Phenogroup-based stratification of cardiovascular risk in obstructive sleep apnea

Background Obstructive sleep apnea (OSA) exhibits substantial phenotypic heterogeneity in sleep clinic populations; however, its prognostic relevance in patients with coronary artery disease (CAD) undergoing revascularization remains uncertain. Methods We pooled two prospective cohorts of patients with CAD (total n = 2318) treated with percutaneous coronary intervention (n = 1311) or coronary artery bypass grafting (n = 1001). All patients underwent baseline overnight sleep studies and detailed phenotyping. Major adverse cardiac and cerebrovascular events (MACCE: cardiovascular death, non-fatal myocardial infarction, non-fatal stroke, and unplanned revascularization) were recorded over a mean follow-up of 1.9 ± 1.1 years. Latent class analysis using nine demographic and clinical variables (age, sex, body mass index [BMI], ethnicity, diabetes, prior myocardial infarction, excessive daytime sleepiness, left ventricular ejection fraction, and revascularization type) was performed to identify phenogroups. Cox regression models were used to evaluate associations between OSA (apnea–hypopnea index ≥15 events/h) and MACCE overall and within each phenogroup. Results Phenogroup 1 (42.2%) comprised younger, symptomatic individuals (mean age 54 years; ESS >10 in 15.8%). Phenogroup 2 (30.9%) was characterized by a high burden of cardiometabolic comorbidities (diabetes 68.7%, prior myocardial infarction 67.5%). Phenogroup 3 (26.9%) consisted predominantly of older, lean Chinese patients (mean age 67 years; BMI 24 kg/m²). The prevalence of OSA was similar across phenogroups (45.5%, 50.7%, and 47.8%, respectively). During follow-up, 11.4% of patients experienced MACCE. The crude incidence of MACCE was higher in patients with OSA than in those without OSA in phenogroup 1 (2-year estimate: 13.5% vs. 6.4%; P = 0.028) and phenogroup 2 (15.9% vs. 9.3%; P = 0.003), but not in phenogroup 3 (12.7% vs. 10.1%; P = 0.171). Overall, OSA was associated with an increased risk of MACCE (adjusted hazard ratio [HR] 1.71, 95% confidence interval [CI] 1.34–2.19; P < 0.001). However, associations differed by phenogroup: significant in phenogroup 1 (adjusted HR 1.63, 95% CI 1.06–2.50; P = 0.027) and phenogroup 2 (adjusted HR 1.83, 95% CI 1.19–2.83; P = 0.006), but not in phenogroup 3 (adjusted HR 1.44, 95% CI 0.88–1.94; P = 0.151). Conclusions In revascularized patients with CAD, OSA is associated with increased cardiovascular risk, and this association varies by phenogroup. Younger, sleepy patients and those with cardiometabolic comorbidities appear to be at highest risk, whereas risk is attenuated in older, lean, predominantly Chinese patients. Phenogroup-based approaches may enhance risk stratification and inform targeted OSA management in CAD populations.Central illustrationlca--fig--3.png