Abstract
Purpose Obstructive Sleep Apnoea (OSA) is a highly prevalent disease often treated with CPAP therapy, but long-term adherence remains a challenge. This study aimed to explore which phenotypes of OSA patients are more or less likely to achieve high adherence to CPAP therapy, based on pre-treatment characteristics. Methods We analysed demographic, socioeconomic, and clinical characteristics across predefined CPAP adherence groups (high, low, and no adherence) in a multicentre cohort of newly diagnosed OSA patients. Multivariate logistic regression was used to assess predictive accuracy, and observed patterns were synthesised into rule-based clinical phenotypes. Results Among 2845 patients, 41% were highly adherent, 20% had low adherence, and 39% were non-adherent after one year. High adherence was associated with cohabitation, higher education, employment, and greater disease severity. Low or no adherence was linked to comorbidity burden and social vulnerability. The predictive model yielded an accuracy of 60%. Based on group-level differences and model outputs, three clinically reasoned phenotypes were identified: (1) Symptomatic severe OSA with social support, (2) Multimorbidity-dominated, and (3) Non-employed working-age adults. Conclusion Distinct demographic, socioeconomic, and clinical profiles were associated with CPAP adherence. The identified phenotypes offer a practical framework for understanding adherence variation and may support more individualised approaches to treatment and follow-up.
| Originalsprog | Engelsk |
|---|---|
| Artikelnummer | 108822 |
| Tidsskrift | Sleep Medicine |
| Vol/bind | 141 |
| Antal sider | 10 |
| ISSN | 1389-9457 |
| DOI | |
| Status | E-pub ahead of print - 3 feb. 2026 |
Fingeraftryk
Dyk ned i forskningsemnerne om 'Clinical phenotypes and predictors of adherence to continuous positive airway pressure therapy in obstructive sleep apnoea: A multicentre cohort study'. Sammen danner de et unikt fingeraftryk.Citationsformater
- APA
- Standard
- Harvard
- Vancouver
- Author
- BIBTEX
- RIS