Hypoxia parameters, physical variables, and severity of obstructive sleep apnea

Abstract

Objective:To determine the relation between hypoxia and physicalparameters in patients who had different levels of severity of obstructive sleep apnea (OSA). Methods:This was a retrospective, cross-sectional study of 259 menwho were evaluated with overnight polysomnography. Severity ofOSA was graded based on the apnea-hypopnea index (AHI):normal/simple snoring (n=31); mild OSA (n=70); moderate OSA(n=63); severe OSA (n=95). Patients with different severity weredivided into subgroups, based on having the lowest or highest valuesof the total sleep time with oxygen saturation <90% (ST90) or minimum oxygen saturation (min SaO2). Results: Median AHI was 20.4 events/hour. Univariate analysisshowed that ST90was correlated with AHI (r=0.772; p≤ 0.001) andEpworth sleepiness scale (ESS) (r=0.344; p≤ 0.001), and min SaO2wasinversely correlated with AHI (r=-0.748; p≤ 0.001) and ESS (r=-0.319;p≤0.001). Multivariate linear regression showed that ST90was independently associated with AHI, ESS, and neck circumference, and minSaO2was independently inversely associated with AHI, ESS, and bodymass index (BMI). In patients who had severe OSA, the subgroupswhich had lowest and highest min SaO2differed significantly in BMI,modified Mallampati score, neck and waist circumferences, and retroglossal Müller grade. In patients with percentage of sleep time withoxygen saturation below 90% (CT90) <10%, the upper limit of ST90was 36 minutes and corresponded to 70% lower limit of min SaO. Conclusion:Hypoxia parameters show significant variation in OSAseverity categories. None of the physical parameters had clinically useful relations with hypoxia parameters in OSA patients except patientswho had severe OSA