BACKGROUND AND PURPOSE:Human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CM) have gained interest as a pharmacological model but their immaturity leads to uncertainty regarding translation. We studied the key player in maintaining ionic homeostasis, the Na+/K+-ATPase (NKA), in hiPSC-CM. EXPERIMENTAL APPROACH:Atrial and ventricular engineered heart tissues (EHTs) were produced from hiPSC-CMs. For comparison, adult atrial and ventricular tissues were obtained from patients undergoing open heart surgery. We measured NKA gene expression, NKA pump currents (INKA) and ouabain effects on action potentials and contractility. Computational modelling was used to further investigate the direct and indirect impacts of NKA inhibition. KEY RESULTS:The mRNA abundance of the major NKA isoforms was higher in ventricular than atrial cardiomyocytes (adult and hiPSC-CM). Consistently, INKA was also higher in ventricular than in atrial hiPSC-CM and higher in ventricular than in atrial adult cardiomyocytes. Ouabain potency to block INKA did not differ between hiPSC-derived and adult cardiomyocytes. Ouabain shortened plateau phase in EHT and adult tissue. However, lower ouabain concentrations depolarized diastolic potential and depressed force more in EHTs, demonstrating a higher integrated sensitivity to NKA inhibition. Computational modelling indicated that weaker IK1 increases the susceptibility to depolarisation by NKA block. CONCLUSIONS AND IMPLICATIONS:HiPSC-EHTs express NKA with biophysical characteristics and ouabain-sensitivity that are not different from human tissue. Action potential duration responses in EHT reproduces chamber-specific response pattern as seen in human heart. The higher sensitivity of EHTs to depolarize under ouabain-induced block of NKA needs to be considered when employing hiPSC-CM in drug research.

AIMS:Atrial fibrillation (AF) is conventionally classified as paroxysmal or persistent. AF burden might better reflect response to rhythm-control interventions. This study evaluated the association between estimated pre-procedural time spent in AF and recurrences after AF ablation, compared with conventional AF classification. METHODS AND RESULTS:In patients scheduled for AF ablation, clinical characteristics were collected before ablation, and 60-second single-lead ECGs were recorded three times daily (with additional symptom-triggered recordings) for four weeks. Pre-procedural time spent in AF was estimated as the number of days with detected AF divided by total monitoring days. The primary endpoint was AF recurrence between 3 and 12 months post-ablation. Of 302 patients (mean age 64 ± 9 years; 33% female), 201 (67%) had paroxysmal AF and 101 (33%) had persistent AF. After 12 months, recurrence rates were higher in persistent than in paroxysmal AF (37.6% vs. 24.4%, P & 0.01) and in patients with a higher (>32%) vs. lower (≤32%) percentage of time spent in AF (36.4% vs. 24.0%, P < 0.01). Notably, paroxysmal AF patients with a higher percentage of time in AF had recurrence rates comparable to those of patients with persistent AF. Pre-procedural percentage of time in AF independently predicted arrhythmia recurrence at 12-month follow-up (HR: 1.06; 95% CI: 1.0-1.1; P & 0.025). CONCLUSION:A higher pre-procedural percentage of time spent in AF derived from 60-second single-lead ECGs is independently associated with atrial arrhythmia recurrence after ablation. Assessing the percentage of pre-procedural time spent in AF might help identify paroxysmal AF patients with a high AF recurrence risk in future studies.

Abstract Background The triglyceride-glucose (TyG) index and its obesity-related derivatives have emerged as surrogate markers of insulin resistance associated with cardiovascular outcomes. However, whether these indicators influence atrial fibrillation (AF) risk through cardiac structural and functional remodeling remains unclear. Methods This is a post-hoc analysis of a prospective cohort study of 32,500 UK Biobank participants free of AF who underwent baseline cardiac magnetic resonance (CMR) imaging. The TyG index and its obesity-related derivatives (TyG-body mass index (BMI), TyG-waist circumference (WC), and TyG-waist-to-height ratio (WHtR)) were calculated at baseline. Multivariable Cox regression models were used to assess associations with incident AF, and mediation analyses quantified the contribution of CMR-derived cardiac parameters to these associations. Results Over a median follow-up of 13.61 years, 1,288 incident AF cases occurred. The TyG index alone showed no independent association with AF risk. In contrast, all TyG obesity-related derivatives were significantly associated with incident AF, with TyG-WC demonstrating the strongest association (HR = 1.245, 95% CI 1.169–1.325), followed by TyG-BMI (HR = 1.223, 95% CI 1.158–1.293) and TyG-WHtR (HR = 1.190, 95% CI 1.122–1.262). Mediation analyses identified left atrial maximum volume (LAVmax) as the predominant mediator, accounting for 70.63%, 47.83%, and 40.79% of the associations for TyG-BMI, TyG-WHtR, and TyG-WC, respectively. Conclusions TyG obesity-related derivatives, particularly TyG-WC, were independently associated with incident AF. Cardiac structural remodeling, especially LA enlargement, appeared to be a key mediating pathway. These findings support the importance of early metabolic intervention to prevent adverse atrial remodeling and reduce AF susceptibility. Graphical Abstract

