Correlation analysis between sleep quality and cardiac function indicators in patients with chronic heart failure

Abstract Objective To assess sleep quality and its influencing factors in patients with chronic heart failure (CHF). Methods A study of 112 CHF patients was conducted using the Pittsburgh Sleep Quality Index (PSQI). Clinical and socio‐demographic data were collected. Relationships were analyzed via multiple linear regression (with LASSO for variable selection), and feature importance was evaluated with a random forest model, validated by training‐test splitting and SHAP analysis. Results PSQI scores were significantly elevated in CHF patients. Multiple regression identified NYHA class and NT‐proBNP as independent predictors of poor sleep, while longer heart failure duration was inversely associated with PSQI in the adjusted model—a finding that warrants further investigation. LASSO selected these three variables. Random forest indicated NT‐proBNP had the highest feature importance, confirmed by SHAP, though the model's predictive performance was limited ( R 2 = 0.049). Conclusion Impaired sleep is prevalent in CHF and is significantly associated with disease severity markers NT‐proBNP and NYHA class, highlighting sleep management as a potential component of comprehensive care.