Simultaneous STN-LFP and EEG recordings in Parkinson's disease patients revealed distinct subcortical electrophysiological signatures in sleep-dependent regulation of neural arousal, with STN-LFP aperiodic features alone classifying sleep stages above chance and outperforming scalp EEG.
Key Findings
Results
A significant interaction between recording modality and sleep stage was found in spectral slope (aperiodic exponent) between STN-LFP and scalp EEG across PSG-defined sleep stages.
Study included 19 Parkinson's disease patients undergoing deep brain stimulation (DBS) who underwent simultaneous PSG and STN local field potential recording.
During wakefulness and REM sleep, STN-LFP exhibited a significantly steeper spectral slope compared to scalp EEG.
The dissociation between modalities diminished with sleep depth during N2.
During N3 (deep sleep), the relationship reversed, with EEG becoming steeper than LFP.
Results
STN-LFP maintained higher aperiodic exponents overall compared to scalp EEG across sleep stages.
Aperiodic 1/f spectral parameters (exponent and offset) were analyzed as markers of arousal across sleep stages.
The aperiodic component showed 'robust state dependence and modality differences' across subjects.
STN-LFP maintained higher exponents overall relative to scalp EEG.
These findings reflect distinct subcortical electrophysiological signatures in sleep-dependent regulation of neural arousal.
Results
STN-LFP aperiodic features alone classified PSG-defined sleep stages above chance and outperformed scalp EEG classification.
Classifiers were trained on aperiodic features (1/f exponent and offset) to test whether 1/f structure alone could recover PSG-defined sleep stages.
STN-LFP aperiodic features alone successfully classified sleep stages above chance.
STN-LFP classification outperformed scalp EEG using the same aperiodic features.
STN-LFP performance approached the ceiling achieved by the combined-modality (EEG + LFP) classifier.
Methods
The study investigated correlated cortical-subcortical spectral power between STN-LFP and EEG across PSG-defined sleep stages in Parkinson's disease patients.
19 patients undergoing DBS for Parkinson's disease were enrolled (ClinicalTrials.gov: NCT04620551).
Simultaneous polysomnography (PSG) and STN local field potential recordings were obtained.
Spectral power correlation between STN-LFP and EEG was evaluated during PSG-defined wakefulness, REM, N2, and N3 sleep stages.
Aperiodic 1/f spectral parameters (exponent and offset) were analyzed as markers of arousal across sleep stages.
Discussion
The findings were interpreted as offering potential biomarkers for both invasive and non-invasive closed-loop neuromodulation strategies targeting sleep dysfunction in Parkinson's disease.
The STN is described as 'highly interconnected with human sleep circuitry' and is identified as 'a current target of interest for sleep neuromodulation.'
Distinct subcortical electrophysiological signatures were identified in sleep-dependent regulation of neural arousal across brain networks in PD.
The authors propose these insights offer potential biomarkers for closed-loop neuromodulation.
Both invasive (STN-LFP based) and non-invasive (EEG based) neuromodulation strategies are considered as targets.
What This Means
This research suggests that the brain's deep structures involved in Parkinson's disease — specifically the subthalamic nucleus (STN), a region targeted by deep brain stimulation — show distinct electrical activity patterns during different stages of sleep that differ meaningfully from what is measured on the scalp with standard EEG. By simultaneously recording both scalp EEG and electrical signals directly from the STN in 19 Parkinson's disease patients during overnight sleep studies, researchers found that the STN and scalp EEG behave differently depending on the sleep stage: during wakefulness and REM sleep, the STN's electrical signal had a steeper frequency slope than EEG, but during deep (N3) sleep, this relationship flipped, with EEG becoming steeper than the STN signal.
A key part of the study focused on a mathematical property of the electrical signals called the 'aperiodic' or '1/f' component — essentially the overall slope of the brain's electrical activity across frequencies — which is thought to reflect how alert or aroused the brain is. The researchers found that this aperiodic property tracked sleep stages robustly in both the STN and EEG, but the STN signals were actually better at distinguishing sleep stages than scalp EEG alone, nearly matching the performance achieved when both signals were combined.
This research suggests that the STN plays an active role in regulating sleep-wake states in Parkinson's disease, and that its electrical signals carry information about sleep stages that goes beyond what scalp EEG alone captures. These findings could be important for the development of 'closed-loop' deep brain stimulation systems — devices that could automatically detect sleep stages and adjust stimulation accordingly — potentially offering a new approach to treating the severe sleep problems that many people with Parkinson's disease experience.
Check Your Own Numbers
Upload your bloodwork. We'll cross-reference your results against this study and 4,700 others.
Hirt L, Martini R, Tang S, Summers M, West L, Kushida C, et al.. (2026). Shared Spectral Parameters Between EEG and Basal Ganglia LFP Distinguish Sleep-Specific Oscillations in Parkinson's Disease.. European journal of neurology. https://doi.org/10.1111/ene.70733