The main findings of the present study are that (1) there is less tonic short-latency intracortical inhibition in the dominant motor cortex than in the non-dominant motor cortex, (2) an established experimental paradigm (PAS) induces motor cortical plasticity in both the dominant and non-dominant hemispheres, and to a similar degree, (3) performance improved following short-term practice on a simple ballistic training task for both dominant and non-dominant hands. However, the magnitude of performance improvement was greater for the left, non-dominant, hand.

There are clear anatomical differences between dominant and non-dominant hemispheres. For example, the depth of the central sulcus is greater in the dominant hemisphere of both right- than left-handed subjects (Amunts et al. 1996). Additionally, the volume of the hand representation in the motor cortex is greater in the dominant hemisphere of right- than left-handed subjects (Volkmann et al. 1998). Also, the volume of cortex occupied by nerve cell bodies is relatively smaller in the dominant hemisphere and this might present greater opportunity for intracortical connections (Amunts et al. 1996). Evidence for electrophysiological differences between the hemispheres is much more equivocal. For example, some studies have reported that resting motor thresholds to TMS are lower for the dominant hemisphere (e.g. Macdonell et al. 1991; Triggs et al. 1997; Ilic et al. 2004). However, many other studies have failed to demonstrate significant differences in thresholds (e.g. Cicinelli et al. 1997; Civardi et al. 2000; Semmler and Nordstrom 1998; Garry et al. 2004). Likewise, it has been reported that the duration of the cortical silent period is shorter in the dominant hemisphere (Priori et al. 1999). This finding is in contrast to an earlier study that reported no interhemispheric differences in the duration of cortical silent period (Cicinelli et al. 1997).

A small number of studies have examined whether there are hemispheric differences in SICI, with several reporting no hemispheric asymmetry (Cicinelli et al. 2000; Maeda et al. 2002). However, these studies used stimulation paradigms that are optimal for evoking SICI and may, therefore, be insensitive to subtle changes in SICI. A more recent study (Ilic et al. 2004) which employed a wider range of stimulus parameters designed to provide a more sensitive assessment of SICI, reported a small but significant hemispheric difference in SICI, with inhibitory tone being slightly less in the dominant hemisphere. This finding suggested a subtle difference in γ-aminobutyric acid type A (GABA A )-mediated inhibition between the hemispheres. In contrast to this study, Hammond et al. (2004) reported more SICI, at an ISI of 3 ms, in the dominant hand compared to the non-dominant hand. Again, a range of conditioning intensities was employed and the difference between hands was only seen at lower conditioning intensities of between 0.2 and 0.5 times resting threshold. However, some important details were not reported in this study, such as the amplitude of test MEPs. Additionally, relatively small numbers of trials were used to examine SICI which would have lead to increased variability in these measures. The findings of the present study are in agreement with those of Ilic et al. (2004) and, again, suggest that there is reduced SICI in the dominant hemisphere at lower conditioning stimulus intensities. Therefore, the present results add to those of a small number of other studies that provide electrophysiological evidence of hemispheric differences in the motor cortices. The reasons for the apparent discrepancy between studies examining for electrophysiological evidence of hemispheric asymmetry is not clear. However, the use of non-optimised testing parameters and differences in subject characteristics are likely to be important. For example, the present results reinforce the importance of employing a range of SICI conditioning intensities when investigating subtle differences in the level of SICI.

In vitro studies have shown that cortical plasticity can be facilitated by a reduction in GABA inhibition (e.g. Hess et al. 1996). These findings lead Ilic et al. (2004) to suggest that a reduction in GABA A -mediated SICI might position the dominant hemisphere to undergo plastic change more easily. The experiments described here were designed to investigate this hypothesis. The results from the current PAS studies, which demonstrated that PAS induced a similar excitability change in dominant and non-dominant hemispheres, provide no support for this hypothesis. It should be noted that Stefan et al. (2002) demonstrated that PAS-induced MEP facilitation was not in itself accompanied by a significant change in SICI. This result might suggest little relationship between PAS-induced facilitation and changes in the level of SICI. However, the fact that PAS-induced MEP facilitation is not accompanied by significant changes in SICI does not necessarily mean that PAS facilitation is not influenced by the basal level of SICI.

