Journal of Neurological Sciences (Turkish) 2016 , Vol 33 , Num 3
Association between Hypoxia Parameters with White Matter Hyperintensity and Silent Cerebral Infarcts on Brain Magnetic Resonance Images in Patients with Obstructive Sleep Apnea
Aynur YILMAZ AVCI1, Suat AVCI2, Hüseyin LAKADAMYALI3, Hatice LAKADAMYALI4, Ufuk CAN5
1Başkent Üniversitesi Tıp Fakültesi, Nöroloji, Antalya, Turkey
2Başkent Üniversitesi Tıp Fakültesi, Kulak Burun Boğaz, Antalya, Turkey
3Başkent Üniversitesi Tıp Fakültesi, Göğüs Hastalıkları, Antalya, Turkey
4Başkent Üniversitesi Tıp Fakültesi, Radyoloji, Antalya, Turkey
5Başkent Üniversitesi Tıp Fakültesi, Nöroloji, Ankara, Turkey

Summary

Objective: This study evaluated the association between hypoxia parameters with white matter hyperintensity (WMH) and silent cerebral infarcts (SCI) on brain magnetic resonance (MR) images of patients with obstructive sleep apnea (OSA).

Methods: In this retrospective study, the study group was composed of 453 patients who were evaluated by overnight polysomnography (PSG). Data on hypoxia parameters, such as total sleep duration with oxygen saturation < 90% (ST90), percentage of cumulative time with oxygen saturation < 90% (CT90), and the lowest oxygen saturation (min SaO2), were obtained from PSG. The presence of WMH and SCI was evaluated in all participants using brain MR images.

Results: Hypoxia parameters, such as ST90, CT90, and min SaO2, were significantly associated with WMH (P < 0.001). The multiple regression analysis showed that CT90 was independently associated with SCI (P = 0.038). In addition, when participants were divided into two groups according to CT90 < 10% and CT90 ≥ 10%, age (P = 0.002), sex (P = 0.015), body mass index, Apnea-Hypopnea Index score, Epworth Sleepiness Scale score, and the presence of WMH, hypertension, and diabetes mellitus were significantly higher in the CT90 ≥ 10% group compared with the CT90 < 10% group (P < 0.001 for all parameters). CT90 ≥ 10% increased the risk of WMH 2.34-fold (95% confidence interval, 1.44–3.85; P = 0.006).

Conclusion: The severity of nocturnal intermittent hypoxia may contribute to the pathogenesis of WMH and SCI in patients with OSA.

Introduction

Obstructive sleep apnea (OSA) is a common syndrome characterized by repetitive episodic collapse of the upper airway and intermittent hypoxia during sleep. Disturbances in gas exchange lead to oxygen desaturation, hypercapnia, and fragmented sleep, which contribute to the metabolic, neurocognitive, and cardiovascular effects of OSA(16). OSA is a common syndrome, affecting 3–7% of the general population(10,22). Population-based studies suggest that up to 19% of middle-aged men and 15% of women may suffer from hypopnea and apnea(28). OSA is an important risk factor for vascular diseases. However, the complex mechanisms underlying OSA and vascular diseases are not well understood(8).

The increased use of imaging techniques has allowed white matter hyperintensity (WMH) and silent cerebral infarcts (SCI) to be more commonly observed in asymptomatic patients(25). The clinical importance of WMH constitutes a very important public health problem, as its association with incident stroke, dementia, and mortality has been highlighted in previous studies(7) Therefore, early detection of an SCI is important because it is associated with higher rates of mortality and subsequent clinical cerebral infarction(3,26). Previous studies have shown that patients with moderate to severe OSA have a 2-fold increased risk of developing WMH and SCI(18,21). Hypoxic processes and extreme changes in the vascular system that accompany apneic periods may injure the brain(15).

OSA severity can be determined by the Apnea-Hypopnea Index (AHI), which represents the frequency of apnea and hypopnea episodes per hour of sleep regardless of duration or morphology(2). However, the AHI does not completely reflect the pathophysiological characteristics or severity of hypoxia(19). Moreover, patients with similar AHI scores may have different clinical symptoms and outcomes(1,4). Chronic intermittent hypoxia is usually defined as repeated episodes of hypoxia interspersed with periodic reoxygenation(20). Total sleep duration with oxygen saturation < 90% (ST90), percentage of cumulative time with oxygen saturation < 90% (CT90), and the lowest oxygen saturation (min SaO2) are directly associated with the duration and severity of hypoxia(4,19). This study was conducted to determine the relationship between hypoxia parameters, such as ST90, CT90, and min SaO2, and WMH and SCI on brain magnetic resonance (MR) images in patients with OSA.

