Abstract: Objective To test the hypothesis that there would be a subgroup of patients who have rapid eye movement(REM) predominance of obstructive events and that these subjects share certain clinical characteristics and to assess factors influencing the distribution of events by sleep state in patients with obstructive sleep apnea(OSA). Methods Polysomnography recordings of 417 adults with OSA were reviewed. Apnea hypopnea index(AHI) and duration of respiratory events were calculated separately for REM and non-REM(NREM) sleep and compared. A REM predominance index(RPI) was determined using REM/NREM AHI and its possible influencing factors were investigated. REM sleep related OSA was defined as RPI>or=2. Their polysomnography data and clinical characteristics were compared with other subjects. Results There was no significant difference between AHI in REM vs NREM sleep(t=-0.361, P=0.718). But the difference in apnea-hypopnea time index(AHTI) was significant at the 0.01 level(t=5.784). Higher AHI in NREM than REM sleep was found in 229(55.0%) patients; 286 patients(68.6%) had a higher AHTI in REM sleep than NREM sleep. Analysis showed that RPI was significantly correlated with AHI(r=-0.234, P=0.000). Sixty-seven subjects were REM sleep related OSA patients(16.7%). Differences existed in AHI(t=10.708), arousal index(t=6.070), nadir SaO₂(t=-5.989), proportion of obstructive apnea index to AHI(t=4.282), ESS score(t=2.772) and BMI(t=2.858) between REM sleep related OSA patients and the other subjects. There was no significant difference between sleep state distribution and age. Conclusion The duration of respiratory events was longer in REM sleep than in NREM sleep while AHI was similar. REM sleep-related OSA is a problem in only the patients with mild and moderate disease.

Key words: sleep apnea, obstructive, rapid eye movement, sleep stage