Association of Epicardial Adipose Tissue With Bradyarrhythmias: A Matched Case–Control Study

Abstract

Background

Epicardial adipose tissue (EAT) is implicated in exerting potential proarrhythmic effects. The relationship between EAT and tachyarrhythmias is well documented. However, the connection between EAT and bradyarrhythmias has not been comprehensively explored. This study aimed to investigate the association between EAT and bradyarrhythmias.

Methods

We retrospectively quantified the volume and density of EAT using chest computed tomography scans from patients with bradyarrhythmias and case‐matched controls. Measurements were obtained through manual pericardial contour tracing with a standardized Hounsfield unit threshold of −200 to −50 Hounsfield units.

Results

A total of 652 patients were included, comprising 326 patients with bradyarrhythmias (age 74.00 [interquartile range, 64.00–81.00] years, 46.01% female) and 326 matched controls (age 72.50 [interquartile range, 63.00–80.00] years, 46.01% female). Compared with the matched control group, the bradyarrhythmia group had a significantly greater volume (119.13 [interquartile range, 87.30–151.68] cm³ versus 93.00 [interquartile range, 67.87–116.71] cm³, P<0.001) and a lower density of EAT (−92.02±4.00 Hounsfield units versus −90.68±3.73 Hounsfield units, P<0.001). Conditional logistic analysis demonstrated that the volume rather than the density of EAT is a significant influencing factor for bradyarrhythmias. Subgroup analysis indicated a progressive increase in average EAT volume from groups with first‐degree to third‐degree atrioventricular block. In participants with first‐degree atrioventricular block, there was a positive correlation between PR interval and EAT volume (Spearman’s correlation coefficient: 0.328, P=0.012).

Conclusions

Patients with bradyarrhythmia demonstrated significantly higher EAT volumes compared with matched controls, with a progressive increase observed across advancing grades of AVB. This dose‐dependent relationship between EAT burden and conduction system impairment underscores its potential role as a modifiable influencing factor in arrhythmogenesis. Further investigations are warranted to determine whether targeted EAT reduction could mitigate conduction abnormalities.

Registration

URL: https://www.chictr.org.cn; Unique Identifier: ChiCTR2400088446.

Nonstandard Abbreviations and Acronyms

AVB

atrioventricular block

EAT

epicardial adipose tissue

SND

sinus node dysfunction

Clinical Perspective.

What Is New?

• This is the first matched case–control study demonstrating that patients with bradyarrhythmias have significantly higher epicardial adipose tissue volume and lower density than matched controls.

• Epicardial adipose tissue volume, but not density, is identified as an independent influencing factor for bradyarrhythmias after multivariable adjustment; a progressive increase in epicardial adipose tissue volume is observed across advancing grades of atrioventricular block, from first degree to third degree.

What Are the Clinical Implications?

• Epicardial adipose tissue volume serves as an independent influencing factor for bradyarrhythmias, supporting its clinical utility in risk stratification and establishing it as a potential therapeutic target for cardiac conduction disorders.

Bradyarrhythmias represent a major global health burden, driving millions of pacemaker implantations annually. ¹ Bradyarrhythmias can be compartmentalized broadly into 2 categories, including sinus node dysfunction (SND) and atrioventricular block (AVB). The clinical manifestations of bradyarrhythmias range from subtle, insidious symptoms to episodes of pronounced syncope. ² SND is commonly associated with age‐related, progressive and degenerative fibrosis of the sinus nodal tissue and surrounding atrial myocardium. ³ , ⁴ , ⁵ Meanwhile, AVB is generally attributed to idiopathic fibrosis of the conduction system. ⁶ , ⁷

Epicardial adipose tissue (EAT) is a unique type of visceral fat situated between myocardium and the visceral layer of epicardium, vascularized by branches of the coronary arteries. ⁸ This structural intimacy confers significant pathophysiological relevance in cardiovascular diseases. ⁹ Accumulating evidence implicates EAT volume in promoting atrial fibrillation incidence, severity, and postablation recurrence. ¹⁰ Additionally, our previous study has discovered that a higher volume and a lower density of EAT are significantly associated with frequent premature ventricular complexes. ¹¹ EAT has garnered considerable attention regarding its potential proarrhythmic effects. Despite its recognized proarrhythmic roles in tachyarrhythmias, such as atrial fibrillation and premature ventricular complexes, the relationship between EAT and bradyarrhythmias remains unexplored.

