position significantly influences respiratory disturbance. The effect of sleep position on sleep apnea was first described by Cartwright1 and Lloyd and Cartwright2 in the 1980s.
Positional dependency has been studied by various authors, mostly in Western countries,3- 8 and the prevalence of patients with positional apnea has varied according to each article. The reason for a diverse prevalence might be the small number of patients in each study. Oksenberg et al9 recently reported that 53.8% of 2077 patients in a population-based study had positional-dependent OSA.
There have been only a few studies on positional dependency in Asian countries.10,11 It is well known that Asians have different craniofacial characteristics and more severe OSA than whites when matched for age, sex, and body mass index (BMI).12,13 Because positional dependency is influenced by the severity of OSA and BMI,7 the characteristics of positional dependency for Asians could be different from those of whites. To our knowledge, there have been no large population studies on positional dependency in Asian countries.
We conducted an analysis of a large number of Asians with OSA and determined the prevalence of positional dependency in various subgroups of patients with OSA. In addition, the anthropometric and polysomnographic differences between patients with positional OSA (PPs) and those with nonpositional OSA (NPPs) were analyzed, and the possible determining factors that affect positional dependency were investigated.
Obstructive sleep apnea and hypertension frequently coexist. There is growing evidence that OSA contributes to the genesis of hypertension through several mechanisms,14 and the prevalence of hypertension is strikingly high among patients with OSA (approximately 50%).15 Hence, we also investigated the prevalence of hypertension in patients with OSA and the association among OSA, hypertension, and positional dependency in the Asian population.
The patients were all referred to our Department of Otorhinolaryngology in an academic tertiary referral center from February 1, 2004, through October 31, 2008, for treatment of snoring and/or witnessed apnea. A total of 1916 adults who were older than 18 years underwent nocturnal in-laboratory full-night polysomnography. Among them, 1474 patients received a diagnosis of OSA (apnea hypopnea index[AHI] score, ≥5), and 314 patients who slept less than 5% of the time in either the supine or lateral position were excluded; this resulted in a final population of 1170 patients. This study was approved by the institutional review board of Seoul National University Bundang Hospital, and we obtained written informed consent from all participants.
All patients completed 2 questionnaires related to the Epworth Sleepiness Scale and the Pittsburgh Sleep Quality Index just before polysomnography. The full-night polysomnography was then performed (Embla N7000 system; Embla Systems, Reykjavik, Iceland) with standard electrodes and sensors. Electroencephalography electrodes were applied at C3-M2, O1-M2, C4-M1, and O2-M1, and 2 electro-oculography electrodes were applied at the lateral canthus of both eyes. Submental electromyography electrodes were applied at the submentalis muscle, and electromyography of both anterior tibialis muscles recorded limb movements during sleep. Strain gauges were used for assessing chest and abdominal respiratory movements. A nasal pressure cannula was used to record airflow, and a thermistor was applied to differentiate oral breathing from nasal breathing. Arterial oxygen saturation was measured using a pulse oximeter applied to the patient’s index finger. Based on the criteria of Rechtschaffen and Kales,7 every epoch of 30-second nocturnal polysomnography was scored.
Apnea was defined as cessation of airflow for at least 10 seconds. Hypopnea was defined as a substantial reduction in airflow (>50%) for at least 10 seconds or a moderate reduction in airflow for at least 10 seconds associated with electroencephalographic arousal or oxygen desaturation (≥4%).8 The AHI was defined as the total number of apnea and hypopnea events per hour of sleep, and the oxygen desaturation index was calculated as the number of oxygen desaturation episodes (≥4%) per hour of sleep. Obstructive sleep apnea was defined as an AHI of 5 or higher. The severity of OSA was categorized as mild (5 ≤ AHI < 20), moderate (20 ≤ AHI < 40), or severe (AHI ≥ 40).
Sleep efficiency was defined as the ratio of time spent asleep (total sleep time) to the amount of time spent in bed. Wake time after onset was defined as the amount of awake time after the first onset of sleep.
The sleeping position was analyzed by a position sensor placed on the chest, and this was confirmed by direct observation by a technician using a low-light camera. The AHI of the supine or lateral sleep position was recorded. The participants were classified as PP or NPP according to Cartwright’s criteria.1 Positional dependency of OSA was defined as an AHI that was at least twice as high in the supine position than in the lateral position.
Anthropometric and polysomnographic data were collected and fully analyzed. Blood pressure was checked at each visit to the sleep clinic; hypertension was defined as a systolic blood pressure value above 140 mm Hg or diastolic blood pressure value above 90 mm Hg occurring on more than 3 measurements. Participants who were taking antihypertensive medication were included in the hypertension group. All parametric results are expressed as mean (SD). The unpaired t test was used to compare mean values between the PPs and NPPs. The χ2 test was performed to evaluate the difference in the proportions of the values in each category. Multivariate logistic regression analysis was used to reveal the factors related to positional dependency. Statistical significance was assumed at P <.05 for all factors. Standard statistical methods were used (SPSS 12.0 for Windows; SPSS, Inc, Chicago, Illinois).
