No evidence for an effect of M1 cTBS on schema-mediated motor sequence learning

This study found no causal evidence that inhibitory continuous theta burst stimulation (cTBS) of the left primary motor cortex (M1) affects schema-mediated motor sequence learning, as neither corticospinal excitability nor behavioral performance showed significant differences between cTBS and sham conditions.

The Gist

The Big Question: Can We "Hack" Our Brain's Learning Speed?

Imagine you are learning to play a new song on the piano. If you already know how to play scales and chords (your schema), you can learn that new song much faster than if you were starting from scratch. Your brain uses your old knowledge as a shortcut to build the new memory.

Scientists already knew that the Left Primary Motor Cortex (M1)—the part of the brain that controls your hand movements—lights up when you use these shortcuts. But they didn't know if this part of the brain was essential for the shortcut to work. Was it the engine, or just the dashboard lights?

To find out, the researchers tried to temporarily "turn off" this engine using a special magnetic trick called cTBS (a type of brain stimulation) and see if the learning still happened.

The Experiment: The "Two-Day" Challenge

The researchers set up a game for 48 healthy young adults that took place over two days.

Day 1: Building the Foundation

• Participants learned a specific pattern of finger taps (like a secret code: 4-7-3-8-6-2-5-1).
• Think of this as learning a specific dance routine. By the end of the day, their brains had built a "mental map" or schema for this type of movement.

Day 2: The Test

• About 24 hours later, the participants came back.
• The Twist: Half the group got a "real" brain zap (cTBS) on the left side of their motor cortex. The other half got a "fake" zap (sham) that felt the same but did nothing.
• Immediately after the zap, they had to learn a new finger pattern.
• The Catch: This new pattern was 75% similar to the old one. It used the same "dance moves" but in a slightly different order. It was designed to be a perfect fit for the mental map they built on Day 1.

The Goal:
If the Left M1 is the "engine" for using these shortcuts, turning it off with the zap should have made the new pattern very hard to learn. The participants should have stumbled, been slower, or made more mistakes compared to the group that got the fake zap.

The Results: The Engine Didn't Stop

Here is where the story takes a surprising turn.

1. The Brain Zap Didn't Work (Physically)
First, the researchers checked if the "zap" actually turned off the brain engine. They measured electrical signals in the muscles (like checking if the car engine was actually off).

• Result: Surprisingly, the zap didn't seem to lower the brain's activity in the group as a whole. About half the people responded to the zap, but the other half didn't. It was like trying to dim a lightbulb, but the bulb kept shining just as bright for half the room.

2. The Learning Was Unaffected (Behaviorally)
Because the brain activity didn't change much, the researchers looked at the performance.

• Result: The group that got the "real" zap learned the new pattern exactly as well as the group that got the "fake" zap.
• They were just as fast.
• They were just as accurate.
• They remembered the old moves and learned the new moves perfectly.

The Conclusion: A Dead End?

The study concluded that disrupting the Left M1 did not stop people from using their mental shortcuts to learn new motor skills.

This is a bit like trying to stop a chef from cooking a new recipe by unplugging the blender, only to find out the chef was using a hand mixer the whole time. The researchers expected the Left M1 to be the "blender" essential for this specific type of learning, but the data suggests it might not be the only tool in the kitchen.

Why Did It Fail? (The "Plot Twist")

The authors spent a lot of time wondering why the experiment didn't show the expected effect. They considered a few possibilities:

• The Wrong Tool: Maybe the "zap" wasn't strong enough or wasn't applied at the exact right moment to actually turn off the brain region.
• The Wrong Map: Maybe the brain has a backup system. If the Left M1 is busy, maybe the Right M1 or another part of the brain stepped in to help, acting like a backup generator.
• The Task Was Too Easy: Because the new pattern was so similar to the old one, the brain might have been able to learn it so easily that it didn't need the specific part of the brain they were trying to disrupt.

The Takeaway for You

This study is a great example of how science works. Even when an experiment doesn't prove what you hoped it would, it still teaches us something valuable.

It tells us that our brains are incredibly resilient and flexible. Even when we try to "break" a specific part of the brain that we think is crucial for learning, the brain often finds a way to keep the learning going. It suggests that the "shortcut" for learning new skills might be more distributed across the brain than we previously thought, rather than relying on just one single switch.

Technical Summary

1. Problem Statement and Objective

The Problem:
Memory consolidation is accelerated when new information is compatible with a pre-existing cognitive schema. While the role of the medial prefrontal cortex (mPFC) in schema-mediated declarative memory is well-established, the neural mechanisms supporting schema-facilitated non-declarative (motor) learning are less understood. Previous neuroimaging studies (Reverberi et al., 2025) suggested that the left primary motor cortex (M1) is recruited during the encoding of novel motor sequences that are compatible with a previously learned schema. However, causal evidence for the necessity of M1 in this specific process was lacking.

The Objective:
The study aimed to provide causal evidence for the role of the left M1 in schema-mediated motor learning. Specifically, the researchers sought to determine if transiently disrupting the left M1 using inhibitory continuous theta-burst stimulation (cTBS) prior to learning would impair:

1. The retention of previously learned motor transitions.
2. The integration of novel motor transitions into a compatible schema.

