Exploring the incremental utility of circulating biomarkers for robust risk prediction of incident atrial fibrillation in European cohorts using regressions and modern machine learning methods

Betül Toprak; Stephanie Brandt; Jan Brederecke; Francesco Gianfagna; Julie K K Vishram-Nielsen; Francisco M Ojeda; Simona Costanzo; Christin S Börschel; Stefan Söderberg; Ioannis Katsoularis; Stephan Camen; Erkki Vartiainen; Maria Benedetta Donati; Jukka Kontto; Martin Bobak; Ellisiv B Mathiesen; Allan Linneberg; Wolfgang Koenig; Maja-Lisa Løchen; Augusto Di Castelnuovo; Stefan Blankenberg; Giovanni de Gaetano; Kari Kuulasmaa; Veikko Salomaa; Licia Iacoviello; Teemu Niiranen; Tanja Zeller; Renate B Schnabel

doi:10.1093/europace/euac260

Exploring the incremental utility of circulating biomarkers for robust risk prediction of incident atrial fibrillation in European cohorts using regressions and modern machine learning methods

Betül Toprak, Stephanie Brandt, Jan Brederecke, Francesco Gianfagna, Julie K K Vishram-Nielsen, Francisco M Ojeda, Simona Costanzo, Christin S Börschel, Stefan Söderberg, Ioannis Katsoularis, Stephan Camen, Erkki Vartiainen, Maria Benedetta Donati, Jukka Kontto, Martin Bobak, Ellisiv B Mathiesen, Allan Linneberg, Wolfgang Koenig, Maja-Lisa Løchen, Augusto Di CastelnuovoStefan Blankenberg, Giovanni de Gaetano, Kari Kuulasmaa, Veikko Salomaa, Licia Iacoviello, Teemu Niiranen, Tanja Zeller, Renate B Schnabel

Abstract

AIMS: To identify robust circulating predictors for incident atrial fibrillation (AF) using classical regressions and machine learning (ML) techniques within a broad spectrum of candidate variables.

METHODS AND RESULTS: In pooled European community cohorts (n = 42 280 individuals), 14 routinely available biomarkers mirroring distinct pathophysiological pathways including lipids, inflammation, renal, and myocardium-specific markers (N-terminal pro B-type natriuretic peptide [NT-proBNP], high-sensitivity troponin I [hsTnI]) were examined in relation to incident AF using Cox regressions and distinct ML methods. Of 42 280 individuals (21 843 women [51.7%]; median [interquartile range, IQR] age, 52.2 [42.7, 62.0] years), 1496 (3.5%) developed AF during a median follow-up time of 5.7 years. In multivariable-adjusted Cox-regression analysis, NT-proBNP was the strongest circulating predictor of incident AF [hazard ratio (HR) per standard deviation (SD), 1.93 (95% CI, 1.82-2.04); P < 0.001]. Further, hsTnI [HR per SD, 1.18 (95% CI, 1.13-1.22); P < 0.001], cystatin C [HR per SD, 1.16 (95% CI, 1.10-1.23); P < 0.001], and C-reactive protein [HR per SD, 1.08 (95% CI, 1.02-1.14); P = 0.012] correlated positively with incident AF. Applying various ML techniques, a high inter-method consistency of selected candidate variables was observed. NT-proBNP was identified as the blood-based marker with the highest predictive value for incident AF. Relevant clinical predictors were age, the use of antihypertensive medication, and body mass index.

CONCLUSION: Using different variable selection procedures including ML methods, NT-proBNP consistently remained the strongest blood-based predictor of incident AF and ranked before classical cardiovascular risk factors. The clinical benefit of these findings for identifying at-risk individuals for targeted AF screening needs to be elucidated and tested prospectively.

