What is the clinical evidence from this study?

Study design: Cross-Sectional. Population: Atrial fibrillation (n=2031). Intervention: PROPHECG-Age Single (AI-ECG age gap). Primary outcome: Prevalent atrial fibrillation (OR 1.03, 95% CI 1.01-1.04).

What does this research mean for the field?

A 1-year increase in AI-derived ECG age gap is associated with a 3% increased odds of atrial fibrillation and a 0.8 percentage point rise in AF burden. Novelty: ClaimNovelty.NOVEL_FINDING. Consensus alignment: ConsensusAlignment.NEUTRAL.

What question did this study set out to answer?

To investigate the relationship between AI-derived electrocardiographic age and atrial fibrillation risk.

January 20, 2026Open Access

Wearable device derived electrocardiographic age and its association with atrial fibrillation

SPSeung Hyun ParkYonsei University JJJu Hyun JinYonsei University JKJongwoo KimUnisys (United States)

Key Result

Each 1-year increase in the AI-ECG age gap derived from wearable single-lead ECGs was associated with a 3% higher odds of prevalent atrial fibrillation (OR 1.03).

Key Points

To investigate the relationship between AI-derived electrocardiographic age and atrial fibrillation risk.
Developed AI model PROPHECG-Age Single for estimating ECG age from single-lead wearable ECGs.
Trained the model using synthetic signals converted from 12-lead ECGs.
Validated the model in two independent wearable cohorts.
Mean absolute errors in ECG age predictions were 10.01 and 11.88 years across cohorts.
Each 1-year ECG age gap correlates with a 1.03 odds ratio for atrial fibrillation presence.
A 1-year increase in ECG age gap is linked to a 0.8 percentage point rise in AF burden.

Study Design

Type

Cross-Sectional (n=2,031)

Multicenter

Yes

Structured PICO

Does the AI-derived ECG-age gap from wearable single-lead ECGs associate with atrial fibrillation presence and burden in adults?

Population

Adults aged 20-90 years undergoing wearable single-lead ECG monitoring. The study included a training cohort of 1,008,566 12-lead ECGs from Severance Hospital, an internal validation cohort of 1,502 participants from the S-Patch registry, and an external validation cohort of 529 participants from the Memo Patch registry.

Intervention

PROPHECG-Age Single, an artificial intelligence model estimating ECG age from wearable single-lead ECGs to calculate the ECG-age gap (predicted minus chronological age).

Comparator

Chronological age (for age gap calculation) and participants without atrial fibrillation.

Outcome

Association of the AI-ECG age gap with atrial fibrillation presence and burden.surrogate

An AI model estimating ECG age from wearable single-lead ECGs demonstrated that a higher ECG-age gap is significantly associated with increased atrial fibrillation presence and burden, highlighting its potential as a digital biomarker.

Main Result

Effect estimate: OR 1.03 (95% CI 1.01-1.04)

Limitations

High mean absolute error (≈10–12 years) due to structural information loss inherent to single-lead settings
Cross-sectional design cannot establish causality or temporality
Quasi-episodic high-frequency sampling rather than real-time, beat-to-beat continuous tracking
Small sample size for the AF burden analysis in the external cohort
Single East Asian ethnicity limits generalizability to other populations
Lack of direct comparison between single-lead and standard 12-lead AI-ECG age gap within the same participants
Lack of uncertainty estimates limits the ability to differentiate true physiological variability from model prediction error

Abstract

Artificial intelligence (AI)-derived electrocardiographic (ECG) age is a promising marker of atrial fibrillation (AF) risk. We developed PROPHECG-Age Single-an AI model estimating ECG age from wearable single-lead ECGs-and examined whether the ECG-age gap (predicted minus chronological age) is associated with AF presence and burden in real-world self-monitoring context. One million 12-lead ECGs from a hospital were converted to synthetic single-lead signals via Cycle-Consistent Generative Adversarial Network and used to train a residual network-based model. Validation in two independent wearable cohorts (S-Patch ClinicalTrials.gov: NCT05119725, registered November 2021; Memo Patch ClinicalTrials.gov: NCT05355948, registered May 2022) showed mean absolute errors of 10.01 and 11.88 years, respectively. The pooled association with AF presence was significant (odds ratio 1.03 per 1-year gap), and for AF burden, each 1-year gap increase corresponded to a 0.8 percentage point rise. These findings support wearable-based AI-ECG age as a potential digital biomarker for proactive cardiovascular monitoring.

KI fragen

Bookmark

View Full Paper