Multivariate pattern analysis reveals resting-state EEG biomarkers in fibromyalgia

Fibromyalgia (FM) involves widespread musculoskeletal pain and hypersensitivity, often accompanied by neurological, cognitive, and affective disturbances. Resting-state electroencephalography (EEG) studies have revealed abnormal brain activity in affected individuals, with anxiety and symptom duration potentially exacerbating these alterations. This study applied multivariate pattern analysis to differentiate resting-state EEG signals between FM patients and healthy controls across frequency bands associated with pain processing, while incorporating state and trait anxiety scores. It also examined differences between patients with short- and long-duration symptoms and identified the most relevant scalp regions contributing to the models. Fifty-one female participants (25 FM patients, 26 controls; aged 35-65) were included. Patients were classified into short-term (12) and long-term (13) groups. EEG was recorded and power spectral density values were extracted and normalized to be used to train machine learning classifiers. Anxiety scores were included in the patient-control analysis. Haufe transformed weights were computed to determine key scalp contributions. The models successfully distinguished patients from controls and between FM subgroups, with accuracies exceeding 0.70 across all frequency bands. Including anxiety scores substantially improved classification, with accuracies reaching 0.99. These findings highlight significant differences in resting-state EEG activity between patients and controls, and between FM subgroups, underscoring the relevance of emotional and neural factors. Frequency-specific alterations in pain-related scalp regions support the view of disrupted pain processing in FM. Resting-state EEG, combined with multivariate pattern analysis, may support the development of biomarkers to enhance diagnosis and guide treatment strategies in clinical settings.