New treatment strategy for chronic low back pain with alpha wave neurofeedback

Research Participants

Table 1 shows the attributes of research collaborators. The present study targeted 97 patients with chronic low back pain but no surgical history of low back pain who were referred to our institution’s orthopedic/anaesthesiology department and pain center for low back pain after April 2020. These patients were also recommended to receive CBT and PT since they were diagnosed as no primary organic cause of low back pain based on MRI and neurological symptoms by four spine surgeons. Patients were also resistant to standard orthopedic treatments such as drugs (i.e. NSAIDs, opioids, gabapentinoids, and antidepressants) and various block injections. For the registered cases, senior physicians confirmed that the MRI results were inconsistent with the patients’ symptoms at a Pain Center conference. The application of CBT and PT was assessed with the Brief Rating Scale for Psychiatric Problems in Orthopedic Patients (BS-POP)15 and Locomo 2516. Standard orthopedic treatments were continued during the intervention trial, but there was no change in oral medication.

Table 1 Demographic characteristics of study participants.

The study was conducted in accordance with the Declaration of Helsinki on Ethical Principles for Medical Research Involving Human Subjects. The research protocol was approved by the Chiba University Ethics Committee and all reviews were conducted in accordance with these guidelines and regulations. We provided a detailed explanation of the study to all patients and received written informed consent before starting the study.

Research design, neurofeedback and EEG analysis

Figure 1 shows a schematic of the neurofeedback intervention and training schedule in this study and the neurofeedback system. We performed a prospective longitudinal study to assess treatment effects in each group using EEG and psychosocial factors as indicators. A researcher who was unaware of the status of each collaborator implemented a fully randomized design through the calculation of random numbers with patient identifiers. After randomization, our groups included 20 cases for the control group, 18 for the CBT (TCC) group, 13 for the exercise (PT) group, 20 for the neurofeedback training (NFT) group, 16 for the CBT+NFT (CBT − NFT) and 10 cases for the group PT+NFT (PT−NFT).

Figure 1

A diagram of the neurofeedback intervention and training schedule in this study and the neurofeedback system.

Each group visited our institution once a week and received therapeutic intervention and neurofeedback for a total of eight sessions over a 2 month period. The NFT group also visited our institution during the first and eighth weeks and conducted training for about 10 minutes three times a day at home. Patients in the CBT-NFT and PT-NFT groups received a device and performed neurofeedback training for approximately 10 minutes three times a day at home, in addition to measurement at the time of consultation.

To perform the measurements and neurofeedback seamlessly without any delay to the therapeutic intervention, we measured the EEG with a portable electroencephalograph and then performed real-time neurofeedback via a smartphone app, ALPHA SWITCH ver. 1.3.1 (Mediaseek Inc., Tokyo, Japan,, available on the App Store). Since the application was developed for alpha wave auditory neurofeedback while listening to music and adopted a system to perform with eyes closed, it is possible to prevent EEG noise contamination caused by the movement of eyes, as this is a major problem in EEG measurement.

EEG measurement and neurofeedback were performed with eyes closed and at rest in a quiet room. Recording and analysis for EEG and real-time feedback of analysis results were performed with the default functions of ALPHASWITCH as follows. Prior to the feedback session, a 30 second calibration was implemented to measure the baseline. After the voltage data of four electrodes was received from Muse2 256 times/s, the voltage information was stored in the fixed-length circular buffer. We applied the DC offset to the time-series voltage signals (768 gratings) and then calculated the signal’s spectrum tilt and the cumulative noise percentage. To generate power for the alpha frequency band, we performed a Hilbert transformation by applying a bandpass filter from 8 to 13 Hz, then performed a logarithmic transformation for the mean after calculating the magnitude of 768 arrays. For values ​​after transformation, outliers were rejected by the Smirnov-Grubbs test. We also calculated the mean and standard deviation of the rejected amplitude values ​​to obtain the amplitude of the alpha wave (alpha power) at the time of calibration.

A three-minute feedback session was conducted after the calibration session, and the duration of the feedback session ranged from 3 to 30 min17,18,19. Thus, we performed the measurement for the shortest time of 3 min to avoid sleeping during the feedback session. After processing the signal similarly to the calibration session and performing a logarithmic transformation on the average of the alpha wave amplitude, we calculated and set the Z-score as normalized alpha power (nAP) using the mean of the amplitude, the mean of the calibration amplitude and the standard deviation after transformation. nAP is an index of magnitude by which alpha can be increased by neurofeedback relative to calibration. We also performed signal processing and analysis using MATLAB ver. 9.10.0 (The Mathworks, Natick, MA).

