Abstract
Background
Chronic kidney disease (CKD) is a key component of the cardiovascular–kidney–metabolic framework. However, in patients undergoing atrial fibrillation ablation, the distribution of CKD across Kidney Disease: Improving Global Outcomes risk categories and its prognostic value remain unclear.
Methods
Patients undergoing atrial fibrillation ablation with preprocedural urinalysis were studied. Patients were classified into Kidney Disease: Improving Global Outcomes CKD risk categories (low, moderate, high, very high) using the estimated glomerular filtration rate and predicted urine albumin–creatinine ratio derived from dipstick proteinuria. We compared postdischarge heart failure hospitalization incidence across risk categories and assessed cardiovascular–kidney–metabolic multimorbidity (all 3 components present) and the use of kidney‐protective therapies.
Results
Among 942 patients, 505, 277, 118, and 42 were classified as low, moderate, high, and very high risk, respectively; 46.3% met the Kidney Disease: Improving Global Outcomes definition of CKD (≥moderate). Over a median follow‐up of 5.1 years (interquartile range, 3.1–7.1), 72 patients (7.6%) were hospitalized for heart failure. Heart failure hospitalization incidence increased across CKD risk categories (p<0.001), with a marked rise from moderate to high risk (1.90 [95% CI, 1.26–2.74] versus 7.74 [95% CI, 5.66–10.30] events per 100 person‐years). Cardiovascular–kidney–metabolic multimorbidity occurred in 42% of the high‐ and 67% of the very‐high‐risk patients. The use of sodium–glucose cotransporter 2 inhibitors was 19% in the moderate‐risk group and 27% in the high‐risk group, compared with 62% in the very high‐risk group.
Conclusions
Kidney Disease: Improving Global Outcomes CKD risk categories stratified long‐term heart failure hospitalization risk after atrial fibrillation ablation, with a sharp increase at high risk.
Keywords: atrial fibrillation, cardiovascular‐kidney‐metabolic syndrome, catheter ablation, chronic kidney disease
Subject Categories: Atrial Fibrillation, Heart Failure, Metabolic Syndrome

Nonstandard Abbreviations and Acronyms
- CKM
cardiovascular–kidney–metabolic
- KDIGO
Kidney Disease: Improving Global Outcomes
- RAS
renin–angiotensin system
- SGLT2i
sodium–glucose cotransporter 2 inhibitor
- UACR
urine albumin‐creatinine ratio
Clinical Perspective.
What Is New?
In a real‐world cohort undergoing index atrial fibrillation ablation, 46.3% met Kidney Disease: Improving Global Outcomes chronic kidney disease risk moderate or higher, and 10.6% had albuminuria A2 or higher based on dipstick‐derived predicted urine albumin‐creatinine ratio.
The Kidney Disease: Improving Global Outcomes chronic kidney disease risk categories stratified long‐term heart failure hospitalization after ablation, with a sharp increase in the high‐risk category.
What Are the Clinical Implications?
Kidney Disease: Improving Global Outcomes high risk identified a subgroup with substantially higher heart failure hospitalization risk after atrial fibrillation ablation; at the time of index ablation, the implementation of chronic kidney disease‐ and heart failure‐directed therapies differed across Kidney Disease: Improving Global Outcomes risk categories, supporting the value of reviewing comorbidity management in patients in the high‐risk stratum.
Recently, the American Heart Association defined cardiovascular‐kidney‐metabolic (CKM) syndrome to highlight the interplay among metabolic risk factors, kidney disease, and cardiovascular complications. 1 In this framework, chronic kidney disease (CKD) risk categorization, staged according to the estimated glomerular filtration rate (eGFR) and urine albumin‐creatinine ratio (UACR) recommended by the Kidney Disease: Improving Global Outcomes (KDIGO) guidelines, represents a key pathway linking metabolic dysregulation to cardiovascular disease 1 , 2
In the American Heart Association construct for CKM staging, individuals with atrial fibrillation (AF) are generally classified as having stage 4 CKM syndrome, reflecting the presence of established heart disease. 1 Nevertheless, even within the population with AF, the coexistence of atherosclerotic cardiovascular disease, metabolic disease, and CKD (CKM multimorbidity) has been associated with a stepwise increase in the risk of adverse clinical outcomes 3 , 4 , 5 Accordingly, among individuals with AF, systematic assessment of CKM comorbidities remains clinically important to refine risk stratification and optimize care. CKD is a strong predictor of outcomes and has become an important treatment target due to recent therapeutic advances 2 , 4
In patients undergoing catheter ablation for AF, long‐term heart failure (HF) events, rather than thromboembolic events, constitute the predominant clinical concern. 6 , 7 , 8 , 9 , 10 However, the distribution of KDIGO CKD risk categories based on eGFR and UACR, and their prognostic value for long‐term HF events after ablation, have not been well evaluated. Furthermore, the burden of overlapping CKM comorbidities and contemporary use of guideline‐directed CKD therapies in this setting have not been fully characterized.
Therefore, we aimed to (1) describe CKD risk categories in patients undergoing AF ablation and assess their ability to predict the risk of HF hospitalization during long‐term follow‐up, (2) quantify the overlap of CKM components across risk categories, and (3) examine the implementation of guideline‐directed CKD therapies.
METHODS
Study Setting and Population
This single‐center, retrospective cohort study included consecutive patients who underwent index catheter ablation for AF between October 2011 and March 2024. Eligible patients were required to have undergone urinalysis within 1 week before the procedure. The Institutional Review Board approved the study protocol and waived the requirement for written informed consent using an opt‐out approach. The study was conducted in accordance with the Declaration of Helsinki and reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology guidelines. 11 The corresponding author (H.I.) had full access to all data in the study and takes responsibility for its integrity and data analysis. Deidentified data and analytical codes are available from the corresponding author upon request.
