Agreement Between The Left Ventricular Hypertrophy Diagnosed By Electrocardiography And By Echocardiography: The Northern Shanghai Study

Objective: Echocardiographic measurement was recommended as the golden standard to detect left ventricular hypertrophy (LVH), whereas LVH could also be assessed by electrocardiograph. There are several criteria to define LVH by electrocardiograph, including Sokolow-Lyon-Rappaport (SLR), Cornell and Cornell Product criteria (CP). However, it remains unclear which electrocardiographic criteria is superior over others in detecting LVH. We therefore was aimed to investigate the agreement between the left ventricular hypertrophy diagnosed by electrocardiography and by echocardiography in a community-based cohort. Design and Method: We conducted echocardiographic and electrocardiographic measurement to define LVH in 1599 elderly Chinese aged over 65 years old in communities located in the northern Shanghai. Echocardiographic LVH (ECHO-LVH) was defined by left ventricular mass indexed for body surface area (LVM/BSA) or indexed for body height2.7 (LVM/height2.7). Electrocardiographic LVH (ECG-LVH) was defined by SLR, Cornell and CP criteria. Correlation analysis and Chi-squared test, as well as Receiver operating characteristic (ROC) curve, were applied to compare the agreement between ECG-LVH defined by three different criteria and ECHO-LVH. LVH1, LVH2 and LVH3 was respectively defined by LVM/BSA ≥ 125 g/m2 in males, ≥110 g/m2 in females and LVM/BSA ≥ 115 g/m2 in males, ≥95 g/m2 in females and LVM/height2.7 ≥50 g/m2.7 in males, ≥47 g/m2.7 in females. Results: In the correlation analyses, CP had the highest correlation coefficient, as compared with SLR and Cornell, in males, in females and in total population. In Chi-squared analyses, only LVH diagnosed by SLR was significantly associated with LVH diagnosed either by LVM/BSA or by LVM/height2.7, indicating considerable disagreement. In ROC analyses, CP criterion had highest area under curve (AUC) of ROC than Cornell and SLR criteria (0.62 vs. 0.58 & 0.54 in LVH1, 0.62 vs. 0.55 & 0.51 in LVH2 and 0.62 vs. 0.57 & 0.51 in LVH3). Conclusions: CP criterion has greater agreement than SLR index and Cornell criterion with LVH diagnosed by echocardiograph.