Prediction of 10-year cardiovascular disease risk by diabetes status and lipoprotein-a levels; the HellenicSCORE II+
Demosthenes B. Panagiotakos, Christina Chrysohoou, Christos Pitsavos, Konstantinos Tsioufis
Abstract
BACKGROUND: The aim of this study was to develop an updated model to predict 10-year cardiovascular disease (CVD) risk for Greek adults, i.e., the HellenicSCORE II+, based on smoking, systolic blood pressure (SBP), total and high-density-lipoprotein (HDL) cholesterol levels, and stratified by age group, sex, history of diabetes, and lipoprotein (Lp)-a levels. METHODS: Individual CVD risk scores were calculated through logit-function models using the beta coefficients derived from SCORE2. The Attica study data were used for the calibration (3,042 participants, aged 45 (14) years; 49.1% men). Discrimination ability of the HellenicSCORE II+ was assessed using C-index (range 0-1), adjusted for competing risks. RESULTS: The mean HellenicSCORE II+ score was 6.3% (95% confidence interval (CI) 5.9% to 6.6%) for men and 3.7% (95% CI 3.5% to 4.0%) for women (p < 0.001), and were higher compared to the relevant SCORE2; 23.5% of men were classified as low risk, 40.2% as moderate, and 36.3% as high risk, whereas the corresponding percentages for women were 56.2%, 18.6%, and 25.2%. C-statistic index was 0.88 for women and 0.79 for men when the HellenicSCORE II+ was applied to the Attica study data, suggesting very good accuracy. Stratified analysis by Lp(a) levels led to a 4% improvement in correct classification among participants with high Lp(a). CONCLUSION: HellenicSCORE II+ values were higher than SCORE2, confirming that the Greek population is at moderate-to-high CVD risk. Stratification by Lp(a) levels may assist in better identifying individuals at high CVD risk.