Feedback was delivered to each subject aurally through headphones while listening to healing music, “Sunbeams”. In the feedback session, white noise was overlaid such that noise level was inversely correlated with nAP. Along with the sigmoid function, the white noise volume was set to zero and maximum when nAP was 2 and −−2, respectively. Peak music and white noise volume levels were around 60–70 and 60 dB, respectively. Participants were asked to minimize the noise level by increasing the alpha power as much as possible. Participants were asked to sit down, close their eyes and meditate, for example creating a relaxing image at the start of the neurofeedback session.

EEG device

EEG was measured by Muse2 (InteraXon Inc., Toronto, Canada,, a wearable headband-type EEG device that can be strapped to the forehead with the ends of the band over both ears. Although EEG can be easily measured without any special pretreatment on the scalp or forehead, EEG paste was applied for accurate measurement in the present study. Muse2 has four active electrodes and one reference electrode; two active silver chloride electrodes are located on both sides of the forehead, and another two active electrodes with conductive silicone rubber are located on the two dorsal sides of the headset to prevent artifacts caused by eye movements, while A reference electrode is located between the two active electrodes on the forehead. Muse2 is based on the international 10–10 system, the electrodes are arranged in four locations: TP9, AF7, AF8 and TP10. The sample rate was fixed at 256 Hz and the recorded data was immediately transferred to a tablet (iPad, Apple Inc., San Francisco, CA, USA) via Bluetooth.

Review Items

We performed the following examination items for all cases in terms of pain and treatment satisfaction at the time of consultation: Visual Analog Scale (VAS); the Japanese version of the Oswestry Disability Index (ODI)20.21; the Japanese version of the Hospital Anxiety and Depression Scale (HADS)22.23; the Japanese version of the Pain Catastrophic Scale (PCS)24.25; and the Pain Self-Efficacy Questionnaire (PSEQ)26.27.

statistical analyzes

IBM SPSS27 Statistics® (IBM, Armonk, NY) was used to analyze the results. The purpose of this study is to investigate whether the therapeutic effect can be enhanced by combining alpha wave neurofeedback training with CBT and PT according to the previous study. It was also pointed out that it is important to perform CBT and TP at an early stage, so that statistical analysis is done on stage (early chronic/late chronic) and treatment (Controls, CBT, PT, NFT, TCC-NFTs, PT-NFTs) 2-factor 8-level ANOVA was performed to examine the low back pain score. And we looked at the mean difference in pain and psychological scores before/after the therapeutic intervention between the groups. We examined the mean difference in pain and psychological scores before/after the therapeutic intervention in each group by performing a U-test without assuming a normal distribution when considering the dispersive deviation. Correlation analysis was also performed on the relationship between alpha wave intensity and low back pain scores.


The protocol is shown in Table 2. Three clinical psychologists with more than 10 years of experience conducted the CBT. CBT techniques have adopted psychoeducation, cognitive reframing, relaxation (abdominal breathing and progressive muscle relaxation), stress management, stimulation and behavioral activation in common use28.29. Due to the limited number of reservations, we have fixed a session of 50 min/week, for a total of eight sessions.


The PT program was a combination of individual PT with one session of 50 minutes/week for a total of eight sessions and daily home exercises. The content of the exercise prescription was a multimodal exercise program consisting of lower extremity muscle strengthening exercises, trunk motor control training, stretching, and aerobic exercises.30,31,32. Exercise was set at 40 min/session with an exercise intensity of approximately 12-13 on the Borg scale consistent with the equivalent of METs 4-633.

After a demonstration of the exercise was provided to the participants in the first session, each participant performed the exercise independently. A physiotherapist helped the participant confidently perform their exercise at home every day for 8 weeks, verifying correct techniques with individualized instruction as needed. For the assessment of the participant’s exercise compliance at home, we used a 5-point Likert scale and the question “How often did you do the exercises at home?” “.

About Dawn Valle

Check Also

Interview of Rina Sawayama on the new album

Following her on social media can make it seem like you’re traveling together. Although Sawayama …