Abstract 13949: Biomarkers for Prediction of Atrial Fibrillation in European Cohorts
Introduction: Blood-based biomarkers are objective tools to assess susceptibility for atrial fibrillation (AF).
Hypothesis: Whether novel biomarkers improve risk prediction based on clinical risk factors and N-terminal pro B-type natriuretic peptide (Nt-proBNP) remains to be shown.
Methods: In individuals from two European community studies (N=32,318) we examined 11 circulating biomarkers representing lipids (total cholesterol, LDL cholesterol, HDL cholesterol, triglycerides, lipoprotein(a), apolipoprotein A1, apolipoprotein B), inflammation (C-reactive protein), renal function (cystatin C), myocardial damage (high sensitivity assayed troponin I (hsTnI)), and vitamin D in relation to incidence of AF in addition to clinical risk factors and Nt-proBNP. Multivariable-adjusted Cox regression models were computed for models comprising classical cardiovascular risk factors and biomarkers. Net reclassification improvement (NRI) was computed.
Results: The median age was 53.2 (interquartile range 44.0, 62.9) years, age range 24.1-97.6 years, 51.9% women. Over a median follow-up of 4.9 years N=893 AF cases occurred. In multivariable-adjusted models there was evidence for an association between one standard deviation increase in biomarker for total cholesterol (hazard ratio (HR) 0.90, 95% confidence interval (CI) 0.83-0.97;P=0.0042), LDL cholesterol (HR 0.89, 95% CI 0.82-0.96; P=0.0015), apolipoprotein B (HR 0.89, 95% CI 0.83-0.96; P=0.0033), cystatin C1/3 (HR 1.19, 95% CI 1.11-1.28; P<0.001), hsTnI1/3 (HR 1.09, 95% CI 1.04-1.13;P<0.001), besides Nt-proBNP (HR 1.99, 95% CI 1.86-2.13;P<0.001). Only Nt-proBNP increased the C-index markedly to 0.872 (95% CI 0.844-0.901) compared to the model based on clinical risk factors C-index 0.836 (95% CI 0.809-0.863). Similarly, only Nt-proBNP improved reclassification (NRI 0.219, 95% CI 0.159-0.279).
Conclusions: Several known and novel biomarkers including blood lipids (total and LDL cholesterol, lipoprotein B), hsTnI and cystatin C are related to incident AF in large European cohorts. None proved to be superior to Nt-proBNP in C-statistic and net reclassification analyses. A possible benefit in the clinical application of Nt-proBNP for AF risk prediction needs to be shown.
Author Disclosures: R.B. Schnabel: None. T. Niiranen: None. F. Gianfagna: None. J.K. Vishram: None. F.M. Ojeda: None. S. Costanzo: None. S. Soederberg: None. E. Vartiainen: None. M. Donati: None. G. Pasterkamp: None. M. Hughes: None. J. Kontto: None. M. Bobak: None. A. Schillert: None. H.M. den Ruijter: None. E. Mathiesen: None. W. König: None. W. König: None. W. König: None. S. Blankenberg: None. G. de Gaetano: None. T. Joergensen: None. K. Kuulasmaa: None. V. Salomaa: None. L. Iacoviello: None. T. Zeller: None.
- © 2016 by American Heart Association, Inc.