Atrial fibrillation (AF) increases energy demand in atrial myocytes, yet the mitochondrial mechanisms underlying this stress remain poorly defined. Using previously published proteomic data from left atrial tissue of AF and sham-operated goats, we performed organelle-specific bioinformatic analyses of the mitochondrial fraction. Over-representation and consensus pathway analyses consistently highlighted enrichment of oxidative phosphorylation (OXPHOS) subunits. Gene set enrichment and network analyses implicated Heat Shock Protein Family A Member 9 (HSPA9) as a potentially central regulatory hub coordinating the dysregulation of Complex I and III subunits, with 69% of regulatory relationships showing pathway concordance. These results indicate a coordinated, system-wide mitochondrial adaptation in AF, integrating energy production, proteostasis, and respiratory chain regulation.

Atrial fibrillation (AF) is the most common cardiac arrhythmia, linked to greater risk of heart failure, stroke and death. Inflammation has been connected to AF emergence, however mechanisms of inflammation-caused AF remain thus far elusive, leading to a lack of mechanism-based treatments. An isogenic, 3D tissue model containing hiPSC-derived atrial-like cardiomyocytes (aCM), cardiac fibroblasts (cfb), and cardiac macrophages was engineered using custom injection-molded pillar devices. Electrophysiological changes were examined via sharp electrode recordings, calcium imaging, and multi-electrode assays. Gene function was interrogated using siRNA knock-down, lentiviral overexpression, and pharmacological modulation. In silico tissue and whole-heart models validated findings under simulated stress and heterogeneous conditions. Activation of M1 macrophages led to a 50% reduction in contraction amplitude, action potential spike amplitude (aCM+cfb+M1: 61.3 mV ±13.9 vs control: 71.6 mV ±14.5, p < 0.01) and increased beat irregularity (M1: 150.7% ± 388.9 vs control, p < 0.001). Calcium transient amplitude was reduced (12.3 a.u. ± 14.7, p < 0.05) and upstroke velocity slowed. SCN5A knock-down reduced contraction amplitude (-51.9% ± 37.2, p < 0.01) without inducing arrhythmias, whereas combined GJA5 and ATP1A1 knock-down induced significant irregularity (403% ± 371.3, p < 0.001), increased conduction heterogeneity (+18%), and reduced velocity (-52.4%). In silico modeling confirmed that paired 50% downregulation of sodium-potassium pump and tissue conductivity induced AF under tachycardia even without ectopic activity. This work reveals a novel, inflammation-driven mechanism for AF initiation. Combined downregulation of GJA5 (connexin 40) and ATP1A1 (NaK ATPase) disrupted intercellular connectivity and ion flux, establishing a substrate for arrhythmogenesis. These results were robust across in vitro, genetic/pharmacological, and in silico models, defining new avenues for translational intervention.

Aims Symptom severity in atrial fibrillation (AF) influences healthcare use, but conventional assessments lack temporal precision. We evaluated the feasibility of smartphone-based symptom severity assessment after AF catheter ablation and explored predictors and variability of symptom severity.Methods and results In this analysis of the ISOLATION study, 140 patients used a photoplethysmography-supported smartphone application to record heart rhythm, symptoms, and symptom severity three times daily and when symptomatic for 7 days at 3, 6, and 12 months after ablation. Symptom severity was classified using the modified European Heart Rhythm Association scale. A total of 8575 recordings were analysed, of which 11% showed AF. Median adherence was 67%. Symptom severity was higher during AF, in women, and with higher heart rate. Nearly half of AF recordings were asymptomatic.Conclusion Smartphone-based symptom severity assessment after AF ablation is feasible and reveals substantial individual variability beyond rhythm status alone.

