Polysomnographic and clinical characteristics of positional obstructive sleep apnea patients

Obstructive sleep apnea is a condition characterized by the complete or partial obstruction of the upper airway during sleep. This study aimed to compare the clinical and polysomnographic characteristics of our obstructive sleep apnea patients according to their positional and non-positional features. Two hundred eighty patients were included in the study. One hundred two patients (36.43%) were female, while 178 patients (63.57%) were male. While 88 (31.43%) of these patients were defined as positional patients, 192 (68.57%) were defined as non-positional patients. The mean age of the positional patients (46.78 ± 9.66) was lower than the mean age of the non-positional patients (50.90 ± 10.96) (p 0.001). Similarly, the mean body mass index of the positional patients (29.39 ± 3.80) was lower than the mean body mass index of the non-positional patients (33.30 ± 6.45) (p < 0.001). Neck circumference values of the positional patients (40.36 ± 2.65) were lower compared to the non-positional patients (43.32 ± 2.54) (p < 0.001). Sleep values were compared based on the presence of positional sleep apnea. In the positional patients, sleep duration, sleep efficiency (percentage), duration of stage N3, minimum, and mean saturation values were found to be higher compared to the non-positional patients, while nightlong apnea hypopnea index, apnea index, percentage of sleep time with oxygen saturation below 90%, oxygen desaturation index, mean heart rate, and periodic limb movement index values were found to be lower (p < 0.05). The rate of severe sleep apnea (7.95%) in the positional patients was lower than the non-positional patients (53.65%) (p < 0.001). In the light of these data, positional OSA is a very important condition presented in 31.43% of OSA patients and it was determined that these patients were younger, had less body mass index, and shorter neck circumference. The rate of severe disease was found to be lower in positional OSA patients


Background
Obstructive sleep apnea syndrome (OSA), which is observed at the rate of 2-4% in the general population, is a chronic and complex clinical syndrome characterized by snoring, periodic apnea, hypoxemia during sleep, and excessive daytime sleepiness [1,2]. Physiological changes that emerge as a result of recurrent apneas cause the development of pulmonary, metabolic, and neurological diseases, especially cardiovascular diseases [3].
Abnormalities in the pharynx anatomy, physiology of the upper respiratory tract dilator muscles, and stability of respiratory control are important causes of recurrent pharyngeal collapse during sleep [2]. However, this event occurring with a complex structural and neuromuscular interaction is still not fully known [4]. Factors reducing the width of the upper respiratory tract and facilitating the collapse increase the tendency towards OSA. Known major risk factors for OSA include advanced age, male gender, obesity, short and thick neck structure, smoking, alcohol, and sedative drug use [5].
Sleeping position is a mechanical factor that may contribute to the development of OSA. In individuals with and without apnea, the airway cross-sectional area significantly reduces in the supine position, which can cause OSA [6]. The severity of breathing abnormalities in the general OSA population is generally associated with body position during sleep. The supine position increases the number and severity of apnea events and affects the emergence of respiratory disorder. During all stages of sleep, upper airway collapse is higher in the supine position compared to the lateral position. Positional OSA occurs when apnea hypopnea index (AHI) in the supine position is two times higher than AHI in the lateral position [7]. The absence of upper respiratory tract obstruction in the side-lying position or significantly less than obstruction in the supine position in some patients suggests that some characteristics of these patients are different [4]. A study revealed that there were fewer and less severe respiratory abnormalities in the positional OSA group compared to the nonpositional OSA group, and it was observed to have an inversely proportional relationship with body mass index (BMI) and age [8]. In recent years, it is important to define different clinical types of sleep apnea and to determine appropriate treatment approaches. However, there were not enough studies about positional OSA. In this study, it was aimed to investigate the possible clinical and polysomnographic differences in the patients with positional OSA and patients with non-positional OSA.

