Observing without acting: a balance of excitation and suppression in the human corticospinal pathway

Transcranial magnetic stimulation (TMS) studies of human primary motor cortex (M1) indicate an increase corticospinal excitability during the observation of another's action. This appears to be somewhat at odds with recordings of pyramidal tract neurons in primate M1 showing that there is a balance of increased and decreased activity across the population. TMS is known to recruit a mixed population of cortical neurons, and so one explanation for previous results is that TMS tends to recruit those excitatory output neurons whose activity is increased during action observation. Here we took advantage of the directional sensitivity of TMS to recruit different subsets of M1 neurons and probed whether they responded differentially to action observation in a manner consistent with the balanced change in activity in primates. At the group level we did not observe the expected increase in corticospinal excitability for either TMS current direction during the observation of a precision grip movement. Instead, we observed substantial inter-individual variability ranging from strong facilitation to strong suppression of corticospinal excitability that was similar across both current directions. Thus, we found no evidence of any differential changes in the excitability of distinct M1 neuronal populations during action observation. The most notable change in corticospinal excitability at the group level was a general increase, across muscles and current directions, when participants went from a baseline state outside the task to a baseline state within the actual observation task. We attribute this to arousal- or attention-related processes, which appear to have a similar effect on the different corticospinal pathways targeted by different TMS current directions. Finally, this rather non-specific increase in corticospinal excitability suggests care should be taken when selecting a "baseline" state against which to compare changes during action observation