Echocardiographic assessment of peak atrial longitudinal strain in paroxysmal atrial fibrillation patients compared to age‐matched controls

Aim To investigate left atrial (LA) function assessed by peak atrial longitudinal strain (PALS) in an elderly population with silent paroxysmal atrial fibrillation (PAF) compared to age‐matched control group. Methods From the STROKESTOP2‐study, patients with newly diagnosed silent PAF were enrolled together in a randomly selected age‐matched control group. All patients had NT‐proBNP levels between 125 and 900 ng/L. Echocardiographic examinations were performed and analyzed offline using dedicated software. In addition to conventional echocardiographic variables, PALS was measured using a two‐dimensional speckle tracking (2D‐ST) biplane. Results The study population (mean age 76 ± 0.3) demonstrated significantly reduced PALS in patients with PAF ( N = 69) compared to controls ( N = 94) (23.9% ± 6.5% vs 30.7% ± 7.6%, p < 0.001). In the PAF group, 29 (42%) patients had normal LA volume index (LAVI) defined as 34 mL/m 2 . When subgroups with normal LAVI were compared, PALS was reduced in PAF compared to controls (25.1 ± 8.0 vs 31.1 ± 7.8, p = 0.001). NT‐proBNP levels did not differ between the PAF group (median 228 ng/L) and the control group (median: 217 ng/L) ( p = 0.16). In multiple linear regression analysis, left ventricular global longitudinal strain, stroke volume index and LAVI were independent predictors of PALS. Receiver operating characteristic curve identified PALS threshold <22% for predicting risk of silent PAF. Conclusion PALS was significantly reduced in PAF patients even in the absence of significant LA remodeling. PALS <22% predicts risk for silent PAF in general elderly population. In high‐risk patients PALS could be used for AF screening enrichment.