Original language	English
Journal	Europace : European pacing, arrhythmias, and cardiac electrophysiology : journal of the working groups on cardiac pacing, arrhythmias, and cardiac cellular electrophysiology of the European Society of Cardiology
ISSN	1099-5129
DOIs	https://doi.org/10.1093/europace/euac260
Publication status	E-pub ahead of print - 4 Jan 2023

Access to Document

10.1093/europace/euac260

Cite this

Toprak, B., Brandt, S., Brederecke, J., Gianfagna, F., Vishram-Nielsen, J. K. K., Ojeda, F. M., Costanzo, S., Börschel, C. S., Söderberg, S., Katsoularis, I., Camen, S., Vartiainen, E., Donati, M. B., Kontto, J., Bobak, M., Mathiesen, E. B., Linneberg, A., Koenig, W., Løchen, M-L., ... Schnabel, R. B. (2023). Exploring the incremental utility of circulating biomarkers for robust risk prediction of incident atrial fibrillation in European cohorts using regressions and modern machine learning methods. Europace : European pacing, arrhythmias, and cardiac electrophysiology : journal of the working groups on cardiac pacing, arrhythmias, and cardiac cellular electrophysiology of the European Society of Cardiology. https://doi.org/10.1093/europace/euac260

Exploring the incremental utility of circulating biomarkers for robust risk prediction of incident atrial fibrillation in European cohorts using regressions and modern machine learning methods. / Toprak, Betül; Brandt, Stephanie; Brederecke, Jan et al.

In: Europace : European pacing, arrhythmias, and cardiac electrophysiology : journal of the working groups on cardiac pacing, arrhythmias, and cardiac cellular electrophysiology of the European Society of Cardiology, 04.01.2023.

Research output: Contribution to journal › Journal article › Research › peer-review

Toprak, B, Brandt, S, Brederecke, J, Gianfagna, F, Vishram-Nielsen, JKK, Ojeda, FM, Costanzo, S, Börschel, CS, Söderberg, S, Katsoularis, I, Camen, S, Vartiainen, E, Donati, MB, Kontto, J, Bobak, M, Mathiesen, EB, Linneberg, A, Koenig, W, Løchen, M-L, Di Castelnuovo, A, Blankenberg, S, de Gaetano, G, Kuulasmaa, K, Salomaa, V, Iacoviello, L, Niiranen, T, Zeller, T & Schnabel, RB 2023, 'Exploring the incremental utility of circulating biomarkers for robust risk prediction of incident atrial fibrillation in European cohorts using regressions and modern machine learning methods', Europace : European pacing, arrhythmias, and cardiac electrophysiology : journal of the working groups on cardiac pacing, arrhythmias, and cardiac cellular electrophysiology of the European Society of Cardiology. https://doi.org/10.1093/europace/euac260

Toprak B, Brandt S, Brederecke J, Gianfagna F, Vishram-Nielsen JKK, Ojeda FM et al. Exploring the incremental utility of circulating biomarkers for robust risk prediction of incident atrial fibrillation in European cohorts using regressions and modern machine learning methods. Europace : European pacing, arrhythmias, and cardiac electrophysiology : journal of the working groups on cardiac pacing, arrhythmias, and cardiac cellular electrophysiology of the European Society of Cardiology. 2023 Jan 4. https://doi.org/10.1093/europace/euac260

Toprak, Betül ; Brandt, Stephanie ; Brederecke, Jan et al. / Exploring the incremental utility of circulating biomarkers for robust risk prediction of incident atrial fibrillation in European cohorts using regressions and modern machine learning methods. In: Europace : European pacing, arrhythmias, and cardiac electrophysiology : journal of the working groups on cardiac pacing, arrhythmias, and cardiac cellular electrophysiology of the European Society of Cardiology. 2023.

@article{70c617b8cbc04e6583c4186f0eedd882,

title = "Exploring the incremental utility of circulating biomarkers for robust risk prediction of incident atrial fibrillation in European cohorts using regressions and modern machine learning methods",

