AI-estimated heart age using 12-lead electrocardiograms significantly predicted both mortality and major adverse cardiovascular events, according to a new study.

Researchers developed the AI-derived heart age, referred to as “AI electrocardiograms (ECG)-heart age,” to predict both mortality and cardiovascular outcomes. They trained a deep neural network using 425,051 ECGs collected from January 2006-December 2021. The model was validated on a separate dataset of 97,058 ECGs, demonstrating a mean absolute error of 5.8 years.

The researchers’ findings indicated that an AI ECG-heart age exceeding chronological age by 6 years or more was significantly associated with higher all-cause mortality (hazard ratio [HR], 1.60) and major adverse cardiovascular events (HR, 1.91). Conversely, an AI ECG-heart age 6 years younger than chronological age was associated with reduced mortality (HR, 0.82) and fewer major adverse cardiovascular events (HR, 0.78).

The analysis also revealed notable changes in ECG features, including the PR interval, QRS duration, QT interval, and corrected QT interval (QTc), as AI ECG-heart age increased. “This suggests that as individuals biological age, there may be changes in their cardiac conduction system and ventricular depolarization and repolarization, which are reflected in ECG changes," wrote the researchers.

The study, published in Frontiers in Cardiovascular Medicine, is limited by its retrospective design and the homogeneous sample of predominantly healthy Koreans. The researchers noted that further studies are needed to evaluate, improve, and create more accurate aging biomarkers, understand the AI-ECG age according to sex differences, and examine the usefulness of the AI-ECG heart age model in clinical practice.

Although more research needs to be conducted, the researchers noted the importance of the findings. "This concept can provide patients and their physicians with further tailored cardiac health information in engaged and motivated patients, which would contribute to the earlier implementation of a better, healthy lifestyle."

All authors are employed by DeepCardio Inc. and the study received funding from Inha University Research Grant and the Korean Government's Institute of Information & Communications Technology Planning & Evaluation.