Purpose The purpose of this study is to investigate the predictive value of the mitral E/A ratio for morbidity and mortality in a large population-based cohort of middle-aged African Americans.
Methods This is a prospective population-based cohort study design. Between 1993 and 1996, participants enrolled in the Jackson (all African American) cohort of the Atherosclerotic Risk in Communities Study received transthoracic echocardiograms, underwent interview and had blood pressure (BP), body mass index (BMI), heart rate (HR), serum lipids and glucose measured. For this study, three subgroups of E/A ratio were defined: E/A ratio < 0.7, E/A ratio = 0.7-1.5, and E/A ratio > 1.5. Study end points included all-cause mortality and CVD events (coronary events and strokes). Cox proportional hazard models were used to assess prognosis of participants with E/A ratio < 0.7 and E/A ratio > 1.5 compared to those with E/A ratio 0.7-1.5, adjusting for age and gender alone and with BP, HR, BMI, diabetes, lipid status, tobacco use, left ventricular mass and ejection fraction < 40%. The mean follow-up period was 6.8 ± 1.3 years.
Results The 2,211 participants (65.2% women, mean age = 62 years) had a high prevalence of hypertension, obesity, diabetes and echocardiographic left ventricular hypertrophy (LVH). Adjusting for age and gender, an E/A ratio > 1.5 was predictive of both all-cause mortality (HR 1.74, CI 1.05-2.90; p = .03) and incident CVD (HR 1.75, CI 1.08-2.83; p = .02). An E/A ratio > 1.5 was also independently predictive of all-cause mortality (HR 2.19, CI 1.20-4.02; p = .01) in the multivariable model. An E/A ratio < 0.7 was predictive of all-cause mortality (HR 1.79, CI 1.17-2.73; p = .007) and incident CVD (HR1.91, CI 1.29-2.83; p < .0001) after adjusting for age and gender but was not independently predictive in the multivariable model (p < .05).
Conclusion In a population-based cohort of middle-aged African Americans where hypertension, obesity and diabetes are epidemic, an E/A ratio > 1.5 independently predicts all-cause mortality. The E/A ratio > 1.5 and an E/A ratio < 0.6 were both predictive of incident CVD when adjusted for age and gender alone but were not independently predictive in the multivariable analysis.