Cardiovascular Mortality Associated With 5 Leading Risk Factors: National and State Preventable Fractions Estimated From Survey Data

We recently reported that approximately half of national cardiovascular deaths among U.S. Americans aged 45-79 years could be attributed collectively to elevated cholesterol levels, diabetes, hypertension, obesity, and smoking—five modifiable biomedical risk factors for cardiovascular disease (1). Contrary to the conclusion by Grant et al., we in fact controlled for race/ethnicity, along with educational attainment, in estimating the input hazard ratios for the attributable fraction calculations (please see Methods and Statistical Appendix) (1).

We also estimated preventable fractions of cardiovascular mortality associated with each of the five risk factors individually. We found that hypertension and smoking were associated with the largest preventable fractions of cardiovascular mortality nationally among both men and women. In contrast, elevated cholesterol individually were associated with preventable fractions that could not be statistically distinguished from zero among women.

The focus of our analysis was the independent contribution of modifiable biomedical risk factors for cardiovascular disease on cardiovascular mortality. We wholeheartedly endorse investigation into the well-established “upstream” social determinants of cardiovascular health—such as race/ethnic background, education, and income—to better inform public health practice and health policy. Public-health oriented efforts to reduce the onset of biomedical risk factors in the population and social welfare policies to improve socioeconomic conditions in which risk factors and disease more broadly arise are complementary goals.

The paper by Patel and colleagues used data from 50 states plus the District of Columbia to develop a model to explain risk factors for cardiovascular disease (CVD) mortality in 2009-2010 (1). Their model included elevated cholesterol level, diabetes, hypertension, obesity, and current smoking. Since the data from NHANES used in their model were provided, we decided to reexamine their model. The data were analyzed using SPSS 20.0 (IBM, Armonk, New York). In single linear regression analyses, only three of the factors were significantly correlated with CVD mortality rates, with hypertension most important, followed by current smoking, and obesity. In a multiple-linear regression analysis, obesity was not significantly correlated with CVD. The adjusted R2 for hypertension and current smoking was 0.81.

However, one important factor was omitted explicitly from their model: ethnic background. African Americans have much higher CVD mortality rates than Asians, Hispanics or whites, and comprise large fractions of populations of many states (2). When fraction of population African American was run with hypertension, and current smoking, obesity, the best model results were obtained (adjusted R2= 0.88). However, the risk factors are not independent. Hypertension is highly correlated with current smoking, African-American population, and obesity, in that order, but not with cholesterol or diabetes.

The primary mechanism increasing blood pressure appears to be oxidative stress from reactive oxygen species (2). The Taiwan Society of Cardiology and the Taiwan Hypertension Society for the management of hypertension have issued guidelines for the management of hypertension "starting with life style modification (LSM) including S-ABCDE (Sodium restriction, Alcohol limitation, Body weight reduction, Cigarette smoke cessation, Diet adaptation, and Exercise adoption)." (3).

Another recent paper found that lower income and educational level were strongly correlated with CVD mortality rates and that minority and low socioeconomic groups explained 44% of the variation in U.S. CVD mortality rates (4). This finding suggests that even if the important CVD risk factors were identified, many who might die from CVD would be unable to change lifestyle due to economic and educational level constraints.

While other studies indicate that cholesterol is a risk factor for CVD, targeting cholesterol may not be wise. An observational study in Wales, UK involving 1773 middle-aged men followed for an average of 15.4 years found a sub-hazard ratio related to cholesterol for CVD mortality of 1.20 (95% CI, 1.05-1.37) but 0.81 (0.72-0.90) for non-CVD mortality (5).


References
1. Patel SA, Winkel M, Ali MK, Narayan KM, Mehta NK. Cardiovascular mortality associated with 5 leading risk factors: National and state preventable fractions estimated from survey data. Ann Intern Med. 2015;163:245-53.PMID:26121190
2. Montezano AC, Dulak-Lis M, Tsiropoulou S, Harvey A, Briones AM, Touyz RM. Oxidative stress and human hypertension: vascular mechanisms, biomarkers, and novel therapies. Can J Cardiol. 2015;31:631-41. PMID: 25936489
3. Chiang CE, Wang TD, Ueng KC, Lin TH, Yeh HI, Chen CY, et al. 2015 guidelines of the Taiwan Society of Cardiology and the Taiwan Hypertension Society for the management of hypertension. J Chin Med Assoc. 2015;78:1-47. PMID: 25547819
4. Gebreab SY, Davis SK, Symanzik J, Mensah GA, Gibbons GH, Diez-Roux AV. Geographic variations in cardiovascular health in the United States: contributions of state- and individual-level factors.J Am Heart Assoc. 2015;4:e001673. PMID:26019131
5. Patterson CC, Blankenberg S, Ben-Shlomo Y, Heslop L, Bayer A, Lowe G, et al. Which biomarkers are predictive specifically for cardiovascular or for non-cardiovascularmortality in men? Evidence from the Caerphilly Prospective Study (CaPS).Int J Cardiol. 2015;201:113-118. PMID:26298350