This was a prospective, multicenter, observational cohort study. Patients with atrial fibrillation were consecutively recruited from the Cardiovascular Research Platform. After providing written informed consent, research assistants scheduled and assigned the patients to the participating centers of the BEAT-AF TE project according to their individual circumstances. Each center conducted follow-up management in accordance with the investigator's handbook, with a follow-up duration of 24 months. The patients were managed according to local clinical practice, and no study-related interventions were administered in this study; only follow-up data were recorded. This real-world data analysis aimed to investigate, based on the primary exposure variable (atrial fibrillation burden), primary outcome variables (ischemic stroke, systemic embolism, etc.), and secondary outcome variables (transient ischemic attack, hospitalization for heart failure, major bleeding events, all-cause mortality, etc.), whether there is a risk inflection point between atrial fibrillation burden and stroke, whether there is a reasonable threshold for the benefit of anticoagulation therapy, and whether there is a time-dependent risk of atrial fibrillation burden prior to stroke/thromboembolic events.