Methods

Study design and patients
The study group in this retrospective study was composed of 856 consecutive patients suspected of having OSA who underwent a complete polysomnography (PSG) evaluation in our accredited sleep disorder center from April 2008 to October 2015. In total, 453 patients who underwent both PSG and brain MR imaging (MRI) were included in this study. The patient inclusion criterion for the study was no medical history of cerebrovascular diseases, such as ischemic stroke, transient ischemic attack, or intracerebral hemorrhage. Demographic and PSG data and information regarding age, sex, body mass index (BMI), and PSG parameters, such as total sleep duration, sleep efficiency, arousal index, AHI score, longest duration apnea episode, total apnea duration, ST90, CT90, and min SaO2, were all recorded after obtaining approval from the institutional review board. Data on the duration of hypertension, smoking status, diabetes mellitus, coronary heart disease, and hyperlipidemia were mainly documented from the patients' medical records.

Patients with the following were excluded from the study: those who had central sleep apnea syndrome; narcolepsy; underwent previous treatment for OSA using continuous positive airway pressure, surgery, and/or an oral device; cerebrovascular disease; chronic obstructive pulmonary disease; bronchial asthma; dementia; renal failure; hepatic damage; malignancy; head trauma; brain tumor or malignancy; or were <18 years of age.

Informed consent could not be obtained due to the retrospective nature of this study. The study was evaluated by the Baskent University Institutional Review Board.

Polysomnography
All participants underwent PSG in a sleep laboratory using a computerized PSG device (E series, 44 channels; Compumedics, Victoria, Australia). Sixteen channels were used to document the following parameters: four-channel electroencephalogram, electro-oculogram, submental and leg electromyogram, electrocardiogram, nasal airflow using a nasal pressure cannula, air flow at the nose and mouth (thermistors), chest and abdominal respiratory movements, oxygen saturation (pulse oximetry), snoring via a microphone, and body position. Data from all participants were evaluated by a sleep specialist who was blinded to all information about the participants. Apnea was defined as cessation of air flow for at least 10 s with continued effort (obstructive) or lack of effort (central) to breathe. Hypopnea was defined as a >50% decrease in a valid air flow measure without the requirement for associated oxygen desaturation or arousal and with less reduction of air flow in association with oxygen desaturation >3% or arousal for at least 10 s. The AHI score was the number of apnea and hypopnea episodes per hour. ST90 was recorded in minutes. The min SaO2 and CT90 were recorded as percentages. Sleep stages were evaluated according to the American Academy of Sleep Medicine criteria(2). An AHI score < 5 was considered normal or simple snoring, 5 to <15 was considered mild OSA, 15 to <30 was considered moderate OSA, and ≥30 was considered severe OSA.

Magnetic resonance imaging
The presence of SCI and WMH was assessed by whole-brain MRI. The orbitomeatal line was considered the reference for all MRI brain scans (1.0 Tesla, Siemens Magnetom Vision Plus; Siemens, Munich, Germany). The scans included sagittal T1-weighted, axial T2-weighted, and axial fluid attenuated inversion recovery images. The slice thickness was 5 mm, the gap was 1 mm, and no intravenous contrast was used. All MRI scans were reviewed and scored by a radiologist who was blinded to the clinical details. The scans were visually assessed to determine the presence of WMH and SCI.

Statistical analysis
The data analysis was performed with IBM SPSS Statistics for Windows v. 21.0 statistical software (IBM Corp., Armonk, NY, USA). Continuous variables are presented as mean ± standard deviation (SD) or median (range). Categorical variables are presented as numbers or percentages. The normality of the continuous variable distributions was evaluated by the Kolmogorov–Smirnov test. Similarities between groups were assessed using Levene's variance test. Differences in continuous variables between the two groups were evaluated by independent sample t-tests or the Mann–Whitney test. Comparisons between more than two groups were evaluated by a one-way analysis of variance (ANOVA). Pairwise comparisons were evaluated with Tukey's test. Categorical variables were compared using Pearson's chi-square test. Factors affecting ST90, CT90, and min SaO2 were determined by multiple logistic regression. A P-value ≤ 0.05 was considered significant.

