Association between 3D left atrial strain and device-detected atrial fibrillation in a high-risk general population: results from the LOOP study

Background Measures of left atrial (LA) function are strongly associated with atrial fibrillation (AF). Three-dimensional echocardiography (3D) enables more precise geometric assessment than conventional 2D echocardiography and may offer incremental value to standard measures. Purpose To investigate the association between 3D LA strain and AF in a high-risk population. Methods We used data from the randomized controlled LOOP study which included elderly (≥70 years) with at least one stroke risk factor (hypertension, diabetes, heart failure, or prior stroke), who were randomized to implantable loop recorder (ILR) monitoring or usual care. Participants who had an ILR implanted and underwent 3D echocardiography were included. 3D-reservoir strain (3D-LASr), conduit strain (3D-LAScd), and contraction strain (3D-LASct) was quantified. The endpoints were incident device-detected AF ≥6 minutes and ≥24 hours. Associations between strain and AF were evaluated using Cox regression. Multivariable analysis was adjusted for the CHARGE-AF risk score (age, height, weight, blood pressure, smoking, hypertension, diabetes, congestive heart failure, history of myocardial infarction). Incremental prognostic value was assessed using C-statistics. Results The study included 819 participants, all underwent 3D echocardiography and had an ILR implanted. Mean age was 74.2±3.9 years, 638 (53%) were male. During a median follow-up time of 4.3 years (IQR 2.2-5.5 years), 236 (28.8%) participants had device-detected AF ≥6 minutes, and 34 (4.2%) AF ≥24 hours. Mean 3D-LA 3D-LASr was 20.7%, 3D-LAScd was 8.9%, and 3D-LASct was 11.9%. In univariable analysis 3D-LASr (HR 1.03 (95% CI 1.01-1.05); p=0.015, per 1 % decrease) and 3D-LASct (HR 1.05 (95% CI 1.02-1.07); p<0.001), per 1 % decrease), but not 3D-LAScd (HR 0.98 (95% CI 0.96-1.01); p=0.15, per 1% decrease) were associated AF≥6 minutes (Figure 1). These findings persisted after multivariable adjustment: 3D-LASr (HR 1.03 (95% CI 1.00-1.05); p=0.017, per 1% decrease) and 3D-LASct (HR 1.05 (95% CI 1.02-1.07); p<0.001, per 1% decrease). Similar findings were observed for ≥24-hour device detected AF in unadjusted analysis (Figure 2), and after multivariable adjustment (3D-LASr: HR 1.12 (95% CI 1.05-1.18); p=0.001, per 1% decrease; 3D-LASct: HR 1.16 (95% CI 1.08-1.24); p<0.001, per 1% decrease). When the continuous 3D variables were added to the CHARGE-AF score, 3D-LASct increased the C-statistic for both AF≥6 minutes (change in C-statistic 0.528 to 0.584, p =0.011) and AF≥24 hours (0.61 to 0.71, p =0.042) whereas 3D-LASr did not. Given association between 3D-LASr and outcomes were non-linear (p for nonlinearity<0.05), further studies are needed to evaluate whether 3D-LASr adds incremental value. Conclusion In a high-risk population, 3D reservoir and contraction strain were independently associated with device-detected AF, and 3D contraction strain adds incremental value to clinical risk scores.