Neuromodulatory effects of offline low-frequency repetitive transcranial magnetic stimulation of the motor cortex: A functional magnetic resonance imaging study

Abstract

Repetitive transcranial magnetic stimulation (rTMS) of the primary motor cortex (M1) can modulate cortical excitability and is thought to influence activity in other brain areas. In this study, we investigated the anatomical and functional effects of rTMS of M1 and the time course of after-effects from a 1-Hz subthreshold rTMS to M1. Using an "offline" functional magnetic resonance imaging (fMRI)-rTMS paradigm, neural activation was mapped during simple finger movements after 1-Hz rTMS over the left M1 in a within-subjects repeated measurement design, including rTMS and sham stimulation. A significant decrease in the blood oxygen level dependent (BOLD) signal due to right hand motor activity during a simple finger-tapping task was observed in areas remote to the stimulated motor cortex after rTMS stimulation. This decrease in BOLD signal suggests that low frequency subthreshold rTMS may be sufficiently strong to elicit inhibitory modulation of remote brain regions. In addition, the time course patterns of BOLD activity showed this inhibitory modulation was maximal approximately 20 minutes after rTMS stimulation.

Figure 1. Group one sample t-test results (p < 0.05, FDR-corrected for multiple comparisons at the whole brain level).

Activation (motor task > rest) was presented in yellow and deactivation (motor task < rest) was presented in blue. (A) pre-rTMS, (B) post1, (C) post2, and (D) post3. After rTMS of the left M1, within group analysis of the right finger-tapping task showed that the extent of motor activation was reduced at post1, post2 and post3, although the activities of ipsilateral motor networks slightly recovered at post3. The most striking effect of rTMS was significant deactivation in sensory and non-motor brain areas. These deactivations were most significant at post2.

Figure 2

(A) One-way within subject ANOVA for the four conditions (pre, post1, post2, and post3). The SPM{F}s had a threshold of p < 0.05, with FDR-correction for multiple comparisons at the whole brain level. (B) Post-hoc two sample t-tests showed that the difference between the pre and post2 conditions (pre > post2) was responsible for the difference shown in ANOVA. The SPM{t}s had a threshold of p < 0.05, with a Bonferroni adjusted alpha level of 0.0084, and FDR-correction for multiple comparisons at the whole brain level. Activation differences were presented in yellow and deactivation differences were presented in blue.

Figure 3

Results of the right finger-tapping task during four BOLD fMRI scans before and after 1-Hz rTMS of the left (A) and right M1 (B). Group one sample t-test maps had no threshold to show time course of changes in whole brain neural activity. (A) A time course of BOLD activity in the left M1 showed positive BOLD signal (activation) during right finger-tapping in four conditions (pre, post1, post2, and post3). The mean percent signal change was slightly reduced at post1 and post2, but returned to the pre-rTMS value. The reduction in percent signal change at post1 and post2 was not statistically significant. (B) Time course of BOLD activity in the right M1 showed negative BOLD signal (deactivation) during right finger-tapping at four conditions (pre, post1, post2 and post3). Deactivation was slightly increased at post1, post2, and post3, but there was no statistical significance. Plots showed means (lines) and SEMs (shading) of percentage signal changes.

Figure 4

Results of the right finger-tapping task during four BOLD fMRI scans before and after 1-Hz rTMS of the left (A) and right SMA (B). Group one sample t-test maps had no threshold to show time course of changes in whole brain neural activity. (A) A time course of BOLD activity in the left SMA showed changes from positive responses (activation) from pre-stimulation to negative responses (deactivation) at post2, then back to positive responses (activation) at post3. (B) Time course of BOLD activity in the right SMA showed changes from positive responses (activation) at pre-stimulation to negative responses (deactivation) at post2 and post3. (*) P < 0.05 and (**) P < 0.01. Plots showed means (lines) and SEMs (shading) of percentage signal changes.

Figure 5

Results of the right finger-tapping task during four BOLD fMRI scans before and after 1-Hz rTMS of the left (A) and right inferior frontal cortex (B). Group one sample t-test maps had no threshold to show time course of changes in whole brain neural activity. A time course of BOLD activity in the left and right inferior frontal cortex showed changes from positive responses (activation) from pre-stimulation to negative responses (deactivation) at post1, post 2 and post3. (*) P < 0.05 and (**) P < 0.01. Plots showed means (lines) and SEMs (shading) of percentage signal changes.

Figure 6. Experimental design.

Subjects underwent four BOLD-fMRI scans before and after 1-Hz real- or sham-rTMS. Simple finger-tapping tasks were performed during fMRI. After pre-rTMS fMRI, 600 biphasic stimuli of 1 Hz rTMS were given over the left M1 hand area. All subjects received 10-minute trains of 1-Hz rTMS delivered outside the fMRI scanner. Post-rTMS fMRI was performed three times starting 10 minutes after application of 10-minute trains of 1-Hz rTMS. Real- and sham-rTMS were performed on separate days.