Objective: Atrial fibrillation (AF) is an important risk factor for cardiovascular disease (CVD). Left atrial (LA) dilatation detected by ultrasonic cardiography (UCG) has been reported to be associated with the incidence of AF in Western countries. However, the relationship between LA dilatation and the AF incidence has not been well investigated in Asian countries, with lower incidence of AF and obesity and with higher incidence of hypertension and stroke. To investigate the relationship between LA dilatation and the AF incidence in Japanese general population, we estimated the multivariate-adjusted hazard ratios (HR) of LA dilatation assessed by UCG for the incidence of AF.
Methods: Participants were 1,585 individuals without AF (665 men, 920 women) who are the participants of a prospective cohort study for CVD incidence in urban Japanese general population (the Suita study). The mean±SD age of the participants were 67.8±7.0 years. Transthoracic UCG was performed from July in 2007 to June in 2013, and LA dimension was determined in accordance with American Society of Echocardiography recommendations by M-mode echocardiograms. After the UCG, the electrocardiograms (ECG) were repeatedly performed until March in 2016. The physicians coded the ECG using Minnesota Code (MC), and AF code is defined as 8-3-1. The HRs of LA dilatation for the incidence of AF with 95% confidence interval (CI) were estimated by Cox proportional hazard model with the adjustment for the followings at the UCG test; age, sex, systolic blood pressure, body mass index, the presence of heavy drinking (more than 2 drinks/day), non-HDL cholesterol, and the presence of the moderate or severe mitral or aortic valve abnormality detected by UCG.
Results: The mean follow-up period from UCG to the last ECG was 4.6±1.8 years, and 19 AF incidences were detected. The multivariate-adjusted HR of the LA dilatation (>40 mm) for the incidence of AF was 8.19 (95% CI: 2.47-27.14). And the multivariate-adjusted HR of 5 mm increment of LA dimension was 2.95 (1.65-5.29), and that of the highest quartile of LA dimension compared to the lowest was 3.37 (0.64-17.78). Among 674 participants with left atrial volume index (LAVI) data (5 AF cases), the multivariate-adjusted HR of LAVI ≥ 32 (ml/m2) for AF incidence was 3.93 (95%CI: 0.34-45.86), and it showed the highest HR among the co-adjusted factors.
Conclusion: LA diameter assessed by UCG is a simple and non-invasive clinical data, and could be a useful factor for the prediction of AF incidence among Japanese general population. Although the follow-up period was not enough compared to the previous studies in Western population, the present study might indicate that individuals with LA dilation detected by UCG are the important candidate for careful and continuous check-ups about CVD risk factors, ECG and UCG in Asian population.