Prognostic value of electrocardiographic criteria for left ventricular hypertrophy.

BACKGROUND Many electrocardiographic (ECG) criteria for left ventricular hypertrophy (LVH) exist, but few studies have compared their relative prognostic value for predicting cardiovascular (CV) mortality. METHODS We analyzed the first ECG on 46950 consecutive veterans. We targeted male outpatients with a body mass index > 20 to avoid confounding by complicating catabolic illnesses and further excluded those with conduction abnormalities. Using Cox regression models adjusted for age, heart rate, and body mass index, we compared the hazard ratios (HRs) for CV mortality obtained from seventeen commonly used ECG criteria for LVH. RESULTS During a mean follow-up of 7 +/- 4 years, in a total population of 19434 patients (mean age 54 +/- 14 years), 1254 (6%) patients died of CV causes. The adjusted HR for CV mortality ranged from 1.4 (95% CI 1.2-1.6) to 3.7 (95% CI 2.7-5.0) among the various criteria. Left ventricular hypertrophy defined by composite criteria was generally associated with higher adjusted HRs compared with voltage-only criteria. Among patients with ECG-LVH, the presence of a left ventricular strain pattern or increased negative P-terminal force were most predictive of CV mortality (HR 3.9 and 3.5, 95% CI 3.3-4.6 and 2.8-4.2). CONCLUSIONS Compared with voltage-only criteria for detecting LVH, composite ECG criteria are more strongly predictive of CV mortality. By applying these ECG criteria into routine clinical practice, individuals with LVH who are at higher risk for CV mortality can be identified and appropriately treated.