Relationship between electrocardiography and echocardiography for left ventricular mass change in mild systemic hypertension (TOMHS)

Correlation between ECG and echocardiographic (Echo) estimates of left ventricular mass (LVM) are modest in mild hypertension. However, the correlation between ECG and Echo for change in LVM is unknown. The correlation of 8 ECG criteria sets with an Echo standard for level and change in LVM in subjects with mild hypertension were determined. ECGs were computer processed to define the following criteria sets: 1) Casale/Devereux (CD); 2) Cornell product (CP); 3) Cornell voltage (CV); 4) 12-lead voltage product, (Σ12P); 5) sum of 12-lead (Σ12); 6) Rautaharju (R); 7) Sokolow-Lyon (SL); 8) Romhilt-Estes (RE). Echo LVM was determined by Devereux formula and expressed in g, g/m and g/m2. ECGs and Echocardiograms were recorded at baseline, 3 months, and annually through 4 years. ECG and Echo data were correlated for individual average levels over all recordings and for change in ECG and Echo data (the average of all follow-up minus the baseline). Results Conclusions 1) Correlations between ECG criteria set and Echo measures of LVM are more highly correlated for level than for change. 2) No single ECG criterion set is best in all comparisons. 3) Because level and change in heart size may be associated with risk of cardiovascular death and the ECG and Echo determined measures are only modestly correlated, both may contribute different sets of information about risk, and future studies may consider combining these as predictors of cardiovascular outcomes.