Prevalence and Risk Factors for Paroxysmal Atrial Fibrillation and Flutter Detection after Cryptogenic Ischemic Stroke

Introduction Long-term cardiac monitoring with implantable loop recorders (ILRs) has revealed occult paroxysmal atrial fibrillation and flutter (PAF) in a substantial minority of cryptogenic ischemic stroke (CIS) patients. Herein, we aim to define the prevalence, clinical relevance, and risk factors for PAF detection following early poststroke ILR implantation. Materials and methods A retrospective study of CIS patients (n = 100, mean age 65.8 years; 52.5% female) who underwent ILR insertion during, or soon after, index stroke admission. Patients were prospectively followed by the study cardiac electrophysiologist who confirmed the PAF diagnosis. Univariate and multivariate analyses compared clinical, laboratory, cardiac, and imaging variables between PAF patients and non-PAF patients. Results PAF was detected in 31 of 100 (31%) CIS patients, and anticoagulation was initiated in almost all (30 of 31, 96.8%). Factors associated with PAF detection include older age (mean [year] 72.9 versus 62.9; P  = .003), white race (odds ratio [OR], 4.5; confidence interval [CI], 1.8-10.8; P  = .001), prolonged PR interval (PR > 175 ms; OR, 3.3; CI, 1.2-9.4; P  = .022), larger left atrial (LA) diameter (mean [cm] 3.7 versus 3.5; P  = .044) and LA volume index (mean [cc/m 2 ]; 30.6 versus 24.2; P  = .014), and lower hemoglobin (Hb)A1c (mean [%] 6.0 versus 6.4; P  = .036). Controlling for age, obesity (body mass index > 30 kg/m 2 ; OR, 1.2; CI, 1.1-1.4; P  = .033) was independently associated with PAF detection. Discussion PAF was detected with high prevalence following early postcryptogenic stroke ILR implantation and resulted in significant management changes. Older age, increased PR interval, LA enlargement, and lower HbA1c are significantly associated with PAF detection. Controlling for age, obesity is an independent risk factor. A larger prospective study is warranted to confirm these findings.