P142 Head flexion has the greatest impact on OSA severity during REM sleep

Recent work has shown that head flexion has a modest worsening effect and head rotation has a modest protective effect on OSA severity. However, there is substantial variability both within and between individuals. In this analysis we aimed to identify if this variability is explained by sleep-state, BMI, age or sex. Methods 28 participants provided informed consent and were studied using diagnostic polysomnography with the addition of a customised, accelerometry based, head posture measurement device. For each epoch during supine sleep, the sleep state (NREM/REM), average head flexion (degrees) and average head rotation (degrees) were recorded. A logistic mixed effects model was fit across all epochs with the anthropometrics (BMI, sex, age), sleep state, average head flexion and average head rotation as explanatory variables with the absence/presence of one or more respiratory event(s) as the binary outcome variable. Results In total, 2122 of 5369 supine sleep epochs had a respiratory event. Three participants had no supine sleep. There were significant interaction effects for flexion-rotation, BMI-rotation and REM-flexion. The REM-flexion interaction effect was the strongest interaction effect with an odds ratio per 5 degrees of head flexion in REM sleep of 1.47 (95% CI: 1.13 – 1.86). Discussion Head flexion related worsening of OSA severity is greatest during REM sleep. This may be explained by attenuated upper airway neuromuscular activation in REM sleep compared with NREM sleep.