[Skip to Content]
[Skip to Content Landing]
July 6, 1994

Left Ventricular Mass and Risk of Stroke in an Elderly CohortThe Framingham Heart Study

Author Affiliations

From the Framingham (Mass) Heart Study, (Drs Bikkina, Levy, Benjamin, Wolf, Castelli, and Larson, and Ms Evans); the National Heart, Lung, and Blood Institute, Bethesda, Md (Drs Levy and Castelli); the Divisions of Epidemiology and Preventive Medicine (Drs Levy, Wolf, and Larson, and Ms Evans), Cardiology (Dr Benjamin), and Neurology (Dr Wolf), Boston (Mass) University School of Medicine; and the Divisions of Cardiology and Clinical Epidemiology, Beth Israel Hospital, Boston, Mass (Dr Levy).

JAMA. 1994;272(1):33-36. doi:10.1001/jama.1994.03520010045030

Objective.  —To evaluate the association of echocardiographically determined left ventricular mass (LVM) with incidence of stroke or transient ischemic attack in an elderly cohort.

Design.  —Cohort study with a follow-up period of 8 years.

Setting.  —Population-based sample.

Subjects.  —Elderly original cohort subjects of the Framingham Heart Study who were free of cerebrovascular disease and atrial fibrillation at the 16th biennial examination and who had adequate echocardiograms. This group consisted of 447 men (mean age, 67.8 years; range, 60 to 90 years) and 783 women (mean age, 69.2 years; range 59 to 90 years).

Main Outcome Measures.  —Age-adjusted 8-year incidence of stroke was examined as a function of baseline quartiles of LVM-to-height ratio. Proportional hazards regression was used in multivariate analyses to assess risk of stroke as a function of LVM-to-height ratio quartile, adjusting for age, sex, systolic blood pressure, hypertension treatment, diabetes, cigarette smoking, and blood lipid levels.

Results.  —Among the 1230 subjects eligible, 89 cerebrovascular disease events (62 strokes and 27 transient ischemic attacks) occurred during follow-up. In men, 8-year age-adjusted incidence of cerebrovascular events was 18.4% in the highest quartile of LVM-to-height ratio and 5.2% in the lowest quartile. Corresponding values in women were 12.2% and 2.9%. The hazard ratio for cerebrovascular events comparing highest to lowest quartile of LVM-to-height ratio was 2.72 (95% confidence interval [CI], 1.39 to 5.36) after adjusting for age, sex, systolic blood pressure, hypertension treatment, diabetes, cigarette smoking, and the ratio of total cholesterol to high-density lipoprotein cholesterol. After adjusting for age, sex, and cardiovascular disease risk factors, the hazard ratio for cerebrovascular events was 1.45 (95% CI, 1.17 to 1.80) for each quartile increment of LVM-to-height ratio.

Conclusions.  —Echocardiographically determined LVM-to-height ratio offers prognostic information beyond that provided by traditional cerebrovascular disease risk factors. Echocardiography provides information that facilitates identification of individuals at high risk for stroke and transient ischemic attack.(JAMA. 1994;272:33-36)