The pathogenic mechanisms underlying bradyarrhythmias remain unclear. Prior evidence suggests EAT may promote arrhythmogenesis through anatomical insulation of conduction pathways and paracrine secretion of profibrotic mediators. ¹² , ¹³ Given the established link between EAT and tachyarrhythmias but the unexplored relationship with bradyarrhythmias, we hypothesized that EAT characteristics may associate with bradyarrhythmias. Our study sought to characterize the potential relationship between different types of bradyarrhythmias and characteristics of EAT underlying chest computed tomography (CT).

METHODS

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Study Population and Study Design

Adult patients diagnosed with bradyarrhythmias at Zhongda Hospital affiliated with Southeast University, between December 1, 2019, and June 30, 2024, were included in this study. This was a case–control study including patients with bradyarrhythmias and a control group of patients who had no history of any bradyarrhythmia. Control participants were individually selected from a large pool of eligible individuals without bradyarrhythmias and matched to each case in a 1:1 ratio using a direct nearest‐neighbor matching algorithm. The matching was performed without replacement (meaning each control could only be used once) based on the following prespecified criteria to ensure close comparability: exact matching for sex, history of hypertension, coronary artery disease (CAD), and diabetes. Caliper matching with a tolerance of ±2 years for age and ± 1.5 kg/m² for body mass index (BMI). This process ensured that for each case, the selected control was the closest available match from the pool according to these criteria, thereby minimizing potential confounding from these variables. According to classic classification criteria, ² the population with bradyarrhythmia were divided into a group with SND and a group with AVB. To comprehensively compare the associations of EAT with 3 AVB subtypes, the group with AVB were divided into 3 categories, namely first‐degree, second‐degree, and third‐degree AVB (complete AVB). Each participant underwent standard 12‐lead ECG examination and had clear CT images of the chest within 3 months. All CT scans were clinically indicated (eg, preoperative assessment, symptom evaluation) and performed using standardized protocols. ECG measurements and diagnoses were checked by at least 2 well‐trained physicians. Bradyarrhythmias were diagnosed in accordance with American College of Cardiology/American Heart Association/Heart Rhythm Society practice guidelines. ²

Participants were excluded from this study with the following conditions: reversible factors such as hypothyroidism, hyperkalemia, or use of medications affecting heart rate or atrioventricular conduction (beta blockers, nondihydropyridine calcium channel blockers, ivabradine, digoxin, Class I/III antiarrhythmics); acute myocardial infarction; prevalent heart failure, infection status, a clear history of cardiac surgery, atrial fibrillation, malignant neoplasms, or chronic kidney disease receiving dialysis or awaiting renal transplant. Participants lacking complete data on height, weight, BMI, or ECGs were also excluded. As CT imaging is not routine in clinical practice, patients without scans were excluded. The Ethics Committee Review Board of Zhongda Hospital affiliated to Southeast University, China, approved this study (Approval No: 2024ZDSYLL283‐P01) and waived the requirement for informed consent due to its retrospective nature.

Data Collection

Data retrospectively collected from the hospital information system (Zhongda Hospital) included age, sex, BMI (kg/m²), comorbidities (hypertension, diabetes, and CAD), triglycerides, total cholesterol, low‐density lipoprotein cholesterol, high‐density lipoprotein cholesterol, creatinine, and estimated glomerular filtration rate, which was calculated from creatinine following the Chronic Kidney Disease Epidemiology Collaboration formula. ¹⁴

Chest CT and EAT Measurement

Chest CT images were obtained employing a 64‐section multidetector CT system (Discovery CT 750 HD, GE). The experiment was conducted at 120 kVp, with an autoregulated tube current ranging from 150 mAs to 400 mAs and using a collimation of 64 mm × 0.625 mm, a rotation time of 0.33 seconds, and a pitch of 1 mm.

