Abstract P320: Discordance of Framingham Risk Score and Reynolds Risk Score to Global Vascular Risk Score: Results from the Hispanic Community Health Study/Study of Latinos

Abstract

Introduction: The Framingham Risk Score (FRS) is a coronary heart disease (CHD) risk model established using an ethnically homogeneous population that predicts 10-year hard CHD events, myocardial infarction (MI) and coronary death. The Reynolds Risk Score (RRS) and Global Vascular Risk Score (GVRS) are validated CHD risk models that, in addition to hard CHD events, predict stroke and other CHD outcomes. In addition to major CHD risk factors, RRS adds systemic inflammation and family history of MI as GVRS adds behavioral and anthropometric measures. This study aims to compare agreement of RRS and GVRS with FRS among Hispanic/Latino adults and to describe discordance in RRS and GVRS with FRS categories, by socio-demographic characteristics. Methods: HCHS/SOL is a population-based cohort study of Hispanics/Latinos in four US communities. The analytic sample includes 6,058 non-diabetic participants 45-74 years of age with no past history of CHD and stroke who underwent comprehensive baseline examination. 10-year hard CHD risk score was calculated; participants were categorized as low (<10%), moderate (10-<20%), and high (≥20%) risk. Kappa scores were calculated to compare agreement of RRS and GVRS with FRS. Socio-demographic characteristics of concordance and discordance were characterized overall; multinomial logistic regression models was used to examine age-sex-adjusted likelihood of in discordance by these factors. Results: Mean age of the participants was 55 (SE=0.15) years, 54.3% were women, 41% had family history of CHD, and 90% were foreign born. Overall, 4,805 (74%) had low FRS, 1,143 (24%) had moderate FRS, and 110 (2%) had high FRS. There was poor agreement between RRS and FRS (Kappa=0.16, P<0.01) and fair agreement between GVRS and FRS (Kappa=0.36, P<0.01). In age-sex-adjusted analyses, RRS and GVRS were both more likely to classify persons of moderate and high risk who are between the ages of 60-74; GVRS classified more moderate and high risk women than the FRS. RRS and GVRS discordance with FRS was not associated with nativity and length of time in US. Conclusion: Significant discordance was observed between RRS and GVRS compared to FRS. Among Hispanic/Latino adults, use of RRS or GVRS may be more inclusive in classifying older age adults and women at high 10-year CHD risk.