Pulse rate variability predicts atrial fibrillation and cerebrovascular events in a large, population-based cohort

Abstract

BackgroundMany patients with atrial fibrillation (AF) present with stroke as their first clinical manifestation and since improved AF screening methods are thus required, we investigated whether pulse rate variability parameters predict future AF and cerebrovascular events. MethodsIn an observational cohort study of 5000 community-resident adults (58% male; 50–84 years), the beat-to-beat variability of suprasystolic brachial blood pressure waveforms was measured with root mean square of successive differences (RMSSD) and irregularity index (IrrIx). Based on outcome-oriented and previously validated thresholds for detecting AF, RMSSD and IrrIx were dichotomised at 100 ms and 7.7%, respectively. Participants were followed up for 4.6 years (median), accruing 249 AF and 120 cerebrovascular events in the total sample (n = 5000), and 133 AF and 90 cerebrovascular events among those without prior AF diagnosis (n = 4296). ResultsIn multivariable-adjusted analyses, an elevated RMSSD (>100 ms) or IrrIx (>7.7%) was strongly associated with a higher risk of AF (hazard ratios (HRs) = 2.00–2.95) and cerebrovascular events (HRs = 1.91–2.28), even among people without prior AF diagnosis: HRs for AF = 1.70–2.05 and cerebrovascular events = 2.00–2.28. These associations were strongest in the highest RMSSD tertile >100 ms or IrrIx tertile >7.7%: HRs for AF = 2.32–4.47 and cerebrovascular events = 2.43–3.69. Among those without prior AF diagnosis, the highest categorical net reclassification improvement for 5-year cerebrovascular risk was 14% (95% confidence interval: 7–21%). ConclusionsElevated RMSSD or IrrIx values indicative of the presence of AF predict future AF and cerebrovascular events; more so with increasing pulse irregularity and even among those without prior AF diagnosis.