This article is more than 5 years old. Information may no longer be current.

Socioeconomics may be bigger driver of atherosclerotic CVD risk than previously thought

A risk model for predicting major atherosclerotic CVD–related events systematically underestimates the negative cardiovascular effects of living in a socioeconomically disadvantaged neighborhood, according to a study published in Annals of Internal Medicine.

“Atherosclerotic CVD remains the leading cause of death for most Americans,” Jarrod E. Dalton, PhD, from Cleveland Clinic, and colleagues wrote. “Even modest reductions in cardiovascular health disparities have the potential to substantially improve the health and well-being of socioeconomically challenged populations.”

“Risk varies by race and socioeconomic position; however ... [socioeconomic position] generally is not considered in cardiovascular risk assessment,” they added. “In 2014, the American College of Cardiology/American Heart Association Task Force on Practice Guidelines released the Pooled Cohort Equations Risk Model (PCERM) for 10-year ASCVD risk ... Although the goal of the PCERM was to establish more demographically representative models for [atherosclerotic] CVD events, it did not incorporate variation in risk directly related to [socioeconomic position].”

Dalton and colleagues performed an observational cohort analysis of geocoded longitudinal electronic health records to assess the spatial association between neighborhood disadvantage and major atherosclerotic CVD–related events, as well as the degree to which neighborhood-level variation in atherosclerotic CVD event rates can be explained by neighborhood disadvantage and physiologic risk. They also evaluated the predictive accuracy of PCERM in regard to neighborhood socioeconomic position.

The study included 109,793 patients who had an outpatient lipid panel drawn at the Cleveland Clinic Health System between 2007 and 2010. The researchers measured the time from baseline (date of first qualifying lipid panel) to first major atherosclerotic CVD event, such as myocardial infarction, stroke or CV death within 4 years. They created a neighborhood disadvantage index that represented several factors of neighborhood socioeconomic position and served as a specific measure of neighborhood disadvantage across northeastern Ohio.

The researchers found that among patients in disadvantaged neighborhoods, the atherosclerotic CVD event risk was systematically underpredicted by the PCERM. Compared with the PCERM, the neighborhood disadvantage index accounted for more than three times the amount of census tract–level variation in atherosclerotic CVD event rates (32% vs. 10%). Stroke, acute MI and CV death were the most commonly observed events. 

“Neighborhood [socioeconomic position] appears to be an important determinant of PCERM accuracy,” Dalton and colleagues concluded. “Efforts are needed to enhance risk prediction by incorporating aspects of neighborhood [socioeconomic position] and discerning its systemic effects on individuals. Such efforts are particularly important in the context of health disparities in [atherosclerotic] CVD, whereby the mechanisms involved in [atherosclerotic] CVD progression may differ qualitatively among subpopulations defined according to social strata. In addition to supplemental risk models and clinical screening criteria, a collective approach is needed to develop grass-roots and policy-oriented approaches to ameliorate the deleterious effects of neighborhood conditions on health outcomes.”

In an accompanying editorial, Sandro Glea, MD, DrPH, from Boston University, and Katherine M. Keyes, PhD, from Columbia University, wrote that these findings are a “sobering reminder” that predictive models may fail to account for base prevalence and co-occurring factors.

“Even if clinicians recognize the conceptual challenges behind prediction models, they may have difficulty factoring in those limitations when considering the neat percentage risk for a particular disease that a predictive model offers,” they concluded. “Of course, if that risk is misestimated, patient management may be suboptimal. The field of predictive models in health is overdue for a conceptual and methodological rethink, building on the principles of population health science mentioned earlier. The current study by Dalton and colleagues is a good first step in this direction.” – by Alaina Tedesco

Disclosure: Dalton reports receiving grants from NIH/NCATS. Please see full study for complete list of all other author’s relevant financial disclosures.