Abstract
Introduction: In-hospital pneumonia is a serious complication of acute stroke, linked with longer stay, worse outcomes, and higher mortality. Sleep apnea, common yet underdiagnosed in stroke, may increase pneumonia risk through impaired airway clearance, hypoxia, aspiration, and systemic inflammation. While post-stroke pneumonia is often attributed to dysphagia and immune dysfunction, whether sleep apnea independently increases pneumonia risk in stroke patients remains unexamined in large real-world cohorts.
Methods: We conducted a retrospective cohort study using the Florida Stroke Registry, which includes 170+ hospitals in the American Heart Association's Get With The Guidelines-Stroke program. Adult patients discharged between January 2010 and January 2025 with complete data on sleep apnea and pneumonia were included. The primary outcome was in-hospital pneumonia; the main exposure was history of sleep apnea. Multivariable logistic regression adjusted for demographic (age, sex, race/ethnicity, insurance), behavioral (smoking), clinical (BMI category, NIHSS score, stroke subtype, thrombolysis [IV/IA tPA], hypertension, diabetes, dysphagia, coronary artery disease, atrial fibrillation, dyslipidemia, deep vein thrombosis/pulmonary embolism [DVT/PE]), and system-level factors (arrival mode, stroke center type, and region). Adjusted odds ratio (aOR) with 95% confidence intervals (CI) are reported. Model diagnostics showed good fit (mean VIF = 2.10).
Results: Among 189,757 stroke patients, 2.1% had sleep apnea, and 4.6% developed pneumonia. Pneumonia occurred in 7.9% of patients with sleep apnea vs. 4.5% without. Sleep apnea was independently associated with higher in-hospital pneumonia risk (aOR=1.76, 95% CI: 1.55-2.00). Other predictors included male sex, atrial fibrillation, diabetes, dyslipidemia, depression, and smoking. Stroke severity showed a graded association, with severe stroke carrying the highest odds. Dysphagia and DVT/PE had particularly strong associations (aORs >2-3). Comprehensive and primary stroke centers reported higher pneumonia rates (Figure 1).
Conclusion: In acute stroke care, sleep apnea independently predicts in-hospital pneumonia, likely due to aspiration risk, respiratory compromise, and systemic inflammation. Given the challenges of formal screening in acute care, efforts should focus on recognizing pre-existing sleep apnea and supporting adherence to therapies (e.g., CPAP/BiPAP) alongside standard pneumonia prevention protocols.