According to the results of the "Measuring Effects on intima-media thickness: Evaluation of Rosuvastatin (METEOR) 21" trial, the average annual increase in IMT at 12 sites of the carotid artery was 0.012 mm (SD, 0.058). Assuming no significant difference in the annual increase in mean IMT in the Tong Xin Luo group, at least 491 patients in the Tong Xin Luo and placebo groups required 90% power to detect a difference of 0.012 mm (SD, 0.058) in mean carotid IMT changes between groups with an alpha level of 0.05. Assuming a 20% dropout rate in each group, 590 patients would be required in each group. The outcome data for all subjects who were randomized were analyzed according to the intention-to-treat principle. Primary and secondary results were analyzed using a multilevel repeated measurement linear mixed effects model, which was specified in terms of the fixed effects of group, time, and the interaction between group and time. The random effects in the model are the intercepts and slopes of individual participants and sites within participants. All continuous values are expressed as mean ± SD, except for the result values of changes from baseline to 12 or 24 months, which are expressed as least square (LS) averages and 95% confidence intervals. Disaggregated data are expressed as percentages. A sensitivity analysis of the main results was performed using a mixed model of multiple pulses and modes. The time of the first major cardiovascular event was estimated using Kaplan-Meier method, and the difference between the two groups was determined using logarithmic rank test. To test the reproducibility of ultrasound measurements, two independent researchers re-measured four key variables, namely IMT, plaque area in the long - and short-axis visual fields, and RI, in 50 randomly selected patients. Inter-observer variability was assessed by two investigators and intra-observer variability by one investigator at different times. Bland-Altman plots are used to analyze inter-observer and intra-observer variability and to calculate intra-group correlation coefficients. The test is two-tailed and always displays a nominal P-value without adjustment. All statistical analyses involved the use of SAS 9.2 (SAS Inst., Inc., Cary, NC, USA).