Stretch versus shortening contractions subsequently decrease versus increase neural drive to the human tibialis anterior

This study demonstrates that following active muscle stretch or shortening, neural drive to the human tibialis anterior is respectively decreased or increased through adjustments in motor unit recruitment and discharge rates to account for residual force enhancement and depression, with discharge rate modulation being specific to contraction intensity.

The Gist

The Big Picture: The "Ghost" in the Muscle Machine

Imagine your muscles are like a car engine. Usually, if you press the gas pedal (send a signal from your brain) a certain amount, the car goes a certain speed (produces a certain amount of force). Scientists have spent decades trying to build a perfect map that says: "If the brain sends signal X, the muscle produces force Y."

However, this paper discovered that the map is broken when the car is moving up or down a hill.

The researchers found that muscles have a "memory." If you stretch a muscle while it's working, it suddenly becomes stronger than expected. If you shorten it while working, it becomes weaker. Because of this, the brain has to change how hard it pushes the gas pedal to keep the car moving at the same speed.

The Problem: If we try to guess how strong a muscle is just by looking at its electrical signals (like listening to the engine noise), we will get it wrong because the brain is secretly adjusting the pedal based on that "memory."

---

The Experiment: The Ankle Gymnastics

To figure this out, the researchers asked 17 people to do a specific exercise with their ankles (using the Tibialis Anterior muscle, which lifts your toes up).

They asked the participants to hold their ankle at a specific angle and lift with a specific amount of force (like holding a heavy box steady). But they did this in three different ways:

1. The Reference (The Baseline): Just hold the weight steady.
2. The Stretch (The Rubber Band): First, the machine pulled their foot down (stretching the muscle), and then they had to hold the weight steady.
3. The Shortening (The Squeeze): First, the machine pushed their foot up (shortening the muscle), and then they had to hold the weight steady.

The Goal: They wanted to see how the brain changed its electrical signals (the "neural drive") to keep the weight steady in all three scenarios.

The Findings: How the Brain Adapts

The researchers used high-tech sensors to listen to individual "workers" inside the muscle (called motor units) to see how fast they were firing. Here is what they found:

1. The Stretch Effect (Residual Force Enhancement)

The Analogy: Imagine you are pulling a heavy rubber band. Once you stretch it, it snaps back with extra energy.

• What happened: When the muscle was stretched, it naturally produced more force than the brain expected.
• The Brain's Reaction: The brain realized, "Whoa, I'm getting too much force! I need to relax." So, it turned down the volume on the electrical signals. It recruited fewer workers and told the existing workers to fire more slowly.
• Result: The muscle stayed strong, but the brain sent fewer signals to do it.

2. The Shortening Effect (Residual Force Depression)

The Analogy: Imagine you are running up a hill and then suddenly hit a patch of mud. Your legs feel sluggish and heavy.

• What happened: When the muscle was shortened, it naturally produced less force. It felt "tired" or "depressed."
• The Brain's Reaction: The brain realized, "Oh no, I'm not getting enough force! I need to work harder."
• The Twist: How the brain reacted depended on how heavy the load was:
  • Light Load (20% effort): The brain just hired more workers. It recruited new muscle fibers to help out, but the existing workers didn't speed up.
  • Heavy Load (40% effort): The brain was already using almost all the workers. It couldn't hire more, so it shouted louder. It told the existing workers to fire much faster to make up for the weakness.

Why This Matters: The "Broken" GPS

The main takeaway is that you cannot accurately predict how strong a muscle is just by measuring its electrical activity (EMG) if the muscle has just been stretched or shortened.

• Before this study: Scientists thought, "If the electrical signal is low, the muscle is weak."
• After this study: They realized, "Wait, if the muscle was just stretched, the electrical signal might be low, but the muscle is actually super strong because of the stretch memory."

The Real-World Analogy:
Imagine you are trying to guess how much a person is lifting by listening to their breathing.

• If they just lifted a heavy box and held it, they might be breathing calmly (low signal) because the box is "stuck" to their hands (stretch effect).
• If they just dropped a box and tried to lift a new one, they might be gasping for air (high signal) even though the new box is light, because their arms feel "sluggish" from the previous movement (shortening effect).

The Conclusion

The human body is incredibly smart. It constantly adjusts its "neural drive" (the brain's instructions) to compensate for the muscle's history.

• After a stretch: The brain relaxes because the muscle helps out.
• After a shortening: The brain works overtime because the muscle is struggling.

This means that for doctors, athletes, and engineers trying to measure muscle strength or design better prosthetics, they need to account for this "muscle memory." Otherwise, their calculations will be off, leading to inaccurate predictions of how much force a person can actually generate.

Technical Summary

1. Problem Statement

Accurately predicting muscle force from electromyography (EMG) signals is a fundamental challenge in biomechanics, crucial for injury prevention, rehabilitation, and estimating joint loading. Current models often fail because they do not account for Residual Force Enhancement (rFE) and Residual Force Depression (rFD).

• rFE: Following active muscle stretch, steady-state force is higher than predicted for a given activation level.
• rFD: Following active muscle shortening, steady-state force is lower than predicted.

To maintain a specific torque level despite these history-dependent force changes, the central nervous system (CNS) must modulate neural drive (the sum of motor unit recruitment and discharge rates). However, the extent and mechanisms of this modulation remain unclear. Conventional bipolar surface EMG cannot accurately isolate neural drive because EMG amplitude is influenced by peripheral muscle properties (e.g., fiber conduction velocity, geometry). Therefore, there is a need to assess individual motor unit discharge patterns to understand how the CNS compensates for rFE and rFD.

