Baseline brain activity predicts response to neuromodulatory pain treatment

Abstract

Objectives: The objective of this study was to examine the associations between baseline electroencephalogram (EEG)-assessed brain oscillations and subsequent response to four neuromodulatory treatments. Based on available research, we hypothesized that baseline theta oscillations would prospectively predict response to hypnotic analgesia. Analyses involving other oscillations and the other treatments (meditation, neurofeedback, and both active and sham transcranial direct current stimulation) were viewed as exploratory, given the lack of previous research examining brain oscillations as predictors of response to these other treatments.

Design: Randomized controlled study of single sessions of four neuromodulatory pain treatments and a control procedure.

Methods: Thirty individuals with spinal cord injury and chronic pain had their EEG recorded before each session of four active treatments (hypnosis, meditation, EEG biofeedback, transcranial direct current stimulation) and a control procedure (sham transcranial direct stimulation).

Results: As hypothesized, more presession theta power was associated with greater response to hypnotic analgesia. In exploratory analyses, we found that less baseline alpha power predicted pain reduction with meditation.

Conclusions: The findings support the idea that different patients respond to different pain treatments and that between-person treatment response differences are related to brain states as measured by EEG. The results have implications for the possibility of enhancing pain treatment response by either 1) better patient/treatment matching or 2) influencing brain activity before treatment is initiated in order to prepare patients to respond. Research is needed to replicate and confirm the findings in additional samples of individuals with chronic pain.

Keywords: Chronic Pain; Electroencephalography; Nonpharmacological Treatments; Spinal Cord Injury.

Figure 1

Box plot of pretreatment relative theta power in hypnosis responder group vs nonresponders (responder ≥30% pre to postsession reduction in pain).

Figure 2

Box plot of presession pretreatment relative alpha power in meditation responder group vs nonresponders (responder ≥30% pre to post-session reduction in pain).

Figure 3

Associations between baseline EEG bandwidth activity at each electrode site and pain reduction with each treatment procedure. Moderate positive coefficients (r ≥ 0.30) are indicated by a yellow fill, and moderate negative coefficients (r ≤ −0.30) by a blue fill. Coefficients at P < 0.05 are indicated by a greater weight (thicker line) of the border around each fill for each electrode. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]