The correlation between oxygen saturation indices and the standard obstructive sleep apnea severity

• Main Authors: Siraj Omar Wali, Bahaa Abaalkhail, Ibrahim AlQassas, Faris Alhejaili, David W Spence, Seithikurippu R Pandi-Perumal
• Format: Article
• Language: English
• Published: Wolters Kluwer Medknow Publications 2020-01-01
• Series: Annals of Thoracic Medicine
• Subjects: apnea–hypopnea index; hypopnea; obstructive sleep apnea; oxygen saturation; polysomnography
• Online Access: http://www.thoracicmedicine.org/article.asp?issn=1817-1737;year=2020;volume=15;issue=2;spage=70;epage=75;aulast=Wali

OBJECTIVE: Currently accepted guidelines categorize obstructive sleep apnea (OSA) severity according to the Apnea–Hypopnea Index (AHI). However, it is unclear how to best define OSA severity. The present study sought to evaluate the concurrent validity of the widely accepted AHI by correlating it with various oxygen saturation (SpO2) and polysomnographic parameters. METHODS: The study utilized the data of a previous survey concerning the prevalence of OSA among a middle-aged Saudi population (n = 2682). Among the 346 individuals who underwent polysomnography, 178 had total sleep times of at least 240 min with rapid eye movement (REM) sleep and were included in the study. The standard classification of OSA severity was compared with different SpO2and polysomnographic parameters. RESULTS: The study found that there were correlations between the standard OSA severity based on AHI severity classification and different SpO2and polysomnographic parameters, including the desaturation index (DI), the sum of all desaturations, desaturation below 90%, the average duration of respiratory events, and indices of total arousals and respiratory arousals. All of these parameters correlated directly with OSA severity classification (P < 0.001 for each). However, REM sleep duration and SpO2nadir were inversely correlated with OSA severity (P < 0.003 and < 0.001, respectively). In addition, only the DI, SpO2nadir, and respiratory arousal index were predictors of OSA severity, as determined through a multiple logistic regression analysis. CONCLUSION: Our findings support the clinical reliability of the currently used standard classification of OSA severity based on the AHI.