The aim of this study was to examine the neurophysiologic effects of 30 Hz TBS on left M1. We demonstrated that 30 Hz iTBS safely increased while cTBS decreased MEP amplitudes.
The development of noninvasive transcranial magnetic TBS for human research was based on in vitro
animal studies showing that high-frequency theta burst electrical stimulations can induce LTP and LTD (Hess et al., 1996
). Modeling this stimulation pattern, high frequency transcranial magnetic TBS was developed using M1, yielding similar LTP-like and LTD-like effects (Huang et al., 2005
). Although the initial TBS was performed with 50 Hz stimulation bursts, others have modified the stimulation protocol using 30 Hz bursts (Nyffeler et al., 2008
; Nyffeler et al., 2006
). These 30 Hz TBS studies demonstrated that cTBS had inhibitory effects on the underlying cortical regions. However, the regions examined were outside of M1 and the effects of 30 Hz TBS on the primary motor cortex are unknown. Based on other repetitive TMS studies, we hypothesized that the excitatory and inhibitory effects of 30 Hz TBS would be similar over M1 when compared to nonmotor regions. For instance, in post-stroke nonfluent aphasic patients, inhibitory 1 Hz rTMS over the right hemisphere language homologue area produces speech improvement and increased left hemispheric language activation (Naeser et al., 2012
). Whereas speech improvement has also been reported by using excitatory repetitive magnetic stimulation over the residual left hemispheric Broca's area after a stroke (Szaflarski et al., 2011
). Hence our pilot data confirms our hypothesis by showing that 30 Hz iTBS and cTBS have the expected neurophysiologic effects on M1 and these effects are similar to those of the conventional 50 Hz TBS (Huang et al., 2005
There are currently several repetitive TMS methods that produce cortical excitability changes outlasting the duration of the stimulation (Hoogendam et al., 2010
). Of these, TBS has the following potential advantages – lower stimulation intensity and shorter stimulation duration. Conventional rTMS at set frequencies (e.g., 1 Hz, 25 Hz) has been in use since the 1990s. Low frequency inhibitory rTMS takes longer than inhibitory cTBS (i.e., 600 pulses of cTBS takes 40 seconds while 600 pulses of 1 Hz rTMS takes 10 minutes). Compared to low frequency conventional rTMS, high frequency excitatory rTMS takes less time. However, this time is still significantly longer than iTBS (190 seconds). For example, based on the most recent safety recommendations requiring pause intervals between stimulation trains (Rossi et al., 2009
), it would take at least 17 minutes to deliver 600 pulses of 25 Hz rTMS. Other forms of patterned repetitive TMS (Paired Associative Stimulation, Repetitive Paired-Pulse TMS, Quadripulse Stimulation) also requires longer stimulation time (Hoogendam et al., 2010
), often lasting up to 30 minutes. In general, these non-TBS techniques usually use higher stimulation intensities when compared to TBS (Hoogendam et al., 2010
). Therefore, the shorter stimulation duration of TBS makes this technique practical especially for participants who have difficulty staying still during the study (e.g., children, movement disorder patients). Furthermore, the lower stimulation intensity required for TBS also makes it useful for persons with higher motor thresholds (e.g., children, patients taking certain neuropsychiatric medications).
Although TBS may be easier to deliver compared to other forms of rTMS, it still carries the risk of inducing seizures. TBS research is much less prevalent compared to conventional rTMS. The crude risk of seizure per session of TBS is estimated to be approximately 0.02% (Oberman et al., 2011
). Seizure induction rate for conventional rTMS has only been calculated in the epilepsy population, with a crude risk per participant of ~1.4% (Bae et al., 2007
). To date, virtually all TBS publications are adult studies. However, we recently published our data involving 40 kids showing that TBS is relatively safe in children (Wu et al., 2012
). Unfortunately, at the most recent international meeting for TMS researchers, a consensus for TBS stimulation parameters was not reached (Rossi et al., 2009
). The consensus group stated that several TBS factors still need to be evaluated for safety: numbers of stimulation pulses, interval between repeated TBS sessions, number of multiple applications and stimulation intensity (Rossi et al., 2009
There are several limitations with this study. First, we did not measure MEP beyond ten minutes after TBS. Therefore, we do not know if the duration of the after-effects is comparable to 50 Hz TBS. Second, the sample size is small (n = 9 for each group), although this is the same size as the initial 50 Hz TBS report (Huang et al., 2005
). Third, we measured AMT before applying TBS. Different groups have shown that muscle activity prior to TBS can change or eliminate TBS after-effects (Gentner et al., 2008
; Goldsworthy et al., 2012
). Finally, we did not directly compare, within the same individuals, the magnitude of the iTBS and cTBS effects at 30 Hz vs. 50 Hz. However, the 30 Hz effect size we identified () is comparable both to 50 Hz TBS studies in our lab as well as to those published elsewhere.
In conclusion, the modified 30 Hz TBS safely produced the expected neurophysiologic changes in left M1. These changes are similar to 50 Hz TBS. However, when delivering 30 Hz TBS with Magstim® SuperRapid2, one can use higher stimulation intensities. Therefore, 30 Hz TBS may be useful in persons with higher motor thresholds.