Pallidal Spectral and Phase-Amplitude Coupling Differences in Parkinsons Disease Locomotor States

This study characterizes distinct spectral and phase-amplitude coupling patterns in the Globus Pallidus internus across locomotor states in Parkinson's disease, revealing that specific neural modulations correlate with clinical motor scores and freezing-of-gait while demonstrating an inverse relationship to subthalamic nucleus activity.

The Gist

Imagine your brain's movement control center as a bustling, high-tech orchestra. In a healthy brain, the musicians (neurons) play together in perfect harmony, switching seamlessly between a slow, resting melody and a fast, energetic dance tune when you decide to walk.

In Parkinson's Disease, however, the orchestra gets stuck. The musicians start playing a loud, repetitive, and sluggish rhythm (called "beta waves") that locks the body in place, making it hard to start moving or to walk smoothly. This is why patients often experience stiffness, slowness, or freezing in their tracks.

To fix this, doctors often perform a procedure called Deep Brain Stimulation (DBS). Think of this as installing a "smart conductor" (an electrode) deep inside the brain to listen to the orchestra and send electrical signals to help the musicians play in time again.

For years, scientists have focused on one specific section of the orchestra, the Subthalamic Nucleus (STN), to understand these rhythms. But this new study decided to listen to a different, equally important section: the Globus Pallidus Internus (GPi).

Here is what the researchers found, explained simply:

1. The "Sit, Stand, Walk" Test

The researchers asked six patients with Parkinson's to do three simple things while their brain activity was recorded:

• Sit (Resting)
• Stand (Preparing to move)
• Walk (Moving through an obstacle course)

They wanted to see how the brain's "music" changed between these states.

2. The Big Discovery: Two Different "Volumes"

The study found that the brain uses two different types of volume controls to switch between these states:

• The "High-Volume" Switch (Beta & Gamma): When the patients went from sitting to walking, the "loud, stuck" beta rhythm quieted down, and a fast, energetic "gamma" rhythm turned up. This is like the orchestra realizing, "Okay, we're moving now!" and switching from a slow dirge to a fast-paced march.
• The "Low-Volume" Switch (Delta): When the patients went from standing to walking, a very slow, deep rhythm (delta) got louder. This seems to be the specific signal that tells the body, "We are no longer just standing still; we are actually moving forward."

3. The "Teamwork" Mystery (Phase-Amplitude Coupling)

This is the most fascinating part. Imagine the brain's rhythms as a drummer (low frequency) and a violinist (high frequency). In a healthy brain, the drummer's beat tells the violinist exactly when to play a loud note. This teamwork is called Phase-Amplitude Coupling (PAC).

• In the STN (the old target): When people walk, the drummer and violinist get more synchronized. They team up harder to drive movement.
• In the GPi (the new target): The researchers found the opposite happened! When the patients walked, the teamwork between the low and high rhythms actually broke down.

The Analogy:
Think of the GPi as a strict traffic cop.

• When you are sitting or standing, the traffic cop is very active, directing traffic with strict hand signals (high teamwork/coupling) to keep cars (movement) from crashing.
• When you start walking, the traffic cop steps back and stops signaling so strictly (low teamwork/coupling). This "release" allows the cars to flow freely.

The study suggests that for the GPi to let you walk, it actually needs to stop micromanaging the different brain rhythms. It's like a manager who needs to stop giving constant instructions so the team can just get the job done.

4. Why Does This Matter?

The researchers also looked at how these brain signals matched the patients' real-world symptoms.

• The change in the slow rhythm (Delta) when walking was strongly linked to how severe a patient's overall Parkinson's symptoms were.
• The change in the teamwork (PAC) when walking was linked to how often patients experienced "freezing of gait" (suddenly feeling like their feet are glued to the floor).

The Takeaway

This study is like finding a new map for a different part of the brain. It tells us that the GPi doesn't just copy the STN; it does the exact opposite to help us move.

Why is this exciting?
Currently, "smart" DBS devices (Adaptive DBS) are being built to automatically adjust stimulation based on brain signals. Most are programmed to listen to the STN. This paper suggests that if we program these devices to listen to the GPi instead, we might need to set the rules differently. Instead of turning up the volume when the brain tries to move, we might need to turn down the strict control signals.

By understanding these unique "dance moves" of the GPi, we can build better, smarter devices that help people with Parkinson's walk more naturally and with fewer side effects.

Technical Summary

1. Problem Statement

Deep Brain Stimulation (DBS) of the Globus Pallidus internus (GPi) is a standard therapy for Parkinson's Disease (PD), particularly for managing levodopa-induced dyskinesias. While the Subthalamic Nucleus (STN) has been extensively studied regarding its local field potential (LFP) signatures (specifically beta-band power and phase-amplitude coupling) across different motor states, comparable data for the GPi is lacking.

• The Gap: Existing adaptive DBS (aDBS) systems rely heavily on STN biomarkers. However, the GPi exhibits distinct and potentially reciprocal oscillatory dynamics compared to the STN. There is a critical need to characterize GPi spectral power and cross-frequency coupling (specifically Phase-Amplitude Coupling, or PAC) during naturalistic locomotor states (sitting, standing, walking) to identify viable biomarkers for closed-loop control in GPi-DBS.

