ABC | Volume 111, Nº1, July 2018

Original Article Pereira et al. Genes and coronary artery disease risk Arq Bras Cardiol. 2018; 111(1):50-61 Supplementary Table 2 – Logistic regression with respective ORs and ROC curves with respective AUCs of the GRS models GRS models OR (95% CI) P value 1 AUC (95% CI) Sensitivity (%) Specificity (%) P value 2 Multiplicative 1.78 (1.52 – 2.10) < 0.0001 0.61 (0.59 – 0.62) 54.0 62.3 < 0.0001 Additive 1.06 (1.04 – 1.09) < 0.0001 0.56 (0.54 – 0.58) 58.7 50.5 < 0.0001 Weighted (Best model OR) 1.02 (0.94 – 1.10) 0.660 0.57 (0.55 – 0.59) 41.0 70.3 < 0.0001 Weighted (Beta) 2.23 (1.88 – 2.65) < 0.0001 0.60 (0.58 – 0.61) 43.0 71.5 < 0.0001 Weighted (Literature OR) 1.35 (1.12 – 1.62) 0.001 0.54 (0.52 – 0.55) 53.4 54.1 0.008 Classic weighted 3.01 (2.32 – 3.89) < 0.0001 0.59 (0.57 – 0.61) 59.4 54.4 < 0.0001 OR: Odds ratio; ROC: Receiver Operating Characteristic; AUC: Area under curve; GRS: Genetic Risk Score; CI: Confidence interval; P value 1 : Obtained by logistic regression to evaluate the significance of the odds ratio; P value 2 : Obtained by the ROC Curve to verify the significance of the area under the curve; Statistically significant for p < 0.05. 61