This cohort study evaluates changes in global and domain-specific cognitive function after catheter ablation for atrial fibrillation.
Key Points
Question
Is catheter ablation for atrial fibrillation associated with changes in global or domain-specific cognitive function?
Findings
In this cohort study of 75 patients undergoing catheter ablation and 40 patients in a control group, ablation was not associated with cognitive decline at 6 months. Global cognitive performance assessed by the Montreal Cognitive Assessment remained stable, while domain-specific testing with the novel REMEDES for Alzheimer (R4Alz) tool identified modest improvements in short-term working memory and perceptual inhibition.
Meaning
These findings suggest that catheter ablation for atrial fibrillation is cognitively safe in the short term, and domain-specific cognitive assessment may detect subtle changes not captured by global screening tools.
Abstract
Importance
Atrial fibrillation (AF) is associated with cognitive impairment, but the cognitive effects of catheter ablation (CA) using contemporary techniques remain uncertain.
Objective
To evaluate changes in global and domain-specific cognitive function after CA for AF.
Design, Setting, and Participants
This prospective cohort study included consecutive patients undergoing CA for AF and a nonablation AF control group at a tertiary university hospital in Greece between January 2022 and June 2024. Cognitive assessments were performed at baseline and at 6-month follow-up.
Exposures
CA using cryoballoon or pulsed field ablation (PFA).
Main Outcomes and Measures
Changes from baseline to 6 months in Montreal Cognitive Assessment (MoCA) scores and the novel REMEDES for Alzheimer (R4Alz) total and subdomain scores. Brain magnetic resonance imaging was performed before and within 48 hours after ablation to detect new cerebral infarcts.
Results
Among 75 patients (40 [53.3%] female) undergoing ablation (mean [SD] age, 61.9 [10.4] years; 54 cryoballoon, 21 PFA) and 40 in a nonablation control group (26 [65.0%] female; mean [SD] age,61.4 [9.2] years), baseline cognitive scores were similar between groups. No new cerebral infarcts were detected after ablation. At 6 months, MoCA scores in the ablation group did not change significantly (mean change, 0.27; 95% CI, −0.08 to 0.62), whereas R4Alz total scores increased (mean change, 2.81; 95% CI, 1.55 to 4.07), mainly associated with gains in short-term working memory and perceptual inhibition. No significant changes were observed in the control group. Compared with the control group, the ablation cohort demonstrated significantly greater improvements in both MoCA (mean difference in change, 0.74; 95% CI, 0.17 to 1.31; P = .01) and R4Alz total score (mean difference in change score, 3.18; 95% CI, 0.87 to 5.49; P = .007). In adjusted analyses accounting for baseline cognitive score and clinical covariates, ablation was independently associated with higher follow-up MoCA (β = 0.88; 95% CI, 0.31 to 1.44) and R4Alz scores (β = 3.80; 95% CI, 1.70 to 5.90).
Conclusions and Relevance
In this prospective observational cohort study of 75 patients undergoing CA for AF, ablation was not associated with cognitive decline and was instead associated with modest short-term improvements in selected cognitive domains. Domain-specific assessment detected changes not captured by global cognitive screening, suggesting that modern ablation techniques may be associated with early subtle cognitive gains.
Introduction
Atrial fibrillation (AF) is associated with a higher risk of cognitive impairment and dementia, even in the absence of overt stroke.1 Several mechanisms have been proposed, including silent cerebral infarction, cerebral hypoperfusion due to irregular rhythm and impaired cardiac output, systemic inflammation, and microembolic events.2
Catheter ablation (CA) is now an established therapy for AF, improving rhythm control and quality of life.3 However, its impact on cognitive function remains uncertain. Some studies have shown that restoration of sinus rhythm and improved hemodynamics may benefit cognition, while others have reported transient cognitive decline after ablation, possibly linked to silent cerebral emboli.4,5
The Montreal Cognitive Assessment (MoCA) is the most widely used screening tool for mild cognitive impairment, but it may not adequately capture domain-specific cognitive impairments.6 The REMEDES for Alzheimer (R4Alz) battery is a novel computerized tool originally developed for the early detection of Alzheimer disease and is designed to evaluate working memory, attention control, and executive functions using REMEDES pads, which present visual and auditory stimuli requiring rapid motor-cognitive responses with the goal to be less affected by cultural and educational factors.7 It may therefore provide a more sensitive measure of cognitive performance in patients with AF undergoing ablation. This study evaluated cognitive function before and 6 months after CA using both MoCA and the R4Alz battery, hypothesizing that R4Alz would detect minor changes in cognitive function compared with MoCA.