It might be argued that PAS is an experimental paradigm that is inappropriate for examining hemispheric differences in plasticity induction. However, there are several reasons to suggest that PAS induces plasticity in functionally relevant motor cortical circuits. Firstly, PAS has a number of features that suggest its effects are due to long term potentiation (LTP)-like process within the motor cortex (Stefan et al. 2000, 2002; Wolters et al. 2003). Secondly, there is evidence that the circuits modified by PAS are important for motor learning. In the rat motor cortex strong evidence has been provided that motor skill training involves changes in efficacy of intracortical connections brought about by LTP (Rioult-Pedotti et al. 1998, 2000). Evidence has been presented that similar LTP-like processes are involved in human task-dependent plasticity. Butefisch et al. (2000) demonstrated that motor cortical plasticity induced by performance of repeated thumb movements could be blocked by the application of dextromethorphan, an N-methyl-d-aspartate receptor blocker. In an intriguing series of experiments Ziemann et al. (2004) demonstrated that a period of MT prevented subsequent PAS-induced excitability change. These findings suggest two things. Firstly, that motor learning in human subjects involves LTP-like processes. Secondly, that PAS induces reorganisation in functionally relevant motor circuits.

The significance of the MT data is somewhat more equivocal. We hypothesised that if the dominant hemisphere was placed at some advantage for functionally relevant plasticity because of decreased inhibitory tone, performance improvement on a training task with the dominant hand might be greater than that of the non-dominant hand. In the present study, subjects performed a ballistic thumb movement repeatedly over a period of 30 min. This task has been shown to result in significant performance improvements which are thought to be due to an LTP-like process in the motor cortex (Ziemann et al. 2004). Our results demonstrated that performance on the task improved for both dominant and non-dominant hands. However, contrary to our hypothesis, performance of the non-dominant hand improved significantly more than that of the dominant hand. This finding suggests that use-dependent plasticity may be greater in the non-dominant hemisphere. However, the interpretation is made more difficult by the fact that baseline motor performance was significantly better in the dominant hand. This finding is in agreement with previous studies that have reported that baseline performance on tasks such as finger tapping (Triggs et al. 1997; Brouwer et al. 2001) and the Purdue pegboard (Brouwer et al. 2001; Garry et al. 2004) is better for the dominant hand of right-handed subjects. The fact that performance improved more in the non-dominant hand might be due to the lower level of baseline performance. Obviously, there are physical limitations to performance of such simple ballistic tasks and therefore performance improvement may saturate more quickly in the dominant hand. Given the significantly greater baseline performance of the dominant hand, it is interesting to note that Ilic et al. (2004) speculated that reduced SICI may provide some advantage for the “readiness and ease to carry out movements with the dominant hand”.

Further evidence that the level of SICI does not exert a strong influence over plasticity induction in the motor cortex, at least under these experimental conditions, comes from the regression analyses. These regression analyses demonstrated that there was no significant relationship between the level of SICI (at a conditioning intensity of 80% AMT) and the amount of MEP facilitation seen with PAS or MT for either hand. The decision to base these analyses on the SICI data at an intensity of 80% AMT was made for two reasons. Firstly, at this conditioning intensity there was a moderate level of inhibition and this would minimise the possibility of inhibitory saturation affecting the result. Secondly, at this conditioning intensity there was a significant difference in the level of SICI between the two hemispheres. This difference may have allowed hemispheric differences to be seen in these correlations.

In conclusion, we have confirmed a previous finding of reduced SICI in the dominant hemisphere (Ilic et al. 2004). Also, we have demonstrated that PAS is similarly effective at inducing motor cortical plasticity in the two hemispheres. Additionally, performance improvement on a simple ballistic MT task was greater for the non-dominant hand. Taken together these findings do not support the hypothesis that reduced SICI in the dominant hemisphere positions it favourably for use-dependent plasticity. Also, the data presented question whether SICI has any role in influencing plasticity induction in the motor cortex, at least under these experimental conditions.