Results

Table 1 shows the patient characteristics. The median age of the patients was 51 years (range, 22–84 years), and most were males (69.2%). Table 2 shows the statistical differences in the parameters between the AHI groups. Male predominance was observed in all OSA groups (Table 2). The hypoxia parameters, such as ST90 and CT90, were significantly higher but min SaO2 was significantly lower in the OSA group compared with those in the control group (Table 2). WMH was more frequent in the OSA group than that in the control group (Table 2). Furthermore, SCI was observed more frequently in the severe OSA group compared with that in the control group (Table 2). Additionally, the odds ratio (OR) of WMH was 2.53-fold higher (95% confidence interval CI, 1.60–3.98; P = 0.0001) in the OSA group (AHI ≥ 5) compared with that in the control group (AHI < 5). The OR of SCI was 3.41-fold higher (95% CI, 1.19–9.76; P = 0.02) in the OSA group (AHI ≥ 5; 345 patients) than that in the control group (AHI < 5; 108 patients).

Table 1: Characteristics of study patients with obstructive sleep apnea*

Table 2: Comparison of the Apnea-Hypopnea Index in patients with obstructive sleep apnea*

The univariate analysis revealed a significant correlation between ST90 and age, BMI, AHI, Epworth Sleepiness Scale (ESS) score, CT90, min SaO2, hypertension, diabetes mellitus, and WMH (Table 3). Significant correlations were also revealed between CT90 and age, AHI score, BMI, ESS score, ST90, min SaO2, hypertension, coronary heart disease, hyperlipidemia, diabetes mellitus, and WMH (Table 3). Significant inverse correlations were observed between min SaO2 and age, BMI, AHI score, ESS score, ST90, CT90, hypertension, coronary heart disease, hyperlipidemia, diabetes mellitus, and WMH (Table 3). Strong correlations were detected between the AHI score and ST90 and CT90 and min SaO2, but the CT90 values varied, particularly among patients with severe OSA (Figure 1).

Table 3: Univariate analysis of factors affecting total sleep duration with oxygen saturation < 90% and the lowest oxygen saturation in patients with obstructive sleep apnea*

Figure 1: Distribution of the percentages of cumulative time with oxygen saturation below 90% (CT90) values within different Apnea-Hypopnea Index (AHI) severity groups. OSA, obstructive sleep apnea. CT90 (%), AHI (events/h).

The multiple linear regression analysis revealed that AHI score (P < 0.001), age (P < 0.001), CT90 (P < 0.001), and min SaO2 (P = 0.016) remained independent parameters affecting ST90 after adjusting for other confounders (Table 4). The multiple linear regression analysis also showed that AHI score (P < 0.001), age (P < 0.001), ST90 (P < 0.001), and SCI (P = 0.038) were independent parameters influencing CT90 after adjusting for other confounders (Table 4). AHI score (P < 0.001), age (P < 0.001), ST90 (P < 0.001), and BMI (P = 0.002) remained independent parameters influencing min SaO2 after adjusting for other confounders (Table 4).

Table 4: Multiple linear regression analysis of factors affecting total sleep duration with oxygen saturation < 90% and the lowest oxygen saturation in patients with obstructive sleep apnea*

Additionally, the participants were divided into two groups according to CT90 < 10% (367 patients) and CT90 ≥ 10% (86 patients). Four (3.9%) of 103 patients in the mild OSA group had CT90 ≥ 10%; 4 (4.7%) of 86 patients in the moderate OSA group had CT90 ≥ 10%; and 78 (50%) of 156 patients in the severe OSA group had CT90 ≥ 10% (Table 5 and Fig. 2). Seventy-eight of 86 patients (90.7%) who had CT90 ≥ 10% were in the severe OSA group, and half of patients (78 of 156 patients) in the severe OSA group had CT90 ≥ 10% (Table 5). The incidence rates of WMH, hypertension, and diabetes mellitus were significantly higher in the CT90 ≥ 10% group than those in the CT90 < 10% group. SCI tended to be detected more frequently in the CT90 ≥ 10% group than that in the CT90 < 10% group. These findings may be due to the limited number of patients rather than a lack of an effect of CT90 ≥ 10% on SCI (Table 5). Age, AHI score, and ST90 values were significantly higher but the min SaO2 value was significantly lower in the CT90 ≥ 10% group than those in the CT90 < 10% group (Table 5). In addition, the OR of WMH was 2.35-fold higher (95% CI, 1.44–3.85; P = 0.0006) in the CT90 ≥ 10% group than that in the CT90 < 10% group.