BACKGROUND AND AIMS:Atrial cardiomyopathy (AtCM) is associated with atrial fibrillation (AF). While histological studies describe underlying mechanisms of AtCM, these insights remain disconnected from clinical AtCM. This study determines the histological basis of non-invasive AtCM markers and assess their prognostic value in patients undergoing cardiac surgery. METHODS AND RESULTS:Left and right (LA/RA) atrial tissue samples were obtained in patients undergoing cardiac surgery (n = 136) in the RACE V Tissue Bank study. Pre-operative rhythm history, digital ECG, transthoracic echocardiography (TTE) and biomarker levels were used to group patients in AtCM stages: AF: history of paroxysmal, persistent or permanent AF, Severe (LAVi>50 mL/m2 or LAEF<35%), Moderate (LAVi 34-50 mL/m2, or LAEF 35-50% AND NT-proBNP >250 pg/mL), Mild (no AF, LAVi≤34 mL/m2 AND P-terminal Force V1 > 5 mV*ms OR P wave duration > 120 ms), and no AtCM. Tissue was stained (WGA/CD31/Vimentin) to quantify fibrosis, fibroblast density, myocyte diameter and vascularization. Post-operative rhythm was monitored continuously (2.5 years) using implantable loop recorders. Patients with severe AtCM had extended endomysial fibrosis (LA: +0.82µm, pFDR,LA < 0.001; RA: +0.64µm, pFDR,RA = 0.010) and larger left atrial cardiomyocytes (LA: +0.92µm, pFDR,LA = 0.026). Moderate AtCM patients showed LA myocyte hypertrophy (LA: +1.35µm, pFDR,LA = 0.007). Mild AtCM patients had similar histological features to patients without AtCM. Moderate (β = 3.56, pFDR = 0.045) and severe (β = 3.74, pFDR = 0.034) AtCM patients had a higher AF burden late after cardiac surgery. CONCLUSION:Clinical AtCM markers reflect pro-fibrotic and pro-hypertrophic remodelling in the moderate to severe AtCM stages and are associated with a higher AF burden late after surgery.

Background: Atrial fibrillation (AF) is characterized by a heterogeneous presentation of symptoms. AF ablation reduces symptom burden. However, persistent symptoms following AF ablation are common independently of AF recurrence. Objectives: This study sought to perform a cluster analysis to identify clinically relevant AF subphenotypes based on persistent symptoms following AF ablation and evaluate their associations with clinical characteristics and AF recurrence. Methods: Patients were instructed to perform smartphone app-based simultaneous symptom and photoplethysmography heart rhythm monitoring 3 times daily for 1 week at the 3-month follow-up after AF ablation. A two-step cluster analysis including 7 categorical symptoms variables was performed in symptomatic patients. Results: In total, half of all patients (n = 313 of 614 [51%]) reported symptoms. Five symptom clusters were identified: nonspecified symptoms (n = 52 [17%]), AF with sparse symptoms (n = 93 [30%]), palpitations (n = 47 [15%]), fatigue with comorbidities (n = 63 [20%]), and sinus rhythm with severe symptoms (n = 58 [19%]). Frequency (P < 0.001) and pattern (P < 0.001) of symptom reporting as well as AF recurrence (P < 0.001), AF load (P < 0.001), AF pattern (P = 0.002 and P = 0.005), and symptom-rhythm correlation (P < 0.001) differed between clusters. Furthermore, age (P < 0.01), N-terminal pro-B-type natriuretic peptide levels (P < 0.01), CHA(2)DS(2)-VA score (P < 0.001), and left atrial volume index (P = 0.01) differed between clusters. Conclusions: Half of all patients report symptoms after AF ablation. Using cluster analysis, 5 symptom-based AF subphenotypes were identified, each with distinct clinical characteristics, biomarker profiles, AF recurrence, AF pattern, AF and symptom burden, and symptom-rhythm correlation. Symptom clusters empowered by digital health may facilitate individualized AF management strategies following AF ablation.

