Formula for predicting OSA and the Apnea-Hypopnea Index in Koreans with suspected OSA using clinical, anthropometric, and cephalometric variables

Abstract

This study developed formulas to predict obstructive sleep apnea (OSA) and the Apnea-Hypopnea Index (AHI) in Korean patients with suspected OSA using clinical, anthropometric, and cephalometric variables. We evaluated relevant variables in 285 subjects with suspected OSA. These included demographic characteristics, sleep-related symptoms, medical history, clinical scales, anthropometric measurements including facial surface measurements, and cephalometric measurements. All participants underwent full-night laboratory polysomnography. The prediction formula for the probability of OSA was created by logistic regression analysis and confirmed by the bootstrap resampling technique. The formula for predicting the AHI was developed using multiple linear regression analysis. The probability of having OSA was as follows: p = 1 / (1 + exponential (exp)(-f) ), where f = -16.508 + 1.445 x loudness of snoring 4 + 0.485 x loudness of snoring 3 + 0.078 x waist circumference + 0.209 x subnasale-to-stomion distance + 0.183 x thickness of the uvula (UTH) supine + 0.041 x age. The AHI prediction formula was as follows: -112.606 + 3.516 x body mass index + 0.683 x mandibular plane-hyoid supine + 10.915 x loudness of snoring 4 + 6.933 x loudness of snoring 3 + 1.297 x UTH supine + 0.272 x age. This is the first study to establish formulas to predict OSA and the AHI in Koreans with suspected OSA using cephalometric and other variables. These results will contribute to prioritizing the order in which patients with suspected OSA are referred for polysomnography.