The interaction of a single rTMS-session with emotional brain processing

Research output: Chapter in Book/Report/Conference proceedingConference paper


Abstract: Repetitive transcranial magnetic stimulation (rTMS) has been introduced as a therapeutic tool for depression. Although a reduction of depressive symptoms in patients after rTMS-treatment has been reported, the interaction with the brain still remains unclear. In this study we performed an fMRI-experiment just before and 1 hour after one rTMS-session on healthy female volunteers in a sham-controlled single-blind study design. During the fMRI-experiment, the volunteers had to look at pictures of attractive, aversive and blurred baby faces used as positive, negative and neutral stimuli respectively. They were asked to empathize with the emotional content as much as possible. For all fMRI-experiments the contrast maps ''positive versus neutral" and "negative versus neutral" were calculated. For both contrasts, we performed a 2-way repeated measures ANOVA with group (active, sham) and repetition (before, after) as variables. We found interactions in the prefrontal and cingulate gyri. These regions are known to be involved in processing emotions and controlling social-emotional behaviors.

Introduction: In depression, repetitive transcranial magnetic stimulation (rTMS) over the left dorsolateral prefrontal cortex (DLPFC) has been introduced as a therapeutic tool. In this therapy, patients receive an rTMS-session once a day during 10 days. Although a reduction of depressive symptoms has been reported after treatment [1], only a few studies investigated the corresponding interactions of rTMS with the brain [2,3,4]. In these studies, they focused on metabolic changes [2] after a complete treatment or effects on a memory [3] or a cued reaction time (Posner) [4] task after 1 rTMS-session. Although depression is a mood disorder, we didn't find any study investigating the interaction of rTMS on emotional brain processing in depression. As a first step to investigate the interaction of rTMS with emotional processing in the brain, we performed an fMRI-experiment on healthy volunteer's just before and 1 hour after one rTMS-session.

Material and methods: 19 non-depressed female volunteers (25?5 years) who conformed to the current guidelines for MRI and rTMS research were included. 10 underwent active high frequency rTMS-stimulation (10Hz, 40 trains of 3.9s duration, intertrain interval=26.1s) over the left DLPFC while 9 underwent sham stimulation. Just before and 1 hour after the rTMS-session, an fMRI-experiment consisting of the presentation of 20s duration blocks of attractive (4 blocks with 4 pictures each), aversive (4 blocks with 4 pictures each) and blurred (5 blocks with 4 pictures each) baby faces as positive, negative and neutral stimuli respectively, were performed. All volunteers were asked to focus on their emotions during the experiment. We measured 127 EPI volumes (TR/TE=3000/35ms, 18 slices, matrix size=64x64) with a Philips 1.5T Achieva MRI scanner.
Preprocessing in SPM5 consisted of a rigid body registration as realignment step, an affine transformation to the MNI-template as normalization step and a smoothing step with an isotropic 8mm Gaussian kernel were performed. A multi-linear regression analysis of each fMRI-experiment was performed using the general linear model in SPM5. For each fMRI-session, contrast maps "positive versus neutral" and "negative versus neutral", named positive and negative respectively, were calculated. All maps were organized according to 3 variables: repetition (before, after), group (active, sham) and valence (positive, negative). To investigate treatment effects, a 2-way repeated measures ANOVA "repetition x group" and paired T-tests between repetitions for each group were performed in SPM5 for each valence.

Results: The ANOVA results were investigated with F-statistical maps while the results of the paired T-tests were investigated with T-statistical maps, significance corresponding to an uncorrected pNo main effect of group was found for either valence.
We found main effects of repetition in the prefrontal gyrus bilateral (left: F28.52; right: F=24.01), the left anterior cingulate (F=25.23), the left insula (F=17.35), the right parahippocampus (F=24.28) and the right caudate head (F=18.08) for the positive valence. For the negative valence we found main effects in the prefrontal gyrus bilateral (left: F=26.19; right: F=18.04), the left parahippocampus (F=24.50), the left putamen (F=18.83), the right anterior cingulate (F=17.45) and the caudate head bilateral (left: F=26.53; right: F=17.88). These regions are known to be involved in recognition of facial emotions and memory processes [5].
When looking at the interaction results we found interactions in the temporal gyrus bilateral, the prefrontal gyrus bilateral (left: F=28.01; right: F=21.12), the cingulate gyrus bilateral (left: F=19.32; right: F=17.85), the left hippocampus (F=18.92) the right parahippocampus (F=15.69) and the right thalamus (F=22.40) for the positive valence. The paired T-test showed that in the sham group, the activity in these regions was decreased after the rTMS-session which was not seen in the active group. Moreover, there was an increased activity after the active rTMS-session in the left prefrontal gyrus. For the negative valence, we found interactions at the left temporal gyrus (F=20.04), the left hippocampus (F=16.96), the cingulate gyrus bilateral (left: F=15.63; right: F=20.38), the right precuneus (F=19.63) and the right claustrum (F=52.96). The paired T-test showed that the activity in the cingulate gyrus bilateral decreased after an active rTMS-session and not after a sham session. In the prefrontal regions the activity decreased for both groups.

Conclusions: Although previous studies with mood questionnaires on volunteers were unable to demonstrate any subjective mood change after one rTMS-session [6], we were able to find effects of the rTMS-stimulation on the emotional processing in the brain. It is possible that these different results are due to a limited sensitivity of the questionnaires to small subjective mood changes. Nevertheless, it would be speculative to interpret our results in terms of real changes in mood states or social behavior, at this point. The observed interactions in the prefrontal and cingulate gyri are of interest as these regions are known to be involved in controlling social-emotional behaviors [7]. Compared to normal subjects, depressed patients showed an increased activity in the limbic and prefrontal regions when looking at negative stimuli [8] while they showed a lower neural response in these regions to positive faces [9]. Our results are in line with earlier papers reporting that antidepressive treatments may influence neural activity in these regions [3,8,9].

[1] Schutter D.J. (2008): Psychological medicine; 30:1-11
[2] Luborzewski A. (2007): Journal of Psychiatric Research; 41:606-615
[3] Solé-Padullés C. (2006): Cerebral Cortex; 16 :1487-1493
[4] Rounis E. (2006): The Journal of Neuroscience; 26:9626-9638
[5] Ranganath C. (2006): Neuroscience; 139:277-289
[6] Baeken C. (2008): Clinical Neurophysiology; 119:568-575
[7] Kesler M.L. (2001): Cognitive brain research; 11:213-226
[8] Fu C.H.Y. (2004): Archives of General Psychiatry; 61:877-889
[9] Fu C.H.Y. (2007): American Journal of Psychiatry; 164:599-607
Original languageEnglish
Title of host publicationBHPA
Publication statusPublished - 2009


  • fMRI
  • TMS


Dive into the research topics of 'The interaction of a single rTMS-session with emotional brain processing'. Together they form a unique fingerprint.

Cite this