Abstract
BACKGROUND: Resting heart rate in atrial fibrillation (AF) is a commonly used therapeutic target, but no studies have evaluated whether resting heart rates in AF are reliable over repeated measurements. We aimed to determine the reliability of resting heart rate measurements during AF. METHODS: We included 832 US patients with persistent AF recorded by a mobile cardiac telemetry device equipped with an accelerometer (PocketECG, MEDICALgorithmics, Warsaw, Poland). Resting heart rates were derived from repeatedly sampling the 10-second mean heart rate after ≥10 minutes of rest, as assessed by the accelerometer, with at least 20 minutes between each sample. The overall resting heart rate was defined as the intra-individual mean of all samples. The reliability of a single resting heart rate measurement was shown using Bland Altman plots and computed as the 5-95% reference interval for the overall resting heart rate given a randomly sampled resting heart rate measurement. The reference intervals for the resting heart rates were derived by comparing randomly chosen resting heart rate measurements to the overall mean using bootstrap with 10,000 replications. RESULTS: The mean age was 74 (±12) years, 44% were females. The mean recording duration was 14 (±8.5) days, and the average number of samples was 289 (range 46-600). The mean overall resting heart rate was 85 beats/min (±17). A random resting heart rate often differed substantially from the overall resting heart rate (Figure 1). Table 1 presents the 5-95% interval within which the overall resting heart can be expected, based on a random resting heart rate measurement. Single resting heart rate measurements as low as 90 beats per min were associated with overall resting heart rates ≥110 beats/min in approximately 5% of individuals. CONCLUSIONS: There is a substantial intra-individual variability in resting heart rate during AF, implying that longer duration recordings or repeated measurements may be needed when an accurate measurement of resting heart rate is needed to guide treatment decisions. [Figure: see text] [Figure: see text]