Abstract / Summary
Abstract Background Chronic obstructive pulmonary disease (COPD) is managed primarily with respiratory-based algorithms, even though cardiovascular comorbidity is a major determinant of adverse outcomes. We aimed to identify and characterise cardiopulmonary phenotypes defined by exacerbator profile and cardiovascular risk (CVR) in patients with COPD managed in primary care. Methods Cross-sectional, multicentre, retrospective study of 1,307 patients with COPD across 52 primary care centres in six Spanish regions. Phenotypes cross-classified the exacerbator profile (GOLD) with cardiovascular burden (established cardiovascular disease [CVD] / high risk / low-to-moderate risk). We used mixed-effects logistic regression, intraclass correlation, ROC curves, collinearity analysis and a preliminary unsupervised clustering analysis (machine learning). Results Median age was 72.6 years; 65.7% were men, 58.8% had moderate airflow obstruction, 31.2% were exacerbators and 36.8% had established CVD. The most frequent phenotype was the non-exacerbator with high CVR (40.6%), whereas the exacerbator phenotype with established CVD carried the greatest burden of hospitalisations and disease severity. We identified a hidden-risk phenotype—non-exacerbators with CVD—with a higher risk of hospitalisation and poorer clinical control (adjusted OR 1.99; 95% CI 1.21–3.25; p = 0.006 vs non-exacerbator phenotypes; sensitivity analysis). Clustering confirmed the clinical findings. Conclusions In COPD, jointly considering exacerbations and cardiovascular burden defines clinically relevant phenotypes, including a hidden-risk group of non-exacerbators with CVD. These practical, readily available axes complement the eosinophil count and support an integrated cardiopulmonary assessment.