Prevalence and covariates of electrocardiographic left ventricular hypertrophy in the Hypertension in the Very Elderly Trial

OBJECTIVE: To estimate the prevalence and covariates of electrocardiographic left ventricular hypertrophy (LVH) in the Hypertension in the Very Elderly Trial.

METHODS: A total of 2993 hypertensive people aged at least 80 years with technically codable ECGs without pacing, bundle branch block, or ECG myocardial infarction were studied. LVH was assessed using Sokolow-Lyon (SL-LVH), Cornell voltage (CV-LVH), and Cornell product criterion (CP-LVH).

RESULTS: The prevalence of LVH varied from 2.4 to 17.5% depending on sex, race, and ECG criterion. The highest prevalence of SL-LVH (12.0%) was in Chinese men and in white women for both CV-LVH (17.5%) and CP-LVH (12.9%). Increasing SBP was an independent covariate of the presence of LVH in Chinese women independently of the criterion used (β = 0.052-0.069, P < 0.001), of SL-LVH in Chinese men (β = 0.047 P = 0.006). In white women, CP-LVH was associated with increasing age (β = 0.055, P = 0.027) and SBP (β = 0.023, P = 0.040). Increasing BMI was associated inversely with SL-LVH; the association in white men only was not significant. In white men, a history of diabetes was directly and history of antihypertensive drug treatment inversely related to CV-LVH and CP-LVH. SL-ECG was associated inversely to serum uric acid concentration in Chinese women and to serum hemoglobin concentration in Chinese men.

CONCLUSION: Prevalence and covariates of electrocardiographic LVH varied by sex, race, and ECG criterion. CP-LVH may prove to be the most useful measure of LVH in this study owing to its close relationship to SBP, at least in women, and independence from BMI.