2025 American Academy of Neurology Abstract Website

Objective:

To investigate the predictive ability of artificial intelligence (AI)-electrocardiogram (ECG) models on adverse vascular events in patients with migraine.

Background:

Migraine, especially migraine with aura, is associated with an increased risk of cardiovascular events. Our institution has developed AI-ECG models that could estimate the probability of atrial fibrillation (AF) and patients’ age from normal sinus rhythm (NSR) ECG, and delta age (estimated age – actual age) has also been shown to reflect cardiovascular health.

Design/Methods:

Adult patients with migraine who had at least one digital, standard 12-lead ECG from 2000 to 2020 were identified. The first NSR ECG was defined as the index ECG. Those with adverse vascular events of interest or AF prior to or within 3 days after index ECG were excluded. The vascular events of interest included acute myocardial infarction, acute ischemic stroke, venous thromboembolism and death. AI-ECG models were implemented to estimate AF probability and delta age from index ECG.

Results:

A total of 29928 patients were included. The mean age was 44.3±14.5 years and the median follow-up time was 54 months. Multivariate Cox regression model demonstrated that a higher estimated AF prediction model output was associated with a higher risk of composite vascular outcomes (HR: 1.15 [95% CI: 1.12–1.18] for every 10% increase) after adjustment with age, sex, presence of aura and vascular risk factors. The same was also true for delta age (HR: 1.16 [1.12–1.21] for every 10-year increase). Moreover, an estimated AF probability of ≥1% (HR: 1.45 [1.36–1.54]) and delta age ≥-1 years (HR: 1.09 [1.02–1.16]) were found to be the optimal cutoffs to dichotomize patients into high- and low-risk groups.

Conclusions:

AI-ECG AF prediction model output and delta age could potentially be used to identify migraine patients at risk for adverse vascular outcomes.