Electrocardiographic left atrial abnormality in patients presenting with ischemic stroke

• Published in: Journal of stroke and cerebrovascular diseases 2020-09, Vol.29 (9), p.105086-105086, Article 105086
• Main Authors: Kwon, Younghoon, McHugh, Stephen, Ghoreshi, Kayvon, Lyons, Genevieve R, Cho, Yeilim, Bilchick, Kenneth C., Mazimba, Sula, Worrall, Bradford B., Akoum, Nazem, Chen, Lin Y, Soliman, Elsayed Z.
• Format: Article
• Language: English
• Subjects: Aged; Aged, 80 and over; Atrial Function, Left; Atrial Remodeling; Brain Ischemia - diagnosis; Brain Ischemia - etiology; Brain Ischemia - physiopathology; Cardioembolic; Electrocardiography; Female; Heart Atria - physiopathology; Heart Diseases - complications; Heart Diseases - diagnosis; Heart Diseases - physiopathology; Heart Rate; Humans; Ischemic; Male; Middle Aged; P wave terminal force; Predictive Value of Tests; Registries; Retrospective Studies; Risk Assessment; Risk Factors; Stroke; Stroke - diagnosis; Stroke - etiology; Stroke - physiopathology
• ISSN: 1052-3057; 1532-8511
• DOI: 10.1016/j.jstrokecerebrovasdis.2020.105086

P wave indices represent electrocardiographic marker of left atrial pathology. We hypothesized that P wave would be more abnormal in patients presenting with ischemic stroke than a comparable group without ischemic stroke. We compared P wave terminal force in V1 (PTFV1) between patients admitted with ischemic stroke (case) and patients followed in cardiology clinic (control) at a single medical center. Using logistic regression models, we tested for an association between abnormal PTFV1 (> 4000 µV ms) and ischemic stroke. We also defined several optimal cut-off values of PTFV1 using a LOESS plot and estimated odds ratio of ischemic stroke when moving from one cut-point level to the next higher-level. A total of 297 patients (case 147, control 150) were included. PTFV1 was higher in patients with vs. those without ischemic stroke (median 4620 vs 3994 µV ms; p=0.006). PTFV1 was similar between cardioembolic/cryptogenic and other stroke subtypes. In multivariable analyses adjusting for sex, obesity, age, and hypertension, the association between abnormal PTFV1 and ischemic stroke ceased to be significant (OR 1.53 [0.95, 2.50], p=0.083). Increase to the next cutoff level of PTFV1 (900, 2000, 3000, 4000, 5000, and 6000 µV ms) was associated with 18% increase in odds of having ischemic stroke (vs. no ischemic stroke) (OR 1.18 [1.02, 1.36], p=0.026). Patients presenting with acute ischemic stroke are more likely to have abnormal PTFV1. These findings from a real-world clinical setting support the results of cohort studies that left atrial pathology manifested as abnormal PTFV1 is associated with ischemic stroke.