Clear chest CT images of each patient were analyzed to evaluate the volume and density of EAT using dedicated software tools (GE AW4.6 Volume render11.3) on a specialized clinical workstation. Continuous 1 to 1.5 mm cardiac slices from the bifurcation of the pulmonary artery to the apex of the left ventricle were analyzed. When required, the incorporation of coronal and sagittal views could significantly enhance the precision of manual tracing. A threshold of −200 to −50 Hounsfield units (HU) was then applied to separate fat‐containing voxels, and the volume of EAT was automatically measured in cubic centimeters (cm³). All measurements were conducted by 2 professional researchers under the condition of the clinical status blinding. Interobserver variability was assessed in 50 randomly selected scans via intraclass correlation coefficients.

Statistical Analysis

The Shapiro–Wilk test was used to examine the normality of continuous variables. Normally distributed data were presented as mean ± SD, and nonnormally distributed data were reported as median values and interquartile ranges (IQR). The chi‐square test was used for the descriptive comparison of the matched variables (sex, hypertension, diabetes, CAD) between cases and controls to demonstrate the balance achieved by matching. For all other categorical variables (eg, smoking, drinking status) and for any paired comparisons not involving the prematched covariates, the McNemar test was applied to account for the matched study design. For paired continuous variables, if the differences followed a normal distribution, paired‐sample t tests were used for analysis. Otherwise, the Wilcoxon signed‐rank test was used for comparison. In this case–control study, conditional logistic regression analysis was employed to explore the association between bradyarrhythmia and EAT. Before conducting the multivariable conditional logistic regression, univariable conditional logistic regression analysis was carried out. Additionally, a collinearity check was performed on the independent variables. To further investigate the relationship between EAT and different types of AVB, ANCOVA was performed to compare the differences in EAT volume and density among the 3 AVB subgroups. Spearman’s correlation analysis was used to examine the relationship between EAT volume and PR interval within the case group with first‐degree AVB. All analyzed variables contained <15% missing values. For these, multiple imputation was performed via chained equations with 20 imputations. For logistic analysis, multicollinearity was assessed via variance inflation factors (<5 acceptable). Model fit was evaluated using Hosmer–Lemeshow goodness‐of‐fit test (P>0.05 indicates adequate fit). For subgroup comparisons among 3 AVB grades, ANCOVA was performed with Bonferroni‐adjusted pairwise comparisons. All tests were 2 sided, with P values <0.05 indicating statistical significance. All analyses were performed using STATA (version 18.0, StataCorp).

RESULTS

Baseline Characteristics

A total of 652 participants were enrolled in this study, comprising 326 patients with bradyarrhythmias and 326 matched controls. The baseline characteristics of the 2 groups are presented in Table 1. These 2 groups were well matched in terms of age, sex, BMI, history of hypertension, diabetes, and CAD. Patients with bradyarrhythmias exhibited a lower estimated glomerular filtration rate compared with the matched control group; however, other laboratory parameters showed no significant differences between the 2 groups (Table 1).

Table 1.

