Determinants of sensitivity and specificity of electrocardiographic criteria for left ventricular hypertrophy

Abstract

Numerous electrocardiographic criteria, which are largely dependent on fixed voltage thresholds, have been proposed for the diagnosis of left ventricular hypertrophy (LVH). Electrocardiographic criteria for LVH were examined in 4,684 subjects of the Framingham Heart Study who underwent echocardiographic study for LVH. Echocardiographic LVH was detected in 290 men (14.2%) and 465 women (17.6%). Electrocardiographic features of LVH were present in 2.9% of men (60/2,042) and 1.5% of women (39/2,642). The overall sensitivity of the electrocardiographic diagnosis of LVH was 6.9%, whereas specificity was 98.8%. Sensitivity of the electrocardiogram (ECG) for LVH was marginally lower in women than in men (5.6% vs. 9.0%, p=0.075). Obesity was inversely associated with sensitivity (p<0.05, both sexes combined, sex-adjusted). Smoking was also inversely related to sensitivity (p=0.001, both sexes combined, sex-adjusted). In contrast, sensitivity of the ECG increased with age (p<0.001, both sexes combined, sex-adjusted). These findings suggest that electrocardiographic detection of LVH can be improved by incorporating information about noncardiac factors that impact on electrocardiographic sensitivity for LVH, presumably by attenuating QRS voltage. New strategies that take into consideration sex, age, smoking status, and obesity might improve the sensitivity of the ECG without diminishing specificity.