Risk detection in patients with obstructive sleep apnea syndrome based on cardiovascular time series analysis^.

Obstructive sleep apnea represents the most common form of sleep-disordered breathing and has a high prevalence in patients with cardiovascular disease. Disturbed sleep is a potential risk factor for cardiovascular disorders such as arterial hypertension, cardiac ischemia, sudden cardiac death, and stroke. In this study we investigated polysomnographic records and analyzed the ECG, plethysmogram, respiration and SpO2 time series during wakefulness (WK), NREM, and REM sleep stages applying variability and coupling analyses methods. We enrolled 55 patients with obstructive sleep apnea syndrome (OSAS) and 29 healthy control subjects (CON: 45.9±14.9 years, 21 male) in this study. OSAS patients were subdivided into a low- and high-risk group (LR: 50.8±14.1 years, n=29, 21 male; HR: 57.2±13.4 years, n=26, 19 male) based on the Apnea-Hypopnea Index (AHI) (CON: 0-5 AHI, LR: 5-15 AHI, HR: >15 AHI). We could demonstrate the presence of an altered autonomic function in OSAS patients, differing from healthy controls. This altered autonomic function was mainly based on heart rate-, respiratory-, SpO2- and plethysmogram variability and their couplings. The discriminant analysis showed that the optimal set consisting of two autonomic indices revealed a high classification power (ACC=86.7%, AUC=90.3%, SENS=89.5% and SPEC=84.6%) when comparing low-risk and high-risk OSAS patients during WK. These results were slightly improved when analyzing REM sleep stages. Based on these results it seems to be possible to perform optimal risk stratification for OSAS patients based on autonomic indices. Based on our findings it is possible to differentiate between high-risk OSAS patient and low-risk OSAS patient at an early stage and in a promising manner allowing to set up therapy strategies for those patients in an early stage.