Objective: We investigated the behavior of LFPs in 57 patients across therapeutic states using a data-driven method to find personalized physiomarkers.
Background: The power of beta oscillations (13-30 Hz) in subthalamic nucleus (STN) local field potentials (LFPs) fluctuates with levodopa and deep brain stimulation (DBS) in patients with Parkinson’s disease (PD), which may inform adaptive DBS (aDBS). However, it is unclear if beta power alone may reliably indicate the medication state during DBS, because a recent feasibility study (1) suggests that gamma oscillations (60-90 Hz) entrained at half the stimulation frequency may be more suitable.
Method: LFPs were recorded for one minute in 57 PD patients (109 hemispheres) with STN DBS across four therapy states: ON/OFF stimulation (median 1.9 mA, IQR: 1.6-2.4) and ON/OFF medication (median 150 mg, IQR: 100-200). One-second LFP epochs were converted into power spectra with 1 Hz resolution. We used cluster-based permutation testing (p < 0.01) and Cohen’s d to assess the stimulation and medication effect and their interaction on spectral power within each hemisphere. In each hemisphere, we selected the frequency bin with the highest effect size as personalized physiomarker.
Results: OFF-stimulation, we found a medication-related beta decrease in 59% of hemispheres. ON stimulation, the number of hemispheres with a significant effect was reduced to 32%. This difference was reflected in a significant interaction effect between medication and stimulation in 51% of hemispheres. OFF-stimulation we found a medication-related narrowband gamma increase (median frequency: 70 Hz) in 10% of hemispheres and for ON stimulation in 17% of hemispheres (median frequency: 62 Hz). We found stimulation-entrained gamma oscillations in 8% of hemispheres during the ON medication state and in 0 hemispheres during the OFF medication state.
Conclusion: In the majority of patients, personalized physiomarkers of the medication state were found in the beta frequency range even though they interacted with DBS. Only in some patients, when stimulation was applied, gamma frequencies reacted strongest to medication. These results are relevant for the programming and development of aDBS algorithms.
References: 1. Oehrn CR, Cernera S, Hammer LH, Shcherbakova M, Yao J, Hahn A, et al. Chronic adaptive deep brain stimulation versus conventional stimulation in Parkinson’s disease: a blinded randomized feasibility trial. Nature Medicine. 2024.
To cite this abstract in AMA style:
M. de Neeling, C. Oehrn, B. Keulen, M. Stam, D. Hubers, R. de Bie, P. Schuurman, B. van Wijk, A. Buijink, M. Beudel. Individual physiomarkers of levodopa-induced local field potential fluctuations interact with deep brain stimulation in Parkinson’s disease [abstract]. Mov Disord. 2025; 40 (suppl 1). https://www.mdsabstracts.org/abstract/individual-physiomarkers-of-levodopa-induced-local-field-potential-fluctuations-interact-with-deep-brain-stimulation-in-parkinsons-disease/. Accessed October 5, 2025.« Back to 2025 International Congress
MDS Abstracts - https://www.mdsabstracts.org/abstract/individual-physiomarkers-of-levodopa-induced-local-field-potential-fluctuations-interact-with-deep-brain-stimulation-in-parkinsons-disease/