AI electrocardiogram-derived age: a step toward more physiological risk stratification in coronary artery bypass patients

The recently published article by Sawma and colleagues reported that advanced physiological age, as identified by artificial intelligence (AI) electrocardiogram (ECG)-derived age, is independently associated with worse operative outcomes and increased long-term mortality after isolated coronary artery bypass grafting (CABG) (1). AI-derived ECG age may therefore serve as a simple and robust tool for preoperative screening and risk stratification in patients undergoing CABG.

As the authors noted, currently used risk prediction models—such as the Society of Thoracic Surgeons (STS) score and the European System for Cardiac Operative Risk Evaluation (EuroSCORE)—do not incorporate frailty or overall physiological health status, despite their strong association with postoperative outcomes. Prior studies have demonstrated that ECG-based AI algorithms can estimate a patient’s chronological age, and discrepancies between AI-predicted and actual age may reflect underlying physiological health (2). Thus, leveraging the ubiquitous, noninvasive, and inexpensive 12-lead ECG to incorporate physiological health into risk prediction models is highly appealing. Notably, this line of research originated at the Mayo Clinic, which has led extensive and innovative work in AI-driven ECG analysis. Numerous studies have confirmed the feasibility and clinical acceptance of ECG-enabled AI algorithms for community-based screening of cardiovascular conditions—including atrial fibrillation and cardiac dysfunction (3-5)—as well as noncardiac diseases such as chronic liver disease (6). Integration of AI-derived ECG age into established models, such as the STS score and EuroSCORE II, will require overcoming challenges related to software implementation, regulatory considerations, and interpretability.

In the study by Sawma et al., patients with advanced AI-predicted ECG age (defined as an age gap >5 years) had higher body mass index and lower left ventricular ejection fraction, and were more likely to have congestive heart failure, renal dysfunction, prior myocardial infarction, diabetes mellitus, and a history of smoking. An increased ECG age relative to chronological age was also significantly associated with peripheral and cerebrovascular disease. These findings provide a plausible mechanistic explanation for the observed association between the AI-ECG age gap and increased all-cause mortality after CABG. Further studies are warranted to validate AI-driven ECG analysis across other cardiac and aortic surgeries requiring cardiopulmonary bypass.

Finally, several important questions remain. Although an AI-ECG age gap greater than five years appears useful for identifying high-risk patients, it remains a preoperative metric. How was this threshold determined, and to what extent does it reflect conventional ECG abnormalities such as conduction disturbances or repolarization changes? If intensive medical interventions—such as optimized pharmacotherapy or structured rehabilitation—are applied, is the AI-ECG age gap modifiable?

Moreover, given that this parameter can be repeatedly measured using standard ECGs, it may have potential as a dynamic marker for monitoring perioperative or postoperative treatment effects. However, short-term intra-patient variability is a critical limitation; prior data suggest that AI-ECG age may fluctuate by nearly nine years over short intervals due to transient physiological changes (7). As highlighted by Vander Heyde et al., statistical adjustments may be required to account for this variability.

In conclusion, AI-enhanced ECG represents a promising, noninvasive, and clinically impactful tool, and its continued development and validation are eagerly anticipated.

Acknowledgments

None.

Footnote

Provenance and Peer Review: This article was commissioned by the editorial office, Journal of Medical Artificial Intelligence. The article has undergone external peer review.

Peer Review File: Available at https://jmai.amegroups.com/article/view/10.21037/jmai-2026-0051/prf

Funding: None.

Conflicts of Interest: The author has completed the ICMJE uniform disclosure form (available at https://jmai.amegroups.com/article/view/10.21037/jmai-2026-0051/coif). The author has no conflicts of interest to declare.

Ethical Statement: The author is accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Sawma T, Arghami A, Schaff HV, et al. Risk stratification of coronary artery bypass patients using an artificial intelligence electrocardiogram-derived age. J Thorac Cardiovasc Surg 2026;171:201-209.e3. [Crossref] [PubMed]
2. Attia ZI, Friedman PA, Noseworthy PA, et al. Age and Sex Estimation Using Artificial Intelligence From Standard 12-Lead ECGs. Circ Arrhythm Electrophysiol 2019;12:e007284. [Crossref] [PubMed]
3. Attia ZI, Kapa S, Lopez-Jimenez F, et al. Screening for cardiac contractile dysfunction using an artificial intelligence-enabled electrocardiogram. Nat Med 2019;25:70-4. [Crossref] [PubMed]
4. Attia ZI, Noseworthy PA, Lopez-Jimenez F, et al. An artificial intelligence-enabled ECG algorithm for the identification of patients with atrial fibrillation during sinus rhythm: a retrospective analysis of outcome prediction. Lancet 2019;394:861-7. [Crossref] [PubMed]
5. Yao X, Rushlow DR, Inselman JW, et al. Artificial intelligence-enabled electrocardiograms for identification of patients with low ejection fraction: a pragmatic, randomized clinical trial. Nat Med 2021;27:815-9. [Crossref] [PubMed]
6. Olofson A, Lennon R, Kassmeyer B, et al. Detection of Undiagnosed Liver Cirrhosis via Artificial Intelligence-Enabled Electrocardiogram (DULCE): Rationale and design of a pragmatic cluster randomized clinical trial. Contemp Clin Trials Commun 2025;45:101494. [Crossref] [PubMed]
7. Vander Heyde L, Dujardin K, Anné W, et al. Artificial intelligence-derived electrocardiographic age gap as a predictor of mortality after coronary revascularization: prognostic value and short-term intra-patient variability. Eur Heart J Digit Health 2026;7:ztag039. [Crossref] [PubMed]