In collapsible biologic conduits, occlusion and cessation of flow occur when upstream pressure falls below a critical pressure (Pcrit). To examine the relationship between Pcrit and the development of upper airway occlusion, we examined the relationship between maximal inspiratory airflow and nasal pressure in seven normal subjects during sleep. At varying levels of subatmospheric pressure applied to a nasal mask during non-rapid-eye-movement (NREM) sleep, maximal inspiratory airflow decreased in proportion to the level of nasal pressure. When nasal pressure fell below a Pcrit, subjects demonstrated upper airway occlusions terminated by arousals. In these normal subjects, the upper airway Pcrit was found to be -13.3 +/- 3.2 (SD) cmH2O. In four subjects who sustained sleep while nasal pressure remained below the Pcrit, recurrent occlusive apneas were demonstrated. The relationship between maximal inspiratory airflow and nasal pressure in each subject was fit by linear regression and demonstrated upper airway Pcrit at the zero-flow intercept that were not significantly different from those observed experimentally. These data demonstrate that the normal human upper airway during sleep is characterized by a negative Pcrit and that occlusion may be induced when nasal pressure is decreased below this Pcrit.

Recent studies suggest that arousal is the dominant factor acutely increasing blood pressure in obstructive sleep apnea and that neither stimulation of chemoreceptors nor mechanical factors associated with large negative swings in intrapleural pressure substantially contribute to the rise in blood pressure associated with each obstructive apneic event. A canine model of obstructive sleep apnea was used to examine the relative contributions of these mechanisms in the blood pressure response to induced airway obstruction during non-rapid-eye-movement sleep. In part A of the study, the arousal response was eliminated from an obstructive event by restoring airway patency just before the expected arousal, allowing blood pressure responses to be compared between obstructive events with and without arousal. In part B of the study, the protocol of Part A was repeated after pharmacological blockade of the autonomic nervous system with hexamethonium (20 mg/kg iv), eliminating neurally mediated responses due to arousal, stimulation of chemoreceptors, or other reflexes, while maintaining any mechanical effects on blood pressure related to swings in intrapleural pressure. The results of part A (n = 4 dogs) show that obstructive apneic events of 28.5 +/- 3.1 s duration, with arterial hemoglobin desaturation to 92.9 +/- 0.8% and airway pressure swings of -37.6 +/- 6 mmHg, significantly increased mean arterial pressure (MAP) by 13.8 +/- 1.5 mmHg in the absence of arousal (P < 0.005). In comparison, when arousal was allowed to occur, MAP increased by a further 11.8 +/- 1.2 mmHg (P < 0.01). In part B (n = 3 dogs), there was no change in MAP during the obstructive apneic event, and MAP fell by > 10 mmHg in the postobstruction period whether or not arousal occurred (P < 0.05). We conclude that neural reflexes, but not mechanical factors, substantially contribute to the acute blood pressure response to an obstructive apneic event and that arousal produces a separate, additional acute hypertensive response.

We hypothesized that upper airway collapsibility is modulated dynamically throughout the respiratory cycle in sleeping humans by alterations in respiratory phase and/or airflow regimen. To test this hypothesis, critical pressures were derived from upper airway pressure-flow relationships in six tracheostomized patients with obstructive sleep apnea. Pressure-flow relationships were generated by varying the pressure at the trachea and nose during tracheostomy (inspiration and expiration) (comparison A) and nasal (inspiration only) breathing (comparison B), respectively. When a constant airflow regimen was maintained throughout the respiratory cycle (tracheostomy breathing), a small yet significant decrease in critical pressure was found at the inspiratory vs. end- and peak-expiratory time point [7.1 +/- 1.6 (SE) to 6.6 +/- 1.9 to 6.1 +/- 1.9 cmH(2)O, respectively; P < 0.05], indicating that phasic factors exerted only a modest influence on upper airway collapsibility. In contrast, we found that the inspiratory critical pressure fell markedly during nasal vs. tracheostomy breathing [1.1 +/- 1.5 (SE) vs. 6.1 +/- 1.9 cmH(2)O; P < 0.01], indicating that upper airway collapsibility is markedly influenced by differences in airflow regimen. Tracheostomy breathing was also associated with a reduction in both phasic and tonic genioglossal muscle activity during sleep. Our findings indicate that both phasic factors and airflow regimen modulate upper airway collapsibility dynamically and suggest that neuromuscular responses to alterations in airflow regimen can markedly lower upper airway collapsibility during inspiration.