Baseline Characteristics of All Patients With Bradyarrhythmias and the Controls

	Bradyarrhythmias (n=326)	Controls (n=326)	P value
Sex, n (%)			1.000
Male	176 (53.99)	176 (53.99)
Female	150 (46.01)	150 (46.01)
Age, y	74.00 (64.00–81.00)	72.50 (63.00–80.00)	0.265
Body mass index, kg/m2	24.28±3.19	24.02±2.95	0.281
Hypertension, n (%)	203 (62.27)	203 (62.27)	1.000
Diabetes, n (%)	85 (26.07)	85 (26.07)	1.000
Coronary artery disease, n (%)	97 (29.75)	97 (29.75)	1.000
Current smoker, n(%)	107 (32.82)	94 (28.83)	0.270
Current drinker, n (%)	87 (26.69)	71 (21.78)	0.144
Triglycerides, mmol/L	1.25 (0.92–1.75)	1.28 (0.90–1.81)	0.848
Total cholesterol, mmol/L	3.88 (3.27–4.73)	3.83 (3.04–4.46)	0.075
High‐density lipoprotein cholesterol, mg/dL	1.13 (0.94–1.39)	1.19 (0.96–1.41)	0.139
Low‐density lipoprotein cholesterol, mg/dL	2.19 (1.69–2.72)	2.20 (1.72–2.76)	0.905
Estimated glomerular filtration rate, mL/min per 1.73 m2	86.51 (67.65–96.82)	90.77 (77.88–96.91)	<0.001
EAT volume, cm3	119.13 (87.30–151.68)	93.00 (67.87–116.71)	<0.001
EAT density, Hounsfield units	−92.02±4.00	−90.68±3.73	<0.001

Values are expressed as mean±SD, as % (n), or as median (interquartile range). EAT indicates epicardial adipose tissue.

Within the bradyarrhythmia group, there were 98 patients with SND and 228 patients with AVB. The baseline characteristics of these 2 subgroups and their respective matched control groups are shown in Table 2. In the AVB subgroup, there were 58 participants with first‐degree AVB, 79 with second‐degree AVB, and 91 with third‐degree AVB. Detailed baseline characteristics of each AVB subgroup and matched controls are further detailed in Table S1.

Table 2.

Baseline Characteristics of Patients With SND and AVB and Matched Control Groups

SND				AVB
	Cases (n=98)	Controls (n=98)	P value	Cases (n=228)	Controls (n=228)	P value
Sex, n (%)			1.000			1.000
Male	41 (41.84)	41 (41.84)		135 (59.21)	135 (59.21)
Female	57 (58.16)	57 (58.16)		93 (40.79)	93 (40.79)
Age, y	70.35±10.66	69.58±10.77	0.618	75.00 (65.50–82.00)	74.00 (64.50–81.00)	0.282
Body mass index, kg/m2	23.82±3.20	23.70±2.99	0.791	24.48±3.17	24.16±2.92	0.262
Hypertension, n (%)	53 (54.08)	53 (54.08)	1.000	150 (65.79)	150 (65.79)	1.000
Diabetes, n (%)	20 (20.41)	20 (20.41)	1.000	65 (28.51)	65 (28.51)	1.000
Coronary artery disease, n (%)	25 (25.51)	25 (25.51)	1.000	72 (31.58)	72 (31.58)	1.000
Current smoker, n (%)	34 (34.69)	31 (31.63)	0.649	73 (32.02)	63 (27.63)	0.306
Current drinker, n (%)	26 (26.53)	21 (21.43)	0.403	61 (26.75)	50 (21.93)	0.230
Triglycerides, mmol/L	1.20 (0.85–1.78)	1.32 (0.94–1.77)	0.940	1.30 (0.94–1.73)	1.26 (0.89–1.81)	0.753
Total cholesterol, mmol/L	4.04±0.97	3.87±1.03	0.240	3.83 (3.27–4.73)	3.81 (3.03–4.58)	0.105
High‐density lipoprotein cholesterol, mg/dL	1.23 (1.01–1.45)	1.19 (0.97–1.46)	0.434	1.10 (0.93–1.35)	1.19 (0.95–1.41)	0.010
Low‐density lipoprotein cholesterol, mg/dL	2.23±0.68	2.21±0.67	0.809	2.19 (1.69–2.72)	2.26 (1.68–2.78)	0.765
Estimated glomerular filtration rate, mL/min/1.73 m2	89.33 (74.46–99.30)	92.16 (81.20–99.25)	0.073	84.94 (64.17–95.73)	89.96 (76.74–95.93)	0.001
EAT volume, cm3	116.98 (81.49–145.66)	91.48 (67.47–108.84)	<0.001	120.66 (89.57–154.39)	93.54 (68.90–119.93)	<0.001
EAT density, Hounsfield units	−92.96±3.91	−90.49±3.42	<0.001	−91.62±3.97	−90.76±3.87	0.017