Table 5: Distribution of the parameters according to percentage of cumulative time with oxygen saturation < 90% below and above 10%*

Discussion

Our findings suggest a relationship between intermittent hypoxia and the presence of WMH and SCI in patients with OSA. The increased risk of WMH was 2.5-fold higher and the increase in the SCI risk was 3.4-fold higher in patients with OSA compared with those in the control group. Hypoxia parameters, such as ST90, CT90, and min SaO2, were associated with WMH (Table 3). The multiple regression analysis showed that CT90 was independently associated with SCI (P = 0.038) (Table 4). Furthermore, CT90 ≥ 10% led to a 2.35-fold increase in WMH risk compared with that of having CT90 < 10% (OR, 2.35; 95% CI, 1.44–3.84; P = 0.0006). These data suggest that increased severity of hypoxia may contribute to the pathogenesis of WMH and SCI in patients with OSA. This is the first report demonstrating a relationship between hypoxia severity and WMH and SCI in patients with OSA.

Previous studies have also reported that WMH and SCI occur more frequently in patients with OSA(5,18,21). Eguchi et al. demonstrated that nocturnal hypoxia assessed by overnight pulse oximetry was independently associated with the prevalence of SCI among a high-risk community-dwelling Japanese population(11). The mean age of participants in their hypoxia group (70 years old) was older than that of our patients (51 years old), and the male ratio was lower (31%) compared with that in our study (69%). The high female ratio and older age may have affected the findings(11). Conflicting results have been reported in previous studies. Most previous studies failed to demonstrate a relationship between OSA and WMH and SCI(6,8,17). The sample group was quite small in one study, and the mean age of participants in another study was older than that in the current study. Additionally, hypoxia was not considered in these studies. Robbins et al. reported that only central sleep apnea but not OSA was associated with the development of WMH(23). These participants (mean age, 77 years), already had higher basal levels of WMH on their primary examinations. The negative findings may be explained by a survival bias or high rates of vascular risk factors, which are also commonly present in older adults. The participants in the current study were younger (mean age, 51 years) compared with patients in other studies.

Furthermore, Zhang et al. reported a study of 119 patients with OSA who underwent velopharyngeal surgery. CT90 < 10% (grade 1 category) was an independent predictor of high surgical success(27); however, the surgical success rate decreased dramatically when the hypoxia threshold was exceeded. In our study, only 4 (3.8%) of 103 patients with mild OSA and 4 (4.6%) of 86 patients with moderate OSA were in the CT90 ≥ 10% category (Table 5). Among the patients in the CT90 ≥ 10 % group, 90.7% were in the severe OSA group. Half of the patients with severe OSA (78 of 156 patients) were in the CT90 ≥ 10% group, and AHI score, ESS, BMI, age, WMH, hypertension, and diabetes mellitus increased significantly above this threshold (CT90 ≥ 10%) (Table 5). The presence of SCI tended to be higher in the CT90 ≥ 10% group, but the difference was not significant. This lack of an association may be due to the limited number of events rather than the lack of an effect. In addition, the risk of WMH increased 2.35-fold in the CT90 ≥ 10% group compared to that in the CT90 < 10% group (Table 5).

The AHI score only indicates the number of apnea and hypopnea episodes per hour(2). However, the AHI score does not reflect the actual duration or severity of hypoxia or disease outcome. Patients with similar AHI scores can have different durations and depths of cessation of breathing and oxygen desaturation. These differences mostly affect the symptoms and consequences of the disorder. Prolonged duration apnea and hypopnea episodes can paradoxically lead to a decrease in the AHI score, although it is commonly assumed that more severe health consequences are evident compared with those in patients with shorter events. Severe hypoxia despite a similar AHI score may cause severe physiological stress, cardiovascular consequences, or death(4,19).

Chronic intermittent hypoxia is a prominent characteristic of the pathophysiology of OSA. Repetitive episodes of hypoxia and reoxygenation occur more than 60 times per hour (hundreds of times per night) in cases of severe OSA and can last for years when hemoglobin desaturation reaches 50%(13). This high frequency of hypoxia and reoxygenation cycles is similar to that observed in a patient with ischemia-reperfusion injury, resulting in increased production of reactive oxygen species and oxidative stress, which may contribute to OSA-associated cardiovascular pathologies(14). Stopping breathing can lead to CO2 retention and hypoxia, which disturbs the autonomic and hemodynamic responses during sleep(25). Repetitively occurring apnea episodes are often followed by increased sympathetic activity, which can further lead to vasoconstriction of peripheral blood vessels(10).

