Obstructive sleep apnea (OSA) is a common disorder, with roughly 1 in 5 adults estimated to have at least mild OSA and 1 in 15 estimated to have at least moderate OSA (Young 2002). Furthermore, OSA causes a number of adverse cardiovascular, neurocognitive, and daytime functional consequences. As a result, understanding the pathophysiology and clinical consequences of OSA is a major health concern.

Routine clinical variables obtained from polysomnography (PSG), such as apnea-hypopnea index (AHI) poorly predict the success of the non-positive-airway-pressure (non-PAP) therapies, mainly due to the different phenotype traits that cause sleep apnea. Also, they poorly predict the adverse health consequences of sleep apnea, possibly due to heterogeneity of cardiorespiratory responses to individual apneas and hypopneas.