Atrioventricular block can be used as a risk predictor of clinical atrial fibrillation

Background Atrial fibrillation (AF) is the most common cardiac arrhythmia, with its incidence making up nearly one‐third of all hospital admissions. Atrioventricular block (AVB) is a conduction abnormality along the atrioventricular node or the His‐Purkinje system. The relationship between atrioventricular conduction block and AF is controversial. Hypothesis This study is designed to observe whether there is a correlation between AVB and AF, and which type of AVB has the most obvious correlation with AF. Methods This study retrospectively reviewed 1345 patients. We classified the AVB according to the AVB classification criteria. One hundred and two patients were excluded, and the final total sample size was 1243 patients, including 679 patients in the AF group (378, 55.7% males) and 564 patients in the non‐AF group (287, 50.8% males). AF group and non‐AF group were compared to observe the relationship between AVB and AF. Results The I AVB have a relative statistical risk of 1.927 (95% confidence interval [CI]: 1.160‐3.203, P < 0.05) with the occurrence of AF. II AVB occupied the largest proportion, accounting for 67 cases (9.87%), and the statistical risk of II AVB in AF is 16.845 (95% CI: 6.099‐46.524, P < 0.000). III AVB has a comparative statistical risk of 17.599 (95% CI: 4.212‐73.541, P < 0.000). Conclusions The three types of AVB in the AF group were significantly higher than that in the non‐AF group. II AVB has the highest incidence rate compared with other types of AVB in the AF group. AVB can be used as a risk factor for AF occurrence.

which carries a substantial risk of morbidity and mortality. However, the reported results of studies on the relationship between I AVB and AF are inconsistent. It is a controversial argument whether I AVB is a pathological state and related to AF. Some studies suggest that individuals with first-degree AVB are at a substantially increased risk of future AF (~2-fold) compared to individuals without first-degree AVB.
PR prolongation could be a marker of other changes in the cardiovascular system that contributes to a worse prognosis. Prior studies have raised the possibility that slowed intra-atrial or interatrial conduction may directly increase the risk of AF. [12][13][14][15][16][17] An association between PR interval and AF risk has been reported in a separate study that uses Framingham data to construct a clinical risk score for AF, 18 while others claim that first-degree AVB has a benign prognosis. 7,9,10,19,20 First-degree AVB typically occurs in the absence of acute cardiovascular disease. 7,10 The relationship between AVB and AF has always been controversial.
Second-degree AVB is first described in 1899. Second-degree AVB remains poorly understood despite major advances in cardiac electrophysiology in the past three decades. [21][22][23][24][25] Because seconddegree AVB can be present in normal children, young adults, and athletes, 26,27 it does not usually present with any symptoms. However, this is not always the case. The second-degree AVB is also linked to underlying heart diseases such as intrinsic atrioventricular nodal disease, structural cardiomyopathy, myocarditis, endocarditis, acute inferior myocardial infarction, post cardiac surgery, ablation, catheterization procedures, and secondary to hypothyroidism or hyperthyroidism. 28 AF associated with AVB is more likely to induce thrombosis and thromboembolic events as well as heart failure and myocardial ischemia due to a slow ventricular rate and electrical remodeling. 29 Whether AF accompanied with second-degree AVB has clinical significance in the diagnosis of the patient's condition, drug intervention, and treatment guidance, which is worthy of further investigation.
At present, there are not too much controversies about thirddegree AVB. Third-degree AVB is generally accompanied by the pathological state of the disease and is closely related to the occurrence of the disease. Therefore, the current controversy over thirddegree AVB is not too much exalted. The purpose of this study is to analyze the occurrence of atrioventricular conduction block in patients with AF and non-AF, and to explore the relationship between the AF and the three types of atrioventricular conduction block, and find out which one is the most common type of AVB occurred in AF patients. FIGURE 1 The above flow chart shows the design of inclusion and exclusion of atrial fibrillation (AF) and non-AF group. After excluding 102 patients, 679 patients with atrial fibrillation and 564 patients without atrial fibrillation were finally included in the Atrial Fibrillation center of Xi 'an Jiaotong University. To study the relationship between AVB and atrial fibrillation, I AVB,II AVB and III AVB in the atrial fibrillation group and the non-atrial fibrillation group were respectively compared 2 | METHODS 1. The clinical data process of AF and non-AF groups ( Figure 1): Figure 1 shows the whole process of the data collection and analysis.
2. The inclusion and exclusion criteria of AF and non-AF: The inclusion criteria of AF: (1) More than two episodes of AF occurred and recorded by ECG before.
The exclusion criteria of AF: (1) Valvular heart disease and rheumatic heart disease excluded.

