Bringing the Bench to the Bedside

Exploring underlying mechanisms of cranial electrotherapy stimulation

by Randolph S. Marshall, Career Corner Editor

A commentary on the recent Brain and Behavior article, “Effects of crainal electrotherapy stimulation on resting state brain activity”, by Feusner et. al.

This interesting article by Dr. Feusner (K23 recipient) and colleagues addresses a perennial problem in neuroscience – how to verify the scientific validity of an empirically proven therapy. Feusner et. al. set out to explain the underlying mechanism of cranial electrotherapy stimulation (CES), a long-standing empirical treatment for mood alteration which received FDA approval in 1979. CES has been used for a variety of indications including anxiety, insomnia, depression, and pain, but without clear physiological explanation of its effect. Although several studies have reported beneficial results, it has remained unclear how subsensory alternating current delivered to the earlobes at .5 or 100Hz can alter behavioral symptoms. Preclinical work had shown effects of CES on slowing of alpha waves on EEG in monkeys, associated with reduced adverse reactions to stressful stimuli, but it was unclear whether changes in brain waves were a cause or an effect of improved clinical states. Given uncertainty about mechanisms, the authors proposed to look at effects of CES on brain activity by delivering CES to healthy control subjects while in an fMRI scanner.

One of the strengths of this study, and a general principle for successful investigations of this kind, was the generation of plausible a priori hypotheses based on other studies. Clear statement of hypotheses such as the following establish the scientific context of the study, and let the reader know what to look for in interpreting the results. The hypotheses were:

  1. CES would cause a general deactivation in cortical and thalamic regions because of prior evidence that the stimulation reduces alpha power on EEG.
  2. CES would produce alteration in connectivity networks such as default mode network (DMN) because CES 100Hz affects the EEG beta band, which correlates most highly with the DMN.
  3. CES would alter other connectivity networks such as the dorsal fronto-parietal network (FPN) because there was clinical evidence of CES affect on attention, and the sensorimotor network (SMN) because of clinical evidence for CES effect on pain.

Although there was no randomization, the study design was to use baseline (“off”) periods compared to “on” periods during stimulation. Subject blinding was done by forced choice testing prior to scanning to ensure that participants couldn’t tell if the CES machine was on or off. Subjects had no knowledge of status while in the scanner, and thus there would be no behavior confounder for changes in BOLD fMRI activity.

The authors found that CES correlated with a decreased activation in several brain regions – bilateral SMA, right supramarginal gyrus, right superior parietal and left superior frontal. No increases in regional brain activation were found. The connectivity results demonstrated that the 100Hz stimulation altered the DMN. The FPN and SMN showed no change with CES, and only the higher frequency stimulation produced alterations in connectivity.

Based on these results, the authors concluded that the study demonstrated positive proof that there is a biological effect of CES. The electrical current was proposed to reach the cortex where it would disrupt brain oscillation patterns. The reduction in BOLD activity in several brain regions was consistent with previous EEG studies demonstrating reduction in alpha frequency signal. The altered connectivity for 100 Hz and not .5 Hz was consistent with knowledge that 100Hz frequency affects the beta band, which correlates with DMN activity.  The authors closed the discussion with several unanswered questions, including how the CES alteration of brain activation and connectivity translates to clinical effect, and whether the fMRI BOLD effects they observed would be similar in subjects who had a depression, anxiety, insomnia, or pain.


This was an excellent study from the perspective of initiating a line of inquiry. Aside from the results themselves helping to advance our understanding of the effects of external electrical stimulation on brain activity, this is a strategic line of inquiry from a career perspective. Studies such as these are both hypothesis-testing and hypothesis-generating. What better way to justify your next study or grant than to generate an interesting question by answering the one at hand? High-level neuroscience journals are rife with this type of investigation. In addition, from a grants perspective, early phase clinical trial proposals are now expected to include a physiological “surrogate marker” of efficacy. Developing an imaging method to assess underlying mechanisms of clinical phenomena can therefore be a fruitful line of investigation. Congratulations to Dr. Feusner et al, and good luck with your next study!

Questions for discussion:

  1. Did the results as reported in the paper have ecological validity, i.e. seem plausible, interesting, and relevant to clinical practice? What additional studies would need to be done to convince you that the results are valid? Which hypotheses were left unproven and why?
  2. What weaknesses can be identified in the study design? What are the strengths/advantages of this design?
  3. What is the next hypothesis to test? How would the next study be best designed?

Please visit the Brain and Behavior homepage for more high-quality, Open Access papers and to sign up for our new content alerts!