Background and aims:Cryoballoon ablation of atrial fibrillation alters a significant amount of vital myocardium contributing to its electrical activity. However, the effect of pulmonary vein isolation (PVI) on the P-wave is unclear. The study aimed to quantify P-wave changes following individual PVI using signal-averaged ECG analysis. Materials and methods:Unfiltered high-resolution (2,000 samples/second) ECGs were obtained for 5 min pre- and post-isolation of each pulmonary vein in 69 participants using 12 standard and 10 additional leads optimized for atrial electrical activity detection. Signal-averaged ECGs were computed and analyzed using custom software in Matlab. This study was registered at ClinicalTrials.gov (Identifier: NCT06061120). Results:Significant changes in ECG parameters were found. Following complete PVI, P-wave terminal force measured in lead V1 (3.08 ± 1.99 vs. 2.63 ± 1.76 mV*ms, p = 0.003), as well as P-wave complexity (2.05 ± 0.65 vs. 1.79 ± 0.51, p < 0.001) and sample entropy (0.210 ± 0.039 vs. 0.187 ± 0.028 au, p < 0.001), both averaged across all leads, decreased significantly. Ablation of the left pulmonary veins was associated with lower values in P-wave terminal force (3.07 ± 2.05 vs. 2.84 ± 1.88 mV*ms, p = 0.003), P-wave complexity (2.02 ± 0.64 vs. 1.90 ± 0.57, p = 0.003), and sample entropy (0.210 ± 0.039 vs. 0.197 ± 0.031 au, p < 0.001) after ablation. For ablation of the right pulmonary veins, a decline was observed in sample entropy (0.196 ± 0.031 vs. 0.190 ± 0.027 au, p = 0.046). Conclusion:The observed changes in signal-averaged P-wave parameters during cryoballoon PVI may indicate acute intraprocedural effects on atrial electrophysiology. These exploratory findings suggest that high-resolution, non-invasive ECG recordings are capable to detect stepwise changes of electrical atrial activity, offering new perspectives for intraprocedural assessment of ablation effects.

Atrial fibrillation (AF) is associated with an increased risk of stroke and hypercoagulability. Coagulation factors mediate remodeling processes via protease-activated receptors (PARs) in various organs.We evaluated whether inhibition of factor Xa (FXa) via rivaroxaban protects against atrial structural remodeling in goats with persistent AF and explored FXa and thrombin hypertrophic effect on human iPSC-derived cardiomyocytes (hiPSC-CMs).Three groups of goats were tested: CTRL AF (control AF, n = 10), RIVA AF (rivaroxaban treatment during AF, n = 11), and SHAM (no AF, n = 10). Pacing-induced AF was maintained for 16 weeks. AF stability, hemodynamics, and AF complexity were assessed. Atrial samples were collected for histological and gene expression analyses. hiPSC-CM were stimulated with PAR-1 agonist TRAP14, FXa, or thrombin with and without their inhibitors. Pro-hypertrophic and pro-inflammatory gene expression was assessed by qRT-PCR after 24 hours.Rivaroxaban inhibited thrombin generation in RIVA AF goats (baseline: 249 ± 42 nM vs. final: 69 ± 33 nM). Sixteen weeks of AF induced atrial myocyte hypertrophy in CTRL AF (13.5 µm [95% CI: 12.9, 14.0] vs. SHAM: 12.5 µm [95% CI: 12.0, 13.0]) and pro-hypertrophic (NPPA: fourfold; NPPB: 22-fold) and pro-fibrotic (COL1A1: threefold) gene expression. Rivaroxaban fully prevented hypertrophy (12.2 µm [95% CI: 11.7, 12.7]) and downregulated inflammatory signaling without altering hemodynamics and AF stability. In hiPSC-CM, thrombin and TRAP14 induced overexpression of the pro-hypertrophic genes NPPA and NPPB. The PAR1 antagonist, SCH79797, prevented thrombin-induced NPPA and NPPB upregulation.Prolonged rivaroxaban treatment reduces thrombin generation, preventing AF-induced atrial myocyte hypertrophy through inhibition of PAR-1 signaling.

Atrial fibrillation (AF) is both influenced by and contributes to atrial structural remodelling, including atrial enlargement and fibrosis. In this work, we aimed to understand how differences in atrial anatomy affect AF inducibility and dynamics, both without and with fibrosis, through in silico models. Atrial wall anatomies from late gadolinium-enhanced magnetic resonance imaging (LGE-MRI) were used to generate personalised models. Detailed intra- and inter-atrial structures (pectinate muscles, Bachmann’s bundle, fossa ovalis, coronary sinus, and fibre orientations) were mapped onto patient-specific atrial anatomies through a novel strategy using universal atrial coordinates and a highly detailed reference model. Patient-specific endomysial fibrosis was incorporated based on LGE-MRI. We quantified AF and macro-reentrant atrial tachycardia (MRAT) initiation rates in models with and without fibrosis, comparing reentry dynamics using renewal theory. We generated personalised models of 10 patients, all but one with low-fibrotic content (Utah stages ≤ 2). Anatomical variability did not significantly affect AF initiation rates in both non-fibrotic (54.0 ± 13.7