Methods
Study Population and Data Collection
This was a prospective cohort study conducted at a tertiary university hospital in Thessaloniki, Greece. Consecutive patients with atrial fibrillation undergoing CA were enrolled between January 2022 and June 2024. A contemporaneous nonablation AF control cohort was recruited from the same outpatient clinic and time period. Patients in the control cohort did not undergo ablation during the study period owing to patient preference, mild symptom burden, or physician assessment that ablation was not immediately indicated and were treated according to standard clinical practice, including antiarrhythmic drug therapy and cardioversion at the discretion of the treating physician as part of an individualized rhythm-control strategy. The control cohort was selected to approximate the ablation cohort with respect to age, sex, diabetes, and cardiovascular comorbidity profile. Exclusion criteria for both cohorts included (1) prior diagnosis of dementia or major neurocognitive disorder, (2) ischemic stroke or transient ischemic attack within the preceding 12 months, (3) left ventricular ejection fraction of 35% or less, and (4) inability to complete cognitive testing. All participants provided written informed consent. The study was approved by the institutional ethics committee of Aristotle University of Thessaloniki and conducted in accordance with the Declaration of Helsinki. This study follows the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline. Because the R4Alz battery has not previously been applied in an AF ablation population, a formal power calculation was not performed. A target of approximately 35 paired assessments was considered reasonable for an exploratory prospective study, informed by published recommendations for pilot study sample size considerations.8 Final sample size was determined by feasibility and patient availability during the study period. Brain magnetic resonance imaging (MRI) was performed in patients undergoing ablation using a 1.5-Tesla unit (Siemens Symphony Maestro Class 1.5T), before and within 18 to 48 hours after the ablation procedure, and the images were interpreted by 2 board-certified radiologists (A.A. and G.S.) who were blinded to any clinical information.
Cognitive Assessment
Cognitive function was assessed at baseline and at 6 months using MoCA and selected tasks from the R4Alz battery. The R4Alz battery is a computerized neuropsychological tool that assesses 3 main domains: working memory (storage, processing, and updating), attention control (selective, sustained, and divided attention), and executive functions (inhibition, task or rule switching, and cognitive flexibility).7 Given the extensive structure of the complete R4Alz battery, representative tasks from each major cognitive-control domain were selected a priori to preserve domain coverage while maintaining feasibility for repeated assessments in a clinical AF population. Specifically, these included (1) R4Alz 1 (working memory), consisting of R4Alz 1a (short-term working memory storage): immediate recall of the correct sequence of pads’ activation; and R4Alz 1b (working memory processing): immediate reverse recall of the correct sequence of pads’ activation; (2) R4Alz 2 (visual-auditory-divided attention control): manipulation of numbers, sounds, and sequences; and (3) R4Alz 3a (inhibition), consisting of R4Alz 3a.1 (visual identification): animal naming, R4Alz 3a.2 (auditory identification): sound recognition, R4Alz 3a.3 (perceptual inhibition): visual selective attention with auditory inhibition, and R4Alz 3a.4 (perceptual and cognitive inhibition): auditory selective attention with visual inhibition.
A total score was calculated to provide an overall measure of memory and executive functioning. Testing was performed by trained personnel (S.C.Z., S.E.K., M.I.M., and A.Z.) at baseline (preablation) and repeated at 6-month follow-up.
Ablation Procedure and Follow-Up
Patients underwent CA using either cryoballoon ablation (Arctic Front Advance Pro cryoablation system, Medtronic Inc) or pulsed field ablation (PFA; Farapulse system, Boston Scientific), according to operator preference and device availability. The former was performed with conscious sedation (using fentanyl and midazolam as needed), while PFA was performed under general anesthesia, with periprocedural anticoagulation managed according to current guidelines (intravenous heparin was administered intraprocedurally to maintain an activated clotting time of 300 s or more). Transeptal puncture was assisted by intraprocedural imaging (transesophageal or intracardiac echocardiography [TEE/ICE]), which was also employed to ensure absence or development of new visible thrombi (either in the left atrial appendage or related to instrumentation). Pulmonary vein isolation was confirmed in all patients at the end of the procedure using standard entrance block verification with a mapping catheter. All patients were maintained on oral anticoagulation throughout the periprocedural period and during follow-up, in accordance with CHA2DS2-VASc (congestive heart failure; hypertension; age 75 years or older; diabetes; stroke, TIA, or thromboembolism history; vascular disease; age 65 to 74 years; and sex category [female]) score and guideline recommendations. Antiarrhythmic drug therapy before and after CA was managed according to routine clinical practice at the discretion of the treating electrophysiologist and was not standardized by the study protocol. In the ablation cohort, temporary continuation of antiarrhythmic therapy during the postablation blanking period was permitted, whereas subsequent continuation or discontinuation of antiarrhythmic drugs was individualized according to rhythm status, symptom burden, recurrence pattern, and physician judgment. In the control cohort of patients who did not undergo ablation, antiarrhythmic drug therapy and cardioversion were continued according to standard clinical practice at the discretion of the treating physician as part of an individualized rhythm-control strategy. Follow-up visits occurred at 3 and 6 months, including clinical assessment, electrocardiogram (ECG) monitoring, and repeat cognitive testing at 6 months. Recurrence of AF was assessed using ECG recordings obtained during follow-up visits and ambulatory Holter monitoring when clinically indicated and was defined as any atrial tachyarrhythmia lasting 30 or more seconds after the blanking period. All participants completed paired baseline and follow-up assessments. Isolated missing values were present for individual cognitive variables, resulting in less than 5% missingness across primary cognitive outcomes.
Statistical Analysis
Continuous variables are expressed as mean (SD) or median (IQR) depending on distribution, and categorical variables are given as counts and percentages. Group comparisons at baseline (cryoballoon vs PFA) were performed with independent t tests or Mann-Whitney U tests for continuous variables and χ2 or Fisher exact tests for categorical variables. Within-group comparisons of baseline and follow-up MoCA and R4Alz scores were conducted using paired t tests. Between-group differences in change scores (change = follow-up − baseline values) were examined with independent t tests, and 95% CIs are reported for change estimates.
Pearson correlation was used to assess associations between MoCA and R4Alz scores. Linear regression analyses were performed to explore baseline cognitive score and change variables in R4Alz and MoCA scores.
