Performance of primary and derived M-mode echocardiographic measurements for detection of left ventricular hypertrophy in necropsied subjects and in patients with systemic hypertension, mitral regurgitation and dilated cardiomyopathy.

To determine which M-mode echocardiographic (echo) measurement best detects left ventricular (LV) hypertrophy, the sensitivity and specificity of upper normal limits of echo LV anatomic measurements (previously shown to have 97% specificity in living normal subjects) were tested in 60 necropsied patients with anatomic hypertrophy and in 28 necropsied patients with normal left ventricles. The prevalence of hypertrophy by each echo criterion was determined in 165 living patients with systemic hypertension, mitral regurgitation or dilated cardiomyopathy. The best separation between patients with normal vs increased necropsy LV mass was obtained using sex-specific echo LV mass index criteria (overall accuracy = 73 of 88 patients, 83%). Lower overall accuracies for separation of patients with and without hypertrophy were observed for echo cross-sectional area (59 of 88 patients, 67%; p less than 0.05 vs LV mass index) and indexes of LV wall thickness (39 to 51%, p less than 0.001). Among 113 living patients with moderate or severe hypertension, mitral regurgitation or dilated cardiomyopathy, LV mass index was increased in 73%, cross-sectional area index in 58% (p less than 0.02 vs LV mass index), and posterior wall thickness, septal thickness and relative wall thickness in only 11 to 32% (all p less than 0.001 vs LV mass index). Thus, an M-mode echo LV mass index of more than 134 g/m2 in men and more than 110 g/m2 in women detects concentric and eccentric LV hypertrophy accurately by comparison with necropsy and clinical reference standards; cross-sectional area is slightly less useful; and other M-mode echo criteria of LV hypertrophy perform too poorly to be clinically applicable.