A clinic-based cluster analysis in patients with moderate-severe obstructive sleep apnea (OSA) in Chile

Abstract

RationalePatients commonly report differences in either clinical or symptomatic profiles, despite having the same severity of obstructive sleep apnea (OSA). ObjectiveTo identify clinical and symptomatic phenotypes and to evaluate cardiovascular mortality in each phenotype. MethodsData from 1370 participants (788 with moderate-severe OSA and 582 controls as a reference group) were extracted using the SantOSA database. Sixteen variables were analyzed using latent class analysis to define clinical subtypes. The association between subtypes and cardiovascular mortality was evaluated using Kaplan–Meier survival analysis and the Cox proportional hazards model. Adjusted hazard ratios (HRs) with confidence intervals (CIs) were modified by cardiovascular confounders. ResultsThe median observation period was 5.2 years. We found four clusters: cluster #1: symptomatic men with major comorbidities (n = 252); cluster #2: symptomatic women with comorbidities (n = 154); cluster #3: asymptomatic men with comorbidities (n = 143); and cluster #4: symptomatic young men without major comorbidities (n = 239). In cluster #1, mortality was 4.76% and was independently associated with age (HR 1.12; CI 1.07–1.17), type 2 diabetes mellitus (HR 3.37; CI 1.29–8.78) and coronary heart disease (HR 3.85; CI 1.27–11.56); in cluster #2, mortality was 3.89% and was independently associated with age (HR 1.12; CI 1.06–1.19) and the oxygen desaturation index (ODI, HR 1.02; CI 1.01–1.04); and in cluster #3, mortality was 3.49% (HR 3.50; CI 1.03–11.90) and was independently associated with age (HR 1.19; CI 1.10–1.29). In cluster #4, mortality was 1.25% and showed nonsignificant associations. ConclusionIn patients with moderate-severe OSA, we described four phenotypes of patients according to clinical features with different risks of cardiovascular mortality. Study registerISRCTN62293645.