Adjusted multiple linear regression analyses were performed to compare 6-month cognitive outcomes between the ablation and control cohorts after adjustment for baseline cognitive score, age, sex, educational level, and CHA2DS2-VASc score. Sensitivity analyses additionally adjusted for AF recurrence, rhythm status during cognitive testing, and baseline and follow-up amiodarone use. Missing cognitive data were infrequent (<5%), and complete-case analyses were therefore performed.
Results are reported as regression coefficients (β) with 95% CIs and P values. A 2-sided P < .05 was considered statistically significant. All analyses were performed in R version 4.1.1 (R Project for Statistical Computing).9
Results
Baseline Characteristics and Procedural Parameters
The ablation cohort included 75 patients with AF (mean [SD] age, 61.9 [10.4] years; 40 [53.3%] female), of whom 54 underwent cryoballoon ablation and 21 underwent PFA. The control cohort included 40 patients (26 [65.0%] female; mean [SD] age,61.4 [9.2] years) receiving medical therapy alone. All participants completed paired baseline and follow-up assessments. Baseline demographic characteristics, comorbidities, medication use, rhythm status at cognitive testing, and cognitive scores were similar between cohorts (Table 1).
Table 1. Baseline Characteristics.
| Variable | Participants, No. (%) | P valuea | Participants, No. (%) | P valueb | ||
|---|---|---|---|---|---|---|
| Total (N = 75) | Cryo (n = 54) | PFA (n = 21) | Control (n = 40) | |||
| Age, mean (SD), y | 61.95 (10.38) | 61.98 (10.04) | 61.86 (11.47) | .96 | 61.40 (9.21) | .77 |
| Sex | ||||||
| Female | 40 (53.4) | 30 (55.6) | 10 (47.6) | .72 | 26 (65.0) | .22 |
| Male | 35 (46.6) | 24 (44.4) | 11 (52.3) | 14 (35.0) | ||
| Persistent AF | 30 (40.0) | 17 (31.5) | 13 (61.9) | .01 | 18 (45.0) | .64 |
| Atrial flutter | 15 (20.0) | 10 (18.5) | 5 (23.8) | .74 | 7 (17.5) | .72 |
| Previous AF ablation | 13 (17.3) | 9 (16.7) | 4 (19.0) | >.99 | 5 (12.5) | .47 |
| Previous AFl ablation | 5 (6.7) | 2 (3.7) | 3 (14.3) | .13 | 1 (2.5) | .51 |
| BMI, mean (SD)c | 27.57 (5.79) | 27.48 (5.86) | 27.81 (5.74) | .82 | 27.51 (5.40) | .95 |
| Sinus rhythm at baseline cognitive testing | 49 (65.3) | 37 (68.5) | 12 (57.1) | .35 | 27 (67.5) | .82 |
| Sinus rhythm at 6-mo cognitive testing | 66 (88.0) | 43 (79.6) | 17 (81.0) | .90 | 31 (80.0) | .15 |
| Comorbidities | ||||||
| Chronic kidney disease | 3 (4.0) | 2 (3.7) | 1 (4.8) | >.99 | 0 | .20 |
| Coronary artery disease | 13 (17.3) | 11 (20.4) | 2 (9.5) | .33 | 13 (32.5) | .07 |
| Dyslipidemia | 40 (53.3) | 28 (51.9) | 12 (57.1) | .68 | 24 (60.0) | .68 |
| Heart failure | 8 (10.7) | 6 (11.1) | 2 (9.5) | >.99 | 5 (12.5) | .76 |
| History of stroke or TIA | 2 (2.7) | 1 (1.9) | 1 (4.8) | .48 | ||
| Hypertension | 46 (61.3) | 31 (57.4) | 15 (71.4) | .26 | 24 (60.0) | .88 |
| Obstructive sleep apnea | 12 (16.0) | 9 (16.7) | 3 (14.3) | >.99 | 4 (10.0) | .37 |
| Peripheral artery disease | 0 | 0 | 0 | >.99 | 0 | >.99 |
| Pulmonary disease | 7 (9.3) | 4 (7.4) | 3 (14.3) | .39 | 4 (10.0) | .90 |
| Obesity | 38 (50.7) | 28 (51.9) | 10 (47.6) | .74 | 20 (50.0) | .94 |
| Valvular heart disease | 8 (10.7) | 7 (13.0) | 1 (4.8) | .43 | 3 (7.5) | .58 |
| HAS-BLED score | 1.28 (1.06) | 1.24 (1.04) | 1.38 (1.12) | .62 | 1.18 (0.98) | .59 |
| CHA2DS2-VASc score | 1.92 (1.35) | 1.80 (1.28) | 2.24 (1.51) | .24 | 1.93 (1.40) | .98 |
| Medications | ||||||
| Oral anticoagulant | 73 (97.3) | 53 (98.1) | 20 (95.2) | .48 | 35 (87.5) | .23 |
| Amiodarone | 30 (40.0) | 24 (44.4) | 6 (28.6) | .20 | 20 (50.0) | .12 |
| Antiplatelet | 4 (5.3) | 4 (7.4) | 0 | .57 | 2 (5.0) | .86 |
| β-Blocker | 47 (62.7) | 33 (61.1) | 14 (66.7) | .65 | 27 (67.5) | .86 |
| Calcium-channel blocker | 8 (10.7) | 5 (9.3) | 3 (14.3) | .67 | 5 (12.5) | .62 |
| Sotalol | 1 (1.3) | 1 (1.9) | 0 | >.99 | 0 | .52 |
| Statin | 34 (45.3) | 22 (40.7) | 12 (57.1) | .20 | 19 (47.5) | .80 |
| AAD (class I) | 13 (17.3) | 9 (16.7) | 4 (19.0) | >.99 | 8 (20.0) | >.99 |
| Digoxin | 2 (2.7) | 2 (3.7) | 0 | >.99 | 0 | .31 |
| ACEI | 10 (13.3) | 8 (14.8) | 2 (9.5) | .71 | 7 (17.5) | .70 |
