Abstract
Rationale: Apneic individuals have reduced airway caliber during sleep. The biomechanical changes in upper airway anatomy contributing to this airway narrowing are largely unknown. Objectives: We sought to investigate the state-dependent (wake vs. sleep) biomechanical behavior of the upper airway soft-tissue and craniofacial structures. Methods: Upper airway magnetic resonance imaging was performed in 15 sleep-deprived control subjects (apnea-hypopnea index, <5; 0.3 ± 0.5 events per hour) and 12 sleep-deprived apneic subjects (apnea-hypopnea index, ⩾5; 35.2 ± 18.1 events per hour) during wake and sleep and analyzed for airway measures and soft-tissue/mandibular movement. Results: In the retropalatal region, control subjects showed sleep-dependent reductions (P ⩽ 0.037) in average cross-sectional airway area (CSA), minimum CSA, and anteroposterior and lateral dimensions. Apneic subjects showed sleep-dependent reductions (P ⩽ 0.002) in average CSA, minimum CSA, and anteroposterior and lateral dimensions. In the retroglossal region, control subjects had no sleep-dependent airway reductions. However, apneic subjects had sleep-dependent reductions in minimal CSA (P = 0.001) and lateral dimensions (P = 0.014). Control subjects only showed sleep-dependent posterior movement of the anterior-inferior tongue octant (P = 0.039), whereas apneic subjects showed posterior movement of the soft palate (P = 0.006) and all tongue octants (P ⩽ 0.012). Sleep-dependent medial movement of the lateral walls was seen at the retropalatal minimum level (P = 0.013) in control subjects and at the retropalatal and retroglossal minimum levels (P ⩽ 0.017) in apneic subjects. There was posterior movement of the mandible in apneic subjects (P ⩽ 0.017). Conclusions: During sleep, control and apneic subjects showed reductions in retropalatal airway caliber, but only the apneic subjects showed retroglossal airway narrowing. Reductions in anteroposterior and lateral airway dimensions were primarily due to posterior soft palate, tongue and mandibular movement and to medial lateral wall movement. These data provide important initial insights into obstructive sleep apnea pathogenesis.