Cognitive impairment in Parkinson’s disease is associated with enhanced modulation of subthalamic beta-oscillations during DBS neurofeedback

Background Deep brain stimulation (DBS) neurofeedback enables Parkinson’s disease (PD) patients to voluntarily control pathological subthalamic beta-oscillations. While neurofeedback has shown promise in improving motor performance, the clinical factors that influence individual modulation capacity remain unclear.Methods We retrospectively analysed 26 PD patients from three previously published deep DBS-neurofeedback cohorts. Participants were classified as cognitively normal (MoCA score ≥26) or cognitively impaired ( < 26) based on their preoperative scores. Neurofeedback performance was quantified as the percentage reduction in subthalamic beta-peak power during the final training round compared to baseline, z-normalised within each cohort. Group comparisons were conducted using t-tests and non-parametric tests. Correlations and stepwise linear regression were employed to examine associations between neurofeedback performance and clinical variables, including MoCA, age, Hoehn and Yahr stage, levodopa equivalent dose, and MDS-UPDRS scores.Results Participants with cognitive impairment (n = 9) showed significantly better neurofeedback performance compared to cognitively normal participants (p = 0.0108). Additionally, a negative correlation was found between MoCA scores and neurofeedback capacity. Stepwise regression revealed that MoCA was the sole significant predictor of performance (, p = 0.0032), while age, motor symptom severity, and medication status did not contribute independently.Conclusions Cognitive impairment in PD is associated with enhanced modulation of subthalamic beta-oscillations during DBS-neurofeedback. This finding challenges the assumption that preserved cognition is necessary for effective neurofeedback, suggesting that compensatory or disease-related network dynamics may underpin modulation capacity. Future studies should investigate mechanistic explanations and assess whether stratifying patients by cognitive status can optimise personalised neuromodulation strategies.