Methods
The files of 310 patients diagnosed with OSA in the sleep clinic of our hospital between 2017 and 2018 were evaluated retrospectively. Thirty patients, who had cardiac failure, were diagnosed with central sleep apnea and had lung disease affecting oxygen saturation, were excluded from the study. The patients' symptoms, age, gender, Epworth Sleepiness Scale score, neck circumference measured from the cricoid cartilage level in the sitting position, body mass index, arterial blood gas, and additional diseases were recorded. The polysomnographic records of the patients were examined, and polysomnographic findings such as total sleep duration, sleep efficiency, sleep latency, sleep stages, AHI, apnea index (AI), AHI in the supine and lateral position, minimum oxygen saturation, mean oxygen saturation, oxygen desaturation index (ODI), percent sleep time spent with oxygen satur-ation< 90% (TST 90), and periodic limb movement index (PLM) were recorded.

Statistical analyses
Statistical analyses were performed by using the SPSS version 17.0 program. The suitability of the variables for normal distribution was examined with histogram graphs and the Kolmogorov-Smirnov test. Mean, standard deviation, median, and min-max values were used when presenting descriptive analyses. In 2 × 2 cells, they were compared with Pearson's chi-squared and Fisher's exact tests. While evaluating the nonparametric variables in paired groups, Mann-Whitney U test was used. While comparing the variables that did not show the normal distribution in triple groups, the Kruskal-Wallis test was used. In the analysis of the measurement data within each other, the Spearman correlation test was used. Conditions with a p value below 0.05 were accepted as statistically significant results.

Results
Two hundred eighty patients were included in the study. One hundred two patients (36.43%) were female, while 178 patients (63.57%) were male. While 88 (31.43%) of these patients were defined as positional patients, 192 (68.57%) were defined as non-positional patients. Snoring was seen in 277 (98.93%) of the patients, witness apnea in 229 (81.79%), daytime sleepiness in 199 (71.07%), morning headache in 72 (25.71%), and night sweating in 31 (11.07%). The mean age of the positional patients (46.78 ± 9.66) was found to be lower than the mean age of the non-positional patients (50.90 ± 10.96) (p:0.001). Similarly, the mean BMI of the positional patients (29.39 ± 3.80) was found to be lower than the mean BMI of the non-positional patients (33.30 ± 6.45) (p < 0.001). The neck circumference of the positional patients (40.36 ± 2.65) was found to be lower than the neck circumference of the non-positional patients (43.32 ± 2.54) (p < 0.001). No significant difference was found between the positional and non-positional patients in terms of gender.
Sleep values were compared based on the presence of positional sleep apnea. In the positional patients, sleep duration, sleep efficiency (percentage), duration of stage N3, minimum, and mean saturation values were found to be higher compared to the non-positional patients, while nightlong AHI, AI (apnea index), percentage of sleep time with oxygen saturation below 90%, sleep ODI, mean heart rate, and PLM index values were found to be lower (p < 0.05) ( Table 1).
Comorbidities were compared based on the presence of positional sleep apnea. The rate of hypertension (HT) in the positional patients (25%) was found to be lower than the rate of HT in the non-positional patients (38.02%) (p:0.033). The rate of hypothyroidism in the positional patients (1.14%) was lower than the nonpositional patients (6.77%) (p:0.045).The rate of diabetes mellitus (DM) in the positional patients (13.64%) was lower compared to the non-positional patients (23.96%) (p:0.048). There was no significant difference between the two groups in terms of the other comorbidities (p > 0.05).