| ARB | 15 (20.0) | 8 (14.8) | 7 (33.3) | .10 | 8 (20.0) | .80 |
| MRA | 8 (10.7) | 6 (11.1) | 2 (9.5) | >.99 | 8 (20.0) | .24 |
| Baseline cognitive performance score, mean (SD) | ||||||
| MoCA | 27.16 (2.03) | 27.00 (1.99) | 27.57 (2.11) | .53 | 26.85 (2.05) | .39 |
| R4Alz | 175.08 (11.3) | 175.33 (11.71) | 174.43 (10.43) | .12 | 174.51 (11.16) | .81 |
Abbreviations: AAD, antiarrhythmic drug; ACEI, angiotensin-converting enzyme inhibitor; AF, atrial fibrillation; AFl, atrial flutter; ARB, angiotensin receptor blocker; BMI, body mass index; CHA2DS2-VASc, congestive heart failure; hypertension; age 75 years or older; diabetes; stroke, TIA, or thromboembolism history; vascular disease; age 65 to 74 years; and sex category (female); Cryo, cryoballoon ablation; HAS-BLED, hypertension, abnormal kidney or liver function, stroke, bleeding history or predisposition, labile International Normalized Ratio, elderly (aged >65 years), drugs and/or alcohol concomitantly; MoCA, Montreal Cognitive Assessment; MRA, mineralocorticoid receptor antagonist; PFA, pulsed field ablation; R4Alz, REMEDES for Alzheimer; TIA, transient ischemic attack.
Between Cryo and PFA.
Between ablation and control cohorts.
Calculated as weight in kilograms divided by height in meters squared.
Baseline Cognitive Function
At baseline, mean (SD) MoCA and R4Alz scores in the ablation cohort were 27.16 (2.03) and 175.08 (11.3), respectively, and did not differ significantly from the control group (Table 1). Baseline MoCA and R4Alz scores were moderately correlated (r = 0.42; P < .001). In multiple linear regression analyses, increasing age was independently associated with lower baseline R4Alz scores (β = −0.39; 95% CI, −0.47 to −0.06). No clinical variables were independently associated with baseline MoCA scores after adjustment (Table 2).
Table 2. Linear Regression Analyses of Baseline Cognitive Performance Variables.
| Variable | Single regression, β (95% CI) | P value | Multiple regression, β (95% CI) | P value |
|---|---|---|---|---|
| MoCA | ||||
| Chronic kidney disease | −0.51 (−2.91 to 1.88) | .67 | NA | NA |
| AF recurrence | −0.09 (−1.16 to 1.33) | .89 | NA | NA |
| Coronary artery disease | 0.12 (−1.10 to 1.35) | .83 | NA | NA |
| Heart failure | 0.10 (−1.42 to 1.62) | .89 | NA | NA |
| Stroke or TIA | −1.19 (−4.09 to 1.71) | .41 | NA | NA |
| Male sex | −0.04 (−0.98 to 0.89) | .93 | NA | NA |
| CHA2DS2-VASc score | −0.17 (−0.51 to 0.17) | .33 | NA | NA |
| Arterial hypertension | −0.63 (−1.59 to 0.31) | .18 | NA | NA |
| Age | −0.05 (−0.09 to −0.0001)a | .048a | −0.03 (−0.08 to 0.007) | .09 |
| OSA | 0.71 (−0.56 to 1.97) | .27 | NA | NA |
| Diabetes | −0.06 (−0.46 to 0.33) | .73 | NA | NA |
| Dyslipidemia | −0.25 (−1.20 to 0.68) | .59 | NA | NA |
| Educational level | 0.58 (0.01 to 1.15)a | .01a | 0.48 (−0.09 to 1.06) | .10 |
| R4Alz | ||||
| Chronic kidney disease | 4.76 (−5.08 to 14.61) | .33 | NA | NA |
| AF recurrence | −2.87 (−7.95 to 2.21) | .26 | NA | NA |
| Stroke/TIA | −3.86 (−15.88 to 8.15) | .52 | NA | NA |
| Male sex | −4.17 (−0.75 to −0.31) | .33 | NA | .13 |
| Arterial hypertension | −6.67 (−11.80 to −1.52)a | .01a | −1.17 (−7.51 to 0.11) | .09 |
| Age | −0.53 (−0.75 to −0.31)a | <.001a | −0.39 (−0.47 to −0.06)a | <.001a |
| Dyslipidemia | −4.62 (−1.20 to 0.68) | .07 | 1.41 (−4.27 to 7.10) | .62 |
| Coronary artery disease | −1.52 (6.68 to 3.63) | .55 | NA | NA |
| Heart failure | −2.75 (−11.20 to 5.71) | .52 | NA | NA |
| CHA2DS2-VASc score | −3.45 (−5.22 to −1.68)a | .007a | −1.30 (−3.99 to 1.40) | .20 |
| OSA | 1.08 (−4.21 to 6.37) | .69 | NA | NA |
| Diabetes | −0.51 (−2.15 to 1.13) | .54 | NA | NA |
| Educational level | 3.17 (−0.017 to 6.35) | .05 | 1.94 (−1.18 to 7.10) | .62 |
Abbreviations: AF, atrial fibrillation; β, regression coefficient; CHA2DS2-VASc, congestive heart failure; hypertension; age 75 years or older; diabetes; stroke, TIA, or thromboembolism history; vascular disease; age 65 to 74 years; and sex category (female); MoCA, Montreal Cognitive Assessment; NA, not applicable; OSA, obstructive sleep apnea; R4Alz, REMEDES for Alzheimer; TIA, transient ischemic attack.
