Multi-session Transcranial Direct Current Stimulation Over Primary Motor Cortex Facilitates Sequence Learning, Chunking, and One Year Retention.

Brian Greeley,Willem B Verwey,Jonathan S Barnhoorn,Rachael D Seidler

doi:10.3389/fnhum.2020.00075

Brian Greeley, Willem B Verwey + Show 2 more

Open Access

https://doi.org/10.3389/fnhum.2020.00075

Copy DOI

Abstract

Transcranial direct current stimulation (tDCS) over the primary motor cortex (M1) can facilitate motor learning, but it has not been established how stimulation to other brain regions impacts online and offline motor sequence learning, as well as long-term retention. Here, we completed three experiments comparing the effects of tDCS and sham stimulation to the prefrontal cortex (PFC), M1, and the supplementary motor area complex to understand the contributions of these brain regions to motor sequence learning. In Experiment 1, we found that both left and right PFC tDCS groups displayed a slowing in learning in both reaction time and number of chunks, whereas stimulation over M1 improved both metrics over the course of three sessions. To better understand the sequence learning impairment of left PFC anodal stimulation, we tested a left PFC cathodal tDCS group in Experiment 2. The cathodal group demonstrated learning impairments similar to the left PFC anodal stimulation group. In Experiment 3, a subset of participants from the left PFC, M1, and sham tDCS groups of Experiment 1 returned to complete a single session without tDCS on the same sequences assigned to them 1 year previously. We found that the M1 tDCS group reduced reaction time at a faster rate relative to the sham and left PFC groups, demonstrating faster relearning after a one-year delay. Thus, our findings suggest that, regardless of the polarity of stimulation, tDCS to PFC impairs sequence learning, whereas stimulation to M1 facilitates learning and relearning, especially in terms of chunk formation.

Highlights

According to the Cognitive framework for Sequential Motor Behavior (C-SMB), sequence learning depends on communication between a central and a motor processor (Verwey et al, 2015)
The authors of the C-SMB framework posit that learning in the discrete sequence production (DSP) task involves three separate modes: a reaction mode, occurring early in learning when participants are responding to each individual stimulus, a central-symbolic mode in which sequence execution relies on verbal and/or spatial sequence representations, and a chunking mode occurring later in learning when execution is based on motor chunk representations
Hypothesis-driven pairwise comparisons for the first session revealed that anodal stimulation to left prefrontal cortex (PFC) produced a significantly slower rate of change in reaction time (β = 0.16, SE = 0.05, p = 0.001; Figure 4) relative to sham

Summary

Introduction

According to the Cognitive framework for Sequential Motor Behavior (C-SMB), sequence learning depends on communication between a central and a motor processor (Verwey et al, 2015). The motor processor is responsible for executing individual responses early in learning and entire motor chunks after extended practice (Abrahamse et al, 2013; Verwey et al, 2015). The authors of the C-SMB framework posit that learning in the discrete sequence production (DSP) task (see Abrahamse et al, 2013) involves three separate modes: a reaction mode, occurring early in learning when participants are responding to each individual stimulus, a central-symbolic mode in which sequence execution relies on verbal and/or spatial sequence representations, and a chunking mode occurring later in learning when execution is based on motor chunk representations. The posterior striatum and the SMA play key roles in representing motor chunks in the chunking mode by controlling primary motor cortex (M1), but with extensive practice, M1 may exclusively represent the motor chunks (Karni et al, 1998; Verwey et al, 2002; Abrahamse et al, 2013)

Methods

Results

Conclusion