Neurophysiological response and connectivity changes in the STN with a month-long exercise intervention

Parkinsons disease (PD) affects over 1 million people in the U.S., posing challenges in symptom management. Treatments include medications like levodopa, surgical options such as deep brain stimulation (DBS), and rehabilitative approaches like exercise. While the effects of levodopa and DBS on basal ganglia circuits--especially the subthalamic nucleus (STN)--are well studied, exercise-induced changes in STN activity remain underexplored. This study investigates the short-and long-term effects of a month-long motorized cycling intervention on STN activity, using local field potentials (LFPs) recorded from 29 electrodes across 18 STNs in 9 participants. Consistent with our previous work, dorsal STN regions showed stronger 1/f noise, peak oscillatory power, and 8-35 Hz activity than ventral regions. Acute effects were minimal, but long-term data revealed increased peak and alpha/beta power in dorsal STN--distinct from the beta suppression seen with levodopa or DBS, suggesting a unique exercise-driven modulation. We further analyzed dorsal-ventral interactions using phase slope index (PSI) and imaginary coherence (iCOH). Low PSI (< 0.1) indicated minimal crosstalk, implying independent dorsal and ventral oscillators. However, iCOH showed beta coherence initially rose significantly over time. Using a novel System Structure (SStr) algorithm, we assessed causality between dorsal and ventral signals. Preliminary findings suggest both are influenced by a hidden driver, likely the exercise. In summary, exercise induces sustained, distinct changes in STN activity, possibly mediated by extrinsic inputs acting in parallel on dorsal and ventral regions. Further studies are needed to clarify these mechanisms.