Indicates significant result.
CA, Follow-Up, and Cognitive Outcomes
During follow-up, AF recurrence occurred in 16 patients undergoing ablation (23%) and 14 control group patients (35%). Sinus rhythm at the time of 6-month cognitive testing was present in 66 patients undergoing ablation (88.0%) and 31 control group patients (80.0%). Detailed rhythm status and antiarrhythmic drug use during follow-up are presented in eTable 1 in Supplement 1. No new cerebral infarcts were detected on postprocedural MRI.
In the ablation cohort, MoCA scores did not change significantly from baseline to 6 months (mean change, 0.27; 95% CI, −0.08 to 0.62). In contrast, R4Alz total scores increased significantly (mean change, 2.81; 95% CI, 1.55 to 4.07). No significant within-group changes in MoCA or R4Alz scores were observed in the control cohort. Compared with the control group, the ablation cohort demonstrated significantly greater improvements in both MoCA (mean difference in change score, 0.74; 95% CI, 0.17 to 1.31; P = .01) and R4Alz total score (mean difference in change score, 3.18; 95% CI, 0.87 to 5.49; P = .007) (eTable 2 in Supplement 1).
Within the ablation cohort, significant improvements were observed in short-term working memory (mean change, 1.31; 95% CI, 0.78 to 1.84) and perceptual inhibition (mean change, 0.37; 95% CI, 0.01 to 0.73). Changes in other R4Alz subdomains were not statistically significant. No significant differences were observed between cryoballoon ablation and PFA (eTable 2 in Supplement 1).
In adjusted multiple linear regression analyses controlling for baseline cognitive score, age, sex, educational level, and CHA2DS2-VASc score, the ablation cohort demonstrated significantly higher follow-up cognitive scores compared with the control group for both MoCA (β, 0.88; 95% CI, 0.31 to 1.44; P = .003) and R4Alz total score (β, 3.80; 95% CI, 1.70 to 5.90) (Table 3). In the R4Alz model, higher CHA2DS2-VASc score was also independently associated with follow-up R4Alz score (β, 0.94; 95% CI, 0.13 to 1.75). Exploratory adjusted analyses of R4Alz subdomains demonstrated a significant association between ablation and higher short-term working memory scores (R4Alz 1a: β, 1.31; 95% CI, 0.53 to 2.09), while other subdomains did not reach statistical significance (eTable 3 in Supplement 1). The Figure illustrates adjusted estimated 6-month cognitive outcomes according to treatment strategy.
Table 3. Multiple Linear Regression Analyses of 6-Month Cognitive Outcomes.
| Variable | MoCA, β (95% CI) | P value | R4Alz, β (95% CI) | P value |
|---|---|---|---|---|
| Ablation cohort | 0.88 (0.31 to 1.44) | .003a | 3.80 (1.70 to 5.90) | <.001a |
| Baseline MoCA score | 0.74 (0.60 to 0.88) | <.001a | NA | NA |
| Baseline R4Alz score | NA | NA | 0.82 (0.73 to 0.92) | <.001a |
| Age | 0.03 (−0.00 to 0.06) | .08 | −0.08 (−0.19 to 0.03) | .16 |
| Male sex | 0.38 (−0.16 to 0.92) | .17 | 1.22 (−0.81 to 3.26) | .24 |
| Educational level | 0.11 (−0.25 to 0.47) | .55 | 0.94 (−0.39 to 2.27) | .17 |
| CHA2DS2-VASc score | −0.07 (−0.29 to 0.14) | .50 | 0.94 (0.13 to 1.75) | .02a |
Abbreviations: CHA2DS2-VASc, congestive heart failure; hypertension; age 75 years or older; diabetes; stroke, TIA, or thromboembolism history; vascular disease; age 65 to 74 years; and sex category (female); MoCA, Montreal Cognitive Assessment; NA, not applicable; R4Alz, REMEDES for Alzheimer.
Indicates significant result.
Figure. Violin Plot Showing Adjusted 6-Month Cognitive Outcomes According to Treatment Strategy.
Estimated values derived from multiple linear regression models adjusted for baseline score, age, sex, educational level, and CHA2DS2-VASc (congestive heart failure; hypertension; age 75 years or older; diabetes; stroke, TIA, or thromboembolism history; vascular disease; age 65 to 74 years; and sex category [female]) score. Filled circles indicate adjusted mean predicted values; error bars indicate 95% CIs; open circles indicate individual predicted values; and boxes indicate IQR with the center line indicating the median. MOCA indicates Montreal Cognitive Assessment; R4Alz, REMEDES for Alzheimer.
Sensitivity Analyses
Sensitivity analyses additionally adjusting for AF recurrence, rhythm status at cognitive testing, and baseline and follow-up amiodarone use were completed. These analyses demonstrated consistent associations between the ablation cohort and higher follow-up R4Alz and MoCA scores (eTable 4 in Supplement 1).
