Quantifying sleep apnea heterogeneity using hierarchical Bayesian modeling

Obstructive Sleep Apnea (OSA) is a breathing disorder during sleep that affects millions of people worldwide. The diagnosis of OSA often occurs through an overnight polysomnogram (PSG) sleep study that generates a massive amount of physiological data. However, despite the evidence of substantial heterogeneity in the expression and symptoms of OSA, diagnosis and scientific analysis of severity typically focus on a single summary statistic, the Apnea-Hypopnea Index (AHI). To address the limitations inherent in such analyses, we propose a hierarchical Bayesian modeling approach to analyze PSG data. Our approach produces an interpretable vector of random effect parameters for each patient that govern sleep-stage dynamics, rates of OSA events, and impacts of OSA events on subsequent sleep-stage dynamics. We propose a novel approach for using these random effects to produce a Bayes optimal cluster of patients under K-means loss. We use the proposed approach to analyze data from the APPLES study. This analysis produces clinically interesting groups of patients with sleep apnea and a novel finding of an association between OSA expression and cognitive performance that is missed by an AHI-based analysis.