Resting-State EEG Functional Connectivity and Independent Correlates of Cognitive Impairment in Patients with Atrial Fibrillation

Min-Qiang Bao,1,* Dan Xiao,2,* Xiang-Shun She,1 An-Feng Yin,1 Xiao-Ning Sheng,1 Shuang-Shuang Chen,1 Dandan Chu,1 Guo-Liang Gao,3 Yi-Nong Chen,1 Yu Wang21Department of Neurology, The Affiliated Xuancheng Hospital of Wannan Medical University, Xuancheng, Anhui, People’s Republic of China; 2Department of Neurology, The First Affiliated Hospital of Anhui Medical University, Hefei, Anhui, People’s Republic of China; 3Department of Electrocardiography, The Affiliated Xuancheng Hospital of Wannan Medical University, Xuancheng, Anhui, People’s Republic of China*These authors contributed equally to this workCorrespondence: Yu Wang, Department of Neurology, The First Affiliated Hospital of Anhui Medical University, Hefei, Anhui, People’s Republic of China, Email yfy126330@fy.ahmu.edu.cn Yi-Nong Chen, Department of Neurology, The Affiliated Xuancheng Hospital of Wannan Medical University, Xuancheng, Anhui, People’s Republic of China, Email 2chenyinon@163.comObjective: To investigate alterations in resting-state electroencephalography (rsEEG) functional connectivity (FC) in patients with atrial fibrillation (AF) and cognitive impairment, examine its independent association with cognitive dysfunction, and evaluate the potential neurophysiological biomarker value of rsEEG FC in AF-related cognitive impairment.Methods: A total of 141 participants were divided into three groups: controls without AF or cognitive impairment (HC, n = 41), patients with AF without cognitive impairment (PT-WoCI, n = 37), and patients with AF and cognitive impairment (PT-CI, n = 63). All participants underwent rsEEG and cognitive assessment using the Montreal Cognitive Assessment. EEG metrics, including power spectral density (PSD), FC, phase–amplitude coupling (PAC), and sample entropy (EnSA), were derived from rsEEG data. Candidate EEG features were further screened using LASSO regression. Multivariable logistic regression models were constructed to assess independent associations between rsEEG metrics and cognitive impairment. Receiver operating characteristic (ROC) curves were used to evaluate model discriminatory performance.Results: Compared with HC and PT-WoCI groups, the PT-CI group showed significantly reduced alpha-band FC. In multivariable logistic regression analysis, alpha-band wPLI remained independently associated with cognitive impairment in patients with AF (OR = 0.44, 95% CI: 0.24– 0.72, P = 0.002). The model incorporating clinical variables and alpha-band wPLI showed good discriminatory ability (area under the ROC curve: 0.814 [95% CI: 0.729– 0.900]).Conclusion: Impaired alpha-band FC is associated with cognitive impairment in patients with AF. Alpha-band wPLI showed an independent association with cognitive status, supporting the potential clinical utility of rsEEG FC as an objective marker of AF-related cognitive impairment.Keywords: atrial fibrillation, resting-state electroencephalography, functional connectivity, alpha band, cognitive impairment