This study included 1003 men and 167 women with OSA. The mean AHI was 29.8 (21.6). The mean age and BMI (calculated as weight in kilograms divided by height in meters squared) were 50.8 (12.9) years and 25.9 (3.5), respectively.
Of the total patients, 874 belonged to the PP group (74.7%) and 296 patients belonged to the NPP group (25.3%). To investigate the prevalence of OSA with positional dependency, the patients with OSA were divided into subgroups according to their disease severity (AHI), BMI, and age (Table 1). According to the AHI, the prevalence of OSA with positional dependency was 87.0% and 84.2% in the mild and moderate groups, respectively. In contrast, the prevalence of OSA was 43.1% in the severe group, showing a significant difference compared with that of the mild and moderate groups (χ22 = 199.1, P < .001).
Patients were divided into 3 subgroups according to their BMI: nonobese (BMI <25), obese (25 ≤BMI <30), and severely obese (BMI ≥30). The prevalence of OSA with positional dependency was 82.7% and 74.6% in the nonobese and obese groups, respectively. However, in the severely obese group, the prevalence was significantly lower (46.4%), and there was a significant difference from that of the other 2 subgroups (χ23 = 70.8, P < .001). This finding revealed that obesity could decrease the prevalence of positional dependency.
Patients were categorized into 3 subgroups according to age: 20.0 to 39.9, 40.0 to 59.9, and 60.0 years or older. There was no significant difference in the prevalence of positional dependency among these subgroups (χ22 = 4.4, P = .11).
There was a significant difference in age between the PP group and the NPP group; patients in the PP group were approximately 1.8 years older than those in the NPP group (P = .03). There was no significant difference in height. The BMI was significantly lower in the PP vs NPP group (25.6 [3.1] vs 27.3 [3.9], P < .001). Body weight, neck circumference, waist circumference, and hip circumference were also less in the PP group, suggesting a relationship between obesity and positional dependency (Table 2).
The Epworth Sleepiness Scale is a measure of daytime sleepiness, and the score was significantly lower in the PP group (P = .005). However, there was no significant difference in the Pittsburgh Sleep Quality Index. The mean apnea index, the hypopnea index, and AHIs were significantly lower in the PP group (all P < .001, Table 1). The mean AHIs in the PP and NPP groups were 23.9 (15.7) and 47.4 (26.5), respectively. As for the other respiratory measures, the supine sleep AHI, the lateral sleep AHI, the rapid eye movement sleep AHI, and the non–rapid eye movement sleep AHIs were also significantly lower in the PP group (all P < .001, Table 2). Accordingly, the lowest oxygen saturation was significantly higher in the PP group (P < .001).
The PP group had significantly higher percentages of stages N2 and N3 sleep and rapid eye movement sleep, and a lower percentage of stage N1 sleep (P = .002 for rapid eye movement sleep and P < .001 for the others). The arousal index score was significantly lower in the PP group (P < .001). The total sleep time was not significantly different between the 2 groups (P = .52). Sleep latency, sleep efficiency, wake-time after sleep onset, percentage of supine sleep time, and percentage of lateral sleep time did not show significant differences between the groups (all P >> .05).
Both systolic and diastolic blood pressure was lower in the PP group compared with the NPP group (P = .005 and P < .001, respectively). The prevalence of hypertension was 38.3% in the whole population, and it was significantly higher in the NPP group compared with the PP group (49.7% vs 34.4%; χ21 = 21.68, P < .001). Because the higher AHI and BMI in the NPP group could explain the higher prevalence of hypertension, binary logistic regression analysis was performed to reveal the significance of these factors, including positional dependency, on hypertension. Age, AHI score, BMI, and positional dependency were added into the model. The results showed that not only age, BMI, and AHI but also positional dependency were statistically significant contributing factors for hypertension in the order of age (adjusted odds ratio [AOR], 2.344; P < .001), BMI (AOR, 2.252; P < .001), AHI (AOR, 1.906; P < .001), and positional dependency (AOR, 1.398; P = .03) (Table 3).
To reveal the significance of the associated factors, binary logistic regression analysis was performed and age, AHI, and BMI were added into the model. This revealed that AHI was the most significant factor contributing to positional dependency, and the AOR of patients with severe enough OSA (AHI, ≥40) to be NPP was 7.54 compared with those who had mild OSA (AHI, <20). The BMI was also a significant contributing factor; the AOR of a higher BMI (≥30) was 3.25 compared with a lower BMI (<25). Age was not a statistically significant factor (P = .89, Table 4).
