Proactive Inhibition Activation Depends on Motor Preparation: A Single Pulse TMS Study.

In everyday life, environmental cues are used to predict and respond faster to upcoming events. Similarly, in cueing paradigms (where, on cued trials, a cued target requires a speeded response), cues are known to speed up response times (RTs), suggesting that motor preparation has occurred. However, some studies using short cue-target intervals (<300 ms) have found slower RTs on cued, compared to uncued trials (namely, the "paradoxical warning cost"). One explanation of this paradoxical effect is proactive inhibition, a motor gating mechanism that prevents false alarms, also called "the default state of executive control." Alternative hypotheses claim that, with such short cue-target delays, participants cannot fully prepare the motor response, thus producing slower RTs. In studies of action inhibition, it is often assumed that participants prepare a response on each trial, a prerequisite to induce and measure (proactive) motor inhibition. In this study, we psychophysically manipulated stimulus' duration in a simple RT task, and measured a duration threshold at which participants responded on time on 80% of the trials. When participants are tested at their stimulus' duration threshold, they are more likely to prepare the motor response on each trial. Furthermore, we directly measured participants' readiness to respond by recording transcranial-magnetic stimulation (TMS)-elicited motor evoked potentials (MEPs), a direct measure of corticospinal excitability. Participants performed cued and uncued trials on a simple RT task with short cue-target intervals. We expected lower MEPs' amplitude on cued than uncued trials with short cue-target intervals, as it would be predicted by the proactive inhibition account. However, when conditions are equated so that motor preparation is induced both under cued and uncued trials, the paradoxical warning cost disappears, as RTs were always faster on cued than uncued trials. Moreover, MEPs recorded from the task-relevant muscle were never suppressed at target onset compared to baseline, a result that does not support the proactive inhibition hypothesis. These results suggest that proactive inhibition is not active by default and that its activation depends on motor preparation.