Discussion
In this prospective observational study evaluating cognitive performance after CA for AF, we found no evidence that ablation adversely affected cognitive function. On the contrary, we observed modest improvements in domain-specific measures captured by the R4Alz tool, while global cognitive performance assessed by MoCA remained stable over 6 months. The key finding is the demonstration of short-term cognitive safety of CA, with no signal of cognitive decline. R4Alz gains were associated with improvements in short-term working memory and perceptual inhibition. No new cerebral infarcts were identified on MRI, supporting the neurological safety of the ablation procedures.10,11 These findings suggest that sensitive, domain-specific cognitive testing can reveal subtle cognitive dynamics after AF ablation that global screening tools may not detect.
Although the magnitude of the observed cognitive differences was modest, several considerations suggest that these findings may still be clinically relevant. First, the study population was fairly young and cognitively preserved at baseline, limiting the likelihood of large interval cognitive changes over a follow-up period of only 6 months. Second, cognitive changes related to AF may initially manifest as subtle domain-specific abnormalities that are difficult to detect using brief global screening instruments alone. Finally, even modest short-term differences may become more meaningful over longer follow-up periods if sustained reductions in AF burden translate into cumulative neurocognitive benefit. Nevertheless, the exploratory nature of these findings and the absence of established reference thresholds for R4Alz changes require cautious interpretation.
Previous studies have consistently demonstrated that AF is associated with accelerated cognitive decline and an increased risk of dementia, even in the absence of clinical stroke.12,13,14,15 Several mechanisms have been proposed to explain this association, including silent cerebral embolism, chronic cerebral hypoperfusion, endothelial dysfunction, systemic inflammation, impaired cerebral autoregulation, autonomic dysfunction, and microvascular disease.16,17,18,19 Our findings extend these observations by suggesting a possible association between CA and a more favorable cognitive profile in patients with AF by restoring sinus rhythm and consequently improving cerebral hemodynamics, reducing embolic burden, and dampening systemic inflammation.
Notably, no new cerebral infarcts were detected on postprocedural MRI in the ablation cohort. This finding, which may contradict some previous observations,20,21 might be explained by the uninterrupted anticoagulation protocol as well as the short left atrium dwell times (reflecting current laboratory workflows), along with the use of intraprocedural imaging (TEE/ICE), also considering that the applied MRI protocol could not preclude the presence of extremely small microemboli.
One notable finding was the difference in granularity between global and domain-specific cognitive assessments. Although MoCA scores remained stable over follow-up, R4Alz identified significant within-group improvements in specific cognitive domains. This observation suggests that domain-specific tools may provide complementary information beyond global screening instruments, particularly when evaluating subtle cognitive changes over relatively short follow-up periods. Previous studies have suggested that AF-related cognitive dysfunction may preferentially affect executive and attentional domains, whereas brief screening instruments such as MoCA may be less sensitive for detecting modest changes over short follow-up periods.1,22 Similar observations have been reported in prior AF cohorts, where more detailed neurocognitive assessments identified domain-specific differences not captured by global screening measures.1,23
The mechanisms underlying the more favorable cognitive profile observed in the ablation cohort remain uncertain. Although AF recurrence rates were relatively comparable between groups, binary recurrence measures may incompletely reflect overall arrhythmic exposure and rhythm stability.24 CA is known to reduce AF burden even among patients with recurrent atrial arrhythmias, and lower arrhythmic burden may theoretically contribute to more stable cerebral hemodynamics, reduced inflammatory activation, and lower cumulative embolic exposure.24,25,26 However, because AF burden was not systematically quantified in the present study, these potential mechanisms remain speculative. Nevertheless, sensitivity analyses adjusting for rhythm-related variables and antiarrhythmic drug use demonstrated findings generally consistent with the primary analyses.
Interestingly, in adjusted analyses, higher CHA2DS2-VASc score was independently associated with follow-up R4Alz performance. However, this finding should be interpreted cautiously given the relatively low overall CHA2DS2-VASc burden and limited score range in our cohort. Larger studies including broader cardiovascular risk profiles are needed before drawing conclusions regarding potential relationships between baseline vascular risk and cognitive outcomes.
From a clinical standpoint, our results highlight 2 important messages. First, AF ablation appears cognitively safe, with no evidence of short-term deterioration and a pattern of subtle improvements that may accompany restoration of more stable rhythm physiology, even in the absence of radiological evidence of cerebral injury. Second, domain-specific assessments such as R4Alz may offer superior sensitivity compared with global screening tools like MoCA, and their use may enhance the detection of cognitive changes in future AF research. These findings align with recent consensus statements supporting systematic evaluation of cognition as a patient-centered outcome in AF research.27,28
Limitations
Several limitations must be acknowledged in the present study. First, the study was conducted at a single center with a relatively small sample size, particularly in the PFA subgroup, which reduces generalizability and limits statistical power. Second, follow-up was limited to 6 months; longer observation is needed to determine whether the cognitive improvements observed are sustained or whether they translate into reduced long-term dementia risk. Third, AF burden was not systematically quantified using continuous rhythm monitoring, which represents a major limitation of the present study. Because binary recurrence measures may incompletely reflect overall arrhythmic exposure and rhythm stability, the association between AF burden reduction and cognitive outcomes could not be directly assessed in the present study. Fourth, although no new infarcts were detected on MRI, its sensitivity for acute ischemic lesions within 48 hours is approximately 50% to 60%, whereas adding diffusion-weighted imaging identifies lesions in almost 94% of stroke patients. In this context, small or early lesions may have been missed, and microstructural or perfusion abnormalities below the resolution of conventional imaging cannot be excluded. Fifth, while MoCA is a widely used global cognitive screening instrument with scores of 26 or higher generally considered within the normal range, the R4Alz battery remains a novel domain-specific tool originally developed for Alzheimer disease and lacks validated cutoffs for clinically meaningful change.6,7 Nevertheless, the observed cognitive changes were mainly associated with improvements in specific domains, suggesting that relatively small changes in the composite R4Alz score may reflect larger changes within individual cognitive subdomains. Finally, although we adjusted analyses for several clinical covariates, residual confounding cannot be excluded. Factors such as medication changes, practice effects from repeated testing or placebo, and expectation-related effects associated with undergoing an interventional procedure may also have influenced results.
