Background According to the Recommendations for the Standardisation and Interpretation of the ECG of AHA/ACCF/HRS declared in 2009, there are more than 30 criteria for diagnosing left ventricular hypertrophy (LVH). The sensitivity of the various criteria is generally quite low. However, the specificity of ECG in LVH could be 85–90%, acting as a strong predictor for cardiovascular events.

Purpose To investigate the proper criteria in epidemiological screening with different purpose.

Method 5209 qualified ECG out of 6830 people were selected from cross-sectional Handan Eye Study. 2 criteria were chosen: (1) Sokolow-Lyon index—sum of SV1+RV5 or V6≥35 mm; (2) Cornell voltage duration product—men: (SV3+RaVL) ×QRS duration ≥2440 ms; women: (SV3+ (RaVL+8 mV)) ×QRS duration≥2440 ms. The ECG-LVH group contains 829 individuals (12.18%of the cohort), including 607 by Sokolow-Lyon index and 278 by Cornell voltage duration product.

Result After estimated data, we found there is no statistical differences between ECG-LVH population and normal population. While evaluated these data in the group that have Retionopathy without diabetes, the Sokolow-Lyon index and Cornell voltage duration product are statistically different with normal population (p=0.023, p=0.014, respectively). We use the simple MMSE to evaluated the mental status of the population, and found that when the score above 15, there is no statistically different between the positive and negative people in Sokolow-Lyon index (p=0.135); however, while evaluated Cornell voltage duration product, there is statistically difference (p=0.001). When evaluated the data associated with atherosclerotic factors, we find no differences in Cornell Voltage-Duration Product; as for the Sokolow-Lyon index, the differences are shown in systolic blood pressure, total cholesterols and uric acid (p=0.03, 0.04, 0.04, respectively).

Conclusion Different criterion should be chosen for different purposes. For atherosclerotic screening or epidemiological survey of cardiovascular diseases, the Sokolow-Lyon Index might be better. If we use the criterion for epidemiological ophthalmology such as our Handan Eye study, both Sokolow-Lyon Index and Cornell voltage duration Product could be used. As for the evaluation of mental status and its relationship with the LVH risk factors, we might choose Cornell voltage duration product.