The prevalence of REM-related obstructive sleep apnoea is reduced by the AASM 2012 hypopnoea criteria

Purpose

The variations in reported prevalence of rapid eye movement-related obstructive sleep apnoea (REM-OSA) have been attributed to different definitions, although the effect of hypopnoea criteria has not been previously investigated.

Methods

Within this retrospective study, 134 of 382 consecutive patients undertaking polysomnography (PSG) for the suspicion of OSA met the inclusion criteria. PSGs were scored using both the 2007 AASM recommended hypopnoea criteria (AASM_2007Rec) and the 2012 AASM recommended hypopnoea criteria (AASM_2012Rec). For each hypopnoea criteria, REM-OSA patients were grouped as REM-related [either as REM-predominant OSA (rpOSA) or REM-isolated OSA (riOSA)] or non-stage-specific OSA (nssOSA). Outcome measures (SF-36, FOSQ and DASS-21) were also compared between groups.

Results

Incorporation of the AASM_2012Rec criteria compared to the AASM_2007Rec criteria increased the apnoea-hypopnoea index (AHI) for NREM and REM sleep but decreased the AHI_REM/AHI_NREM ratio from 1.9 to 1.3 (p < 0.001). It also decreased the prevalence of riOSA [15.7 vs 2.2% (p < 0.001) for AASM_2007Rec and AASM_2012Rec, respectively]. The prevalence of rpOSA remained the same for each hypopnoea criteria although the prevalence of nssOSA increased with the AASM_2012Rec hypopnoea criteria [53.0 vs 66.4% (p < 0.006) for AASM_2007Rec and AASM_2012Rec, respectively]. There were no differences in clinical symptoms between the groups, irrespective of hypopnoea criteria used.

Conclusions

This study demonstrates that in comparison with AASM_2007Rec, the AASM_2012Rec hypopnoea criteria reduce the prevalence of riOSA but not rpOSA by reducing the ratio of REM respiratory events and NREM respiratory events.