Study End Points
The primary end points were (1) the distribution of KDIGO CKD risk categories among patients undergoing AF ablation and (2) the incidence of hospitalization for HF after discharge across these categories.
Secondary end points included (1) arrhythmia recurrence after ablation across CKD risk categories, (2) prevalence of CKM‐related comorbidities, and (3) implementation of guideline‐directed CKD interventions.
KDIGO CKD Risk Categorization
Patients were stratified into CKD risk categories according to the KDIGO framework, based on eGFR and UACR. 2 The classification grids and risk definitions are provided in Table S1. The eGFR was calculated using the CKD Epidemiology Collaboration creatinine equation. As quantitative albuminuria measurements were not available for all patients, we estimated the UACR from urine dipstick testing using a previously validated prediction model. 12 This model incorporates semiquantitative dipstick proteinuria and clinical covariates, including sex, hypertension, and diabetes, to generate a predicted UACR value. Patients were then classified into 4 KDIGO risk categories (low, moderate, high, and very high) based on the eGFR and predicted UACR.
Clinical Data Collection
Urinalysis was performed within 1 week before the ablation procedure, typically at hospital admission. Urine pH was recorded in 0.5‐unit increments (range, 5.0–9.0), and urine specific gravity was recorded in 0.005‐unit increments (range, 1.005–1.030). 13 Values outside the prespecified ranges were excluded. Baseline clinical variables were obtained from the medical records at admission.
Definitions
HF hospitalization was defined as any unplanned admission requiring treatment for HF after discharge following the index ablation procedure, as previously described. 8 , 9 , 10
For the CKM components, the definitions of the CKM domains were adapted from previous AF registry–based analyses. 4 , 5 The kidney component was defined as KDIGO moderate, high, or very high risk. Cardiovascular comorbidities were defined as any of the following: history of atherosclerotic disease (including coronary artery disease), ischemic stroke, transient ischemic attack, or congestive HF. 4 Metabolic comorbidities were defined as diabetes, dyslipidemia, or body mass index ≥25 kg/m2. 4
The implementation of guideline‐directed core CKD interventions was assessed at the time of index ablation based on the 4 core CKD intervention domains recommended in the current guidelines: (1) use of SGLT2 (sodium–glucose cotransporter 2) inhibitors (SGLT2i); (2) use of renin–angiotensin system inhibitors, (3) use of statins, and (4) achievement of the blood pressure target (<130/80 mm Hg) 2 , 14 Blood pressure target achievement was evaluated using measurements obtained at outpatient visits within 1 month before admission or on the day of admission. Although the guidelines specifically recommend angiotensin‐converting enzyme inhibitors or angiotensin II receptor blockers, 2 we also included angiotensin receptor–neprilysin inhibitors in the renin–angiotensin system inhibitor category because they contain an ARB component. Because SGLT2i were approved for CKD treatment in Japan in late August 2021, analyses of these 4 interventions were restricted to patients treated from September 2021 onward. Because the number of patients with clear indications for the recommended CKD therapies was expected to be limited in the low‐risk group, implementation rates were nonetheless evaluated in the same manner across all KDIGO risk categories to allow contextual comparison between groups.
Ablation Strategy
For the index procedure, pulmonary vein isolation was performed using a 3‐dimensional electroanatomical mapping system, irrespective of the AF type (paroxysmal or persistent). Ablation was conducted using radiofrequency energy or balloon‐based catheters (predominantly cryoballoons). Targeted ablation of the atrial flutter was performed when it was clinically documented or induced during the procedure. Empirical atrial substrate modification was not performed unless otherwise indicated.
Postprocedural Follow‐Up
The patients were generally discharged within a few days after the procedure. At 3 months, all patients underwent standardized follow‐up including 24‐hour Holter monitoring and echocardiography. Thereafter, follow‐up was scheduled every 6–12 months for up to 5 years or until the end of follow‐up. Additional Holter monitoring was actively performed when patients reported palpitations or other symptoms suggestive of arrhythmia recurrence.
Arrhythmia recurrence was defined as any atrial tachyarrhythmia lasting >30 seconds that occurred beyond a 3‐month blanking period. In cases of recurrence, repeat ablation was offered unless the patient declined. Clinical events and outcome data were ascertained at scheduled visits and, when needed, by contacting primary care physicians.
Statistical Analysis
Continuous variables were reported as mean±SD for normally distributed data and as median (interquartile range) for nonnormally distributed data. Categorical variables were presented as counts and percentages.
To account for the competing risk of all‐cause death, cumulative incidence functions for hospitalization due to HF were estimated with 95% CIs and compared using Gray's test. Incidence rates were calculated as events per 100 person‐years with Poisson‐based 95% CIs.
The 1‐year arrhythmia recurrence rate was compared across the groups using the chi‐square test. To assess whether CKD risk category was independently associated with arrhythmia recurrence, we performed a multivariable logistic regression analysis for 1‐year arrhythmia recurrence including clinically relevant covariates, and the results were expressed as adjusted odds ratios (ORs) with 95% CIs. Covariates included age, sex, persistent AF, body mass index, AF duration, cardiac implantable electronic device status, left ventricular ejection fraction, and left atrial diameter.
The association between CKD risk categories and HF hospitalization was examined using a multivariable Fine–Gray subdistribution hazards model, with the low‐risk group as the reference category. Covariates were prespecified based on their clinical relevance to HF and included age, sex, persistent AF, body mass index, prior HF, diabetes, B‐type natriuretic peptide, and left ventricular ejection fraction. The results were reported as adjusted subdistribution hazard ratios (HRs) with 95% CIs.
As missingness was negligible (<0.1%), no imputations were performed. Analyses were conducted using R software (version 4.3.2; R Foundation for Statistical Computing, Vienna, Austria). Two‐sided P values <0.05 were considered statistically significant.
