Abstract 10355: Electrocardiographically Determined Left Ventricular Mass Index and Incident Diabetes
Objective: Electrocardiography (ECG) is a measure of cardiac damage that can predict clinical cardiovascular events. In a post-hoc analysis of the Losartan Intervention for Endpoint Reduction (LIFE) trial, hypertensive patients with ECG-determined left ventricular hypertrophy had a higher incidence of diabetes than their counterparts without hypertrophy. We tested whether ECG-determined LV mass (LVM) was associated with incident diabetes in healthy young adults.
Methods: Participants were from the Coronary Artery Risk Development in Young Adults (CARDIA) study, a longitudinal study of adults aged 18–30 at baseline (1985–86) who had up to 6 follow-up exams through 2005–06. Analyses were restricted to 4724 participants (55% women, 51% black) who were free from diabetes at baseline, attended at least 1 follow-up exam and whose baseline ECGs were coded in the CARDIA ECG ancillary study. LVM (g) was calculated using previously published sex- and race- specific models based on Cornell Voltage and body weight. LVM index (LVMI) was calculated as LVM (g)/height (m)2.7. Incident diabetes was defined by any of following criteria during follow-up: fasting glucose≥126 mg/dL, diabetes control medication use, 2-hr post-challenge glucose≥200 mg/dL or HbA1c≥6.5%.
Results: Over 20 years, 339 participants developed diabetes (7.1%). Mean LVMI was higher (p=0.01) in participants who developed diabetes (36.1, SD=6.4) than in those who did not (32.2, SD=5.3). The unadjusted Cox proportional hazard ratio (HR) for incident diabetes was 1.77 (95% confidence interval [CI]: 1.63–1.92) per SD (5.4) increment in LVMI. Following adjustment for age, race, sex, education, field center, and baseline systolic blood pressure (BP), BP medication, body mass index, physical activity, smoking and fasting glucose, the association between higher LVMI and incident diabetes remained significant (HR=1.23, 95% CI: 1.06–1.42).
Conclusions: Higher LVMI is associated with incident diabetes in healthy young adults even after accounting for metabolic covariates. Further analyses are needed to explore potential biological pathways and to determine the temporality of the observed association.
- © 2010 by American Heart Association, Inc.