LAUSR

Background Left ventricular hypertrophy (LVH) is strongly associated with the development of heart failure. The two-dimensional left ventricular mass (LVM) algorithms suffer from measurement variability that can lead to misclassification of patients with LVH as normal, or vice versa. Among the four echocardiographic measurements required by the two-dimensional LVM algorithms, epicardial and endocardial area have the lowest inter-observer variation and could be used to corroborate LVM calculations. We sought cut-off values able to discriminate between elevated and normal LVM based on endocardial or epicardial area alone. Methods Using data from 664 men enrolled in the Mind Your Heart Study, we calculated the correlation of left ventricular mass index (LVMI) with epicardial area and endocardial area. We then used receiver operator characteristic curves to identify epicardial and endocardial area cut-points that could discriminate among individuals with normal LVM and LVH. Results LVMI was more strongly correlated with epicardial area compared to endocardial area, r = 0.70 versus r = 0.27, respectively. As a screening test for LVH, epicardial area significantly outperformed endocardial area as reflected by a considerably higher area under the ROC curve, 0.90 (95% CI = 0.86 to 0.93) versus 0.63 (95% CI = 0.57 to 0.71). An epicardial area cut-point of ≥ 38.0 cm2 corresponded to a sensitivity of 95.0% and specificity of 54.4% for detecting LVH. The seven participants who had epicardial areas below our cut-point but still met criteria for LVH by LVMI (false negatives), had lower body surface areas and lower left ventricular cavity volumes than the cohort average. Conclusions Epicardial area showed promise as a method of rapid screening for LVH and could be used to validate formal LVM calculations. Figure 1