RESULTS
Study Population
A flow chart of the study is shown in Figure S1. A total of 942 patients were included in this study. Of these, 505 were classified as low risk, 277 as moderate risk, 118 as high risk, and 42 as very high risk. The patient characteristics are summarized in Table 1. Across the increasing CKD risk categories, the patients were older, had persistent AF more frequently, and had higher CHA2DS2‐VASc scores and B‐type natriuretic peptide levels.
Table 1.
Patient Characteristics
| Characteristics | Low risk (n = 505) | Moderate risk (n = 277) | High risk (n = 118) | Very high risk (n = 42) |
|---|---|---|---|---|
| Age, y, mean±SD | 63±10 | 70±8 | 70±6 | 72±8 |
| Female sex, n (%) | 152 (30.1) | 97 (35.0) | 39 (33) | 12 (28) |
| Persistent AF, n (%) | 170 (33.7) | 118 (42.6) | 59 (50) | 24 (57) |
| AF duration (mo), median (IQR) | 12 (4–42) | 10 (4–42) | 11 (3–57) | 11 (3–45) |
| Cardiac implantable electronic devices, n (%) | 15 (2.9) | 22 (7.9) | 18 (15) | 6 (14) |
| Body mass index, kg/m2, mean±SD | 23±3 | 23±3 | 24±3 | 24±3 |
| Systolic blood pressure, mm Hg, mean±SD | 126±16 | 126±17 | 122±19 | 127±17 |
| Diastolic blood pressure, mm Hg, mean±SD | 75±12 | 75±12 | 73±14 | 75±14 |
| CHA2DS2‐VASc score, mean±SD | 1.9±1.4 | 2.6±1.4 | 3.1±1.4 | 3.9±1.5 |
| B‐type natriuretic peptide level, pg/mL, median (IQR) | 63 (24–126) | 102 (41–185) | 145 (59–265) | 167 (99–390) |
| Left anterior descending coronary artery, mm, mean±SD | 40±7 | 41±6 | 44±7 | 44±6 |
| Left ventricular ejection fraction, %, mean±SD | 63±9 | 61±10 | 58±13 | 54±18 |
| Renal‐related parameters | ||||
| Estimated glomerular filtration rate, mL/min/1.73 m2, mean±SD | 73±12 | 55±8 | 43±10 | 29±12 |
| Predicted urine albumin‐creatinine ratio, mg/g, median (IQR) | 10 (8–11) | 11 (10–22) | 15 (10–41) | 57 (12–298) |
| Proteinuria ≥1+, n (%) | 0 (0.0) | 36 (13.0) | 32 (27) | 28 (66) |
| Hematuria ≥1+, n (%) | 73 (14.5) | 59 (21.3) | 26 (22) | 17 (40) |
| Any structural heart disease, n (%) | 69 (13.7) | 55 (19.9) | 38 (32) | 18 (42) |
| Coronary artery disease, n (%) | 29 (5.7) | 19 (6.9) | 13 (11) | 12 (28) |
| HCM, n (%) | 13 (2.6) | 13 (4.7) | 12 (10) | 2 (4) |
| Dilated cardiomyopathy /dilated phase of HCM, n (%) | 5 (1.0) | 2 (0.7) | 1 (1) | 3 (7) |
| Valvular heart disease, n (%) | 12 (2.4) | 13 (4.7) | 4 (3) | 0 |
| Congenital heart disease, n (%) | 2 (0.4) | 2 (0.7) | 1 (1) | 0 |
| Amyloidosis, n (%) | 2 (0.4) | 2 (0.7) | 2 (2) | 0 |
| Other cardiomyopathies, n (%) | 6 (1.2) | 4 (1.4) | 5 (4) | 1 (2) |
| Comorbidities | ||||
| Hypertension, n (%) | 257 (50.9) | 187 (67.5) | 79 (66) | 30 (71) |
| Diabetes, n (%) | 78 (15.4) | 39 (14.1) | 32 (27) | 19 (45) |
| Dyslipidemia, n (%) | 293 (58.0) | 177 (63.9) | 88 (74) | 35 (83) |
| Congestive heart failure, n (%) | 61 (12.1) | 54 (19.5) | 41 (34) | 21 (50) |
| History of ischemic stroke/transient ischemic attack, n (%) | 45 (8.9) | 28 (10.1) | 18 (15) | 9 (21) |
| Procedure‐related parameters | ||||
| Radiofrequency PVI, n (%) | 358 (70.9) | 214 (77.3) | 102 (86) | 37 (88) |
| Balloon‐based PVI, n (%) | 147 (29.1) | 63 (22.7) | 16 (14) | 5 (12) |
| Therapeutic agents | ||||
| direct oral anticoagulants, n (%) | 433 (85.7) | 229 (82.7) | 94 (79) | 27 (64) |
| Angiotensin‐converting enzyme inhibitor/angiotensin receptor blocker, n (%) | 187 (37.0) | 141 (50.9) | 62 (52) | 27 (64) |
| Angiotensin receptor–neprilysin inhibitor, n (%) | 5 (1.0) | 3 (1.1) | 4 (3) | 3 (7) |
| Beta blocker, n (%) | 269 (53.3) | 166 (59.9) | 78 (66) | 33 (78) |
| Mmineralocorticoid receptor antagonist, n (%) | 38 (7.5) | 32 (11.6) | 23 (19) | 10 (23) |
| Statin, n (%) | 124 (24.6) | 90 (32.5) | 48 (40) | 24 (57) |
| Sodium‐glucose cotransporter 2 inhibitor, n (%) | 8 (1.6) | 14 (5.1) | 10 (8) | 8 (19) |
| Glucagon‐like peptide‐1 receptor agonist, n (%) | 1 (0.2) | 0 (0.0) | 1 (1) | 1 (2) |
| Other oral hypoglycemic agents, n (%) | 43 (8.5) | 23 (8.3) | 21 (17) | 11 (26) |
| Insulin, n (%) | 6 (1.2) | 3 (1.1) | 5 (4) | 3 (7) |
| Type I/III antiarrhythmic drugs, n (%) | 174 (34.5) | 110 (39.7) | 63 (53) | 13 (30) |
Numerical data are expressed as mean±SD or median (IQR; first quartile, third quartile), and categorical data are expressed as number (percentage).
