Value of routine heart rate variability parameters for atrial fibrillation detection in ischaemic stroke and...

Area of Science:

• Cardiology
• Neurology
• Medical Diagnostics

Background:

• Undetected atrial fibrillation (AF) is a significant risk factor for recurrent ischemic stroke.
• Current prediction scores for AF lack the integration of heart rate variability (HRV) measures.
• HRV parameters are readily available from standard 24-hour Holter electrocardiograms (ECGs).

Purpose of the Study:

• To evaluate the efficacy of time-domain HRV parameters in predicting incident AF within one year in patients with non-AF ischemic stroke or high-risk transient ischemic attack (TIA).
• To compare the predictive performance of HRV measures against established clinical prediction scores (Brown-ESUS AF and AS5F).

Main Methods:

• The study included 697 patients from the STROKE-CARD Registry (NCT04582825).
• Eight time-domain HRV parameters were assessed.
• Receiver Operating Characteristic (ROC) analyses, logistic regression, and the Youden index were employed to determine optimal cut-offs and compare predictive performance.

Main Results:

• New-onset AF was detected in 4.0% of patients.
• PNN50, rMSSD, and SDSD demonstrated superior discrimination for AF prediction (AUCs 0.711-0.775) compared to clinical scores (AUC ≤ 0.612).
• Optimal HRV cut-offs were identified, and these parameters showed strong associations with AF (ORs 5.34-7.70, p < 0.001), significantly enhancing prediction when added to existing scores.

Conclusions:

• Specific HRV parameters (PNN50, rMSSD, SDSD) derived from routine Holter ECGs improve AF risk prediction in patients following non-cardioembolic stroke or high-risk TIA.
• These HRV metrics can potentially guide targeted monitoring strategies for AF detection.
• Incorporating HRV analysis into post-stroke assessments may enhance the identification of patients at risk for AF.