BACKGROUND AND AIMS:Atrial cardiomyopathy (AtCM) is increasingly recognized as an important substrate for atrial fibrillation (AF). This study aimed to examine potential markers and risk factors of AtCM, and associations with incident AF, heart failure (HF), and stroke. METHODS:Individuals from the UK Biobank with cardiac magnetic resonance imaging and electrocardiographic information were included. Atrial cardiomyopathy markers included left atrial dilation, left atrial mechanical dysfunction, P-wave prolongation, and abnormal P-wave terminal force. Risk factors for AtCM were assessed using logistic regressions. Incident AF, HF, and stroke according to AtCM markers were assessed in multivariable Cox-regression and cumulative incidence models. AF risk according to AtCM markers, clinical and genetic risk factors was evaluated by integrating the HARMS2-AF score and a polygenic risk score for AF. We used net reclassification improvement (NRI) to evaluate reclassification of risk when considering AtCM markers. RESULTS:Among 26 467 individuals, 4145 (15.7%) had ≥1 marker and 619 (2.3%) had ≥2 markers of AtCM. Age, coronary artery disease, and hypertension were consistently associated with AtCM. Having one AtCM marker conferred a hazard ratio (HR) for AF of 1.88 [95% confidence interval (CI): 1.54-2.31; P < .001], with higher rates observed in individuals with ≥2 markers (HR: 4.59; 95% CI: 3.52-5.99; P < .001). Addition of AtCM markers was associated with an NRI of 13.7% (95% CI: 9.2%-18.3%). Integration of clinical and genetic risk factors indicated an additive effect on AF rates. Having ≥2 markers associated with HF (HR: 3.08, 95% CI: 2.03-4.66, P < .001), and stroke (HR: 3.07, 95% CI: 1.78-5.28, P < .001). CONCLUSIONS:One in seven individuals had at least one marker of AtCM. Atrial cardiomyopathy markers were associated with AF, HF, and stroke, supporting AtCM as a common substrate for all three outcomes.

AIMS:Early rhythm control reduces cardiovascular events in patients with atrial fibrillation (AF) and cardiovascular comorbidities. Whether this effect is consistent across severities of different AF-associated conditions is not known. METHODS:In the EAST-AFNET 4 biomolecule study (n = 1586; median age 70 years; 45% women), 14 circulating biomarkers reflecting inflammation, fibrosis, ageing, cardiac strain, and myocardial damage were quantified. Log-transformed, winsorized biomarkers were analysed in quintiles to detect non-linear associations and as continuous parameters. Cox proportional hazards models evaluated associations with the primary outcome (cardiovascular death, stroke, or unplanned hospitalization for heart failure or acute coronary syndrome) and the safety outcome (death, stroke, or major bleeding) of the trial, including interaction terms for biomarkers and randomized treatment [early rhythm control (ERC) vs. usual care (UC)]. Two independent cohorts, BBC-AF resembling UC and a TRUST snapshot resembling ERC were combined for external replication. All tests were exploratory. RESULTS:ERC reduced the primary outcome consistently across biomarker concentrations. No signal for treatment interaction was observed for 13 of 14 biomarkers for the primary outcome. For BMP10, a nominal interaction was observed in categorical analyses, suggesting an attenuated treatment effect of ERC in patients in the lowest BMP10 quintile. No interaction was detected for continuous parameters. Similar patterns were observed in exploratory analyses of the replication cohort. CONCLUSION:ERC therapy is effective across patients with different severities of conditions associated with AF as quantified by biomolecule concentrations. A potentially attenuated effect of ERC with low BMP10 concentrations warrants further research.This analysis is part of the EAST-AFNET 4 trial biomolecule study, the original trial has been registered under the official title: Early Therapy of Atrial Fibrillation for Stroke Prevention Trial (EAST) (EudraCT 2010-021258-20) at https://clinicaltrials.gov/study/NCT01288352.