AF indicates atrial fibrillation; ARNI, HCM, hypertrophic cardiomyopathy; IQR, interquartile range; and PVI, pulmonary vein isolation.
CKD Distribution
Figure S2 shows the distribution of the renal function components. For eGFR, 395 patients (41.9%) had values in category G3a or worse (<60 mL/min/1.73 m2; Figure S2A). Mapping of urine dipstick protein to predicted UACR categories is shown in Figure S2B: trace proteinuria (+/−) was predominantly mapped to A1, whereas most 1+ or 2+ results were mapped to A2. Overall, 100 patients (10.6%) were classified as having predicted albuminuria of category A2 or higher.
Figure 1 summarizes the KDIGO CKD risk categories. The proportion and number of patients with albuminuria (≥A2) increased with risk severity (Figure 1A); albuminuria was present in 37 patients (13.3%) in the moderate‐risk group and 35 patients (30%) in the high‐risk group. According to eGFR strata (Figure 1B), 12.4% of patients with eGFR category G3a (45–59 mL/min/1.73 m2) were classified as high or very high risk.
Figure 1. Distribution of patients across KDIGO CKD risk categories.

A, Number of patients according to KDIGO CKD risk category. B, Prevalence of KDIGO CKD risk categories stratified according to eGFR category. Bar colors indicate the KDIGO CKD risk categories (green, low; yellow, moderate; orange, high; red, very high). The color gradient within each bar denotes the UACR category (light to dark: A1, A2, and A3). CKD indicates chronic kidney disease; eGFR, estimated glomerular filtration rate; KDIGO, Kidney Disease: Improving Global Outcomes; and UACR, urine albumin‐creatinine ratio.
Arrhythmia Recurrence
Over a median follow‐up of 5.1 years (interquartile range, 3.1–7.1), arrhythmia recurrence occurred in 432 patients (45.8%); 256 patients (27.1%) experienced recurrence within 1 year. The 1‐year recurrence rate increased across the CKD risk categories (low 26.5%, moderate 28.5%, high 35%, very high 41%; P<0.001). However, this association was not significant after multivariable adjustment. In logistic regression analysis using the low‐risk group as the reference, the adjusted ORs for 1‐year arrhythmia recurrence were 0.92 (95% CI, 0.64–1.32) for the moderate‐risk group, 1.08 (95% CI, 0.67–1.75) for the high‐risk group, and 1.52 (95% CI, 0.75–3.09) for the very‐high‐risk group (Table S2).
HF Hospitalization According to CKD Risk Category
During follow‐up, 72 patients (7.6%) experienced HF hospitalization (113 total events), with a median time to first event of 2.5 years (interquartile range, 0.9–5.3). In the competing‐risk setting, 61 patients died, with a median time to death of 4.8 years (interquartile range, 3.1–5.9). Figure 2 shows the cumulative incidence of HF hospitalization according to CKD risk category; the incidence increased with higher risk severity (P<0.001), with similarly high rates in the high‐ and very‐high‐risk groups (Table 2). Incidence rates per 100 person‐years were 0.88 (95% CI, 0.57–1.30) in low risk, 1.90 (1.26–2.74) in moderate risk, 7.74 (5.66–10.30) in high risk, and 7.25 (3.96–12.20) in very high risk (P<0.001). In multivariable Fine–Gray models using the low‐risk group as the reference, the adjusted subdistribution HRs for HF hospitalization were 0.99 (95% CI, 0.51–1.92) for the moderate‐risk group, 2.66 (95% CI, 1.36–5.18) for the high‐risk group, and 1.71 (95% CI, 0.68–4.35) for the very high‐risk group (Table 3).
Figure 2. Cumulative incidence of HF hospitalization according to KDIGO CKD risk category.

Curves are shown as cumulative incidence functions for hospitalization owing to HF with 95% CIs. Colors indicate the KDIGO CKD risk categories (green, low; yellow, moderate; orange, high; red, very high). CKD indicates chronic kidney disease; HF, heart failure; and KDIGO, kidney disease: improving global outcomes.
Table 2.
Comparison of the Incidence of HF Hospitalization
| Low risk (n = 505) | Moderate risk (n = 277) | High risk (n = 118) | Very high risk (n = 42) | P value | |
|---|---|---|---|---|---|
| Cumulative incidence at median follow‐up period (5.1 y), % (95% CI) | 2.69 (1.17–4.21) | 5.07 (2.24–7.91) | 18.30 (10.40–26.20) | 19.40 (7.14–31.70) | <0.001 |
| Incidence rate throughout follow‐up, per 100 person‐y (95% CI) | 0.88 (0.57–1.30) | 1.90 (1.26–2.74) | 7.74 (5.66–10.30) | 7.25 (3.96–12.20) | <0.001 |
Cumulative incidence estimates are functions that account for competing risks of all‐cause death.
HF indicates heart failure.
Table 3.
