Sleep-stage-independent obstructive sleep apnea: an unidentified group?

Our clinical experience suggested existence of a third group, stage-independent-OSA besides two known groups: REM-dependent-OSA and NREM-dependent-OSA. This study was planned to compare the characteristics of this group with the other two. All the subjects undergoing diagnostic video-polysomnographies with AHI >5/h were included in this study. Based upon the ratio of AHI during REM and NREM sleep, various groups were formed. REM-dependent-OSA was defined as AHI-REM/AHI-NREM >2; all other subjects were included in Not-REM-dependent-OSA (A-1 analysis). This group was further bifurcated into two groups: Non-REM-dependent OSA (NREM-dependent-OSA) where AHI-NREM/AHI-REM >2 and remaining subjects were included in the sleep-stage-independent-OSA group (A-2 analysis). SPSS v 17.0 was used to calculate independent sample t test (A-1 analysis) and Kruskall-Wallis test (A-2 analysis). Using A-1 approach, REM-dependent-OSA group was found to be suffering from mild-moderate OSA (90 %). REM-dependent OSA group had lower AHI-NREM (P < 0.001; 95 % CI 22.11-36.81) and lower AHI-total (P < 0.001; 95 % CI 15.39-30.73). Surprisingly, AHI-REM and DI-REM were not significantly different between these groups. A-2 analysis showed that overall, REM-dependent-OSA had lowest AHI-total while the stage-independent group had highest (P < 0.001). However, on analysis of REM-dependent-OSA, it was found that few of the subjects from this group had severe OSA (AHI-total > 30/h). The NREM-AHI increased linearly as we moved from REM-dependent-OSA to stage-independent-OSA with a significant difference across groups (P < 0.001). However, similar trend was not observed for AHI-REM. This study showed that a third group, sleep-stage-independent-OSA also exists when OSA is classified according to the proportion of apnea across sleep stages. This classification partially corresponds with the severity of illness.