Values are expressed as mean±SD, as % (n), or as median (interquartile range). AVB indicates atrioventricular block; EAT, epicardial adipose tissue; and SND, sinus node dysfunction.

Association of EAT and Bradyarrhythmias

Excellent reproducibility was observed in EAT measurements, showing intraclass correlation coefficients of 0.92 (95% CI, 0.88–0.95) for volume and 0.90 (95% CI, 0.84–0.93) for density. Patients with bradyarrhythmias exhibited a significantly larger volume of EAT compared with their matched control subjects (119.13 [IQR, 87.30–151.68] cm³ versus 93.00 [IQR, 67.87–116.71] cm³, P<0.001). Furthermore, the EAT density in the bradyarrhythmia group was lower compared with the control group (−92.02±4.00 HU versus −90.68±3.73 HU, P<0.001). Similar trends were observed in both subgroups with SND and AVB, with the case groups generally exhibiting larger EAT volumes and lower EAT densities compared with control populations (Figure 1B, 1C, 1E, 1F).

Figure 1. Comparison of EAT volume and density between the group with bradyarrhythmias and the matched control group.

A through C, respectively, display the comparisons of EAT volume for the group with overall bradyarrhythmias, group with SND, and group with AVB compared with their respective matched controls. D through F present the analysis of EAT density between each case group and its corresponding control in these 3 categories. Boxes represent interquartile range, horizontal lines indicate medians. Patients with bradyarrhythmias exhibited significantly larger EAT volumes and lower densities compared with matched controls (all P<0.05). This pattern was consistently observed in both subgroups with SND and AVB. All P values were calculated using Wilcoxon signed‐rank tests for paired comparisons. AVB indicates atrioventricular block; EAT, epicardial adipose tissue; HU, Hounsfield unit; and SND, sinus node dysfunction.

Association Between EAT and Different Degrees of AVB

After adjusting for covariates including age, sex, BMI, hypertension, diabetes, and CAD, a marked disparity in EAT volume was observed among 3 subgroups with AVB (P=0.001, Figure 2A), whereas the density of EAT showed no significant variation (P=0.307, Figure 2B). Specifically, the comparison of EAT volumes between groups with first‐degree and second‐degree AVB revealed no significant differences (P=0.542). However, the analysis did highlight significant differences in EAT volume between the groups with first‐degree and third‐degree AVB (P=0.001), as well as between the groups with second‐degree and third‐degree AVB (P=0.002). Figure 2A presents the adjusted mean EAT volumes for these subgroups, with the first‐degree subgroup measuring 111.52±5.23 cm³ (adjusted mean±SE), the second‐degree subgroup 115.74±4.52 cm³, and the third‐degree subgroup 135.24±4.22 cm³.

Figure 2. Comparison of EAT among the 3 subgroups of AVB.

Adjusted mean EAT volumes and densities (presented as mean±SE) among populations with first‐ to third‐degree AVB. Covariance analysis adjusts for age, gender, body mass index, hypertension, diabetes, and coronary artery disease. After multivariable adjustment, EAT volume significantly differed among subgroups with AVB (A, P=0.001), with patients with third‐degree AVB showing the highest volumes (135.24±4.22 cm³) compared with groups with first‐degree (111.52±5.23 cm³, P=0.001) and second‐degree (115.74±4.52 cm³, P=0.002) AVB. No significant difference was observed between groups with first‐ and second‐degree AVB (P=0.542). EAT density did not vary significantly across subgroups (B, P=0.307). AVB indicates atrioventricular block; EAT, epicardial adipose tissue; and HU, Hounsfield unit.