Multivariate Analysis to Identify Predictors for HF Hospitalization
| Variables | Adjusted subdistribution HR (95% CI) | P value |
|---|---|---|
| Moderate CKD risk category (vs low risk) | 0.99 (0.51–1.92) | 0.970 |
| High CKD risk category (vs low risk) | 2.66 (1.36–5.18) | 0.004 |
| Very high CKD risk category (vs low risk) | 1.71 (0.68–4.35) | 0.260 |
| Age (per 5‐y increase) | 1.56 (1.31–1.86) | <0.001 |
| Female sex (vs male) | 1.98 (1.21–3.26) | 0.006 |
| Persistent AF (vs paroxysmal AF) | 1.48 (0.88–2.51) | 0.140 |
| Body mass index (per 1 kg/m2 increase) | 0.98 (0.92–1.04) | 0.570 |
| Congestive HF (vs no congestive HF) | 2.31 (1.21–4.37) | 0.01 |
| Diabetes (vs no diabetes) | 1.30 (0.77–2.20) | 0.310 |
| B‐type natriuretic peptide (per 10 pg/mL increase) | 1.01 (1.00–1.01) | 0.004 |
| left ventricular ejection fraction (per 5% increase) | 0.88 (0.79–0.98) | 0.031 |
AF indicates atrial fibrillation; CKD, chronic kidney disease; HF, heart failure; and HR, hazard ratio.
Prevalence of CKM Comorbidities
Figure 3 shows the burden and overlap of the CKM components. The metabolic component was the most common, followed by the kidney and cardiovascular components (Figure 3A); most patients had ≥2 components. Considering the CKD risk category (Figure 3B), metabolic comorbidity alone predominated in the low‐ and moderate‐risk groups, whereas CKM multimorbidity (all 3 components) was more frequent in higher‐risk groups (42% in the high‐ risk and 67% in the very‐high‐risk group).
Figure 3. Prevalence and overlap of CKM components.

A, Venn/Euler diagram showing the overlap of metabolic, kidney, and cardiovascular components in the overall cohort. Colors indicate CKM domains as follows: blue, metabolic; green, kidney; and orange, cardiovascular. B, Stacked bar charts showing, within each KDIGO CKD risk category, the distribution of patients according to the presence of additional nonkidney CKM components: none, metabolic component only, cardiovascular component only, or both metabolic and cardiovascular components. The base color corresponds to the KDIGO CKD risk category (low, green; moderate, yellow; high, orange; very high, red). Within each risk category, shading indicates the burden of additional nonkidney CKM components: gray indicates no additional component, a lighter shade indicates metabolic component only, a darker shade indicates cardiovascular component only, and the darkest shade indicates the presence of both metabolic and cardiovascular components. CKD indicates chronic kidney disease; CKM, cardiovascular–kidney–metabolic; CV, cardiovascular; and KDIGO, kidney disease: improving global outcomes.
Implementation of Guideline‐Directed Core CKD Interventions
Figure 4 summarizes the implementation of guideline‐directed core CKD interventions according to risk category in 195 patients treated after September 2021. SGLT2i use was 19% in the moderate‐risk group and 27% in the high‐risk group (Figure 4A). Blood pressure target attainment exceeded 50% in the groups with moderate or higher risk (Figure 4D). The use of renin–angiotensin system inhibitors was 47% in moderate risk and 50% in high risk, compared with 90% in very high risk (Figure 4B).
Figure 4. Implementation of core interventions across KDIGO CKD risk categories.

A, SGLT2i. B, RAS inhibitors. C, Statins. D, Blood pressure target attainment (<130/80 mm Hg). Numbers above the bars indicate the number of patients receiving the intervention (or achieving the target) out of the total number in each KDIGO risk category. Because clear treatment indications were expected to be less frequent in the low‐risk group, this group is shown primarily as a contextual reference for comparison with the other categories. Colors indicate the KDIGO CKD risk categories (green, low; yellow, moderate; orange, high; red, very high). The analyses were restricted to patients treated from September 2021 onward. CKD indicates chronic kidney disease; KDIGO, Kidney Disease: Improving Global Outcomes; RAS, renin–angiotensin system; and SGLT2i, sodium–glucose cotransporter 2 inhibitor.
DISCUSSION
Overview
In the present study, we characterized the distribution of the KDIGO CKD risk categories and examined their clinical utility and the implementation of guideline‐directed core CKD interventions in patients undergoing catheter ablation for AF. First, 46.4% of patients were classified as having at least moderate KDIGO CKD risk, and 10.6% had moderately increased albuminuria (≥A2), indicating that kidney involvement was common in this cohort. Second, higher CKD risk categories were associated with a higher risk of HF hospitalization after discharge, and crude 1‐year arrhythmia recurrence rates were also higher; event rates for both outcomes were substantially higher in the high‐ and very‐high‐risk groups than in the low‐ and moderate‐risk groups. In multivariable analysis, HF hospitalization risk was independently higher in the high‐risk category than in the low‐risk category. By contrast, the association between CKD risk category and 1‐year arrhythmia recurrence was not significant after multivariable logistic regression adjustment. Third, CKM multimorbidity, defined as concurrent cardiovascular, kidney, and metabolic components, was common and concentrated in the high‐ (42%) and very‐high‐risk (67%) categories. Fourth, although the implementation of core CKD interventions was generally high in the very‐high‐risk category, medication use and blood pressure target attainment in the moderate‐ and high‐risk groups were lower than those in the very‐high‐risk group, particularly for SGLT2i.
The Interaction Between AF and CKD
The coexistence of AF and CKD is common, reflecting not only shared risk factors such as hypertension and diabetes but also evidence for a bidirectional association that is partly supported by genetic studies 15 , 16 In contemporary cohorts with AF, renal dysfunction rarely occurs in isolation and is often associated with cardiovascular and metabolic abnormalities 4 , 5 When the kidney domain is present within the CKM framework, the overall risk increases disproportionately; profiles that include kidney dysfunction show higher mortality and major cardiovascular event rates than those with cardiovascular or metabolic components alone. 4 Recent therapeutic advances, particularly in SGLT2i and nonsteroidal mineralocorticoid receptor antagonists, have increased the focus on CKD as a target for risk modification 17 , 18 These considerations highlight the importance of the timely identification of patients with AF who have kidney involvement and the delivery of kidney‐protective interventions at a modifiable stage.