II AVB of Type II:
It means that the atrial impulse cannot be transmitted to the ventricle, and the ECG is represented as the leakage of QRS interphase.

III AVB:
It means that all atrial excitations cannot be transmitted to ventricles, and the activities of atria and ventricles are independent, respectively, and unrelated to each other.

Case selection
This experiment is designed to study the relationship between AVB and AF. The study retrospectively reviewed 1345 consecutive patients from August 2017 to August 2018, who were diagnosed with AF and non-AF. All patients were enrolled from the Atrial Fibrillation Center, Department of Cardiology, the First Affiliated Hospital of Xi'an Jiaotong university. The clinical data were obtained from medical record review and analyzed in AF center. Patients were divided into the AF group and non-AF group. Among the total of 1345 patients, 102 patients who did not meet the inclusion criteria were excluded:26 cases were diagnosed with Valvular heart disease; 12 cases had dilated heart disease; 6 cases had hypertrophic heart disease; 4 cases had ischemic cardiomyopathy; 17 cases had Congenital heart disease; 18 cases had anemia; 17 cases had hyperthyroidism; and 4 presented with severe left ventricle dysfunction (EF < 35%). Meanwhile, patients in the AF group must comply with the following conditions: (a) more than two episodes of AF occurred and recorded by ECG before and (b) aged from 18 to 80. The final total sample size was 1243 patients, including 679 patients in the AF group (378, 55.7% males) and 564 patients in the non-AF control group (287, 50.8% males).

| STATISTICAL ANALYSIS
Independent sample t test and Pearson χ 2 test were used to compare the continuous variables and the classified variables. Single factor analysis method was used to preliminarily analyze related risk factors.
Statistically significant factors were taken into multiple factors analysis model by using negative binomial regression for the analysis of the related factors. The aim is to evaluate whether there is statistical significance between AF and corresponding covariates in the negative binomial regression model. Finally, the correlation between AVB interval and AF was found.

| The distribution of AVB in AF and non-AF groups
From Table 1, we found that atrioventricular conduction block of degree I, degree II, and degree III increased in the patients with AF compared with patients with non-AF.
The I AVB has a relative statistical risk of 1.927 (95% CI: 1.160-3.203, P < 0.05) with the occurrence of AF. II AVB occupies the largest proportion, accounting for 67 cases (9.87%), and the statistical risk of II AVB of AF is 16.845 (95% CI: 6.099-46.524, P < 0.000) compared with the non-AF group. III AVB has a relative statistical risk of 17.599 (95% CI: 4.212-73.541, P < 0.000) in the AF group compared with the non-AF group.
According to the analysis results of this study, second atrioventricular conduction block occupies the largest percentage,

| Univariate analysis of AF and its related variables are performed
The statistical results are shown in Table 2. We include relevant variables of the AF group and the non-AF group into the univariate

| Negative binomial regression analysis of AF and non-AF groups
We select variables to be included in the multivariate regression model by consulting relevant literature. Variables to be included are selected by DAG diagram and results of univariate analysis. As it is shown in

| DISCUSSION
AF is the most common sustained cardiac arrhythmia in clinical practice, and its incidence is increasing rapidly worldwide accounting for nearly one-third of all hospital admissions. 1 It is also associated with increased risks of stroke, heart failure, cognitive dysfunction, impaired quality of life, and substantial healthcare costs. Furthermore, it eventually contributes to increased risks of thromboembolic events, 2 cardiac, and overall mortalities, 3-6 which may result in different outcomes. AVB is categorized into three types according to the extent of its extension and the characteristics of ECG: first-degree AVB, second-degree AVB, and third-degree AVB.

