We appreciate the comments provided to our recent report 1) , and wish to respond to the letter from Kataoka, et al. 2) 1. 1) We strongly agree with your proposal to use Cox regression, and since we lack temporal data on the occurrence of events, logistic regression was used; 2) We used quartile serum laminin levels in the comparison of baseline information between the MACEs and non-MACEs groups; 3) In our study, NT-proBNP was significant in the univariate analysis, but was excluded from the multifactorial analysis, and we think the possible reasons for this situation are: 1) There are many factors affecting NT-proBNP, and the effect of NT-proBNP on MACE events in patients with AF may be indirect, resulting in the exclusion of NT-proBNP when it is included in the multifactorial model along with these factors. 2) It is caused by insufficient sample size, and in the future we plan to expand the sample size.2. MACEs during the patients' hospitalization were defined as all-cause death, acute heart failure, malignant arrhythmia, and stroke. Malignant arrhythmia includes sustained ventricular tachycardia and ventricular fibrillation, and stroke refers to symptomatic stroke.3. Our study's basic information included comorbidities such as hypertension, diabetes, and coronary artery disease.