Comparison of the prognostic value of left ventricular hypertrophy in African-American men versus women

Echocardiographically determined left ventricular (LV) hypertrophy may be a stronger risk factor of cardiovascular disease (CVD) for women than for men, although it is unclear whether reported gender differences are real or attributable to confounding. We evaluated echocardiographic LV hypertrophy (defined as LV mass/height2.7 ≥51 g/m2.7) collected from the African-American population of the Atherosclerosis Risk in Communities Study. Incident CVD events (57 in men, 62 in women) were determined during a median follow-up of 4.9 years (interquartile range 4.3 to 5.6) and included nonfatal myocardial infarction, cardiac death, coronary revascularization, and stroke. We conducted 2 analyses. First, we created matched samples of 340 men and 812 women who had LV hypertrophy based on propensity score and estimated the gender-specific incidence rate ratios and population-attributable risks. Second, we evaluated the complete cohort (604 men and 1,113 women) with Poisson's regression after adjusting for age, body mass index, hypertension, diabetes mellitus, ratio of total cholesterol to high-density lipoprotein cholesterol, current smoking, and education level. LV hypertrophy was significantly predictive of incident CVD, and the association shown by analyses of matched propensity scores was similar in men and women (incidence rate ratio 1.88 vs 1.92, p = 0.97 for men, population-attributable risk 0.22 vs 0.26, p <0.07 for women). In the multivariate analysis, we found comparable effect estimates for LV hypertrophy (incidence rate ratio 1.66 vs 2.09, p = 0.55 for men; population-attributable risk 0.24 vs 0.32, p <0.07 for women). Thus, LV hypertrophy is a strong predictor of CVD in African-Americans, and the effect of LV hypertrophy on CVD is similar in men and women. Echocardiographically determined left ventricular (LV) hypertrophy may be a stronger risk factor of cardiovascular disease (CVD) for women than for men, although it is unclear whether reported gender differences are real or attributable to confounding. We evaluated echocardiographic LV hypertrophy (defined as LV mass/height2.7 ≥51 g/m2.7) collected from the African-American population of the Atherosclerosis Risk in Communities Study. Incident CVD events (57 in men, 62 in women) were determined during a median follow-up of 4.9 years (interquartile range 4.3 to 5.6) and included nonfatal myocardial infarction, cardiac death, coronary revascularization, and stroke. We conducted 2 analyses. First, we created matched samples of 340 men and 812 women who had LV hypertrophy based on propensity score and estimated the gender-specific incidence rate ratios and population-attributable risks. Second, we evaluated the complete cohort (604 men and 1,113 women) with Poisson's regression after adjusting for age, body mass index, hypertension, diabetes mellitus, ratio of total cholesterol to high-density lipoprotein cholesterol, current smoking, and education level. LV hypertrophy was significantly predictive of incident CVD, and the association shown by analyses of matched propensity scores was similar in men and women (incidence rate ratio 1.88 vs 1.92, p = 0.97 for men, population-attributable risk 0.22 vs 0.26, p <0.07 for women). In the multivariate analysis, we found comparable effect estimates for LV hypertrophy (incidence rate ratio 1.66 vs 2.09, p = 0.55 for men; population-attributable risk 0.24 vs 0.32, p <0.07 for women). Thus, LV hypertrophy is a strong predictor of CVD in African-Americans, and the effect of LV hypertrophy on CVD is similar in men and women.