Identification of Upper Airway Anatomic Risk Factors for Obstructive Sleep Apnea with Volumetric Magnetic Resonance Imaging

We used sophisticated volumetric analysis techniques with magnetic resonance imaging in a case-control design to study the upper airway soft tissue structures in 48 control subjects (apnea-hypopnea index, 2.0 +/- 1.6 events/hour) and 48 patients with sleep apnea (apnea-hypopnea index, 43.8 +/- 25.4 events/hour). Our design used exact matching on sex and ethnicity, frequency matching on age, and statistical control for craniofacial size and visceral neck fat. The data support our a priori hypotheses that the volume of the soft tissue structures surrounding the upper airway is enlarged in patients with sleep apnea and that this enlargement is a significant risk factor for sleep apnea. After covariate adjustments the volume of the lateral pharyngeal walls (p < 0.0001), tongue (p < 0.0001), and total soft tissue (p < 0.0001) was significantly larger in subjects with sleep apnea than in normal subjects. These data also demonstrated, after covariate adjustments, significantly increased risk of sleep apnea the larger the volume of the tongue, lateral pharyngeal walls, and total soft tissue: (1) total lateral pharyngeal wall (odds ratio [OR], 6.01; 95% confidence interval [CI], 2.62-17.14); (2) total tongue (OR, 4.66; 95% CI, 2.31-10.95); and (3) total soft tissue (OR, 6.95; 95% CI, 3.08-19.11). In a multivariable logistic regression analysis the volume of the tongue and lateral walls was shown to independently increase the risk of sleep apnea.