Prognostic markers of all-cause mortality in patients with atrial fibrillation: data from the prospective long-term registry of the German Atrial Fibrillation NETwork (AFNET)

Abstract

Aims: Atrial fibrillation (AF) is associated with a high risk of cardiovascular and non-cardiovascular death, even on anticoagulation. It is controversial, which conditions-including concomitant diseases and AF itself-contribute to this mortality. To further clarify these questions, major determinants of long-term mortality and their contribution to death were quantified in an unselected cohort of AF patients.

Methods and results: We established a large nationwide registry comprising 8833 AF-patients with a median follow-up of 6.5 years (45 345 patient-years) and central adjudication of adverse events. Baseline characteristics of the patients were evaluated as predictors of mortality using Cox regression and C-indices for determination of predictive power. Annualized mortality was highest in the first year (6.2%) and remained high thereafter (5.2% in men and 5.5% in women). Thirty-eight percent of all deaths were cardiovascular, mainly due to heart failure or sudden death. Sex-specific age was the strongest predictor of mortality, followed by concomitant cardiovascular and non-cardiovascular conditions. These factors accounted for 25% of the total mortality beyond age and sex and for 84% of the mortality differences between AF types. Thus, the electrical phenotype of the disease at baseline contributed only marginally to prediction of mortality.

Conclusion: Mortality is high in AF patients and arises primarily from heart failure, peripheral artery disease, chronic obstructive lung disease, chronic kidney disease, and diabetes mellitus, which, therefore, should be targeted to lower mortality. Parameters related to the electrical manifestation of AF did not have an independent impact on long-term mortality in our representative cohort.

Keywords: AF type; Atrial fibrillation; Comorbidities; German AFNET registry; Mortality; Outcome; Prognostic factors.