Supine Sleep, Sleep Apnea and Sleep Architecture: A Cross-sectional Study in Adults
Kai Qing Xiong1,2,3, Trevor Jairam1,2,3, Brian Murray1,2,3, Mark Boulos1,2,3
1Sunnybrook Research Institute, 2Sleep Laboratory, Sunnybrook Health Sciences Centre, 3Department of Medicine, University of Toronto

To quantify the impact of supine sleep on obstructive sleep apnea (OSA) severity and markers of sleep architecture.


OSA is associated with abnormal pauses in breathing during sleep, and has been shown to be independently associated with stroke, motor vehicle collisions, hypertension, diabetes, and increased all-cause mortality. OSA severity is known to worsen in the supine position. However, the physiological degree to which supine sleep worsens OSA and impacts sleep architecture has not been quantified in a dose-dependent manner.


We retrospectively evaluated diagnostic in-laboratory polysomnography recorded during 2010-2015 at our sleep laboratory. Multivariable linear regression models with splines were used to account for non-linear relationships between duration of supine sleep with OSA severity and markers of sleep architecture.

A total of 2848 patients were evaluated (48% male, median age was 53 years). There were linear relationships between supine sleep and the apnea-hypopnea index (AHI), respiratory disturbance index (RDI), and lowest oxygen saturation(p <0.0001, <0.0001, and =0.0017 respectfully); for every 10% increase in duration of supine sleep, there was a 0.66 increase in AHI, 0.65 increase in RDI , and 0.19% decrease in oxygen saturation. The non-linear relationship between supine sleep and N2 was “u”-shaped (p=0.005) and the non-linear relationship between supine sleep and REM was upside-down “u”-shaped (p=0.037). Reduced supine sleep showed a positive nonlinear relationship with both total sleep time and sleep efficiency, and a negative nonlinear relationship with wake after sleep onset (p =0.0001, <0.0001, and <0.0001 respectfully). 

To our knowledge, our study is the first to (1) quantify the dose-dependent relationship between duration of supine sleep and OSA severity and (2) report a non-linear relationship between supine sleep and markers of sleep architecture. These findings further strengthen the recommendation that patients should be advised to avoid supine sleep in order to reduce OSA severity and improve sleep quality.