Objective: To examine periodic and aperiodic EEG correlates of cognition in Parkinson’s disease (PD).

Background: Quantitative EEG metrics are increasingly used to describe variations in brain function associated with cognition in PD.[1,2] Such functional assessment is more repeatable yet complementary to direct brain imaging. A general slowing of EEG spectral frequency reportedly mirrors cognitive decline.[2,3] Classic analysis fails, however, to account for the aperiodic components of EEG signals. Therefore, we used a “Fitting Oscillations and 1/f” (FOOOF) algorithm[4] to assess EEG in a cross-sectional sample of Parkinson’s patients with normal cognition (PD-N), mild cognitive impairment (PD-MCI), or dementia (PDD), relative to controls (Con).

Method: EEG from a sample of 50 PD-N, 32 PD-MCI, 9 PDD, and 29 Con participants was analysed after study exclusions from 164 volunteers (NZBRI Parkinson’s Progression Programme; cognitive status based on MDS Level II criteria). Spontaneous brain activity during awake resting state (rs, eyes closed) was measured using a 64-channel EEG for 3 x 3 minutes (Quick-Cap, Neuroscan, Curry7 software; reference electrode, Cz / CPz; ground, AFz; 500Hz sampling frequency). ICLABEL identified artefacts within ICA components (MNE 1.4.0-Python ICA), with thresholds for brain and eye artefacts (EOG), after bad electrodes substituted by z-score interpolations. The rsEEG components were derived from FOOOF (https://fooof-tools.github.io/fooof/) in Python.

Results: FOOOF accurately fitted models for each group (global mean R-squared across all channels > 0.975 < 0.987). Periodic activity (independent of aperiodic) revealed a progressive slowing of the extended alpha centre frequency (5-13Hz) across Con, PD-N, PD-MCI and PDD; power and bandwidth metrics remained relatively stable. The beta frequency (13-30Hz) revealed decreases in power from Con to PD-N and further decreases in PD-MCI and PDD, but smaller relative differences for beta central frequency and bandwidth. For aperiodic components, the exponent (reduction in power across frequency) and offset (translation of the entire spectrum) showed increases in PD-MCI and PDD relative to PD-N and Con. Positive and negative correlations were found between a global cognition z-score and, respectively, the alpha centre frequency and exponent.

Conclusion: FOOOF analysis provided a more comprehensive perspective on the association between rsEEG and cognition in PD.

References: 1. Geraedts VJ, Koch M, Contarino MF, Middelkoop HA, Wang H, van Hilten JJ, Bäck TH, Tannemaat MR. Machine learning for automated EEG-based biomarkers of cognitive impairment during deep brain stimulation screening in patients with Parkinson’s disease. Clinical Neurophysiology. 2021 May 1;132(5):1041-8.

2. Anjum MF, Espinoza AI, Cole RC, Singh A, May P, Uc EY, Dasgupta S, Narayanan NS. Resting-state EEG measures cognitive impairment in Parkinson’s disease. npj Parkinson’s Disease. 2024 Jan 3;10(1):6.

3. Yassine S, Gschwandtner U, Auffret M, et al. Functional brain dysconnectivity in Parkinson’s disease: a 5-year longitudinal study. Mov Disord 2022;37(7):1444–1453. https://doi.org/10.1002/mds. 29026

4. Donoghue T, Haller M, Peterson EJ, Varma P, Sebastian P, Gao R, Noto T, Lara AH, Wallis JD, Knight RT, Shestyuk A, & Voytek B (2020). Parameterizing neural power spectra into periodic and aperiodic components. Nature Neuroscience, 23, 1655-1665. DOI: 10.1038/s41593-020-00744-x

« Back to 2024 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/periodic-and-aperiodic-eeg-correlates-of-cognition-in-parkinsons-disease/