Conclusions
In this prospective observational cohort study of 75 patients undergoing CA for AF, ablation was associated with favorable cognitive outcomes compared with a control cohort of patients who did not undergo ablation without evidence of measurable neurological injury on MRI. Although the observed cognitive differences were modest, domain-specific assessment identified subtle changes not fully captured by MoCA. Together, these findings suggest that CA does not adversely affect cognitive function and may be associated with subtle early cognitive benefits in selected domains. Larger studies with longer follow-up are warranted to confirm these findings and to determine their implications for long-term cognitive outcomes.
eTable 1. Baseline and Follow-Up Rhythm Status, AF Recurrence, and Antiarrhythmic Drug Therapy of the Ablation and Control Cohorts
eTable 2. Baseline vs Follow-Up Cognitive Outcomes in the Ablation and Control Cohorts
eTable 3. Exploratory Adjusted Multiple Linear Regression Analyses of R4Alz Subdomain Outcomes at 6 Months
eTable 4. Sensitivity Analyses for Adjusted Associations With 6-Month Cognitive Outcomes
Data Sharing Statement
References
- 1.Zhao M, Jiang C, Lai Y, Wang Y, Li S, He L, et al. Association between atrial fibrillation and domain-specific cognitive decline—insights from the Systolic Blood Pressure Intervention trial. Circ J. 2022;87(1):20-6. doi: 10.1253/circj.CJ-22-0224 [DOI] [PubMed] [Google Scholar]
- 2.Kühne M, Krisai P, Coslovsky M, et al. ; Swiss-AF Investigators . Silent brain infarcts impact on cognitive function in atrial fibrillation. Eur Heart J. 2022;43(22):2127-2135. doi: 10.1093/eurheartj/ehac020 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Van Gelder IC, Rienstra M, Bunting KV, et al. ; ESC Scientific Document Group . 2024 ESC Guidelines for the management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS). Eur Heart J. 2024;45(36):3314-3414. doi: 10.1093/eurheartj/ehae176 [DOI] [PubMed] [Google Scholar]
- 4.Gaita F, Caponi D, Pianelli M, et al. Radiofrequency catheter ablation of atrial fibrillation: a cause of silent thromboembolism? Magnetic resonance imaging assessment of cerebral thromboembolism in patients undergoing ablation of atrial fibrillation. Circulation. 2010;122(17):1667-1673. doi: 10.1161/CIRCULATIONAHA.110.937953 [DOI] [PubMed] [Google Scholar]
- 5.Bodagh N, Yap R, Kotadia I, Sim I, Bhalla A, Somerville P, et al. Impact of catheter ablation versus medical therapy on cognitive function in atrial fibrillation: a systematic review. J Interv Card Electrophysiol. 2022;65(1):271-286. doi: 10.1007/s10840-022-01196-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Nasreddine ZS, Phillips NA, Bédirian V, et al. The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment. J Am Geriatr Soc. 2005;53(4):695-699. doi: 10.1111/j.1532-5415.2005.53221.x [DOI] [PubMed] [Google Scholar]
- 7.Poptsi E, Tsardoulias E, Moraitou D, Symeonidis AL, Tsolaki M. REMEDES for Alzheimer-R4Alz battery: design and development of a new tool of cognitive control assessment for the diagnosis of minor and major neurocognitive disorders. J Alzheimers Dis. 2019;72(3):783-801. doi: 10.3233/JAD-190798 [DOI] [PubMed] [Google Scholar]
- 8.Hertzog MA. Considerations in determining sample size for pilot studies. Res Nurs Health. 2008;31(2):180-191. doi: 10.1002/nur.20247 [DOI] [PubMed] [Google Scholar]
- 9.Team RCR. A Language and Environment for Statistical Computing. R Foundation for Statistical Computing; 2021. [Google Scholar]
- 10.Reinsch N, Füting A, Höwel D, Bell J, Lin Y, Neven K. Cerebral safety after pulsed field ablation for paroxysmal atrial fibrillation. Heart Rhythm. 2022;19(11):1813-1818. doi: 10.1016/j.hrthm.2022.06.018 [DOI] [PubMed] [Google Scholar]
- 11.Reddy VY, Gerstenfeld EP, Natale A, et al. ; ADVENT Investigators . Pulsed field or conventional thermal ablation for paroxysmal atrial fibrillation. N Engl J Med. 2023;389(18):1660-1671. doi: 10.1056/NEJMoa2307291 [DOI] [PubMed] [Google Scholar]
