A step over the horse race model: study of inhibitory control of gait initiation by a stop signal task

Walking towards desired locations in space supports the interaction with the environment. Sometimes, unpredictable changes require the cancellation of a planned step. Gait initiation triggers a chain of events leading our body to leave a stable position for an unstable one, in which the body's center of mass (CoM) moves from the center of the base of support (BoS) toward its perimeter, until it leaves the perimeter completely. When the distance between the COM and the centre of BoS in medio-lateral and anterior-posterior increases during the 500 ms of anticipatory postural adjustments (APA) that precede the step onset, a set of muscles is recruited both to lift the foot and to stabilise the body. Sometimes these adjustments need to be revised since external (environmental) changes will require to inhibit the prepared step and maintain the body in the initial position. The study of this form of motor inhibition, intensively analysed in action control of simple movements as a button press tasks, only recently has been extended to more complex actions. A tool used widely for studying the physiological correlates of this form of motor inhibition is the Stop Signal Task (SST). SST is a test requiring executing a movement in response to a go signal (Go trials) and inhibiting it when a stop signal suddenly occurs after the go (Stop trials). The outcome of the task is accounted for by a theoretical race between the Go and Stop processes, triggered by Go and Stop signals respectively. In a stop trial, movements are executed if the Go process runs sufficiently fast. This formulation of the model, mainly developed by studying movements completed in a very short period, such as saccades or finger (key press) movements, approximates the occurrence of movements at the end of the race. However, things can be different for movements evolving on longer timescales, such as gait initiation (GI). As a result, in this thesis we used a version of SST for GI, combined to the overtime recording of the CoM, CoP and muscle activity, to investigate how well the race model accounts for it, and searching for the signature of movements’ inhibition in the recorded signals. Twelve healthy participants were tested in the SST readapted for GI while the position of 33 markers fixed on their body, the ground reaction forces, and the activation of 16 muscles were simultaneously recorded. The recorded signals were used to quantify the temporal evolution of CoM and CoP trajectories, and they were linked to muscle activity during APAs in Go trials and correct (cST) and wrong (wST) Stop trials. All the biomechanical parameters were also used to estimate the reaction time to the go (RT) and stop (SSRT) signals trial by trial. As predicted by the model, even in the GI, the trials that escape inhibition were those with faster RTs, while those with slower RTs are inhibited. Interestingly, in Stop trials an intermediate behavior, called partially wrong stops (pwST), were found. A comparison between of the SSRT derived from the displacement of the CoM, the CoP or muscular contraction revealed that they were significantly lower in wST than in cST and pwST, further supporting the compliance of our measures with the theoretical model. Additionally, SSRT measured from biomechanical parameters correlates strongly with SSRT estimated by the model for cST and pwST. Overall, these findings indicate that if the inhibitory command is delivered sufficiently early in time, the maintenance of the starting position can be achieved. On the contrary, if inhibitory command is delivered later in time, even though it affects the muscle activity, it does not cancel the step initiation because of a higher distance between the CoM and the BoS. This suggests that a biomechanical point of non-return affects the correct inhibition of GI and that the adaptation of the race model to such as complex movement needs to account for the interaction between a set of biomechanical variables. The present results provide important insight for studying the control of complex actions to extend to the field of prevention of injuries and rehabilitation.