Impact of respiratory events associated with autonomic arousals using type 1 sleep laboratory and type 3 ambulatory recordings

Abstract

Introduction The aim was to compare standard apnea-hypopnea indices (AHI) and respiratory disturbance indices with RERA (RDI) to indices including breathing events with autonomic arousals (AA) recorded in suspected sleep-disordered breathing populations using AASM type 1 and 3 recordings. Materials and methods Type 1: 73 subjects underwent overnight in-laboratory polysomnography. The following respiratory outcomes were scored: 4% oxygen desaturation index (ODI), apnea-hypopnea index (AHI), respiratory disturbance index with EEG arousals (RDIe), and RDI with AA (RDIa). Type 3: 89 subjects underwent home recordings for one night with ambulatory monitoring. As above, ODI, AHI, and RDIa were scored (not RDIe). Patients were classified into severity categories as no (0) to mild (1), moderate (2), and severe (3). Analyses performed were: (1) frequency of migration from lower to higher category; (2) Bland–Altman (B-A) to assess agreement between AHI, RDIe, and RDIa (type 1 monitoring), and AHI and RDIa (type 3 monitoring) (limits of agreement defined as the interval of the mean difference ±1.96 ×SD). Results Type 1: A linear severity trend was observed for all mean values of ODI, AHI, RDIe, and RDIa. 29% (21/73) of subjects migrated using RDIe vs. RDIa: 14% from 0 to 1, 11% 1 to 2, and 4% 2 to 3. B-A plots show a 6.63 mean difference (interval [–1.39; 14.66]) for RDIa vs. AHI, and 3.95 difference for RDIa vs. RDIe with interval [–2.83; 10.73]. Type 3: No linear trend was observed for ODI, AHI, and RDIa. 57% (45/79) of subjects migrated using AHI vs. RDIa: 33% from 0 to 1, 14% 1 to 2, 4% 2 to 3, and 4% 0 to 2. B-A plots show a 6.91 mean difference between RDIa and AHI (RDIa-AHI) with limits of agreement [–2.90; 16.73]. Conclusion Autonomic arousal events produce a similar increase in indices over AHI in type 1 and type 3 recordings, with a smaller difference compared to RDIe for type 1. Presumably, AA events captured most of the EEG arousal events not counted in type 3 studies. Unsurprisingly, the impact on AHI severity change was greater in the type 3 study (no EEG). The significance of events with AA remains unclear: their identification in type 3 studies may lead to better concordance with type 1 studies and improved disease definition. Acknowledgements Supported by Canada Res Chair and CHUM.