The pathophysiology of WMH is heterogeneous. The presence of focal myleninolysis, axonal loss, and gliosis associated with vessel wall hyalinosis suggests that chronic hypoperfusion contributes to the development of WMH(12,24). WMH are considered a subclinical stroke, and the same pathogenic mechanism occurs in patients with WMHs and stroke(24,26). The ischemic effect of nocturnal apnea can increase with increased oxidative stress due to intermittent episodes of hypoxemia and reoxygenation. Thus, cerebrovascular endothelial dysfunction increases and autoregulation deteriorates, which preferentially damage small vessels in the brain(25).

Some limitations of our study should be mentioned. A retrospective study has inherent problems, such as biased patient selection and a single-institutional analysis, which can lead to prejudgment and referral bias. In addition, patients with various vascular factors, such as coronary heart disease, hypertension, hyperlipidemia, diabetes mellitus, smoking, and obesity, were not strictly excluded from the study. Comparing the results with different levels of other thrombotic and endothelial dysfunction markers may improve the understanding of OSA pathophysiology. Furthermore, the MRI was not performed immediately after PSG, and we did not evaluate the localization or number of WMH.

In conclusion, the severity of nocturnal intermittent hypoxia may contribute to the pathogenesis of WMH and SCI in patients with OSA. Hypoxia parameters, such as ST90, CT90, and min SaO2, should be screened to detect hypoxemia occurring due to OSA.

Received by: 30 March 2016
Revised by: 20 August 2016
Accepted: 05 September 2016

References

1) Asano K, Takata Y, Usui Y, Shiina K, Hashimura Y, Kato K, Saruhara H, Yamashina A. New index for analysis of polysomnography, “integrated area of desaturation,” is associated with high cardiovascular risk in patients with mild to moderate obstructive sleep apnea. Respiration 2009;78:278–284. doi: 10.1159/000202980.

2) Berry RB, Brooks R, Gamaldo CE, Harding SM, Lloyd RM, Marcus CL, Vaughn BV, American Academy of Sleep Medicine. The AASM Manual for the Scoring of Sleep and Associated Events: Rules, Terminology, and Technical Specifications, Version 2.2. Darien IL: American Academy of Sleep Medicine; 2015.

3) Bernick C, Kuller L, Dulberg C, Longstreth WT Jr, Manolio T, Beauchamp N, Price T. Cardiovascular Health Study Collaborative Research Group: Silent MRI infarcts and the risk of future stroke: the cardiovascular health study. Neurology 2001;57:1222–1229.

4) Bostanci A, Turhan M, Bozkurt S. Factors influencing sleep time with oxygen saturation below 90% in sleep-disordered breathing. Laryngoscope 2014 Apr;125(4) doi: 10.1002/lary.24942.

5) Cho ER, Kim H, Seo HS, Suh S, Lee SK, Shin C. Obstructive sleep apnea as a risk factor for silent cerebral infarction. J Sleep Res 2013; 22:452–458. doi: 10.1111/jsr.12034.

6) Davies CW, Crosby JH, Mullins RL, Traill ZC, Anslow P, Davies RJ. Case control study of cerebrovascular damage defined by magnetic resonance imaging in patients with OSA and normal matched control subjects. Sleep 2001; 24:715–720.

7) Debette S, Markus HS. The clinical importance of white matter hyperintensities on brain magnetic resonance imaging: systemic review and meta-analysis. BMJ 2010;341: c3666. doi: 10.1136/bmj.c3666.

8) Ding J, Nieto FJ, Beauchamp MJ, Harris TB, Robbins JA, Hetmanski JB, Fried LP, Redline S. Sleep-disordered breathing and white matter disease in the brain stem in older adults. Sleep 2004;27:474–479.

9) Dong JY, Zhang YH, Qin LQ. Obstructive sleep apnea and cardiovascular risk: meta-analysis of prospective cohort studies. Atherosclerosis 2013;229:489–495. doi: 10.1016/j.atherosclerosis.2013.04.026.