BACKGROUND:Atrial fibrillation (AF) may progress from paroxysmal AF (PAF) to more persistent forms, but the underlaying mechanisms are not well understood. The aim of this study was to assess the association between atherosclerosis and AF progression in patients with PAF. METHODS:In this substudy of RACE V, 612 patients with PAF underwent extensive phenotyping at baseline and continuous rhythm monitoring. The association between atherosclerosis and AF progression was investigated. RESULTS:The median age was 64 (57-70) years, 257 (42%) were women, and the median CHA2DS2-VA score was 2 (1-3). At baseline, 395 (65%) patients had atherosclerosis, defined by carotid/coronary imaging and/or history of vascular disease. Patients with atherosclerosis were older, had higher waist circumference, more hypertension, and lower eGFR than patients with no atherosclerosis. During a median of 3.4 (2.8-3.7) years follow-up, 108 (18%) patients had AF progression. The presence of atherosclerosis was associated with increased progression (21% vs. 12%; p = .004). In univariable analyses, atherosclerosis was a determinant of AF progression (OR: 2.04; 95% CI: 1.28-3.37; p = .004), and the association persisted following adjustment for established risk factors (OR: 2.23; 95% CI: 1.10-4.89; p = .034). CONCLUSIONS:In patients with paroxysmal AF, 65% of patients had atherosclerosis. Atherosclerosis was a determinant of AF progression after adjustment for established risk factors and comorbidities, suggesting that vascular disease may contribute directly to atrial remodelling and arrhythmia persistence.

Background Postoperative atrial fibrillation (POAF) is common after cardiac surgery. AF in non-surgical patients exhibits distinct circadian patterns (adrenergic/daytime and vagal/nocturnal), but such patterns for new-onset POAF and for postoperative AF recurrence remain undefined. Objective To characterize circadian AF patterns and relate chronotypes to clinical, biochemical, histological, and outcome features. Methods Patients undergoing first cardiac surgery, with and without preoperative AF, were continuously monitored for 2.5 years with an implantable loop recorder. Early-AF was defined as AF within 30 days. Patients with sustained AF (≥20 episodes) were chronotyped via circular statistics and local regression of episode duration and ventricular rate (Vrate), and classified as adrenergic, vagal, mixed, or non-rhythmic. Preoperative imaging, ECG, biomarkers, and paired atrial histology were compared across chronotypes. Results Of 98 patients (71 without and 27 with preoperative AF), 53 (54.1%) developed AF after surgery (3,607 episodes). Patients with prior AF showed a morning onset peak, whereas new-onset POAF showed a uniform 24-hour distribution. Of 26 patients with sustained AF, 20 were rhythmic (9 adrenergic, 7 vagal, 4 mixed) and 6 non-rhythmic; 27 had incidental AF (<20 episodes). Rhythmic AF correlated with larger atrial volumes, prolonged PR intervals, higher BMP10, and greater right-atrial endomysial fibrosis and asymmetry, most pronounced in the mixed chronotype. Adrenergic AF was associated with more arrhythmia-related emergency visits and cardioversions. Conclusion AF after cardiac surgery comprises distinct circadian subtypes. New-onset POAF and AF recurrence display different circadian patterns, and individual chronotypes are associated with varying degrees of atrial remodeling and clinical outcomes.

Abstract Background Ambient air pollution has been linked to atrial fibrillation (AF), yet the underlying metabolic mechanisms remain poorly understood. Methods We analyzed 227,324 UK Biobank participants without baseline AF. We constructed an air pollution score by aggregating all four pollutants (PM2.5, PM10, NO2, NOx). Nuclear magnetic resonance metabolomics identified a pollution-related metabolic signature through elastic net regression. Associations between air pollutants, the metabolic signature and AF were analyzed using Cox models. Mediation analysis was employed to examine the role of the metabolic signature in the association between air pollutants and AF. Results During follow-up, 16,235 participants (7.14%) developed AF. We identified 65-metabolite signature significantly associated with air pollution, predominantly comprising lipoprotein lipid concentrations (32.31%), lipoprotein subclasses (15.38%), fatty acids (13.85%), and amino acids (12.31%). Each standard deviation increase in this metabolic signature was associated with 18% higher AF risk (HR = 1.18, 95%CI:1.03–1.35). The metabolic profile mediated 15.45% of the relationship between air pollution and AF, with lipoprotein parameters showing the strongest mediation effects. Conclusion Air pollution-related metabolic signature was independently associated with AF risk and mediated a significant portion of pollution’s arrhythmogenic effects. These findings provide novel insights into biological mechanisms linking environmental exposures to AF. Graphical abstract