UPPER AIRWAY ANESTHESIA DELAYS AROUSAL FROM AIRWAY OCCLUSION INDUCED DURING HUMAN NREM SLEEP

Six healthy subjects (5 males and 1 female, 26-40 yr old) were studied during non-rapid-eye-movement (NREM) sleep to assess the role of upper airway (UA) afferents in the arousal response to induced airway occlusion. Subjects wore an airtight face mask attached to a low-resistance one-way valve. A valve in the inspiratory circuit allowed instantaneous inspiratory airway occlusion and release; the expiratory circuit remained unoccluded at all times. Each subject was studied during two nights. On one night, occlusions were created during stable stage 2 NREM sleep before and after application of 4% lidocaine to the oral and nasal mucosa. On the other night, the protocol was duplicated with saline ("sham anesthesia") rather than lidocaine. The order of nights was randomized. Occlusions were sustained until electroencephalographic arousal. Three to 12 occlusions were performed in each subject for each of the four parts of the protocol (pre- and post-lidocaine, pre- and post-saline). The auditory threshold for arousal (1,500-Hz tone beginning at 30 dB) was also tested before and after UA lidocaine. For the group, arousal time after UA anesthesia was prolonged compared with preanesthesia arousal time (P < 0.001); arousal time after sham anesthesia did not significantly increase from before sham anesthesia (P = 0.9). The increase in arousal time with UA anesthesia was greater than the increase with sham anesthesia (P < 0.001). The auditory arousal threshold did not increase after UA anesthesia. Inspiratory mask pressure, arterial O2 saturation of hemoglobin, and end-tidal PCO2 during occlusions were similar before and after UA anesthesia. These data suggest a role for afferent information from the UA in the arousal response to inspiratory airway occlusion in humans during NREM sleep.