10) Duran J, Esnaola S, Rubio R, Izteuta A. Obstructive sleep apnea-hypopnea and related clinical features in a population-based sample of subjects aged 30 to 70 yr. Am J Respir Crit Care Med 2001;163:685–689.

11) Eguchi K, Kario K, Hoshide S, Ishikawa J, Morinari, Shimada K. Nocturnal hypoxia is associated with silent cerebrovascular disease in high-risk Japanese community-dwelling population. AHJ 2005;18:1489–1495.

12) Fazekas F, Kleinert R, Offenbacher H, Schmidt R, Kleinert G, Payer F, et al. Pathologic correlates of incidental MRI white matter signal hyperintensities. Neurology 1993; 43(9):1683–1689.

13) Fletcher EC, Costarangos C, Miller T. The rate of fall of arterial oxyhemoglobin saturation in obstructive sleep apnea. Chest 1989; 96:717–722.

14) Fletcher EC. Physiological consequences of intermittent hypoxia: systemic blood pressure. J Appl Physiol 2001; 90:1600–1605.

15) Harper RM, Kumar R, Ogren JA, Macey PM. Sleep-disordered breathing: Effects on brain stucture and function. Respir Physiol Neurobiol 2013;188:383–391. doi: 10.1016/j.resp.2013.04.021.

16) Jordan AS, McSharry DG, Malhotra A. Adult obstructive sleep apnea. Lancet 2014; 383:736–747. doi: 10.1016/S0140-6736(13)60734-5.

17) Kiernan TE, Capampangan DJ, Hickey MG, Pearce LA, Aguilar MI. Sleep apnea and white matter disease in hypertensive patients: a case series. Neurologist 2011;17:289–291. doi: 10.1097/NRL.0b013e31821a25d6.

18) Kim H, Yun CH, Thomas RJ, Lee SH, Seo HS, Cho ER, Lee SK, Yoon DW, Suh S, Shin C. Obstructive sleep apnea as a risk factor for cerebral white matter changes in a middle-aged and older general population. Sleep 2013; 36:709–715. doi: 10.5665/sleep.2632.

19) Kulkas A, Tiihonen P, Julkunen P, Mervaala E, Toyras J. Novel parameters indicate significant differences in severity of obstructive sleep apnea with patients having similar apnea-hypopnea index. Med Biol Eng Comput 2013;51:697–708. doi: 10.1007/s11517-013-1039-4.

20) Neubacher JA. Physiological and genomic consequences of intermittent hypoxia. J Appl Physiol 2001;90:1593–1599.

21) Nishibayashi M, Miyomoto M, Miyamoto T, Suzuki K, Hirata K. Correlation between severity of obstructive sleep apnea and prevalence of silent cerebrovascular lesions. J Clin Sleep Med 2008;4:242–247.

22) Punjabi NM. The epidemiology of adult obstructive sleep apnea. Proc Am Thorac Soc 2008;5:136–143. doi: 10.1513/pats.200709-155MG.

23) Robbins J, Redline S, Ervin A, Walsleben JA, Ding J, Nieto FJ. Association of sleep-disordered breathing and cerebral changes on MRI. J Clin Sleep Med 2005;15:159–165.

24) Schmidt R, Grazer A, Enzinger C, Ropele S, Homayoon N, Pluta-Fuerst A, Schwingenschuh P, Katschnig P, Cavalieri M, Schmidt H, Langkammer C, Ebner F, Fazekas F. MRI-detected white matter lesions: do really matter? J Neural Transm 2011;118:673–681. doi: 10.1007/s00702-011-0594-9.

25) Somers VK, Dyken ME, Mark AL, Abboud FM. Sympathetic-nerve activity during sleep in normal subjects. N Engl J Med 1993;328(5):303–307.

26) Vermeer SE, Hollander M, Dijk EJ, Hofman A, Koudstaal PJ, Breteler MM. Silent brain infarcts and white matter lesions increase stroke risk in the general population. Stroke 2003;34:1126–1129.

27) Zhang J, Li Y, Cao X, Xian J, Tan J, Dong J, Ye J. The combination of anatomy and physiology in predicting the outcomes of velopharyngeal surgery. Laryngoscope 2014;124:1718–1723. doi: 10.1002/lary.24510.

28) Young T, Palta M, Dempsey J, Skatrud J, Weber S, Badr S. The occurrence of sleep-disordered breathing among middle-aged adults. N Eng J Med 1993;328:1230–1235.