Supplementary Material for: Serum microRNA-21 and microRNA-221 as Potential Biomarkers for Cerebrovascular Disease

<b><i>Background/Aims:</i></b> MicroRNA miR-21, miR-221 and miR-145 have been implicated in the cardiovascular system. We aimed to compare the serum levels of the three microRNAs (miRNAs) in different severities of cerebrovascular diseases and evaluate the feasibility of using these miRNAs as biomarkers for stroke. <b><i>Methods:</i></b> We enrolled 167 subjects with ischemic stroke, 66 atherosclerosis subjects with any carotid plaque score and 157 healthy controls. These three types of subjects represent three levels of severity in cerebrovascular diseases. Analysis of covariance was used to evaluate the relationship between miRNAs and disease severity with adjustment for conventional risk factors. To test the prediction for stroke, we built regression models containing the serum miRNA levels and risk factors. Prediction capabilities were compared by the receiver operating characteristic curves. <b><i>Results:</i></b> Stroke patients and atherosclerosis subjects had significantly higher miR-21 and lower miR-221 serum levels than healthy controls, while the miR-145 expression was too low to provide useful information in this regard. The best model showed that miR-21 and miR-221 were independent predictors. There was a 6.2-fold increase for stroke risk when miR-21 levels increase by log<sub>10</sub>2<sup>-ΔCt</sup> = 1, while a 10.4-fold increase was observed as miR-221 decreases by log<sub>10</sub>2<sup>-ΔCt</sup> = 1. <b><i>Conclusions:</i></b> Serum miR-145 was not detected in over 50% of the patients and it may not be an ideal marker to predict stroke. MiR-21 and miR-221 are novel biomarkers for atherosclerosis and stroke.