2. Methodology

Participants & Design:

• Subjects: 17 healthy, recreationally active adults (6 women, 11 men).
• Muscle: Tibialis Anterior (TA).
• Protocol: Participants performed dorsiflexion contractions at 20% and 40% of Maximum Voluntary Torque (MVT).
• Conditions: Three conditions were tested with matched final torque and joint angle (20° plantar flexion):
  1. Stretch-hold (STRhold): 1s ramp-up, 1s pre-activation, 1s active stretch (25°), followed by a 10s hold.
  2. Shortening-hold (SHOhold): 1s ramp-up, 1s pre-activation, 1s active shortening (25°), followed by a 10s hold.
  3. Fixed-end Reference (REF): 1s ramp-up, followed by a 13s hold (no dynamic length change).

Data Acquisition:

• Torque/Angle: Measured via an isokinetic dynamometer (IsoMed2000) at 2 kHz.
• EMG: High-density surface EMG (HDsEMG) using a 64-channel grid (13x5 electrodes) over the TA muscle belly, sampled at 2 kHz.
• Synchronization: Torque, angle, and EMG data were synchronized via TTL signals.

Signal Processing & Decomposition:

• Motor Unit Decomposition: The steady-state phase (6s duration) of valid trials was decomposed using two blind source separation algorithms: DEMUSE (gradient convolution kernel compensation) and MUedit (fast independent component analysis + gCKC).
• Validation: Decompositions were manually edited (DEMUSE) or automatically filtered (MUedit) based on Pulse-to-Noise Ratio (PNR > 28 dB) and Coefficient of Variation (CV < 40%).
• Metrics:
  • Normalized EMG amplitude (RMS).
  • Motor Unit Discharge Rates (DR).
  • Torque steadiness (Coefficient of Variation of torque).
  • Unique motor unit counts (recruitment differences).

Statistical Analysis:

• Linear mixed-effects models were used to analyze fixed effects (contraction intensity, condition) and random effects (participants, motor units).
• Post-hoc pairwise comparisons were conducted with Holm adjustment.

3. Key Contributions

• High-Resolution Neural Drive Assessment: This study utilized HDsEMG decomposition to track a significantly larger population of motor units (~19 matched units per person) compared to previous intramuscular EMG studies (which typically tracked ≤7 units).
• Dual-Algorithm Validation: By employing and comparing two distinct decomposition algorithms (DEMUSE and MUedit), the study ensured the robustness of the motor unit identification.
• Differentiation of Compensation Strategies: The study explicitly differentiated how the CNS compensates for rFE (stretch) versus rFD (shortening) and how these strategies vary by contraction intensity.

4. Key Results

Neural Drive Modulation:

• Following Stretch (rFE): Neural drive was decreased relative to the reference condition.
  • Normalized EMG amplitude was ~2% lower.
  • Discharge rates (DR) were ~1 Hz lower (approx. 6–10% reduction).
  • Fewer unique motor units were detected compared to the reference.
  • Note: This reduction was consistent across both 20% and 40% MVT, suggesting rFE compensation is intensity-independent.
• Following Shortening (rFD): Neural drive was increased relative to the reference condition, but the mechanism was intensity-dependent.
  • At 20% MVT: EMG amplitude was ~1% higher, but DR remained unchanged. The increase was driven primarily by additional motor unit recruitment (more unique motor units detected).
  • At 40% MVT: EMG amplitude was ~3% higher, and DR increased by ~1 Hz. The increase was driven by a combined strategy of additional recruitment and increased discharge rates.

Torque Steadiness:

• Contrary to previous findings (e.g., Jakobi et al., 2020), torque steadiness did not significantly differ between stretch, shortening, and reference conditions. The authors attribute this to the shorter muscle length used in this study, which may have induced less rFE than studies using longer muscle lengths.

Correlations:

• Strong positive correlations were found between normalized EMG amplitude and discharge rates ($r \approx 0.84–0.94$), though the relationship was not perfect due to recruitment changes.

5. Significance and Implications

• Refining Force Prediction Models: The study confirms that EMG-based force predictions are inherently inaccurate following dynamic movements because the CNS alters neural drive to compensate for history-dependent force changes. A simple linear relationship between EMG and force does not hold after stretch or shortening.
• Intensity-Specific Strategies: The CNS employs different strategies to overcome rFD depending on the force level. At low intensities, it relies on recruiting new fibers (which have full force capacity). At higher intensities, it must also increase the firing rate of already-depressed fibers, likely due to constraints in common synaptic input limiting independent motor unit control.
• Clinical and Biomechanical Applications:
  • Rehabilitation: Understanding that neural drive is modulated differently after stretch vs. shortening is vital for designing rehabilitation protocols that account for muscle history.
  • Exoskeletons/Prosthetics: Control algorithms for assistive devices must incorporate contraction history to accurately estimate user intent and required torque.
  • Subject-Specific Calibration: The magnitude of neural drive modulation varies between individuals and contraction intensities, suggesting that future EMG-to-force models require subject-specific and intensity-specific calibration.

In conclusion, the study demonstrates that the human nervous system actively and distinctively modulates neural drive to counteract the mechanical effects of rFE and rFD, with the specific strategy (recruitment vs. rate coding) being dependent on the type of contraction and the intensity of the effort.