Abstract P107: Differences in Cardiovascular Mortality Risk Among African Americans in the Minnesota Heart Survey, 1985-2015, versus African Americans in the Atherosclerosis Risk in Communities Study (ARIC) Cohort: 1987-2015

Abstract

Geographic differences in CVD mortality across the U.S. are well-established, but frequently overlooked. ARIC enrolled African Americans (AA) from Jackson, MS and Forsyth County, NC, areas of the Southeast with some of the highest CVD mortality rates, especially among AAs. The Minnesota Heart Survey enrolled AAs from Minnesota where CVD rates are among the lowest. However, it is not known whether AAs in Minnesota also have low rates. Using these two cohorts, we assessed whether CVD-related mortality risk among AAs differs by region. Baseline measures of CVD risk factors for MHS were taken in 1985 from a population based sample of AAs, ages 45 to 65, living in the Minneapolis-St. Paul metropolitan area. These same measures were made at ARIC visit 1 (1987-89) in AA participants of the same age residing in Jackson, MS and Forsyth County, NC. CVD and total mortality were identified using ICD codes for underlying cause of death from State and National Death Index records in both cohorts. We compared MHS and ARIC on CVD death rates using Poisson regression, prevalence of risk factors, and risk factor hazard ratios using Cox regression. After risk factor adjustment, AA men in MHS had a rate of 5.2 (95% CI: 3.2, 7.2) CVD deaths per 1000 person-years compared to 15.1 (95% CI: 13.1, 17.1) for AA men in ARIC. For AA women, MHS had 4.1 (95% CI: 2.7, 5.5) CVD deaths per 1000 person-years versus 10.2 (95% CI: 9.0, 11.4) in ARIC. CVD mortality rates were higher in Jackson than Forsyth County within ARIC. CVD death rates paralleled risk factor prevalence at baseline. Compared to MHS, ARIC had significantly higher total cholesterol (215 vs. 202 mg/dL), albeit higher HDL cholesterol (55 vs. 53 mg/dL), as well as higher anti-hypertensive medication use (41 vs. 30%), diabetes (13 vs. 11%) and BMI (30 vs. 29 kg/m 2 ), while smoking did not differ. Despite risk factor differences, hazard ratios of CVD death associated with each risk factor did not differ between studies even after inclusion of a competing risk of non-CVD death. In conclusion, the CVD death rate was lower in AAs in MHS than in AAs residing in the Southeast in ARIC largely due to lower risk factor levels, since the hazard of CVD death for each risk factor did not differ. Study differences reflect incompletely identified geographic variation that need further exploration, especially in the context of health disparities, but support maintaining low risk as a key to CVD prevention.