EAT Volume as a Significant Influencing Factor of Bradyarrhythmias

Importantly, our regression models demonstrated robust methodological integrity, with no evidence of multicollinearity (maximum variance inflation factors=2.24) and excellent goodness of fit (Hosmer–Lemeshow P=0.787). In the univariable conditional logistic regression analysis, it was observed that both EAT volume and EAT density demonstrated significant correlations with bradyarrhythmia when analyzed independently (Table 3). However, only EAT volume remained significant when both variables were included in a multivariable conditional logistic regression model (odds ratio [OR], 1.02 [95% CI, 1.01–1.03], P<0.001). This suggests that EAT volume may act as a more prominent influencing factor for bradyarrhythmia, whereas the effect of EAT density appeared to be negligible when adjusted for volume (P=0.060, Table 4). To address the clinical interpretability of EAT volume effects, we implemented a standardized unit approach. Each SD increase (43.34 cm³) in EAT volume conferred a 2.52‐fold higher arrhythmia risk (95% CI, 1.89–3.35, Table S2).

Table 3.

Univariable Conditional Logistic Regression Analysis of Factors Associated With Bradyarrhythmias

Variables	OR (95% CI)	P value
EAT volume	1.02 (1.02–1.03)	<0.001
EAT density	0.91 (0.87–0.95)	<0.001
Triglycerides	1.07 (0.85–1.33)	0.573
Total cholesterol	1.17 (1.00–1.37)	0.054
High‐density lipoprotein cholesterol	0.67 (0.39–1.14)	0.140
Low‐density lipoprotein cholesterol	0.99 (0.78–1.24)	0.905
Estimated glomerular filtration rate	0.98 (0.97–0.99)	<0.001
Current smoker	1.27 (0.89–1.81)	0.181
Current drinker	1.29 (0.89–1.88)	0.183

EAT indicates epicardial adipose tissue; and OR, odds ratio.

Table 4.

Multivariable Conditional Logistic Regression Analysis of Factors Associated With Bradyarrhythmias

Variables	OR (95% CI)	P value
EAT volume	1.02 (1.01–1.03)	<0.001
EAT density	0.95 (0.90–1.00)	0.060
Triglycerides	0.83 (0.62–1.10)	0.191
Total cholesterol	1.47 (1.08–1.98)	0.013
High‐density lipoprotein cholesterol	0.63 (0.31–1.27)	0.195
Low‐density lipoprotein cholesterol	0.63 (0.42–0.95)	0.029
Estimated glomerular filtration rate	0.98 (0.97–0.99)	0.001
Current smoker	1.33 (0.87–2.05)	0.193
Current drinker	1.06 (0.68–1.66)	0.803

EAT indicates epicardial adipose tissue; and OR, odds ratio.

Association Between PR Interval and EAT Volume in First‐Degree AVB

Additionally, to further explore the relationships between EAT volume and density with PR interval, we conducted a correlation analysis in the population with first‐degree AVB. The analysis revealed a statistically significant correlation between EAT volume and the PR interval (Spearman’s correlation coefficient: 0.328, P=0.012, Figure 3A). In contrast, no significant correlation was observed between EAT density and the PR interval (Spearman’s correlation coefficient: −0.246, P=0.063, Figure 3B).

Figure 3. Correlation between EAT volume and PR interval in groups with first‐degree AVB.

In patients with first‐degree AVB (cases only, n=58), EAT volume showed a significant positive correlation with PR interval duration (Spearman’s rho=0.328, P=0.012, A), whereas no significant association was observed between EAT density and PR interval (rho=−0.246, P=0.063, B). AVB indicates atrioventricular block; EAT, epicardial adipose tissue; and HU, Hounsfield unit.

