The Two Stages of Action-Stopping

Post by Shireen Parimoo

The takeaway

Stopping an action that we have already initiated requires inhibitory control. There are two stages of action-stopping: an early detection stage where the brain identifies the need to stop an action, which triggers motor suppression, followed by a specific action-stopping stage.

What's the science?

Response inhibition involves the ability to stop an already initiated action (like reaching for a cup of coffee), often in response to a stop signal. In the time between seeing a stop signal and the stopping response, there is initially widespread motor suppression at 150ms followed by frontal cortex activity at 300 ms, both of which are thought to reflect inhibitory processes. However, these neural signatures may not be specific to response inhibition because the initial motor suppression has been observed in response to salient, non-stopping signals (attentional capture) and the frontal activity may occur too late to have any impact on the stopping response. This week in The Journal of Neuroscience, Tatz and colleagues investigated the time course of response inhibition during action-stopping.

How did they do it?

Two groups of young adults performed a stop-signal task in which they viewed white arrows on the computer screen (Go signal) and had to indicate the direction of the arrow. On a small subset of the trials, the arrow would change colors to magenta (Stop signal) or cyan (Ignore signal) after a variable delay period. Participants were instructed to stop their response upon seeing the Stop signal, but to continue with their response upon seeing the Ignore signal.

In the first experiment, participants (n = 27) responded with foot-pedal presses while transcranial magnetic stimulation was applied to their motor cortex to stimulate muscles in their hand. They applied stimulation at 150ms, 175ms, and 200ms after the Stop/Ignore signal onset. Muscular responses known as muscular evoked potentials (MEP) were recorded from their hands using electromyography, which allowed the authors to observe the magnitude and timing of global motor suppression (i.e., in muscles unrelated to the task). In the second experiment, participants (n = 20) completed the task with manual responses while their brain activity was recorded using electroencephalography. They also recorded partial muscular activity (prEMG) from the hands, which increases in response to the Go signal but rapidly declines when the Stop signal appears. Together, MEP and prEMG recordings allowed them to assess whether motor suppression occurred specifically in response to the Stop signal or if it was also elicited by the salient Ignore signal. Lastly, they used multivariate pattern analysis to determine whether brain activity in response to the Stop and Ignore signals could be decoded from one another.

What did they find?

Participants were slower to respond on Ignore trials than on Go trials, indicating that the Ignore signal was indeed salient and triggered additional processing despite both trials requiring a response. The amplitude of MEPs was larger in response to Go signals than the Stop and Ignore signals, and larger on failed stopping trials compared to successful stopping trials. However, MEP amplitude didn’t differ between failed stopping and Ignore trials, as participants made a response in both cases. Importantly, there was no difference in MEP amplitude between Stop and Ignore trials when stimulation was applied at 150ms, which means that the early global motor suppression is not specific to Stop signals. Peak EMGs were observed less than 200 ms following successful Stop and Ignore signals and their latency did not differ between these trial types, further supporting the idea that early motor suppression is non-specific.

Neural activity on Go trials could be reliably decoded from Stop and Ignore signals immediately after the arrow appeared on the screen. However, neural responses on successful stopping and Ignore trials could not be distinguished from each other until ~180ms following signal onset. This timing coincides with the motor suppression response, suggesting that the inhibitory process associated with action-stopping is distinct from and occurs after the global motor suppression. On the other hand, activity associated with failed stopping and Ignore trials could only be distinguished ~400ms after signal onset, demonstrating that the failure to process the Stop signal elicits a similar neural response as the processing of a salient but non-stopping signal. Thus, Stop signals are initially treated as salient stimuli, and inhibitory mechanisms specific to action-stopping come online at a later stage of processing.

What's the impact?

This study found that action-stopping involves automatic and widespread motor suppression that is then followed by the engagement of selective response inhibition processes. These findings pave the way for future research to investigate alterations in the timescale of inhibitory control and the inhibitory processes that are impacted in populations with deficits in response inhibition (e.g., psychiatric disorders like ADHD).

Tatz et al. Common and unique inhibitory control signatures of action-stopping and attentional capture suggest that actions are stopped in two stages. The Journal of Neuroscience (2021). Access the original scientific publication here.