Effect of induced transient arousal on obstructive apnea duration

Six untreated male patients (age 19-55 yr) with obstructive sleep apnea underwent nocturnal polysomnography with acoustic stimulation to determine the effect of transient arousal on obstructive apneas during sleep. Binaural tone bursts (25-95 dB) were delivered in late expiration during the second obstructive apnea of a cycle consisting of four consecutive apneas. For the group, stimulated apneas were significantly shorter (P < 0.05, Fisher's protected least significant difference test) than were the unstimulated apneas when transient electrocortical arousal was elicited in both non-rapid-eye-movement (non-REM) sleep [mean 17 +/- 7 (SD) vs. 26 +/- 9, 23 +/- 10, and 26 +/- 12 s for 2nd vs. 1st, 3rd, and 4th apnea, respectively, of each cycle] and REM sleep (mean 19 +/- 10 vs. 35 +/- 15, 45 +/- 18, and 39 +/- 20 s). Without electrocortical arousal, the stimulated apnea was significantly shortened in non-REM (23 +/- 9 vs. 25 +/- 7, 24 +/- 8, and 26 +/- 8 s) but not in REM (32 +/- 16 vs. 37 +/- 12, 32 +/- 15, and 30 +/- 16 s). Tones delivered relatively early and late in the apnea were equally likely to be associated with resolution of the apnea. The nadir of arterial oxygen saturation of hemoglobin was inversely proportional to apnea length, with higher saturation nadirs associated with the stimulated apneas. These data indicate that transient arousal, induced by nonrespiratory stimulation, influences the resolution of obstructive apneas during sleep.