DISCUSSION

This study primarily examined the relationship between EAT and bradyarrhythmias, yielding several significant findings. First, compared with matched control groups, patients with bradyarrhythmias exhibited greater volumes of EAT and lower densities. Second, after adjusting for covariates, EAT volume, but not density, was identified as a significant influencing factor for bradyarrhythmias. Third, in subgroup analysis, we observed significant differences in EAT volumes across 3 subgroups with AVB, with a gradual increase in EAT volume from first‐ to third‐degree AVB. Last, we discovered a significant positive correlation between EAT volume and PR interval in the population with first‐degree AVB, suggesting that an increase in EAT volume is likely to be associated with prolongation of PR interval.

EAT and Bradyarrhythmias

To our knowledge, our research is the first study to explore the relationship between EAT and bradyarrhythmias. A previous study by Liu et al ¹⁵ has revealed an increased risk of cardiac conduction blocks associated with obesity, particularly with first‐degree AVB, high‐grade AVB, and left anterior fascicular block. Notably, visceral adiposity demonstrates stronger associations with metabolic derangements than conventional obesity metrics like BMI. ¹⁶ As a specialized visceral fat depot, EAT exhibits distinct arrhythmogenic potential through both anatomical and paracrine mechanisms, ¹⁷ yet its role in bradyarrhythmia pathogenesis remained unexplored before this study. Our investigation has focused on the characteristics of EAT in patients with bradyarrhythmias. Our findings demonstrate these key phenomena: bradyarrhythmia patients exhibit significantly elevated EAT volume, this volume progressively increases with AVB severity and correlates with PR prolongation. These observations align with some interconnected pathophysiological pathways. First, EAT and myocardium do not possess an intervening muscle fascia, thus allowing them to share the same microcirculation and engage in mutual crosstalk. ⁸ EAT can anatomically isolate cardiomyocytes, acting as a barrier to impede electrical signal conduction and consequently, foster the onset of arrhythmias. ¹² Fat infiltration around the atrioventricular node could directly impede electrotonic coupling, mechanistically linking our observed dose‐dependent effect to conduction delay. Increased EAT volume, therefore, might promote electrical disturbances, leading to arrhythmogenesis. Second, the lower EAT density in cases (−92.02 HU versus −90.68 HU, P<0.001) suggests inflammatory activation, ¹⁸ which promotes fibrosis via paracrine release of proinflammatory and profibrotic mediators. ¹⁹ , ²⁰ These mediators disrupt intercellular gap junctions, ²¹ impairing cardiomyocyte electrical coupling. Critically, such fibrotic changes—particularly degenerative myocardial fibrosis ²² and interstitial fibrosis—directly affect impulse formation at the sinoatrial node and conduction through the atrioventricular system, ²³ , ²⁴ thereby exacerbating bradyarrhythmias like SND or AVB. The standardized analysis further demonstrated that each 1 cm³ increment in EAT volume conferred a 2% increased risk. Although this per‐unit effect appears modest, the cumulative risk becomes clinically meaningful across typical EAT ranges. This dose–response relationship underscores EAT volume as a significant influencing factor for bradyarrhythmias. Notably, our retrospective design precluded precise localization of EAT relative to conduction pathways, as non‐ECG‐gated CT limits anatomical resolution. Future studies using ECG‐gated imaging could clarify whether localized perinodal EAT drives this association.

