Abstract EP43: Obesity And Incident Cardiovascular Disease (CVD): The Jackson Heart Study

Abstract

Background: Although obesity is a well-established risk factor for the development of cardiovascular disease (CVD), it is unclear which anthropometric measure(s) of obesity are the best predictors of CVD. Methods: Using data from 3491 African Americans (mean age: 58.3±12.1; 65.2% women) in the Jackson Heart Study, we examined the associations of incident CVD and its individual components (Coronary heart disease (CHD), heart failure, and stroke) with time-dependent obesity measures (body mass index (BMI), waist circumference (WC), waist-to-height ratio (WHtR) and waist-to-hip ratio (WHR)). Time-to-event analyses were conducted using time-dependent Cox regression models. Obesity measures were analyzed both as dichotomous (obese vs. non-obese) and standardized continuous predictors. Covariates included age, sex, systolic blood pressure, diastolic blood pressure, and smoking status. Results: There were 331 incident CVD events, including 122 CHD, 192 heart failure, and 95 stroke events. The median follow-up time is 9.04 (Quartile Range: 1.83) years. Comparing obese to non-obese participants, adjusted hazard ratios (aHR) (95% CI) for CVD were 1.38 (1.10, 1.72), 1.36 (1.05, 1.76), 1.23 (0.80, 1.89), and 1.38 (1.04, 1.82) for BMI, WC, WHtR and WHR, respectively. The corresponding aHRs (95% CI) using standardized continuous BMI, WC, WHtR and WHR measures were 1.27 (1.14, 1.43), 1.32 (1.18, 1.47), 1.31 (1.18, 1.46), and 1.24 (1.12 1.39). The significant associations between CVD and obesity measures were largely driven by heart failure. The associations in younger age group (< 60 yrs) were more significant than those in older age group (>= 60 yrs). Conclusions: Dichotomous and standardized continuous BMI, WC and WHR measures were all associated with incident CVD. For WHtR, only the standardized continuous measure was significantly associated with incident CVD. The risk of CVD associated with obesity is not evident in the elderly.