Introduction

Although much is still unknown about the neural planning that precedes the production of motor responses, it is clear that these preparatory processes are accompanied by changes in neural activation levels in motor‐related cortical structures. For example, neural recordings from nonhuman primates have shown that motor cortical‐related activity increases during the preparatory period prior to execution of a known action (Tanji and Evarts 1976), likely in order to facilitate fast and accurate responses. In humans, measurement of preparatory cortical activity is more challenging as direct measurement is invasive, and neuroimaging techniques often result in poor spatial and/or temporal resolution; thus, transcranial magnetic stimulation (TMS) has been used as an alternative, indirect method to quantify and examine motor preparation processes. TMS is a neurostimulation technique that uses a magnetic pulse to directly activate corticomotor neurons in the primary motor cortex (M1), which then project via the corticospinal tract to motor neurons innervating the target muscle. This descending activation can be measured using electromyography (EMG) as a motor‐evoked potential (MEP). The amplitude of the MEP is thought to reflect the excitability of the corticospinal pathway with larger MEP amplitudes reflecting greater activation in the system (see Reis et al. 2008; Spampinato et al. 2023 for reviews). During a simple reaction time (RT) task where the required response is known in advance and thus can presumably be preprogrammed, it has been shown that MEP amplitudes increase from baseline levels during response preparation, and these levels are maintained until about 300 ms prior to the go‐signal (Kennefick et al. 2014). Interestingly, at this point, MEP amplitudes have consistently been shown to decrease in the foreperiod interval leading up to the go‐signal (Touge et al. 1998; Kamata et al. 2012; Greenhouse et al. 2015; Duque et al. 2017), after which they once again increase due to execution of the prepared response (Rossini et al. 1988; Chen et al. 1998; Leocani et al. 2000).

The relatively large MEP amplitudes observed during the early planning and late response initiation phases of a motor task, which mirror modulation of firing rates directly recorded in nonhuman primates (Riehle and Requin 1989; Churchland et al. 2006), are thought to reflect an expected increase in corticospinal activation that is associated with response preparation and execution, respectively (Davranche et al. 2007; Klein‐Flügge et al. 2013). However, the reason for the observed decrease in MEP amplitude occurring during the foreperiod immediately prior to the imperative signal has been a matter of debate (see Duque et al. 2017; Greenhouse 2022 for reviews). Early explanations considered premovement corticospinal inhibition necessary to either suppress competing movements (i.e., “competition resolution” model; Burle et al. 2004) or to maintain a high level of advance preparation while preventing the premature release of the prepared motor plan (Duque and Ivry 2009, i.e., “impulse control” model: Duque et al. 2010). However, these inhibition‐related explanations have been challenged by the coexistence of decreased corticospinal excitability in task‐irrelevant muscles, which would not be affected if the inhibition was strictly related to suppressing unwanted responses (Greenhouse et al. 2015; Duque et al. 2017). This observation led to an alternative “spotlight control” hypothesis whereby neural inhibition acts to speed movement initiation by increasing the signal‐to‐noise ratio for motor output pathways, thus facilitating response selection and execution (Greenhouse et al. 2015; Lebon et al. 2019). Further supporting an initiation‐based explanation, suppression of corticospinal excitability has been shown to follow a similar time course for both reactive and self‐initiated responses (Ibáñez et al. 2020). These authors argued that movements that are self‐initiated without an external cue should not require any inhibitory processes related to avoiding false starts because, by definition, false starts cannot occur in this circumstance. Instead, they argued that the observed suppression may instead be a function of shifting activation from preparatory processes to response initiation processes. This explanation is also supported by the somewhat paradoxical finding that the magnitude of MEP suppression has been shown to be correlated with response latency, whereby greater inhibition is associated with shorter RTs, suggestive of a role of preparatory inhibition in the facilitation of fast movement initiation (Hannah et al. 2018).

These previous studies have often employed one type of RT paradigm in isolation, or have not directly compared the timeline of MEP suppression between tasks; yet important information may be gleaned by comparing cortical activation dynamics between tasks that result in differences in the timeline of preparation, inhibition, and initiation processes. These types of tasks were originally considered in Donders' “On the Speed of Mental Processes” (1969 translation of original 1868 text), with the use of simple, choice, and go/no‐go RT tasks being used to calculate the durations of the various stages of information processing. In the simple RT task, the participant knows the required response in advance and can actively prepare the movement during the foreperiod before the imperative stimulus. However, although this may enable fast RTs, this ability to prepare to a high level may necessitate greater levels of inhibition prior to the imperative stimulus to prevent premature release of the response. In contrast, when the RT task involves a response that is specified by the imperative stimulus (choice RT), the exact response is unknown in advance, which may result in a lowered state of preparation as response selection and programming are completed following the go‐signal. Lastly, the go/no‐go RT task involves a single known response that is either initiated in response to a “go” imperative cue or withheld in response to a “no‐go” imperative cue, and as such, the go/no‐go task allows for response preparation to occur in advance, similar to a simple RT paradigm. However, the need to inhibit response execution on no‐go trials would be expected to involve additional inhibitory processes once the imperative cue is presented, compared to simple RT in which a response is required every trial. Although the neural processes involved are undoubtedly more complex than summarized above, robust RT differences between the tasks presumably reflect (at least in part) differences in preparatory and inhibitory activation (Filipović et al. 1997; Carlsen et al. 2004), which may manifest in MEP amplitude differences throughout the RT paradigm.

The purpose of the present study was to examine the timeline of corticospinal excitability during the preparatory phase of simple, choice, and go/no‐go RT tasks to provide further insight into the nature of premovement suppression of MEP amplitude. This was done by measuring TMS‐evoked MEP amplitudes in the same effector muscle at various times during the foreperiod in RT tasks in which (1) the effector muscle could be prepared and was certain to be used as a prime mover (i.e., simple RT; right wrist extension); (2) the effector could be prepared, but this prepared action was either initiated or inhibited (i.e., go/no‐go RT; right wrist extension); and (3) the effector could not be fully prepared as the response was unknown and may or may not need to be engaged (i.e., choice RT; right wrist extension or flexion). Although previous studies have examined MEP amplitudes for a choice RT task, they contrasted responses between limbs (e.g., Leocani et al. 2000; Quoilin et al. 2019). In the current study, the use of different RT protocols involving responses within a limb, allowed for comparison of corticospinal suppression between the same musculature under different preparatory conditions. Furthermore, TMS was applied at six time points prior to initiation to gain a more detailed timeline of MEP amplitude changes throughout the foreperiod. If the previously observed reduction in corticospinal excitability relates to high levels of advance preparation that require inhibition to prevent premature/incorrect responses, it would be expected that simple and go/no‐go RT tasks would result in larger MEP amplitude decreases during the foreperiod, as compared to the choice RT task. Conversely, if the reduction in corticospinal excitability is associated with response initiation processes, then suppression would be expected for all tasks, including the choice RT paradigm in which final preparation does not occur until presentation of the go‐signal. It is also worth noting that inhibition versus initiation explanations for premovement suppression are likely not mutually exclusive and do not need to be considered as competing hypotheses, as both processes may contribute to the reduction in MEP amplitudes just prior to response execution.