Neurofeedback as an emerging rehabilitation strategy for Parkinson's disease: a narrative review.

Background: Parkinson's disease (PD) is a progressive neurodegenerative disorder characterized by motor and non-motor symptoms associated with dysfunction of cortico-basal ganglia-thalamo-cortical networks. Because current treatments remain largely symptomatic, there is increasing interest in adjunctive rehabilitation strategies capable of targeting disease-relevant neural activity more directly. Neurofeedback (NF), delivered through electroencephalography (EEG), functional magnetic resonance imaging (fMRI), or deep brain stimulation (DBS)-guided platforms, has emerged as a potential self-regulation-based neuromodulatory approach in PD. Objective: This narrative review aimed to summarize the current empirical evidence on neurofeedback in Parkinson's disease, with particular attention to neurophysiological feasibility, reported clinical effects, and methodological and translational challenges. Methods: A narrative review of the literature was conducted using PubMed, Scopus, Web of Science, and Google Scholar. Empirical studies published in English between 2002 and 2026 investigating EEG-based, fMRI-based, or DBS-guided neurofeedback in patients with idiopathic PD were considered. The review qualitatively synthesized findings related to neural target modulation and motor and non-motor outcomes. Results: Eighteen empirical studies were identified, across EEG-based, fMRI-based, and DBS-guided neurofeedback paradigms. EEG-based protocols represented the largest body of evidence and suggested that at least some patients with PD can achieve partial self-regulation of cortical targets such as sensorimotor rhythms, mu/beta activity, and movement-related cortical potentials. However, associated clinical effects were heterogeneous and generally preliminary. fMRI-based studies supported the feasibility of regulating disease-relevant motor regions and connectivity-based targets, including the supplementary motor area and basal ganglia-related networks, but consistent clinical superiority over active control conditions has not been established. DBS-guided neurofeedback provided the strongest pathophysiological specificity, showing that implanted patients can voluntarily modulate subthalamic beta activity, with preliminary evidence of benefit in selected movement-related outcomes. Across modalities, neurophysiological feasibility appeared more consistently supported than robust clinical efficacy. Conclusion: Neurofeedback represents a promising and mechanistically grounded adjunctive rehabilitation strategy for Parkinson's disease. Current evidence indicates that self-regulation of disease-relevant neural signals is feasible in at least a subset of patients, but clinical effectiveness remains uncertain because of small samples, methodological heterogeneity, and limited replication. Future research should prioritize standardized protocols, adequately powered randomized trials, longer follow-up, and better integration of neurofeedback with established pharmacological, rehabilitative, and device-based treatments.