BACKGROUND: A discordant biological clock could potentially induce sudden cardiac death (SCD). We aimed to evaluate the circadian change of heart rate variability (HRV) and its relationship to the risks of ventricular arrhythmia (VA) and near syncope in patients with chronic kidney disease (CKD). METHODS: In this retrospective study, non-CKD and CKD patients were enrolled and underwent a 24-hour Holter examination for linear and nonlinear HRV analyses. The multiscale entropy (MSE) method was selected for nonlinear HRV analyses. The documented VAs or episodes of near syncope were classified as high-risk SCD group (n=8) and others as low-risk SCD group (n=21). RESULTS: In linear analyses, time and frequency domains revealed no significant difference between groups. In nonlinear analyses with MSE, MSE(5), MSE(6-20), and MSE(slope 5) were significantly lower (p=0.002, p<0.0001, and p=0.013) in the high-risk SCD group, compared to those in the low-risk SCD group, respectively. Comparing between daytime and nighttime within each group, the MSE(5) revealed no difference in the high-risk SCD group (p=0.128), whereas the daytime was significantly higher in the low-risk SCD group (p=0.048). The area under the curve (AUC) analysis revealed MSE(6-20) has the best predictive power associated with VAs and near syncope with a cut-off value of </=24.64 (p<0.001). CONCLUSIONS: Nonlinear analysis with MSE demonstrated the loss of circadian change in CKD patients and was associated with a higher risk for VAs and near syncope. The MSE method demonstrated the diurnal change of rhythm dynamics which identifies potential autonomic dysfunction leading to poor prognosis.