In this study, we investigated the characteristics of positional dependency and its determining factors in 1170 Asian patients with OSA. Sleep position has been recognized as an important factor in OSA from the standpoint of positional dependency; most of these studies were from Western countries.4,7,8 The first large population study7 of positional dependency was conducted in Israel in 1997. The findings showed that 55.9% of 574 patients had positional OSA and that the body position during sleep has a significant effect on the frequency and severity of breathing abnormalities. In our study, 74.7% of the patients had positional OSA, showing a higher prevalence in Asians than in whites. This increased prevalence of Asian PPs might be attributed to the fact that the BMIs of Asians are usually smaller than that of whites. The mean (SD) BMI in our series was 25.6 (3.1) for the PPs and 27.3 (3.9) for the NPPs; in one study,7 the mean BMI of white patients was 29.4 (4.1) for the PPs and 31.9 (4.9) for the NPPs. Because BMI was one of the most significant factors for positional dependency in our study, the lower value in Asians could explain the higher prevalence of PPs in this population.
The mean (SD) AHI in our study was 23.9 (15.7) for the PPs and 47.4 (26.5) for the NPPs. The average Respiratory Distress Index in white patients reported by Oksenberg et al7 was 27.8 (11.0) for the PPs and 44.0 (2.7) for the NPPs; they did not observe a marked difference between Asians and whites. Therefore, the severity of OSA is less likely to explain the difference in the prevalence of positional dependency.
We also identified that AHI and BMI were significant contributing factors for positional dependency, which is compatible with the findings of previous studies.7,9 However, age was also reported as a significant contributing factor for positional dependency in another study,7 and this finding is different from our result. In addition, it was reported that the PPs were significantly younger4,7 and that the prevalence of PPs in the younger age group (<60 years) was significantly higher than that in the older age group (>60 years).4 Those results differ from ours. We found no significant difference in the prevalence of PPs according to age, and age was not a significant contributing factor in the present study. In addition, the mean age of the PPs was slightly higher than that of the NPPs. A recent study by Oksenberg et al9 showed that age was not a significant contributing factor for positional dependency, which is opposite of the earlier findings.7
It is an interesting finding that the PPs had a lower prevalence of hypertension (34.4% vs 49.7% for the NPPs). It is a slightly lower prevalence than those published in Western countries (50%-60%).16- 18 It is possible that confounding variables, such as age, BMI, and AHI, could make a difference in the prevalence of hypertension, so we used binary logistic regression analysis to determine whether positional dependency had a role in the prevalence of hypertension. We identified that positional dependency also had a minor role in the prevalence of hypertension, although the odds ratio was the smallest compared with those of other variables such as age, BMI, and AHI.
In terms of sleep architecture, sleep efficiency did not differ significantly between the 2 groups and the PPs had a longer deep sleep and a lesser arousal index, which is compatible with previous findings.7 The most recent study by Oksenberg et al9 found that sleep efficiency was not significantly different between the 2 groups, which was different from the previous report.7
As we have shown here, the PPs had many more beneficial findings than the NPPs, such as a higher lowest oxygen saturation value, more deep sleep, a lesser arousal index score, less daytime sleepiness, lower blood pressure, and a lower prevalence of hypertension. These beneficial findings and the higher prevalence of PPs among Asians may facilitate the clinical use of positional therapy. A recent study19 reported that positional therapy is equivalent to continuous positive airway pressure at normalizing the AHI for patients with positional OSA, with similar effects on sleep quality and nocturnal oxygenation. However, further well-designed studies are necessary to prove that positional therapy will be a therapeutic alternative for many PPs, and more easily used therapeutic hardware should be introduced.
In conclusion, the prevalence of OSA with positional dependency was 74.7% in Asians, and the PPs had a lower BMI, a lower AHI, longer deep sleep, less daytime sleepiness, and a lower prevalence of hypertension than did the NPPs. The AHI was the most dominant factor for determining positional dependency, followed by the BMI. The higher prevalence of position-dependent OSA in Asians compared with whites might be explained by the lower BMIs of Asians, despite the similar disease severity.
Correspondence: Jeong-Whun Kim, MD, PhD, Department of Otorhinolaryngology, Seoul National University College of Medicine, Seoul National University Bundang Hospital, 300 Gumi-dong, Bundang-gu, Seongnam-si, Gyeonggi-do 464-707, South Korea (firstname.lastname@example.org).
Submitted for Publication: March 26, 2011; final revision received May 3, 2011; accepted May 19, 2011.
Author Contributions: Drs Mo and Lee contributed equally to this study. Drs Mo and Kim had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Mo, Lee, and Kim. Acquisition of data: Rhee, Yoon, and Kim. Analysis and interpretation of data: Mo. Drafting of the manuscript: Mo. Critical revision of the manuscript for important intellectual content: Lee, Rhee, Yoon, and Kim. Statistical analysis: Mo. Obtained funding: Mo. Administrative, technical, and material support: Yoon and Kim. Study supervision: Lee and Kim.
Financial Disclosure: None reported.
Funding/Support: This study was supported by the research fund of Dankook University in 2010.