Circadian rhythm in atrial fibrillation detected by implantable loop recorder: a post-hoc analysis of a randomized clinical trial

Abstract Introduction Previous studies suggest the existence of circadian, weekly, and seasonal rhythms in the occurrence of atrial fibrillation (AF), which may be utilized to improve the yield of screening. The heterogeneity of AF natural history also necessitates better phenotyping to overcome the challenges of identifying candidates for screening and treatment. Purpose To explore the potential circadian rhythm in AF occurrence in persons without known arrhythmia but at high risk of AF and stroke. Methods Post-hoc analysis of the LOOP study (Atrial Fibrillation Detected by Continuous Electrocardiogram Monitoring Using Implantable Loop Recorder to Prevent Stroke in High-Risk Individuals), which randomized 6004 AF naïve participants (≥70 years old with additional stroke risk factors) 1:3 to receive implantable loop recorder screening or usual care. All loop recorder data was extracted to obtain the date and time of AF episode onset, AF duration, mean and max heart rate (HR) during AF, and symptom annotations. New-onset AF episodes lasting ≥6 minutes were adjudicated, and all subsequent episodes were included in the analysis. Results A total of 41,713 AF episodes (median duration 18 (8-64) minutes) were identified among 430 participants (Table 1). The median number of episodes for each participant with AF was 21 (4-82) and the median AF burden was 0.17% (0.04%-0.83%) of time from debut to end of monitoring. The most and least frequent days of AF onset were Sunday (15.3%) and Friday (13.6%), respectively, while AF occurred slightly more often during spring (26%) than in other seasons. The median time of onset was 09:57:15 AM (05:15:30 AM – 03:33:00 PM). The most frequent time of onset was at 09:00 AM (rounded to the nearest hour), which was two times more likely than at 09:00 PM, and 22% of all episodes occurred between 08:00 and 11:00 AM (Figure 1). A total of 25,329 (61%) episodes occurred between midnight and noon (AM). Nocturnal episodes, defined as episodes with onset between 10:00 PM and 07:00 AM, occurred in 368 (86%) AF participants and constituted a median of 40% of all episodes and 37% (14%-63%) of each participant’s episodes. The median duration of nocturnal episodes was 28 (10-104) minutes compared to 14 (8-44) minutes for daytime episodes. The averaged heart rate during AF was 81 (71-96) beats per minute (BPM), while 330 (77%) participants had AF episodes with heart rates equal to or above 100 BPM, and the duration of these episodes was 50 (15-201) minutes compared to 30 (21-85) minutes for slower episodes (below 100 BPM). Conclusion A circadian pattern was observed for screening-detected AF. AF onset occurred predominantly between 08:00 and 11:00 AM, representing more than one in five episodes. Nocturnal AF occurred in most participants with AF and accounted for almost half of all episodes, while these episodes also lasted longer than daytime episodes. This may warrant future studies investigating the timing of AF screening.