Arousals in sleep-disordered breathing: patterns and implications

STUDY OBJECTIVE: To describe the whole spectrum of electroencephalographic (EEG) transients associated with the termination and recovery of obstructed respiratory events and, thus, widen the recognized spectrum of arousal phenomena from sleep. DESIGN: Retrospective review of diagnostic polysomnograms. SETTING: American Academy of Sleep Medicine (AASM)-accredited multidisciplinary sleep disorders center. PATIENTS: 17 patents with obstructive sleep-disordered breathing. INTERVENTIONS: None. MEASUREMENTS AND RESULTS: Nasal airflow using a nasal-cannula-pressure-transducer system and oral flow by a thermistor were used to score apneas and hypopneas; the latter included flow-limitation events. The EEG patterns that crested or occurred within 2 to 3 seconds of respiratory recovery were recorded, and posthoc categories were created for the purpose of tabulation ranging from an AASM 3-second arousal to a single K-complex with no electromyographic increase. Chi-square statistic was calculated to assess the difference in EEG patterns at event termination between apneas and hypopneas. Score-rescore agreement was tested. Apneas were significantly more likely to be associated with a 3-second arousal than were hypopneas, but all types of EEG change were seen with both types of events. Spindles were rarely seen with arousal-linked K-complexes. The majority of events in rapid eye movement sleep were terminated with visible electromyography tone increase. CONCLUSIONS: The spectrum of EEG change associated with the termination of respiratory events identified by using a nasal-cannula-pressure-transducer system is wider than that recognized as arousal phenomena by the 1992 AASM criteria. Scoring arousals with the 3-second rule may falsely minimize the apparent impact of abnormal breathing on sleep. The time may be right to update arousal recognition rules.