Discussion
In this study, where the clinical and polysomnographic characteristics of the patients with positional OSA were examined, it was found that the positional group was younger, had less BMI, and shorter neck circumference. Additional diseases such as HT and DM were less observed in the positional group.
In studies examining the positional OSA rate in the world, the positional OSA rate varies between 27.4% and 70% [10][11][12]. In a study conducted in 2011 on 230 In a study comparing the demographic characteristics of positional OSA and non-positional OSA patients, it was determined that age, gender, weight, height, BMI, and Epworth Sleepiness Scale were similar, and the patients with positional OSA had a shorter neck circumference [10,11]. In a study conducted on 3214 patients, it was revealed that the patients with positional OSA were younger and had lower BMI [13]. Yosunkaya et al. also revealed that the BMI and neck circumference values were lower in positional OSA patients [4]. Similarly, in our study, it was detected that the positional OSA patients were younger, had less BMI, and shorter neck circumference than the non-positional OSA patients. Patients admitted to sleep outpatient clinic wait a long time for polysomnography examination. During this time, position treatment may be recommended not to increase severity of disease.
Obesity is an important factor for OSA, and it has been reported that the frequency of positional OSA decreases with increased BMI [14,15]. In our study, it was determined that BMI was lower in patients with positional OSA. Similarly, Oksenberg et al. detected positional OSA at higher rates in non-obese patients (BMI < 30), while non-positional OSA was found more frequently as weight increased [8]. Similarly, in another study conducted on 278 patients, it was found that the positional group was younger; BMI and neck circumference were also lower compared to the non-positional group [16]. In another study conducted in Korea, BMI was determined to be lower in the positional group compared to the non-positional group [17].
In the study conducted by Mohsenin et al., it was found that the number of males was higher in the nonpositional group compared to the positional group [18]. In the study conducted by Yosunkaya et al., males were more common in both groups; however, there was no difference between the positional group and nonpositional group in terms of gender. Similarly, in our study, there was no difference between the positional group and non-positional group in terms of gender. In another study, it was revealed that the gender distribution in the positional group was similar to the nonpositional group [11].
When the polysomnographic parameters were examined in our study, it was determined that there was a statistically significant difference between the positional group and non-positional group in terms of sleep efficiency, TST, stage 1, stage 2, stage 3, REM period, and ODI. Oksenberg et al. also put forward that night sleep quality was better preserved in positional OSA patients. They also found that sleep efficiency and deep sleep (stages 3-4) percentages were significantly higher, and light sleep (stages 1-2) percentages were significantly lower in positional OSA patients, and there was no significant difference between the two groups in the REM period [8]. In another study conducted in our country, sleep efficiency was found to be significantly higher in patients with positional OSA [4]. Similarly, in our study, sleep efficiency was found to be statistically higher in the positional OSA patients compared to the non-positional patients. Moreover, stage 1 was lower, REM period and stage 3 were higher in the positional group, and the PLM index was also found to be low. In a 52-week retrospective study conducted according to home polysomnography records, it was determined that 54.6% of the patients had positional OSA. These patients were found to be younger, had lower BMI, and oxygen saturation below 90% [19]. In the study conducted by Wang et al. on 372 patients, the ODI value was lower in the positional group compared to the non-positional group, and the percentage of sleep time with saturation below 90% was shorter [20]. In the study conducted by Richard et al., AHI was higher in the non-positional group [21]. In another study, it was revealed that the total AHI values were significantly lower in the positional OSA group, and the positional OSA frequency was higher in the mild OSA group compared to the moderate OSA group [13]. Nonpositional OSA was more commonly seen in patients diagnosed with severe OSA [6,21,22]. Similarly, in our study, it was determined that the severity of the disease was less in the positional group. In a study performed on 292 patients with severe OSA, it was found that 35% was positional, and the severity of the disease was found to be milder in the positional group [23].
The limitations of our study include the absence of a detailed ENT examination and the absence of subtypes in the positional group.

Conclusions
According to the results of this study, positional OSA is a very important condition presented in 31.43% of OSA patients. It was revealed that patients with positional OSA had mild disease, high sleep efficiency, and these patients were young, had less BMI, and short neck circumference. Hence, positional OSA should be evaluated as a different clinical type and appropriate treatment should be initiated accordingly. In clinical practice, if polysomnography appointment is planned long term to young patients who have obvious clinical symptoms and short neck, during this time position treatment may be recommended not to increase severity of disease. In this way, the emergence of complications in the early period can be prevented.