abstract = "AIMS: To identify robust circulating predictors for incident atrial fibrillation (AF) using classical regressions and machine learning (ML) techniques within a broad spectrum of candidate variables.METHODS AND RESULTS: In pooled European community cohorts (n = 42 280 individuals), 14 routinely available biomarkers mirroring distinct pathophysiological pathways including lipids, inflammation, renal, and myocardium-specific markers (N-terminal pro B-type natriuretic peptide [NT-proBNP], high-sensitivity troponin I [hsTnI]) were examined in relation to incident AF using Cox regressions and distinct ML methods. Of 42 280 individuals (21 843 women [51.7%]; median [interquartile range, IQR] age, 52.2 [42.7, 62.0] years), 1496 (3.5%) developed AF during a median follow-up time of 5.7 years. In multivariable-adjusted Cox-regression analysis, NT-proBNP was the strongest circulating predictor of incident AF [hazard ratio (HR) per standard deviation (SD), 1.93 (95% CI, 1.82-2.04); P < 0.001]. Further, hsTnI [HR per SD, 1.18 (95% CI, 1.13-1.22); P < 0.001], cystatin C [HR per SD, 1.16 (95% CI, 1.10-1.23); P < 0.001], and C-reactive protein [HR per SD, 1.08 (95% CI, 1.02-1.14); P = 0.012] correlated positively with incident AF. Applying various ML techniques, a high inter-method consistency of selected candidate variables was observed. NT-proBNP was identified as the blood-based marker with the highest predictive value for incident AF. Relevant clinical predictors were age, the use of antihypertensive medication, and body mass index.CONCLUSION: Using different variable selection procedures including ML methods, NT-proBNP consistently remained the strongest blood-based predictor of incident AF and ranked before classical cardiovascular risk factors. The clinical benefit of these findings for identifying at-risk individuals for targeted AF screening needs to be elucidated and tested prospectively.",

author = "Bet{\"u}l Toprak and Stephanie Brandt and Jan Brederecke and Francesco Gianfagna and Vishram-Nielsen, {Julie K K} and Ojeda, {Francisco M} and Simona Costanzo and B{\"o}rschel, {Christin S} and Stefan S{\"o}derberg and Ioannis Katsoularis and Stephan Camen and Erkki Vartiainen and Donati, {Maria Benedetta} and Jukka Kontto and Martin Bobak and Mathiesen, {Ellisiv B} and Allan Linneberg and Wolfgang Koenig and Maja-Lisa L{\o}chen and {Di Castelnuovo}, Augusto and Stefan Blankenberg and {de Gaetano}, Giovanni and Kari Kuulasmaa and Veikko Salomaa and Licia Iacoviello and Teemu Niiranen and Tanja Zeller and Schnabel, {Renate B}",

note = "{\textcopyright} The Author(s) 2023. Published by Oxford University Press on behalf of the European Society of Cardiology.",

year = "2023",

month = jan,

day = "4",

doi = "10.1093/europace/euac260",

language = "English",

journal = "Europace",

issn = "1099-5129",

publisher = "Oxford University Press",

}

TY - JOUR

T1 - Exploring the incremental utility of circulating biomarkers for robust risk prediction of incident atrial fibrillation in European cohorts using regressions and modern machine learning methods

