Is echocardiographic epicardial adipose tissue thickness measurement a reliable and reproducible method for risk stratification?