Upper airway obstruction induced by negative-pressure ventilation in awake healthy subjects.

Negative-pressure ventilation (NPV) induces sleep-related upper airway obstruction. However, the precise mechanism and site of upper airway obstruction during NPV have not been worked out. We studied seven awake healthy volunteers (23-30 yr old) in an Emerson tank respirator. Subjects had the head outside the iron lung and breathed through a pneumotachograph, which yielded the airflow (V) signal. Supraglottic pressure (Psg) was measured with a catheter with the tip at the retroepiglottic level. Diaphragmatic electromyograms (EMGdi) were obtained from an esophageal bipolar electrode. Tidal volume was measured with an inductance plethysmograph. Measurements were done at -10, -20, and -30 cmH2O. At each pressure run subjects were asked to repeatedly relax or to actively breathe in phase with the respirator. Subjects had been previously trained to relax during NPV. During the relax runs there was no EMGdi activity. Stridor or wheezing occurred in all seven subjects during the relax runs but not during the active runs. Two patterns were associated with NPV during relax runs. One pattern was decreases in both V and Psg followed by zero values of these indexes, which corresponded to an inspiratory narrowing and closure of the glottis. These changes were visualized by fiber-optic bronchoscopy in one subject. The second pattern was a decrease in V and increase in Psg, which corresponded to an inspiratory supraglottic obstruction. In five subjects a supraglottic pattern was observed, whereas in two subjects glottic closure was seen. We conclude that muscular relaxation during NPV produces a decrease in the caliber of the upper airways at the glottic or supraglottic level. An uncoupling of upper airway muscle activity and the diaphragm might be the mechanism responsible for these changes.