Risk factors associated with the severity of obstructive sleep apnea syndrome among adults

Obstructive sleep apnea syndrome (OSAS) is a complete or partial airway obstruction that causes significant physiologic disturbance with various clinical impacts. The aim of the study was to determine the risk factors associated with the severity of OSAS. 303 patients with OSAS and 199 patients without OSAS enrolled in the sleep disorder center from 2017 to 2019 were included in the study. All patients completed physical examination, Epworth sleepiness scale, and polysomnography. By apnea–hypopnea index (AHI), patients were separated into four subsets: AHI < 5 as non-OSAS group, AHI ≧ 5 and < 15 as mild group, AHI ≧ 15 and < 30 as moderate group and AHI ≧ 30 as severe group. AVONA analyses, chi-square test, univariate and multivariate linear regression analyses were conducted to assess the correlation between specific factors and AHI. Our study demonstrated that patients with severe OSAS were associated with higher body mass index, higher systolic blood pressure awake in the morning, lengthened latent period of slow wave sleep, lower level of average oxygen saturation and minimum oxygen saturation (P < 0.05). Our findings provide evidence regarding several potentially useful factors for recognizing OSAS in adults, and physicians should promote the early recognition, diagnosis and intervention of OSAS.


Discussion
Our study indicated that BMI was a clinical predictor of the AHI. The correlation between BMI and OSAS was complex. Most of the literature in the past demonstrated that an increase in BMI was related to an increase in AHI 9,10 , although several studies did not find a significant relationship between BMI and AHI, supporting that truncal obesity was a better factor than BMI as predictors for OSAS 10 . The World Health Organization (WHO) defines 'obesity' as BMI ≧ 30 and this cutoff point provides a benchmark for individual assessment 11 . Obesity predisposed to OSAS, and the prevalence of OSAS was increasing due to the ongoing epidemic of obesity 12 . In addition, the average BMI for patients with severe OSAS was 27.2 kg/m 2 in our study, indicating that overweight, although not obese was still a risk factor for severe OSAS.
Our study found that SBP awake in the next morning was a predicator of severe OSAS. Previous studies reported the link between OSAS and hypertension, and postulated mechanisms connecting elevated blood pressure to OSAS, including increased negative intrathoracic pressure increasing left ventricular transmural pressure 13 , increased sympathetic nervous system activation 14 , and autonomic nervous system dysfunction 15 . It is noteworthy that in our study, higher SBP level awake in the next morning was associated with higher AHI whereas diastolic blood pressure (DBP) level had no such association. The mechanism was still unclear and warranted further studies.
Surprisingly, our study found that lengthened latent period of SWS was associate with higher AHI. Few studies in the past reported the relationship between the severity of OSAS and latent period during the sleep except www.nature.com/scientificreports/ one study reported lengthened latent period of the rapid eye movement (REM) sleep in children with OSAS 16 . Therefore, more studies were warranted to clarify the relationship between AHI and latent period of sleep. Our study found that low level of average oxygen saturation and minimal oxygen saturation during the sleep was associated with the severity of OSAS. Pulse oximetry is a relatively simple, feasible, and inexpensive method  www.nature.com/scientificreports/ that has been extensively used to routinely assess patients' oxygen saturation during sleep. However, the sensitivity and specificity of the oxygen saturation to predict OSAS varied in different studies 17 . We also found that the duration of SWS was not significantly correlated with the severity of OSAS, which was inconsistent with previous findings reporting that AHI was significantly decreased during SWS 18 . Severe OSAS decreased occurrence of SWS and continuous positive airway pressure (CPAP) treatment of OSAS caused SWS rebound 19 . The different sample size and sample population may be a possible reason for the controversial conclusions and further studies need to be conducted.
Our study did not find the relationship between body position and severity of OSAS. The findings in the past literature were mixed. Upper airway collapsibility was greater in supine position compared to lateral position. In supine position, the tongue base narrowed the upper airway by the effect of gravity. Therefore, respiratory events were seen less in side position 20 . Over 60% of the patients with OSAS were considered to have a positional obstructive sleep apnea (POSA), and the degree of severity of OSAS was thought to be mostly associated with the sleep time spent in the supine position 21 . However, in other studies, total sleep time in supine position was not considered as a crucial factor of total AHI 22 and differences in sleep position did not play a major role in nightto-night variability of AHI 23 . Therefore, it is still unclear how much variability in time spent supine contributes to total AHI and more research is warranted to clarity the issue.
Moreover, our study found that there was no significant correlation between ESS scoring and AHI, which was inconsistent with previous studies 24 . ESS was proven to be a reliable self-assessment tool, which could safely predict the presence of OSAS. It might be explained by the fact that although ESS has advantages of being brief and simple to carry out, it has also weaknesses related to subjective scoring.
The limitations of this research are pointed out. Firstly, this is a cross sectional study, therefore, we cannot make any certain conclusion about the correlation between these factors and AHI. In addition, the patients studied were from only one sleep disorder center, which could have resulted in sampling bias. Larger, multicenter studies are required to further investigate the risk factors for OSAS.
Our study demonstrated that the severity of OSAS was correlated with higher BMI, higher SBP awake in the morning, lengthened SWS latent period, lower level of average oxygen saturation and minimum oxygen saturation. Our findings provide evidence regarding several potentially useful factors for recognizing OSAS in adults, and physicians should promote the early recognition, diagnosis and intervention of OSAS. PSG data collection. The gold standard diagnostic method for OSAS is a full-night PSG 26 . The frequency of episodes of apnea and/or hypopnea per hour of sleep, also known as apnea-hypopnea index (AHI), as well as the lowest observed oxyhemoglobin saturation (O 2 SAT) during sleep is used as the main criteria for severity assessment. It is defined as non-OSAS for AHI < 5, mild for AHI ≧ 5 and < 15, moderate for AHI ≧ 15 and < 30, and severe for AHI ≧ 30 27 .

Participants.
All patients had to have a minimum of 8 hours of monitored sleep in the sleep disorder center. They underwent nocturnal PSG monitoring (Philips Alice Version 6, Netherland) using 6 scalp electrodes (C3, C4, F3, F4, O1 and O2 locations), 2 reference electrodes behind the ears (left [A1] and right [A2] mastoid areas), 3 electromyographic electrodes over the submental muscles, 4 electromyographic electrodes over the leg muscles, 2 electro-oculographic electrodes, one ground electrode and nasal flow detector. No ingest alcohol or caffeine, nap, protracted or tedious exercises were informed during the study. Pulse oximeter was used to obtain nocturnal oximetry recordings. Electrocardiogram was used to obtain rhythm of the heart. The AHI documented the number of apnea-plus-hypopnea incidents every hour during sleep. Apnea could be viewed as being without respiration for more than 10 s. Hypopnea could be viewed as the deduction of at least 50% ventilation causing a drop in arterial saturation of at least 3%. OSAS could be viewed as apnea or hypopnea happening at least five times per hour, persisting for more than 10 s.

Statistics.
Statistical analyses were performed with SPSS 26.0 for mac. The summary of descriptive statistics was presented as mean (with SD) for continuous variables and as frequencies (with percentages) for categorical variables. The Pearson chi-square test was used for comparison of qualitative data. Univariate linear regression analyses were used to identify probable factors of severe OSAS. Statistically significant parameters in univariate linear regression analyses were further examined in a multivariate linear regression model. Statistically significant difference was considered if P value < 0.05. Informed consent. Written informed consent was obtained from all participants for the publication of any potentially identifiable images or data included in this article.

Data availability
The data used and analyzed during the current study are available from the corresponding author on reasonable request.