Hypoxemic features of obstructive sleep apnea and the risk of mortality: a cluster analysis

Introduction Patients with moderate to severe obstructive sleep apnea (OSA) have an increased risk of cardiovascular comorbidities and mortality. Although different subtypes of OSA have been described, data about oximetric parameters and their suitability to identify a different phenotype are scant. In this study, we evaluate the association between moderate to severe OSA and oximetric parameters included in the home sleep apnea test (HSAT) and the risks of all-cause mortality, cardiovascular mortality, and cancer mortality. Methods Adult patients with moderate to severe OSA from a clinical cohort in Chile were included (SantOSA study). We developed a latent class analysis (LCA) incorporating oximetric measures commonly reported on HSAT. Differences between the groups were evaluated using ANOVA and the chi-squared test. Survival curves were constructed using a Kaplan-Meier (log-rank) model, and adjusted hazard ratios of mortality were calculated using a Cox regression model following a confounder analysis of cardiovascular comorbidities. Results A total of 889 patients were included in the analysis. LCA identified three different clusters: Cluster 1, "nonhypoxemic" (n = 591); cluster 2, "moderately hypoxemic" (n = 297); and cluster 3, "severely hypoxemic" (n = 115). The mean follow-up was 4.7 years. The hypoxemic groups showed an increased risk of cardiometabolic comorbidities and an independent risk of all-cause mortality (adjusted HR 1.67 (CI 1.0-2.64) p value = 0.027). The moderately hypoxemic group had an adjusted HR of 2.92 (CI 1.00-8.58), p value = 0.05, while the severely hypoxemic group had an adjusted HR of 2.55 (CI 1.08-6.02), p value = 0.031. For cardiovascular mortality, we found an HR of 2.03 (CI 0.50-8.136), p value = 0.31, and for cancer mortality, we found an HR of 5.75 (CI 1.03-32.17), p value = 0.042. Conclusion Oximetric parameters are useful for describing a different phenotype with a high risk of mortality among patients with moderate to severe OSA, beyond the apnea-hypopnea index.