Levodopa’s strongest electrophysiologic effect clustered in the dorsal subthalamic nucleus
Levodopa did not produce a uniform subthalamic response in Parkinson’s disease. In this preprint, the medication preferentially reshaped activity in the dorsal portion of the subthalamic nucleus, where beta power fell far more than in ventral contacts and where multiband state-space measures shifted in a direction consistent with a more marked dopaminergic reset of pathologic motor-network signaling.
The study included twelve participants undergoing bilateral deep brain stimulation of the subthalamic nucleus, yielding thirty recorded contacts: fourteen dorsal and sixteen ventral. Recordings were obtained in both OFF-medication and ON-medication states, then analyzed with spectral parameterization and multiband state-space methods. In six participants, simultaneous electroencephalography added a directional readout of cortico-subthalamic coupling through the phase slope index.
Beta suppression was the clearest signal. Dorsal contacts showed a drop of −2.473 dB, compared with just −0.120 dB ventrally, a difference that reached statistical significance at p=0.0438. For a disease built around excessive beta synchrony, that regional divergence matters biologically: dorsal stimulation fields and dorsal recording contacts appear to sit closer to the circuitry in which levodopa can most effectively interrupt rigid, pathologic oscillatory coupling.
Beta suppression arrived with a broader multiband shift
The electrophysiology did not stop at a single frequency band. Dorsal theta power increased, and baseline normalization suggested a trend in the same direction, although the normalized change did not reach conventional significance. Even so, the pattern is clinically interesting because dopamine replacement often reduces the dominance of beta activity while allowing lower-frequency activity to re-emerge as the network becomes less locked into an akinetic state.
State-space analysis sharpened the same regional asymmetry. Dorsal contacts showed larger theta–beta shifts (+5.904 versus +1.610 dB; p=0.006) and larger gamma–beta shifts (+2.516 versus +0.384 dB; p=0.0188) than ventral contacts. Those changes were driven chiefly by beta suppression rather than by a regionally distinct gamma response, since gamma changes did not differ between dorsal and ventral recordings. In other words, the medication appears to reorganize the network by removing beta-heavy constraints first, rather than by uniformly amplifying faster rhythms across the entire subthalamic structure.
One sentence captures the core finding: levodopa acted like a spatially selective reset switch, not a global volume knob.
Directionality shifted toward cortical leading at selected electrodes
The cortical recordings added a second layer of interpretation. In six participants with simultaneous electroencephalography, dorsal low-beta and beta interactions shifted toward cortical leading at selected electrodes, while ventral responses were less consistent. Directionality is not just a mathematical flourish here. When cortical leading becomes more prominent, it suggests a change in how the cortex and subthalamic nucleus coordinate timing, and that timing may reflect a network that is less trapped in the bidirectional pathological loops typical of Parkinsonian physiology.
This finding is especially relevant for adaptive neuromodulation. Closed-loop systems depend on biomarkers that are not only detectable but anatomically meaningful. If dorsal contacts respond more robustly to levodopa, then contact location may determine which biomarker is most informative for sensing medication state, tracking symptom state, or eventually guiding stimulation algorithms that adapt in real time to network dynamics.
The study’s implication is straightforward: dorsal anatomy appears to carry a disproportionate share of dopaminergic reactivity within the subthalamic target. That is a practical observation for anyone interpreting recordings from implanted leads, because a contact sitting a few millimeters differently may sample a meaningfully different pharmacologic response profile.
Contact anatomy may matter as much as the drug state itself
The preprint positions contact anatomy as a key variable in how dopaminergic network responses are measured. This is not a trivial refinement. Subthalamic deep brain stimulation already depends on precise electrode placement, and levodopa-related electrophysiology may be equally sensitive to whether the contact lies in dorsal motor territory or more ventral subterritory. A biomarker built from pooled contacts could blur a real signal if the dorsal and ventral compartments respond differently.
From a pathophysiologic standpoint, the result fits with the known motor circuitry of Parkinson’s disease. Beta excess has long been linked to bradykinesia and rigidity, while dopaminergic therapy tends to relieve those symptoms partly by reducing exaggerated beta synchronization across the cortico-basal ganglia loop. The stronger dorsal effect suggests that levodopa may preferentially normalize the portion of the subthalamic nucleus most embedded in motor circuitry, leaving ventral regions with a weaker or less consistent response.
The authors call for clinical validation of spatially informed multiband biomarkers for adaptive stimulation. That next step is necessary because this work is still a preprint and involved a small sample. The recordings are compelling, but they do not yet establish how stable the dorsal-ventral distinction will be across a broader Parkinson’s population, different disease stages, or varying medication histories.
Small sample, mechanistic clarity, and an unfinished clinical question
The study is limited by its size and by the fact that only six participants had simultaneous electroencephalography for directionality analyses. The sample is enough to identify a reproducible anatomical pattern, not enough to settle how often it appears, how it varies by phenotype, or whether it predicts motor benefit, dyskinesia risk, or stimulation response.
Even so, the biological message is crisp. Dopaminergic therapy did not merely lower oscillatory activity in a generic way; it reorganized subthalamic dynamics in a spatially ordered fashion, with the dorsal contacts showing the clearest beta suppression and the most prominent multiband shifts. For adaptive stimulation design, that means future biomarker strategies may need to treat the subthalamic nucleus as a set of functionally distinct recording zones rather than as a single interchangeable target.
The remaining task is not to prove that levodopa changes brain rhythms—this study already does that. The harder clinical question is whether dorsal-contact biomarkers can be used reliably enough to steer therapy in real patients, with enough precision to improve symptom control without introducing new instability in stimulation programming.