Apnea-hypopnea index and the polysomnographic risk factors for predicting 5- to 8-year mortality in patients with OSA.

Abstract

The purpose of this study was to investigate the long-term mortality rates of patients withobstructive sleep apnea (OSA) who received an overnight polysomnogram (PSG)for obtaining the diagnosis and to determine the relationship between PSG parameters and overall mortality. Between 2007 and 2013, patients whohad overnight PSG and werediagnosed with OSAwere included in the study. Factors which are thought to influence mortality were assessed for 5-year and overall survival using the log rank test and Kaplan-Meier survival curves. Using multivariable Cox regression analysis, a model was constructed for factors influencing 5-year and overall survival. A total of 762 patients with a mean age of 52.7 (±10.8)and a dominance of men (74.7%)were studied. Gender, OSA severity subgroups, and apnea hypopnea index (AHI) were not statistically significantly associated with either 5-year or overall mortality (p<0.05 for both). Age, having a cardiovascular comorbidity, proportion ofrapid eye movement (%REM), and total sleep time with an oxyhemoglobin saturation of less than 90% (T90) all showed a significant correlation with overall all-cause mortality in the model. For 5-year mortality and overall mortality, the hazard ration (HR) for T90 was 3.6 (95% CI (1.6-8.0) p=0.001) and 3 (95% CI (1.6-5.7) p=0.001), respectively. The study findings suggest that not AHI but PSG parametersof hypoxia, mainly T90, having cardiovascular comorbidity, and %REM sleep weresignificant risk factors for all-cause mortality in patients with OSA. The association of OSA, hypoxia, and mortality is an area that deserves further study.