Home-based EEG Neurofeedback for the Treatment of Chronic Pain: A Randomized Controlled Clinical Trial - 18/09/24

Doi : 10.1016/j.jpain.2024.104651

David A. Rice ^⁎^†,^⁎ , Christine Ozolins ^‡, Riya Biswas ^‡, Faisal Almesfer ^‡, Irene Zeng ^§, Ankit Parikh ^‡, Willis Glen Vile ^‡, Usman Rashid ^⁎, Jon Graham ^¶, Michal T. Kluger ^†,^∥,^⁎⁎
^⁎ Health and Rehabilitation Research Institute, Auckland University of Technology, Auckland, Auckland, New Zealand
^† Waitemata Pain Services, Te Whatu Ora - Health New Zealand Waitematā, Auckland, Auckland, New Zealand
^‡ Exsurgo Limited, Auckland, Auckland, New Zealand
^§ Department of Biostatistics, Faculty of Health and Environmental Sciences, Auckland University of Technology, Auckland, Auckland, New Zealand
^¶ PhysioFunction Ltd., Northampton, Northamptonshire, United Kingdom
^∥ Department of Anaesthesiology and Perioperative Medicine, Te Whatu Ora - Health New Zealand Waitematā, Auckland, Auckland, New Zealand
^⁎⁎ Department of Anaesthesiology, Faculty of Medicine and Health Sciences, University of Auckland, Auckland, Auckland, New Zealand

^⁎Address reprint requests to David A. Rice, Health and Rehabilitation Research Institute, Auckland University of Technology, Private Bag 92006, Auckland 1142, New Zealand.Health and Rehabilitation Research Institute, Auckland University of TechnologyPrivate Bag 92006Auckland1142New Zealand

Sous presse. Épreuves corrigées par l'auteur. Disponible en ligne depuis le Wednesday 18 September 2024

Abstract

This parallel, 2-arm, blinded, randomized controlled superiority trial examined whether, when added to usual care, active-electroencephalography neurofeedback (EEG NFB) was safe and more effective than sham control-EEG NFB for chronic pain. In total, 116 participants with chronic pain were randomly assigned (1:1) to usual care plus ≥32 sessions of active-EEG NFB upregulating relative alpha power over C4 or usual care plus ≥32 sessions of sham control-EEG NFB. Per-protocol analyses revealed no significant between-group differences in the primary outcome, Brief Pain Inventory average pain (mean difference [95% confidence interval]: −.04 [−.39 to .31], P = .90), or any secondary outcomes. However, 44% of participants in the active-EEG NFB group and 45% in the control-EEG NFB group reported at least a moderate (≥30%), clinically important improvement in Brief Pain Inventory average pain. The number of treatment-emergent adverse events were similar in both groups (P = .83), and none were serious. Post hoc analyses revealed similar upregulated relative alpha power in both groups during training, with concordant positive rewards delivered to the active-EEG group 100% of the time and the control-EEG group ∼25% of the time, suggesting a partially active sham intervention. When added to usual care, the active-EEG NFB intervention used in this study was not superior to the sham control-EEG NFB intervention. However, a large proportion of participants in both groups reported a clinically important reduction in pain intensity. A partially active sham intervention may have obscured between-group differences. The intervention was free of important side effects, with no safety concerns identified.

Perspectives

This study is the first attempt at an appropriately blinded, randomized, sham-controlled trial of alpha EEG NFB for the treatment of chronic pain. The findings may interest people living with chronic pain, clinicians involved in chronic pain management, and may inform the design of future EEG NFB trials.

Trial Registration

Australian New Zealand Clinical Trials Registry (ANZCTR): ACTRN12621000667819.

Le texte complet de cet article est disponible en PDF.

Highlights

•	This is the first double-blind randomized sham-controlled trial of its kind for chronic pain
•	The active group successfully upregulated resting relative alpha power over the sensorimotor cortex.
•	There were no significant between-group differences in primary or secondary outcomes.
•	44 to 46% of participants in both groups reported moderate (≥30%), clinically important pain relief.
•	A partially active sham control condition may have influenced the findings.