| I AVB and AF
The relationship between I AVB and arrhythmia has been controversial. The reported results of studies on the relationship between I AVB and AF are inconsistent. Some studies suggest that individuals with first-degree AVB are at a substantially increased risk of future AF (~2-fold) 8 compared with individuals without first-degree AVB. An association between PR interval and AF risk has been reported. 18 PR prolongation could be a marker of worse prognosis in the cardiovascular system. Potential explanations may account for the association of longer PR interval with AF risk. In addition, a prolonged PR interval results in delayed and ineffective mitral valve closure and diastolic mitral regurgitation, 30 especially when the PR interval exceeds 230 milliseconds, 31 while others claim that first-degree AVB has a benign prognosis, 7,9,10,19,20 although these studies are based on young, healthy men in the military. 7,19 First-degree AVB typically occurs in the absence of acute cardiovascular disease. 7,10 The relationship between I AVB and AF has always been controversial.
Through the statistical analysis of the AF data in the Atrial Fibrilla-   Abbreviations: AF, atrial fibrillation; AVB, atrioventricular block. P value less than 0.01 is marked with **; P value less than 0.001 is marked with *** AVB has a relative statistical risk of 1.927 (95% CI: 1.160-3.203, P < 0.05) with the occurrence of AF. I AVB is a risk factor for AF.

| II AVB and AF
Second-degree AVB is often associated with an underlying heart disease such as atrioventricular node disease, structural cardiomyopathy, myocarditis, endocarditis, acute inferior myocardial infarction, post cardiac surgery, ablation, catheterization procedures, and secondary to hypothyroidism or hyperthyroidism. 28 Furthermore, AF associated with AVB is more likely to induce thrombosis and thromboembolic events as well as heart failure and myocardial ischemia due to a slow ventricular rate and electrical remodeling. 29

| III AVB and AF
At present, there are not too much reports and controversies over III AVB and AF. III AVB is usually accompanied with the pathological state of the disease and is closely related to the occurrence of the disease. Therefore, the current controversy over III AVB is not too much great. Through the statistical analysis of the data, we draw the conclusion that: III AVB has a relative statistical risk of 17.599 (95% CI: 4.212-73.541, P < 0.000) in the AF group compared with the non-AF group. III AVB is also an obvious risk factor for AF.
To sum up, the three types of AVB are risk factors for AF. The proportion of three types of AVB in the AF group is, respectively, increased compared with the non-AF group, and all of them have statistical significance. The I AVB has a relative statistical risk of 1.927 (95% CI: 1.160-3.203, P < 0.05) with the occurrence of AF. II AVB takes the largest percentage, accounting for 67 cases (9.87%), and the statistical risk of II AVB in AF is 16.845 (95% CI: 6.099-46.524, P < 0.000) compared with the non-AF group. III AVB has a relative statistical risk of 17.599 (95% CI: 4.212-73.541, P < 0.000) in the AF group compared with the non-AF group. Although I AVB has been controversial, this study agrees that I AVB is a risk factor for AF. The comparatively relevant and highest incidence of AVB is II AVB.
Second-degree AVB is the most correlated type with AF among the three types of AVB. At the same time, third-degree atrioventricular conduction block is also strongly related to the occurrence of AF. The results point out that II and III degree AVB are also risk factors for AF.
The three types of AVB are risk factors for AF occurrence, and II AVB has the highest proportion and high correlation with the occurrence of AF.

| Limitation
This study is an observational research conducted in a single center.
The research center is relatively single, and the sample representation is not enough.

ACKNOWLEDGMENTS
Thanks very much for the guidance from my doctoral supervisor, for the assistance from the Atrial Fibrillation Center of The First Affiliated Hospital of Xi'an Jiaotong University, for the help from my friends in all aspects, for the music accompanying me, and for all the wonderful things in life.