Consistent with this concept, we observed that an increase in the KDIGO CKD risk category was accompanied by a greater burden of CKM multimorbidity and a parallel increase in HF hospitalization. We also observed differences in the implementation of guideline‐directed CKD pharmacotherapy across KDIGO risk categories at the time of index ablation. In particular, the use of SGLT2i and other core CKD therapies was higher in the very‐high‐risk group than in the moderate‐ and high‐risk groups. These findings suggest a stage‐dependent pattern of treatment implementation in this cohort. Given that the moderate‐ and high‐risk‐groups together accounted for 41.9% of the cohort, AF ablation may provide an opportunity to review CKD‐ and HF‐directed management in addition to rhythm and anticoagulation strategies.
Importance of KDIGO CKD Risk Classification
Within the CKM construct, the KDIGO CKD risk classification contributes to staging, with moderate to high risk generally aligned with earlier CKM stages (stage 2) and very high risk aligning with more advanced stages (stage 3). 1 In this context, moderate and high risks may be grouped within a similar CKM stage, representing a potentially modifiable phase characterized by the coexistence of metabolic risk factors and kidney dysfunction.
However, in our cohort of patients undergoing AF ablation, the risk of hospitalization for HF increased abruptly in the high‐risk category (incidence rates: 1.90 versus 7.74 events per 100 person‐years in the moderate‐ and high‐risk groups, respectively). Several factors could explain this pattern. First, the cohort had established cardiovascular disease in the form of AF, implying higher baseline vulnerability than the primary prevention populations in which CKM staging was originally conceptualized. This finding is consistent with the high prevalence of CKM multimorbidity in the high‐risk group (42%). Second, despite this elevated risk, the intensity of guideline‐directed pharmacotherapy in the high‐risk group was broadly similar to that in the moderate‐risk group, particularly for SGLT2i and renin–angiotensin system inhibitors. This suggests that treatment implementation may not have increased in parallel with risk escalation from the moderate‐ to the high‐risk category.
KDIGO high‐risk status has recently been positioned as a category that warrants the timely initiation and intensification of risk‐modifying therapy. 14 In line with this context, our data showed that the high‐risk group carried a markedly elevated HF risk, yet the implementation of key guideline‐directed therapies was not commensurate with this level of risk. This pattern suggests that the periablation period may provide an opportunity to review HF‐ and CKD‐directed management, particularly in the KDIGO high‐risk stratum.
Clinical Implications
Our data support the utility of the KDIGO CKD risk classification in stratifying the risk of postdischarge hospitalization for HF, and crude arrhythmia recurrence rates also differed across categories. Urinalysis is especially informative among patients with eGFR category G3a because a meaningful subset is reclassified to high risk or higher based on the presence of albuminuria (≥A2), whereas eGFR alone would place them in a lower‐risk stratum. In addition, we demonstrated the feasibility of using the predicted UACR derived from dipstick testing. Although quantitative UACR is the standard for CKD risk assignment, its routine use may be limited by cost and logistical barriers 1 , 19 In settings where quantitative UACR is difficult to implement, the dipstick‐based predicted UACR may offer a practical approach to risk stratification.
Our findings also showed differences in the implementation of kidney‐protective therapies across the moderate‐ and high‐risk strata at the time of index ablation. In these groups, the use of kidney‐protective therapies, particularly SGLT2i, was lower than that in the very‐high‐risk group despite higher HF risk than in the low‐risk group. These findings suggest that the periablation period may provide an opportunity to review HF‐ and CKD‐directed management, particularly in the KDIGO high‐risk stratum. Multidisciplinary team‐based CKM care may help support this process when routine implementation is challenging. 20
Finally, CKM multimorbidity was not rare in our ablation cohort; 16.3% of patients met the criteria for all 3 components. This substantial burden underscores the importance of systematically assessing and optimizing the management of CKM‐related comorbidities and risk factors in patients referred for AF ablation. These findings support the concept that AF management should not be limited to the ablation procedure itself but should also include systematic assessment and optimization of CKD/CKM‐related comorbidities, risk factors, and appropriate referrals. This message is consistent with the contemporary guideline recommendations that AF care should prioritize the comprehensive management of concomitant conditions and risk factors rather than focusing on arrhythmia in isolation. 15 , 21
Limitations
This study has several limitations. First, this was a single‐center retrospective analysis of a historical cohort, which restricts the generalizability of the findings. In particular, the study was conducted exclusively in an Asian population, in which body mass index was generally lower than that in Western cohorts with AF, and CKD prevalence was higher. 5 Hence, the absolute distribution of CKD risk categories and CKM multimorbidity patterns may not be generalizable and should be validated in non‐Asian populations. Second, this study was conducted over a long period. Although >70% of the patients were enrolled after 2016 and the distribution of urinalysis findings did not show a clear temporal bias, substantial advances in catheter ablation techniques and HF management over this interval may have influenced procedural success, arrhythmia recurrence, and HF outcomes in ways that could not be fully accounted for. Third, CKD therapeutics have advanced rapidly in recent years, particularly with the introduction of SGLT2i and other kidney‐protective agents. 22 Our analysis of treatment implementation was limited to the early phase of this trend (2021–2024); therefore, clinicians' awareness and adherence to guideline‐directed therapy may already be higher in current practice than in our data set. In addition, emerging agents for diabetic kidney disease and broader populations with CKD have shown promising efficacy, which could further alter risk profiles and management patterns. Therefore, validation of our findings in more contemporary cohorts that reflect these therapeutic advances will be important. Fourth, albuminuria was not measured quantitatively in all patients. Instead, we derived the predicted UACR from urine dipstick proteinuria using a validated conversion approach. However, dipstick‐based classification has limited sensitivity for albuminuria staging, particularly for A2, and the prediction intervals for the converted ACR are wide, potentially limiting the precision of distinguishing A2 from A3 and leading to misclassification of the KDIGO risk categories in some patients. 12 Fifth, treatment indication was not uniform across KDIGO CKD risk categories but varied according to UACR level, CKD risk category, and eGFR. Because of the limited sample size and the cross‐sectional assessment of treatment status at the time of index ablation, we presented simple implementation rates in the current study. Accordingly, these data should not be interpreted as a detailed assessment of treatment appropriateness or underuse. Further studies using indication‐based analyses are warranted. Sixth, although follow‐up data on medication use, blood pressure target achievement, and renal function were available in clinical practice, we did not analyze their longitudinal changes in the present study. Because these postablation trajectories may vary over time according to rhythm status, clinical course, and referral patterns, further investigation is warranted to clarify how these longitudinal changes relate to baseline KDIGO risk classification.
