Risk factors of central sleep apnea syndrome related to heart failure

Objective To analyze the incidence of central sleep apnea syndrome (CSAS) in patients with heart failure, explore the related factors and possible pathogenesis, and establish a multi-index prediction model.

Methods A total of 78 patients with chronic heart failure from the respiratory department and cardiology department were selected as study objects. The patients were divided into left heart failure group (39 cases) and right heart failure group (39 cases). Sleep monitoring index was analyzed, left ventricular ejection fraction (EF), right anterior ventricular wall thickness, tricuspid ring systolic displacement (TAPSE) and tricuspid ring systolic peak flow velocity (S') were determined, and association analysis was performed combining with age, body mass index, gender, smoking history, and blood carbon dioxide partial pressure p_a(CO₂). ROC curves were drawn for indicators with statistically significant differences in regression results.

Results The age in the right heart failure group was higher than the left heart failure group (P < 0.05). The results of multivariate Logistic regression showed that left heart failure, age ≥ 60 years old, and p_a(CO₂) < 39 mmHg were risk factors for CSAS. ROC analysis results showed that the sensitivity and specificity of regression prediction model constructed by combined application of p_a(CO₂) baseline level, left and right heart failure location, and age for CSAS prediction in patients with heart failure were 77.8% and 82.9%, respectively. Patients with apnea in left heart failure group accounted for a higher proportion, those with apnea in right heart failure group had mild TAPSE and S' decrease, while EF was lower in left heart failure group.

Conclusion Compared with the EF, clinical physicians should pay more attention to patients with low p_a(CO₂), because these patients have a higher risk of central sleep apnea. In the right heart failure patients, apnea may be more likely to occur in patients with mild right ventricular dysfunction, but it needs larger sample support for verification.