IMPROVED ELECTROCARDIOGRAPHIC DETECTION OF ECHOCARDIOGRAPHIC LEFT-VENTRICULAR HYPERTROPHY - RESULTS OF A CORRELATED DATA-BASE APPROACH

Objectives. We used electrocardiographic (ECG) and echocardiographic measurements from 3,351 adults from the Framingham Heart Study to evaluate the performance of 10 ECG criteria in detecting left ventricular hypertrophy before and after adjust ment for gender, age and obesity. Background. Significant improvement in the sensitivity and specificity of ECG voltage based criteria for detection of echocardiographic left ventricular hypertrophy using gender-specific criteria adjusted for age and obesity has been demonstrated. Methods. Gender-specific correlation and regression analyses were used to identify the five most sensitive ECG criteria and to adjust them for age and obesity, Standard and truncated receiver operating characteristic curves were used to compare the selected criteria. Results. Linear regression of left ventricular mass on ECG voltages, body mass index and age yielded considerably stronger relations for women than for men because of the greater correlation between ventricular mass and body mass index in women. Obesity and age adjustment of the five voltage criteria produced considerable improvement in their performance. The voltage sum of the R wave in lead aVL and the S wave in lead V-3, alone and in combination with QRS duration, had a sensitivity at 95% specificity of 32% and 39%, respectively, in men and 46% and 51%, respectively, in women after adjustment. Tables of critical voltages for the adjusted criteria and representative performance values are presented. Conclusions. Incorporation of obesity and age in ECG algorithms consistently improves their performance in the detection of hypertrophy. The development of such criteria will enhance the utility of this inexpensive screening test. Such ECG approaches should be used in clinical trials to assess the efficacy of antihypertensive treatment to effect regression of left ventricular hypertrophy. A search for further improvements in the efficacy of ECG criteria for this purpose is warranted. Independent validation of this approach is needed.