Supraventricular arrhythmia, N-terminal pro-brain natriuretic peptide and troponin T concentration in relation to incidence of atrial fibrillation: a prospective cohort study

Background: Frequent supraventricular arrhythmia is associated with increased incidence of atrial fibrillation. However, it is unknown whether the prognostic significance of supraventricular arrhythmia is modified by plasma levels of N-terminal pro-brain natriuretic peptide (NT-proBNP) or troponin T (TnT). This study examined the interrelationships between NT-proBNP, TnT levels and frequent supraventricular arrhythmia, and whether these biomarkers and a measure of frequent supraventricular arrhythmia could improve risk assessment for incidence of AF. Methods: Supraventricular extrasystoles (SVEs) and supraventricular tachycardias were assessed from 24-h electrocardiograph recordings in 373 individuals initially without AF. Elevated NT-pro-BNP, TnT and SVEs was defined as a measurement in the top quartile of the study population distribution. Incident cases of AF were retrieved by linkage with the Swedish National Patient Register. Results: During a mean follow-up of 15.4 years, 88 subjects had a diagnosis of AF. After multivariable adjustment, individuals with both elevated NT-proBNP and frequent SVEs had a significantly increased incidence of AF, compared to subjects without elevated NT-proBNP or frequent SVEs (hazard ratio (HR) 4.61, 95% confidence interval (CI) 2.45–8.69), and compared to individuals with either elevated NT-proBNP or frequent SVEs (both P < 0.05). HRs for frequent SVEs alone or elevated NT-proBNP alone were 2.32 (95% CI 1.33–4.06) and 1.52 (95% CI 0.76–3.05), respectively. The addition of NT-pro-BNP and SVEs to a validated risk prediction score for AF, CHARGE-AF, resulted in improved prediction (Harrell’s C 0.751 (95% CI 0.702–0.799) vs 0.720 (95% CI 0.669–0.771), P = 0.015). Conclusion: Subjects with both elevated NT-proBNP and frequent SVEs have substantially increased risk of AF, and the use of these variables could improve long-term prediction of incident AF.