- 12.Chen LY, Lopez FL, Gottesman RF, et al. Atrial fibrillation and cognitive decline-the role of subclinical cerebral infarcts: the atherosclerosis risk in communities study. Stroke. 2014;45(9):2568-2574. doi: 10.1161/STROKEAHA.114.005243 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Kalantarian S, Stern TA, Mansour M, Ruskin JN. Cognitive impairment associated with atrial fibrillation: a meta-analysis. Ann Intern Med. 2013;158(5 Pt 1):338-346. doi: 10.7326/0003-4819-158-5-201303050-00007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Thacker EL, McKnight B, Psaty BM, et al. Atrial fibrillation and cognitive decline: a longitudinal cohort study. Neurology. 2013;81(2):119-125. doi: 10.1212/WNL.0b013e31829a33d1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.de Bruijn RF, Heeringa J, Wolters FJ, et al. Association between atrial fibrillation and dementia in the general population. JAMA Neurol. 2015;72(11):1288-1294. doi: 10.1001/jamaneurol.2015.2161 [DOI] [PubMed] [Google Scholar]
- 16.Alonso A, Arenas de Larriva AP. Atrial fibrillation, cognitive decline and dementia. Eur Cardiol. 2016;11(1):49-53. doi: 10.15420/ecr.2016:13:2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Hu X, Li W, Ren B, Zeng R. Incidence of silent cerebral events detected by MRI in patients with atrial fibrillation undergoing pulsed field ablation vs thermal ablation: a systematic review and network meta-analysis. Heart Rhythm. Published online April 11, 2025. doi: 10.1016/j.hrthm.2025.04.00 [DOI] [PubMed] [Google Scholar]
- 18.Sepehri Shamloo A, Dagres N, Müssigbrodt A, et al. Atrial fibrillation and cognitive impairment: new insights and future directions. Heart Lung Circ. 2020;29(1):69-85. doi: 10.1016/j.hlc.2019.05.185 [DOI] [PubMed] [Google Scholar]
- 19.Varrias D, Saralidze T, Borkowski P, et al. Atrial fibrillation and dementia: pathophysiological mechanisms and clinical implications. Biomolecules. 2024;14(4):455. doi: 10.3390/biom14040455 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Gaita F, Anselmino M. Silent cerebral events during catheter ablation for atrial fibrillation: not yet to be forgotten. JACC Clin Electrophysiol. 2018;4(12):1610-1612. doi: 10.1016/j.jacep.2018.08.023 [DOI] [PubMed] [Google Scholar]
- 21.Glowniak A, Tarkowski A, Janczarek M, Wysokinski A. Silent cerebral infarcts following pulmonary vein isolation with different atrial fibrillation ablation techniques—incidence and risk factors. Arch Med Sci. 2019;18(3):632-638. doi: 10.5114/aoms.2019.85348 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Petersen M, Chevalier C, Naegele FL, et al. Mapping the interplay of atrial fibrillation, brain structure, and cognitive dysfunction. Alzheimers Dement. 2024;20(7):4512-4526. doi: 10.1002/alz.13870 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Atanasova B, Tokmakova M, Dzhambov AM, Chitak R, Atanassova P. Exploring associations between early cognitive impairment and echocardiographic markers in middle-aged patients with atrial fibrillation and cardiometabolic comorbidities: a pilot study. Clin Pract. 2026;16(5):82. doi: 10.3390/clinpract16050082 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Andrade JG, Yao RRJ, Deyell MW, et al. ; CIRCA-DOSE Study Investigators . Clinical assessment of AF pattern is poorly correlated with AF burden and post ablation outcomes: a CIRCA-DOSE sub-study. J Electrocardiol. 2020;60:159-164. doi: 10.1016/j.jelectrocard.2020.03.008 [DOI] [PubMed] [Google Scholar]
- 25.Poole JE, Bahnson TD, Monahan KH, et al. ; CABANA Investigators and ECG Rhythm Core Lab . Recurrence of atrial fibrillation after catheter ablation or antiarrhythmic drug therapy in the CABANA trial. J Am Coll Cardiol. 2020;75(25):3105-3118. doi: 10.1016/j.jacc.2020.04.065 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Tang SC, Liu YB, Lin LY, et al. Association between atrial fibrillation burden and cognitive function in patients with atrial fibrillation. Int J Cardiol. 2023;377:73-78. doi: 10.1016/j.ijcard.2023.01.007 [DOI] [PubMed] [Google Scholar]
- 27.Zhang W, Liang J, Li C, Gao D, Ma Q, Pan Y, et al. Age at diagnosis of atrial fibrillation and incident dementia. JAMA Netw Open. 2023;6(11):e2342744. doi: 10.1001/jamanetworkopen.2023.42744 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Rivard L, Friberg L, Conen D, et al. Atrial fibrillation and dementia: a report from the AF-SCREEN international collaboration. Circulation. 2022;145(5):392-409. doi: 10.1161/CIRCULATIONAHA.121.055018 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eTable 1. Baseline and Follow-Up Rhythm Status, AF Recurrence, and Antiarrhythmic Drug Therapy of the Ablation and Control Cohorts
eTable 2. Baseline vs Follow-Up Cognitive Outcomes in the Ablation and Control Cohorts
eTable 3. Exploratory Adjusted Multiple Linear Regression Analyses of R4Alz Subdomain Outcomes at 6 Months
eTable 4. Sensitivity Analyses for Adjusted Associations With 6-Month Cognitive Outcomes
Data Sharing Statement
