Abstract P199: Performance of the Atherosclerotic Cardiovascular Disease Pooled Cohort Risk Equations by Socioeconomic Status. The REasons for Geographic And Racial Differences in Stroke (REGARDS) Study

Abstract

Background: The Pooled Cohort risk equations over predict atherosclerosis cardiovascular disease (ASCVD) risk in some contemporary cohorts but not in others. ASCVD risk differs by socioeconomic status (SES) which may influence the performance of the risk equations. We analyzed the performance of the Pooled Cohort risk equations by SES and tested whether adding SES improves the discrimination of the Pooled Cohort risk equations. Methods: We used data from black and white participants enrolled in the REasons for Geographic And Racial Differences in Stroke (REGARDS) study from 2003 to 2007 (all participants ≥45 years of age). The analysis was restricted to participants for whom ASCVD risk may lead to consideration of statin initiation (45-79 years of age without ASCVD or diabetes, LDL cholesterol 70-189 mg/dL, and not taking lipid-lowering medications, N=10,231). SES was assessed using data on living alone, education and household income. Incident ASCVD (fatal/non-fatal stroke, MI or CHD death) was adjudicated through December 2012. Results: The Pooled Cohort risk equations were well calibrated among participants with low SES but over predicted ASCVD risk among participants with higher SES (Table). Discrimination of the equations improved after adding household income (continuous net reclassification improvement: 0.12 [95% CI: 0.03, 0.21]) but not living alone (0.06 [-0.03, 0.14]), or education (-0.04 [-0.13, 0.06]). The continuous net reclassification improvement was 0.10 (95% CI: 0.02, 0.19) after adding the cumulative number of indicators for low SES to the variables in the Pooled Cohort risk equations. Discussion: SES may help explain why the calibration of the Pooled Cohort risk equations has varied across cohorts. Adding information on SES may improve the performance of ASCVD risk equations.