AU - Toprak, Betül

AU - Brandt, Stephanie

AU - Brederecke, Jan

AU - Gianfagna, Francesco

AU - Vishram-Nielsen, Julie K K

AU - Ojeda, Francisco M

AU - Costanzo, Simona

AU - Börschel, Christin S

AU - Söderberg, Stefan

AU - Katsoularis, Ioannis

AU - Camen, Stephan

AU - Vartiainen, Erkki

AU - Donati, Maria Benedetta

AU - Kontto, Jukka

AU - Bobak, Martin

AU - Mathiesen, Ellisiv B

AU - Linneberg, Allan

AU - Koenig, Wolfgang

AU - Løchen, Maja-Lisa

AU - Di Castelnuovo, Augusto

AU - Blankenberg, Stefan

AU - de Gaetano, Giovanni

AU - Kuulasmaa, Kari

AU - Salomaa, Veikko

AU - Iacoviello, Licia

AU - Niiranen, Teemu

AU - Zeller, Tanja

AU - Schnabel, Renate B

PY - 2023/1/4

Y1 - 2023/1/4

N2 - AIMS: To identify robust circulating predictors for incident atrial fibrillation (AF) using classical regressions and machine learning (ML) techniques within a broad spectrum of candidate variables.METHODS AND RESULTS: In pooled European community cohorts (n = 42 280 individuals), 14 routinely available biomarkers mirroring distinct pathophysiological pathways including lipids, inflammation, renal, and myocardium-specific markers (N-terminal pro B-type natriuretic peptide [NT-proBNP], high-sensitivity troponin I [hsTnI]) were examined in relation to incident AF using Cox regressions and distinct ML methods. Of 42 280 individuals (21 843 women [51.7%]; median [interquartile range, IQR] age, 52.2 [42.7, 62.0] years), 1496 (3.5%) developed AF during a median follow-up time of 5.7 years. In multivariable-adjusted Cox-regression analysis, NT-proBNP was the strongest circulating predictor of incident AF [hazard ratio (HR) per standard deviation (SD), 1.93 (95% CI, 1.82-2.04); P < 0.001]. Further, hsTnI [HR per SD, 1.18 (95% CI, 1.13-1.22); P < 0.001], cystatin C [HR per SD, 1.16 (95% CI, 1.10-1.23); P < 0.001], and C-reactive protein [HR per SD, 1.08 (95% CI, 1.02-1.14); P = 0.012] correlated positively with incident AF. Applying various ML techniques, a high inter-method consistency of selected candidate variables was observed. NT-proBNP was identified as the blood-based marker with the highest predictive value for incident AF. Relevant clinical predictors were age, the use of antihypertensive medication, and body mass index.CONCLUSION: Using different variable selection procedures including ML methods, NT-proBNP consistently remained the strongest blood-based predictor of incident AF and ranked before classical cardiovascular risk factors. The clinical benefit of these findings for identifying at-risk individuals for targeted AF screening needs to be elucidated and tested prospectively.

AB - AIMS: To identify robust circulating predictors for incident atrial fibrillation (AF) using classical regressions and machine learning (ML) techniques within a broad spectrum of candidate variables.METHODS AND RESULTS: In pooled European community cohorts (n = 42 280 individuals), 14 routinely available biomarkers mirroring distinct pathophysiological pathways including lipids, inflammation, renal, and myocardium-specific markers (N-terminal pro B-type natriuretic peptide [NT-proBNP], high-sensitivity troponin I [hsTnI]) were examined in relation to incident AF using Cox regressions and distinct ML methods. Of 42 280 individuals (21 843 women [51.7%]; median [interquartile range, IQR] age, 52.2 [42.7, 62.0] years), 1496 (3.5%) developed AF during a median follow-up time of 5.7 years. In multivariable-adjusted Cox-regression analysis, NT-proBNP was the strongest circulating predictor of incident AF [hazard ratio (HR) per standard deviation (SD), 1.93 (95% CI, 1.82-2.04); P < 0.001]. Further, hsTnI [HR per SD, 1.18 (95% CI, 1.13-1.22); P < 0.001], cystatin C [HR per SD, 1.16 (95% CI, 1.10-1.23); P < 0.001], and C-reactive protein [HR per SD, 1.08 (95% CI, 1.02-1.14); P = 0.012] correlated positively with incident AF. Applying various ML techniques, a high inter-method consistency of selected candidate variables was observed. NT-proBNP was identified as the blood-based marker with the highest predictive value for incident AF. Relevant clinical predictors were age, the use of antihypertensive medication, and body mass index.CONCLUSION: Using different variable selection procedures including ML methods, NT-proBNP consistently remained the strongest blood-based predictor of incident AF and ranked before classical cardiovascular risk factors. The clinical benefit of these findings for identifying at-risk individuals for targeted AF screening needs to be elucidated and tested prospectively.

U2 - 10.1093/europace/euac260

DO - 10.1093/europace/euac260

M3 - Journal article

C2 - 36610061

JO - Europace

JF - Europace

SN - 1099-5129

ER -

Exploring the incremental utility of circulating biomarkers for robust risk prediction of incident atrial fibrillation in European cohorts using regressions and modern machine learning methods

Abstract

Access to Document

Fingerprint

Cite this