RISKS OF LONG-TERM MORTALITY AND MAJOR ADVERSE CARDIAC EVENTS ASSOCIATED WITH DIABETES MELLITUS IN PATIENTS HOSPITALIZED FOR ATRIAL FIBRILLATION

Abstract

Diabetes mellitus (DM) is an independent risk factor for cardiovascular diseases and is also a component of clinical predictive rules used to evaluate the stroke risk of patients with atrial fibrillation (AF). However, the degree to which DM influences long-term mortality and other major adverse outcomes of patients with AF remains unclear. We aim to determine whether DM is associated with increased risks of long-term mortality and hospitalizations for stroke, transient ischemic attack (TIA), myocardial infarction (MI) and heart failure (HF) in patients hospitalized for AF. Using the Québec integrated chronic disease surveillance system (QICDSS), the linkage of provincial health administrative databases, we conducted a population-based retrospective cohort study of all patients aged ≥20 years hospitalized in Québec for a primary diagnosis of AF between 1998-2002. A validated definition of DM was used to identify DM. We excluded patients who died during the index hospitalization and those who developed DM after the index hospitalization. Patients with cardiothoracic surgery 30 days prior to index hospitalization and hyperthyroidism, pulmonary embolism, or rheumatic heart disease 365 days before AF diagnosis were also excluded. The primary endpoint was all-cause mortality; the secondary endpoints considered separately were hospitalizations for stroke/TIA, MI and HF. We used multivariate regression Cox proportional hazards to determine the independent risks associated with DM for all-cause mortality and the other secondary endpoints. Covariates entered in the models were age, sex, hypertension, prior HF, coronary artery disease, previous stroke/TIA, renal insufficiency, obstructive lung disease and peripheral vascular disease. A total of 15,031 patients were followed for a mean of 12.1 years; 2,752 patients with DM and 12,279 without. DM patients were slightly older (72±10 years) than non-DM patients (70±13 years, p<0.0001), with more females (53% vs 50%, p=0.0299). DM was independently associated with increased risks of all-cause mortality and hospitalizations for stroke/TIA, MI and HF. Adjusted hazard ratios were 1.48 (95% confidence intervals (CI): 1.40-1.56) for all-cause mortality, 1.38 (95% CI: 1.24-1.54) for stroke/TIA, 1.66 (95% CI: 1.48-1.86) for MI and 1.54 (95% CI: 1.45-1.64) for HF. Our results suggest that DM is associated with approximately 50% increases in long-term risks of all-cause mortality and hospitalizations for stroke/TIA, MI and HF in patients hospitalized for AF. Future studies are needed to determine whether more intensive medical therapies/interventions can improve long-term survival and reduce the risks of hospitalization for stroke/TIA, MI and HF in DM patients hospitalized for AF.