ESC Heart Failure - Machine learning-based prediction of elevated NT-proBNP among US general population

Aug 3, 2026

Natriuretic peptide-based pre-heart failure screening has been proposed in guidelines, yet an effective way to choose screening targets from the US general population remains unclear. This study assesses machine learning prediction models for elevated N terminal pro brain natriuretic peptide (NT-proBNP) levels using nationally representative NHANES data. Models are trained on 1999–2002 data and tested on 2003–2004, showing strong discrimination and stable performance with commonly available variables.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages10

Reading Time25 min

AccessFree to download

Abstract

Aims

Methods and results

Conclusions

Introduction

Screening rationale

Role of machine learning

Evidence gap

ESC HEART FAILURE ESC Heart Failure (2024) Published online in Wiley Online Library ([wileyonlinelibrary.com)](wileyonlinelibrary.com) DOI: 10.1002/ehf2.15056)[ DOI:](wileyonlinelibrary.com) DOI: 10.1002/ehf2.15056)[ 10.1002/ehf2.15056](wileyonlinelibrary.com) DOI: 10.1002/ehf2.15056) ORIGINAL ARTICLE Machine learning-based prediction of elevated N terminal pro brain natriuretic peptide among US general population Yuichiro Mori1 , Shingo Fukuma 1, Kyohei Yamaji2, Atsushi Mizuno3, Naoki Kondo4 and Kosuke Inoue4,5 * 1Department of Human Health Sciences, Graduate School of Medicine, Kyoto University, Kyoto, Japan; 2Department of Cardiovascular Medicine, Graduate School of Medicine, Kyoto University, Kyoto, Japan; 3Department of Cardiovascular Medicine, St. Luke’s International Hospital, Tokyo, Japan; 4Department of Social Epidemiology, Graduate School of Medicine, Kyoto University, Kyoto, Japan; and 5Hakubi Center for Advanced Research, Kyoto University, Kyoto, Japan Abstract Aims Natriuretic peptide-based pre-heart failure screening has been proposed in recent guidelines. However, an effective strategy to identify screening targets from the general population, more than half of which are at risk for heart failure or pre-heart failure, has not been well established. This study evaluated the performance of machine learning prediction models for predicting elevated N terminal pro brain natriuretic peptide (NT-proBNP) levels in the US general population. Methods and results Individuals aged 20–79 years without cardiovascular disease from the nationally representative National Health and Nutrition Examination Survey 1999–2004 were included. Six prediction models (two conventional regression models and four machine learning models) were trained with the 1999–2002 cohort to predict elevated NT-proBNP levels (>125 pg/mL) using demographic, lifestyle, and commonly measured biochemical data. The model performance was tested using the 2003–2004 cohort. Of the 10 237 individuals, 1510 (14 .8%) had NT-proBNP levels > 125 pg/mL. The highest area under the receiver operating characteristic curve (AUC) was observed in SuperLearner (AUC [95% CI] = 0.862 [0.847–0.878], P < 0.001 compared with the logistic regression model) . The logistic regression model with splines showed a comparable performance (AUC [95% CI] = 0 .857 [0 .841–0.874], P = 0 .08) . Age, albumin level, haemoglobin level, sex, estimated glomerular ﬁltration rate, and systolic blood pressure were the most important predictors. We found a similar prediction performance even after excluding socio-economic information (marital status, family income, and education status) from the prediction models. When we used different thresholds for elevated NT-proBNP, the AUC (95% CI) in the SuperLearner models 0 .846 (0 .830–0.861) for NT-proBNP > 100 pg/mL and 0 .866 (0 .849–0.884) for NT-proBNP > 150 pg/mL. Conclusions Using nationally representative data from the United States, both logistic regression and machine learning models well predicted elevated NT-proBNP. The predictive performance remained consistent even when the models incorporated only commonly available variables in daily clinical practice. Prediction models using regularly measured information would serve as a potentially useful tools for clinicians to effectively identify targets of natriuretic-peptide screening. Keywords Machine learning; NHANES; NT-proBNP; Pre-heart failure; Screening Received: 30 November 2023; Revised: 1 August 2024; Accepted: 21 August 2024 *Correspondence to: Kosuke Inoue, Department of Social Epidemiology, Graduate School of Medicine, Kyoto University, Yoshida-Konoecho, Sakyo-ku, Kyoto-shi, Kyoto 6068315, Japan. Email: [inoue.kosuke.](inoue.kosuke.2j@kyoto-u.ac.jp)[2](inoue.kosuke.2j@kyoto-u.ac.jp)[j@kyoto-u.ac.jp](inoue.kosuke.2j@kyoto-u.ac.jp) Introduction Natriuretic peptide-based pre-heart failure (pre-HF) screening is gaining increasing interest as a preventive measu

FAQs

Why is selecting screening targets for pre-heart failure important?

Most people are at risk for heart failure or pre-heart failure, making broad natriuretic-peptide testing resource intensive. Targeting groups with higher probabilities of elevated NT-proBNP improves feasibility.

How were the prediction models developed and evaluated?

Six models were trained on NHANES 1999–2002 data using demographic, lifestyle, and commonly measured biochemical variables, then tested on the 2003–2004 cohort.

Which factors were most important for predicting elevated NT-proBNP?

Age, albumin level, haemoglobin level, sex, estimated glomerular filtration rate, and systolic blood pressure were identified as the most important predictors.

Aug 3, 2026

Upload by River Wang