2. Methodology

The study employed a prospective observational design involving six adults with PD who had undergone GPi-DBS implantation.

• Participants & Hardware:

  • Subjects: 6 PD patients (all reported persistent freezing of gait post-implant).
  • Device: Medtronic Percept PC DBS system implanted in the GPi (unilateral or bilateral).
  • Recording: LFPs were recorded at 250 Hz using the "Indefinite Streaming" mode with stimulation OFF and medication OFF.
  • Channels: Data was analyzed from bipolar channels (C0-C2 and C1-C3) confirmed to be within the GPi via post-operative MRI. For bilateral implants, the hemisphere contralateral to the worse clinical side was selected.

• Experimental Protocol:

  • Participants navigated an obstacle course designed to elicit gait disturbances (including narrow archways, turns, and dual-tasking).
  • States: LFP data was manually segmented into three distinct locomotor states: Sitting, Standing, and Overground Walking.
  • Synchronization: Ankle-mounted IMUs and surface EMG were synchronized with LFPs using a controlled TENS artifact injection to ensure precise temporal alignment.

• Data Analysis:

  • Spectral Power: Power Spectral Densities (PSD) were calculated using Welch's method. Bandpower was analyzed across canonical bands: Delta (0.5–4 Hz), Theta (4–8 Hz), Alpha (8–12 Hz), Beta (13–30 Hz, subdivided into Low 13–20 Hz and High 21–30 Hz), and Gamma (30–100 Hz).
  • Phase-Amplitude Coupling (PAC): Quantified using the Modulation Index (MI) based on Kullback–Leibler divergence. Phase frequencies (1–30 Hz) were coupled against amplitude frequencies (40–100 Hz).
  • Statistics: Pairwise comparisons between states used the Wilcoxon signed-rank test with Benjamini–Hochberg correction.
  • Clinical Correlation: Significant spectral features were correlated with clinical scores (UPDRS-III for motor severity and nFoG-Q for freezing of gait) using robust linear regression.

3. Key Contributions

1. First Detailed GPi Locomotor Characterization: This study provides the first comprehensive characterization of GPi LFP dynamics across sitting, standing, and walking states in PD patients.
2. Reciprocal Dynamics Discovery: It identifies a reciprocal relationship between GPi and STN regarding beta-gamma PAC during walking, challenging the assumption that biomarkers are uniform across basal ganglia targets.
3. Biomarker Identification for aDBS: It proposes specific spectral features (Delta power modulation and Beta-Gamma PAC) as candidate biomarkers for adaptive stimulation in the GPi, distinct from those used in STN.

4. Key Results

• Spectral Power Differences:

  • Walking vs. Sitting: Differentiated by High Beta (21–30 Hz) and Gamma (30–100 Hz) power. Walking was associated with decreased High Beta and increased Gamma power.
  • Walking vs. Standing: Differentiated by Delta (0.5–4 Hz) power. Walking showed increased Delta power compared to standing.
  • Note: The distinction between sitting and walking was not apparent when analyzing the entire beta band (13–30 Hz) as a single unit, highlighting the functional importance of the High Beta sub-band.

• Phase-Amplitude Coupling (PAC):

  • General Trend: PAC between lower-frequency phases (Theta, Alpha, Beta) and Gamma amplitude was significantly reduced during walking compared to sitting and standing (Sit $\approx$ Stand > Walk).
  • Reciprocity with STN: When compared to prior STN data (Farokhniaee et al., 2024), the GPi showed an inverse trend. In the STN, beta-gamma PAC increases during walking; in the GPi, it decreases.

• Clinical Correlations:

  • Delta Power: The modulation of Delta power from standing to walking strongly correlated with UPDRS-III scores ($R = 0.9, p < 0.05$).
  • Beta-Gamma PAC: The modulation of Beta-Gamma PAC from standing to walking highly correlated with Freezing of Gait (nFoG-Q) scores ($R = 0.86, p < 0.05$).

5. Significance and Implications

• Mechanistic Insight: The findings suggest that movement coordination in the GPi relies on a balance of single-frequency mechanisms (Delta/Gamma power) and cross-frequency mechanisms (PAC). The reduction in PAC during walking may reflect a "release" of inhibitory gating, facilitating motor execution, whereas the STN may utilize increased coupling for the same purpose.
• Adaptive DBS (aDBS) Development: These results argue against a "one-size-fits-all" biomarker approach for aDBS.
  • For GPi-DBS, algorithms should consider Delta power (for general motor severity) and Beta-Gamma PAC (specifically for freezing of gait) as control signals.
  • The inverse relationship with STN implies that multi-target DBS systems must account for opposing neural dynamics to avoid conflicting stimulation strategies.
• Future Directions: The study highlights the need for larger cohorts to validate these preliminary correlations and suggests future research should employ simultaneous dual-target (STN and GPi) recordings to directly map the reciprocal oscillatory dynamics during locomotion.

Limitations: The study is limited by a small sample size ($n=6$), which restricts statistical power for clinical correlations. Additionally, the lack of simultaneous STN recordings in this specific cohort prevents direct within-subject comparison of the reciprocal dynamics, though the authors compared their GPi data to existing STN literature.