Association between deep neural network-derived electrocardiographic-age and incident stroke

Area of Science:

• Cardiology
• Neurology
• Artificial Intelligence in Medicine

Background:

• Stroke remains a leading global cause of death and disability.
• Current stroke risk calculators suffer from bias and imprecision.
• Novel stroke predictors are needed to improve patient outcomes.

Purpose of the Study:

• To investigate the association between electrocardiographic-age (ECG-age), estimated using deep neural networks (DNNs), and the risk of incident stroke.
• To explore ECG-age as a potential novel biomarker for stroke risk prediction.

Main Methods:

• Utilized UK Biobank data, including ECGs from 67,757 participants.
• Estimated ECG-age using a DNN applied to raw ECG waveforms.
• Calculated Δage (ECG-age minus chronological age) and assessed its association with incident and prevalent stroke using multivariable Cox regression models.

Main Results:

• A 10-year increase in Δage was associated with a 22% increase in incident stroke risk.
• Accelerated aging (higher Δage) showed a 42% increase in incident stroke risk compared to normal aging.
• Increased Δage was also significantly associated with prevalent stroke.

Conclusions:

• Deep neural network-estimated ECG-age is associated with both incident and prevalent stroke in a large UK Biobank cohort.
• ECG-age demonstrates potential as a novel biomarker for identifying individuals at increased risk of stroke.
• Further research is warranted to validate ECG-age for clinical use in stroke risk assessment.