Temporal Stratification of Atrial Fibrillation Detected After Stroke: Differences in Patient Characteristics and Prognostic Implications

Background Atrial fibrillation detected after stroke (AFDAS) refers to atrial fibrillation (AF) first identified in patients without previously known AF. We evaluated whether AFDAS detection timing identifies distinct phenotypes and outcomes. Methods Retrospective, single‐center study of consecutive acute ischemic stroke patients with newly detected AF, classified as: ECG‐AF, diagnosed on admission 12‐lead ECG, AFDAS‐Early, detected ≤14 days, and AFDAS‐Late, detected >14 days after stroke. Associations with imaging, echocardiographic, laboratory markers, and outcomes (90‐day‐mRS, stroke recurrence, mortality) were assessed using multivariable models. Results We included 328 newly detected patients with AF: median age 77 years, 57.9% women, National Institutes of Health Stroke Scale (NIHSS) 8. Compared with ECG‐AF, AFDAS‐Early exhibited smaller left atria (LA volume‐index: 36 versus 45 mL/m 2 ; adjusted odds ratio [aOR] 0.96, 95% CI 0.94–0.98, P =0.001) and better ejection fraction (<50%: 10.3% versus 26.2%; aOR 0.30, 95% CI 0.14–0.65, P =0.003). Compared with AFDAS‐Late, AFDAS‐Early were older, had more severe strokes, no cardiac differences, but strong association with large vessel occlusion (aOR 1.99, 95% CI 1.09–3.64, P =0.024) and cerebral autonomic network involvement (aOR 1.91, 95% CI 1.02–3.52, P =0.043). AF subtype was not independently associated with 90‐day‐mRS, mortality, or stroke recurrence. Older age (aHR 1.10, 95% CI 1.06–1.15, P <0.001), NIHSS (aHR 1.07, 95% CI 1.04–1.11, P <0.001), coronary artery disease (aHR 2.91, 95% CI 1.72–4.92, P <0.001) were significantly associated with higher mortality, whereas time‐dependent oral anticoagulation was strongly associated with lower stroke recurrence (aSHR 0.22, 95% CI 0.10–0.48, P <0.001). Conclusions AFDAS should be distinguished from ECG‐AF, which likely represents previously undiagnosed AF. Within AFDAS, detection timing reflects different stroke severity and treatment exposure rather than intrinsic risk, supporting early rhythm monitoring and timely anticoagulation.