CONCLUSIONS
In patients undergoing catheter ablation for AF, the KDIGO CKD risk classification, supported by simple urinalysis, identified substantial kidney involvement and a sharp increase in HF risk in the high‐risk category. We also observed variable implementation of kidney‐protective therapies across KDIGO risk categories at the time of index ablation.
Sources of Funding
This work was supported by the Daiwa Securities Health Foundation, Research Grant Program (52nd Research Grant, number 20).
Disclosures
Yasuhiro Yoshiga received lecture fees from Medtronic Inc. Masaki Shibuya received grants from Boehringer Ingelheim Japan, AstraZeneca, and Novartis Pharma K.K. The remaining authors have no disclosures to report.
Supporting information
Tables S1–S2
Figures S1–S2
STROBE Checklist
Acknowledgments
We thank Editage (www.editage.jp) for English language editing and acknowledge the assistance of ChatGPT5 (OpenAI) for initial proofreading of this article. We also thank Mai Okano, a medical student, for their assistance with data collection.
Author Contributions: Conceptualization: Hironori Ishiguchi, Yasuhiro Yoshiga. Data curation: Hironori Ishiguchi, Masakazu Fukuda, Masafumi Fujinaka, Shintaro Hashimoto. Formal analysis: Hironori Ishiguchi. Investigation: Hironori Ishiguchi, Takuya Omuro, Goki Nakashima. Methodology: Hironori Ishiguchi, Shigeki Kobayashi, Motoaki Sano. Project administration: Hironori Ishiguchi, Yasuhiro Yoshiga, Motoaki Sano. Resources: Yasuhiro Yoshiga, Takuya Omuro. Software: Hironori Ishiguchi. Visualization: Hironori Ishiguchi. Validation: Hironori Ishiguchi, Goki Nakashima. Writing‐original draft: Hironori Ishiguchi. Writing–review and editing: Yasuhiro Yoshiga, Masakazu Fukuda, Masaki Shibuya, Shigeki Kobayashi, Motoaki Sano. Supervision: Shigeki Kobayashi, Motoaki Sano, Masaki Shibuya. Guarantor: Hironori Ishiguchi.
This article was sent to Kevin F. Kwaku, MD, PhD, Associate Editor, for review by expert referees, editorial decision, and final disposition.
Supplemental Material is available at https://www.ahajournals.org/doi/suppl/10.1161/JAHA.126.049254
For Sources of Funding and Disclosures, see page 11.
This work was presented in part at the European Society of Cardiology Congress, held August 28–31, 20226, in Munich, Germany.
References
- 1. Ndumele CE, Rangaswami J, Chow SL, Neeland IJ, Tuttle KR, Khan SS, Coresh J, Mathew RO, Baker‐Smith CM, Carnethon MR, et al. Cardiovascular‐kidney‐metabolic health: a presidential advisory from the American Heart Association. Circulation. 2023;148:1606–1635. doi: 10.1161/CIR.0000000000001184 [DOI] [PubMed] [Google Scholar]
- 2. Stevens PE, Ahmed SB, Carrero JJ. KDIGO 2024 clinical practice guideline for the evaluation and management of chronic kidney disease. Kidney Int. 2024;105(4S):S117–S314. doi: 10.1016/j.kint.2023.10.018 [DOI] [PubMed] [Google Scholar]
- 3. Ostrominski JW, Claggett BL, Miao ZM, Mc Causland FR, Anand IS, Desai AS, Jhund PS, Lam CSP, Pfeffer MA, Pitt B, et al. Cardiovascular‐kidney‐metabolic overlap in heart failure with mildly reduced or preserved ejection fraction: a trial‐level analysis. J Am Coll Cardiol. 2024;84:223–228. doi: 10.1016/j.jacc.2024.05.005 [DOI] [PubMed] [Google Scholar]
- 4. Mei DA, Imberti JF, Vitolo M, Romiti GF, Corica B, Mantovani M, Bonini N, Cherubini B, Marin F, Diemberger I, et al. Cardiovascular‐kidney‐metabolic domains and impact on antithrombotic treatment, integrated care and clinical outcomes in patients with atrial fibrillation: results from a prospective European registry. Eur J Intern Med. 2025;145:106512. doi: 10.1016/j.ejim.2025.106512 [DOI] [PubMed] [Google Scholar]
- 5. Romiti GF, Mei DA, Corica B, Proietti M, Boriani G, Olshansky B, Chao TF, Huisman MV, Lip GYH, Investigators GLORIA‐AF. Cardio‐kidney‐metabolic complexity in patients with atrial fibrillation: an analysis from the prospective GLORIA‐AF registry phase III. Cardiovasc Diabetol. 2025;24:395. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Kim D, Shim J, Choi E‐K, Oh I‐Y, Kim J, Lee YS, Park J, Ko J‐S, Park K‐M, Sung J‐H, et al. Long‐term anticoagulation discontinuation after catheter ablation for atrial fibrillation: the ALONE‐AF randomized clinical trial. JAMA. 2025;334:1246–1254. doi: 10.1001/jama.2025.14679 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Verma A, Birnie DH, Jiang C, Heidbüchel H, Hindricks G, Kirchhof P, Healey JS, Wang Y, Dagres N, Deyell MW, et al. Antithrombotic therapy after successful catheter ablation for atrial fibrillation. N Engl J Med. 2025;394:323–332. doi: 10.1056/NEJMoa2509688 [DOI] [PubMed] [Google Scholar]
