aThe model fit was a proportional hazards Cox model with appropriate weighting for the case-cohort design.bP-values obtained from Wilcoxon signed rank (continuous) or Chi-Square (dichotomous) test.cAd
†Non-parametric Kruskal-Wallis test for association was performed between methylation and categorical predictor variables. Spearman's rank correlation was assessed between methylation and continuous p
Plaque characteristics of all patients and divided by CXCL4 quartiles. For each parameter the coefficient and P value with CXCL4 plasma levels is indicated as calculated by Pearson’s correlation. Cont