2. Methodology

Experimental Design:

• Participants: 48 healthy young adults (24 per group) were pseudo-randomly assigned to either an Active Stimulation (STIM) group or a Sham Stimulation (SHAM) group.
• Protocol: A two-session design separated by approximately 24 hours.
  • Session 1 (Day 1): Participants learned a bimanual motor sequence (Sequence 1: 4-7-3-8-6-2-5-1) using a Serial Reaction Time Task (SRTT). Motor-evoked potentials (MEPs) were recorded pre- and post-learning to assess corticospinal excitability changes.
  • Session 2 (Day 2): Participants received either active cTBS or sham stimulation over the left M1 hotspot. Five minutes post-stimulation, they learned a novel sequence (Sequence 2: 4-7-2-8-6-3-5-1).
• Schema Compatibility: Sequence 2 was designed to be highly compatible with Sequence 1. 75% of the elements occupied the same ordinal positions. This created two distinct transition types in Sequence 2:
  • Learned Transitions: 4 transitions already present in Sequence 1 (retention).
  • Novel Transitions: 4 new transitions not present in Sequence 1 (schema integration).

Stimulation Protocol:

• Target: Left M1 (hotspot for the right First Dorsal Interosseous muscle).
• Active Group: Continuous Theta-Burst Stimulation (cTBS) consisting of 600 pulses (3 pulses at 50Hz, repeated every 200ms) at 80% of the resting motor threshold (rMT). This protocol is known to induce long-term depression (LTD)-like suppression of cortical excitability.
• Sham Group: Identical pattern delivered via a coil with a 33mm spacer to mimic the sensation without inducing neural modulation.
• Timing: Stimulation occurred immediately before the learning of Sequence 2.

Measurements:

• Behavioral: Response Time (RT) and Accuracy on the SRTT.
• Physiological: Corticospinal excitability measured via MEP amplitude (120% rMT) collected pre- and post-intervention.
• Controls: Sleep quality, vigilance (PVT), and mood were monitored to ensure no confounding variables.

3. Key Contributions

• Causal Test of M1 in Motor Schemas: This is the first study to attempt to causally disrupt the left M1 to test its necessity in integrating novel motor information into a pre-existing schema.
• Differentiation of Learning Stages: The study specifically isolated the effects of stimulation on retention (learned transitions) versus integration (novel transitions) within a schema-compatible context.
• Investigation of cTBS Variability: The study provides data on the inter-individual variability of cTBS effects on MEPs in a motor learning context, highlighting the "responder vs. non-responder" phenomenon.

4. Results

Behavioral Findings:

• No Stimulation Effect: There was no significant difference between the STIM and SHAM groups in Response Time or Accuracy.
• Retention: cTBS did not impair the performance of previously learned transitions.
• Integration: cTBS did not impair the learning or integration of novel transitions into the schema. Both groups learned Sequence 2 at similar rates.
• Exploratory Analysis: Even when splitting the STIM group into "responders" (those showing MEP suppression) and "non-responders," no significant performance differences were found.

Physiological Findings (MEPs):

• Learning Effect: As expected, MEP amplitudes significantly decreased after learning Sequence 1 in Session 1, indicating learning-induced plasticity.
• Stimulation Failure: Contrary to the hypothesis and previous literature, the cTBS protocol failed to significantly suppress MEP amplitudes in the STIM group compared to the SHAM group.
  • Approximately 50% of participants showed suppression (responders), while the other 50% showed an increase or no change (non-responders).
  • At the group level, the net effect was null.
• Correlations: A significant correlation was found between the magnitude of learning-induced MEP decrease (Session 1) and stimulation-induced MEP decrease (Session 2) only in the STIM group responders, suggesting a subset of individuals is more susceptible to both plasticity and stimulation.

5. Significance and Discussion

Primary Conclusion:
The study failed to provide evidence for a causal role of the left M1 in schema-mediated motor sequence learning. The lack of behavioral impairment suggests that either:

1. The left M1 is not critical for this specific type of schema integration, or
2. The stimulation protocol failed to effectively disrupt M1 function.

Discussion of Null Results:
The authors propose several reasons for the discrepancy between their null results and previous studies showing M1 disruption impairs motor learning:

• Task Differences: Previous studies often used unimanual, probabilistic, or implicit tasks. This study used a bimanual, deterministic, explicit task. The complexity and bimanual nature may engage different neural networks or make the task more robust to M1 disruption.
• Stimulation Efficacy: The most likely explanation is that the cTBS protocol did not successfully suppress M1 excitability in the majority of participants (only ~50% responded). The timing of the task (12 mins post-stimulation) may have also missed the peak window of cTBS effects (often 5–10 mins).
• Schema Robustness: The pre-existing schema might provide a level of robustness that protects the learning process from transient M1 disruption, or the integration process relies more heavily on other regions (e.g., mPFC or hippocampus) when a strong schema is present.

Implications:
These findings challenge the assumption that M1 is universally necessary for all forms of motor learning. They suggest that the role of M1 may be context-dependent (e.g., critical for initial implicit learning but less so for schema-mediated explicit learning) or that individual variability in TMS response is a major confounding factor in such studies. The results highlight the need for more reliable methods to induce state changes in M1 or to identify biomarkers for TMS responsiveness before drawing causal conclusions about motor memory.