- 8. Ishiguchi H, Yoshiga Y, Shimizu A, Ueyama T, Ono M, Fukuda M, Kato T, Fujii S, Hisaoka M, Uchida T, et al. Impact of atrial tachyarrhythmia recurrence on the development of long‐term adverse clinical events following catheter ablation in patients with atrial fibrillation with systolic impairment: a single‐center observational study. J Am Heart Assoc. 2022;11:e023640. doi: 10.1161/JAHA.121.023640 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Ishiguchi H, Yoshiga Y, Fukuda M, Fuji S, Hisaoka M, Hashimoto S, Omuro T, Fukue N, Kobayashi S, Sano M. Comparison of post ablation left atrial volume index versus left atrial reverse remodeling for prognostic events in persistent atrial fibrillation. Sci Rep. 2025;15:27384. doi: 10.1038/s41598-025-13311-w [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Ishiguchi H, Yoshiga Y, Shimizu A, Fukuda M, Omuro A, Hisaoka M, Nakashima Y, Fujita M, Hashimoto S, Omuro T, et al. Novel method for risk stratification of major adverse clinical events using pre‐ and post‐ablation left atrial volume index in patients with persistent atrial fibrillation. Circ Rep. 2024;6:415–423. doi: 10.1253/circrep.CR-24-0062 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP; STROBE Initiative . The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. J Clin Epidemiol. 2008;61:344–349. doi: 10.1016/j.jclinepi.2007.11.008 [DOI] [PubMed] [Google Scholar]
- 12. Sumida K, Nadkarni GN, Grams ME, Sang Y, Ballew SH, Coresh J, Matsushita K, Surapaneni A, Brunskill N, Chadban SJ, et al. Conversion of urine protein–creatinine ratio or urine dipstick protein to urine albumin–creatinine ratio for use in chronic kidney disease screening and prognosis: an individual participant‐based meta‐analysis. Ann Intern Med. 2020;173:426–435. doi: 10.7326/M20-0529 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Jang EC, Park YM, Han HW, Lee CS, Kang ES, Lee YH, Nam SM. Machine‐learning enhancement of urine dipstick tests for chronic kidney disease detection. J Am Med Inform Assoc. 2023;30:1114–1124. doi: 10.1093/jamia/ocad051 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Herrington WG, Judge PK, Grams ME, Wanner C. Chronic kidney disease. Lancet. 2025;407:90–104. doi: 10.1016/S0140-6736(25)01942-7 [DOI] [PubMed] [Google Scholar]
- 15. Chyou JY. ACC/AHA/ACCP/HRS guideline for the diagnosis and management of atrial fibrillation. Circulation. 2024;149:e1–e156. doi: 10.1161/CIR.0000000000001207 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Geurts S, van der Burgh AC, Bos MM, Ikram MA, Stricker BHC, Deckers JW, Hoorn EJ, Chaker L, Kavousi M. Disentangling the association between kidney function and atrial fibrillation: a bidirectional Mendelian randomization study. Int J Cardiol. 2022;355:15–22. doi: 10.1016/j.ijcard.2022.03.004 [DOI] [PubMed] [Google Scholar]
- 17. Olesen JB, Lip GYH, Kamper A‐L, Hommel K, Køber L, Lane DA, Lindhardsen J, Gislason GH, Torp‐Pedersen C. Stroke and bleeding in atrial fibrillation with chronic kidney disease. N Engl J Med. 2012;367:625–635. doi: 10.1056/NEJMoa1105594 [DOI] [PubMed] [Google Scholar]
- 18. Odutayo A, Wong CX, Hsiao AJ, Hopewell S, Altman DG, Emdin CA. Atrial fibrillation and risks of cardiovascular disease, renal disease, and death: systematic review and meta‐analysis. BMJ. 2016;354:i4482. doi: 10.1136/bmj.i4482 [DOI] [PubMed] [Google Scholar]
- 19. Neuen BL, Jun M, Wick J, Kotwal S, Badve SV, Jardine MJ, Gallagher M, Chalmers J, Nallaiah K, Perkovic V, et al. Estimating the population‐level impacts of improved uptake of SGLT2 inhibitors in patients with chronic kidney disease: a cross‐sectional observational study using routinely collected Australian primary care data. Lancet Reg Health West Pac. 2024;43:100988. doi: 10.1016/j.lanwpc.2023.100988 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Barzilay JI, Farag YMK, Durthaler J. Albuminuria: an underappreciated risk factor for cardiovascular disease. J Am Heart Assoc. 2024;13:e030131. doi: 10.1161/JAHA.123.030131 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. McLean P, Bennett J. SGLT2 inhibitors across various patient populations in the era of precision medicine: the multidisciplinary team approach. Npj Metab Health Dis. 2025;61:1903. [Google Scholar]
- 22. Potpara T, Romiti GF, Sohns C. The 2024 European Society of Cardiology guidelines for diagnosis and management of atrial fibrillation: a viewpoint from a practicing clinician's perspective. Thromb Haemost. 2024;124:1087–1094. doi: 10.1055/a-2434-9244 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Tables S1–S2
Figures S1–S2
STROBE Checklist
