Resumen: Background: Evidence regarding the efficacy of positive airway pressure (PAP) therapy in reducing the risk of non-fatal major cardiovascular events (NF-MACE) and mortality in patients with obstructive sleep apnoea (OSA) remains controversial. This study aims to quantify the impact of PAP therapy on these risks and develop a predictive risk estimator.
Methods: We conducted a multicentre, observational, prospective study involving 5358 individuals diagnosed with OSA, with a median (interquartile range (IQR)) follow-up of 14 (10–15) years. We derived and validated a risk estimator of NF-MACE (including myocardial infarction, stroke, revascularisation procedures and congestive heart failure) and all-cause mortality, incorporating PAP adherence alongside clinical and sleep-related data.
Results: The cohort had mean±sd age 55±11 years, body mass index 32.0±5.4 kg·m−2 and apnoea–hypopnoea index (AHI) 35±22 events·h−1; 26% were females and 1467 (37%) were PAP adherent. Over the follow-up period, 754 participants experienced NF-MACE, while 858 deaths were recorded. Significant predictors included prior cardiovascular events, non-high-density lipoprotein cholesterol ≥200 mg·dL−1, COPD diagnosis, AHI >30 events·h−1 and age >60 years. PAP adherence was protective (OR 0.46, 95% CI 0.38–0.56) and the absolute risk reduction varied depending on the baseline risk (median (IQR) 16% (12–18%)). The risk estimator yielded an area under the receiver operating characteristic curve of 0.75 and a Brier score of 0.17, with 64% sensitivity and 75% specificity.
Conclusions: The cohort had mean±sd age 55±11 years, body mass index 32.0±5.4 kg·m−2 and apnoea–hypopnoea index (AHI) 35±22 events·h−1; 26% were females and 1467 (37%) were PAP adherent. Over the follow-up period, 754 participants experienced NF-MACE, while 858 deaths were recorded. Significant predictors included prior cardiovascular events, non-high-density lipoprotein cholesterol ≥200 mg·dL−1, COPD diagnosis, AHI >30 events·h−1 and age >60 years. PAP adherence was protective (OR 0.46, 95% CI 0.38–0.56) and the absolute risk reduction varied depending on the baseline risk (median (IQR) 16% (12–18%)). The risk estimator yielded an area under the receiver operating characteristic curve of 0.75 and a Brier score of 0.17, with 64% sensitivity and 75% specificity. Idioma: Inglés DOI: 10.1183/13993003.00519-2025 Año: 2025 Publicado en: European respiratory journal 66, 5 (2025), 2500519 ISSN: 0903-1936 Tipo y forma: Article (PostPrint) Área (Departamento): Area Medicina (Dpto. Medicina, Psiqu. y Derm.)
Fecha de embargo : 2026-11-25
Exportado de SIDERAL (2026-01-12-11:10:19)