It warrants attention that after adjusting for relevant covariates, we found that EAT volume, rather than density, acts as a significant influencing factor for bradyarrhythmias, even though both of them were significant when analyzed separately. Our research indeed detected a correlation between EAT volume and density (Spearman’s correlation coefficient: −0.348, P<0.001), although no strong collinearity was evident in multiple collinearity analysis. These findings may indicate that the additional contribution of EAT density is diminished after controlling for EAT volume. We speculate that this divergence may be related to the following interconnected factors: First, the density of EAT, quantified through CT attenuation values, partially reflects inflammatory activity, ²⁵ with lower HU values corresponding to metabolically dysfunctional states linked to systemic low‐grade inflammation. ²⁶ Previous studies have demonstrated that EAT density measured by CT attenuation is not only relevant to the adipose tissue itself but is also significantly influenced by the density of surrounding structures. Additionally, factors such as the volume of EAT and the effects of image interpolation can also affect CT attenuation outcomes. ²⁷ We hypothesize that a larger EAT volume may more directly reflect the accumulation of EAT and its direct impact on cardiac electrophysiological functions, possibly through the release of inflammatory factors or by affecting the electrical signal conduction in surrounding cardiac tissues. Conversely, although EAT density can indeed reflect the inflammatory state to a certain degree, its effect may become less significant after considering the volume, given that it is influenced by the volume of EAT itself. Second, methodological constraints reduce detectability of density’s effect. The results might be influenced by the moderate sample size and its limited representativeness. Third, the observational design inherently limits our ability to establish temporal precedence regarding whether density alterations precede or result from conduction disease, while also precluding full adjustment for unmeasured local mediators.

EAT and AVB

In the subgroup analysis, we observed significant differences in the volume of EAT among populations with different types of AVB. Previous research has identified significant associations between AVB and several variables including gender, age, BMI, as well as comorbidities such as hypertension, diabetes, and CAD. ⁷ , ²⁸ , ²⁹ However, few studies have investigated the risk factors associated with the severity of AVB. After adjusting for the relevant covariates above, we discovered a progressive increase in EAT volume from first‐degree to third‐degree AVB. These findings suggest that the volume of EAT is not only associated with the occurrence of AVB but may also influence the severity of AVB. Additionally, research by Chi et al ³⁰ found a linear relationship between CT quantified pericardial fat and thoracic periaortic adipose tissue accumulation with prolonged PR intervals and longer QRS durations. Similarly, studies by Jhou and Hung have confirmed a significant correlation between the accumulation of EAT and prolonged PR intervals. ³¹ , ³² Typically, in the cohort with first‐degree AVB, we observed a correlation between EAT volume and PR interval, corroborating previous research findings, although this association was relatively weak. This may be attributed to the limited sample size and representativeness of our study population.

Regarding these findings, we have speculated on the potential underlying mechanisms. First, an increase in the volume of EAT may exacerbate the inflammatory response and fibrosis in the adjacent myocardial tissues, further affecting the conduction of cardiac electrical signals. Second, as EAT expands, it may act as an anatomical barrier that may further disrupt the electrical conduction in the atrioventricular node and its vicinity. Moreover, the accumulation of EAT might also promote fibrosis in nearby myocardial tissues, thereby disrupting normal electrical pathways and leading to a higher degree of AVB. Further studies are needed to explore the impact of EAT on the severity of AVB and the potential mechanisms involved.

Study Limitations

First, as an observational, retrospective study, our findings can demonstrate only association rather than causation between EAT and bradyarrhythmias. The temporal relationship between EAT accumulation and conduction abnormalities remains unclear. Second, although we carefully matched cases and controls, residual confounding from unmeasured variables may persist. Third, the single‐center design and modest sample size may limit the generalizability of our findings. Larger, multicenter prospective studies would help validate these observations. Fourth, our study lacks mechanistic data to explain how EAT might influence conduction system function. Although we propose several plausible pathways based on existing literature, direct histological or electrophysiological evidence is needed.

CONCLUSIONS

This matched case–control study demonstrates a significant association between EAT and bradyarrhythmias, with EAT volume emerging as a stronger influencing factor than density Within the population with AVB, an increase in EAT volume shows a significant correlation with the severity of AVB. Further research is imperative to unravel the pathophysiological mechanisms through which EAT contributes to these arrhythmias. Clarifying its role may advance preventive and therapeutic strategies, ultimately improving clinical outcomes in cardiovascular care.

Sources of Funding

None.

Disclosures

None.

Supporting information

Tables S1–S2