Objective: To parametrize EEG power-spectra and examine the association of periodic and aperiodic components with cognitive ability in Parkinson’s disease (PD).

Background: PD is associated with cognitive decline, that impacts quality of life and increases caregiver burden. Parametrizing spectral power of EEG into periodic and aperiodic components has been suggested to reflect the physiological aspects of brain function. We characterized spectral power of EEG into components and investigated them in the context of cognitive ability in PD.

Method: We recruited 169 PD and 59 control participants. Resting-state EEG (awake, eyes closed for 9 min; 64 electrode cap) and Level-II neuropsychological testing, to classify cognitive status and to generate a global cognition score, were collected. After preprocessing, artefact-free EEG was available for 65 PD participants with normal cognition (PD-N), 42 PD with mild cognitive impairment (PD-MCI), 11 PD with dementia (PDD), and 40 controls. Power spectra were estimated using the multitaper method from 5-s epochs with a 50% overlap. Periodic (alpha range; 6-14 Hz) and aperiodic components were extracted for frontal, central, posterior, and occipital regions. Multiple linear regression and ANCOVA were used to examine associations with global cognition and compare groups, respectively, with age and sex as covariates.

Results: Alpha dominant frequency was higher in controls than in PD-N across all regions. PD-N showed a higher alpha dominant frequency in posterior and occipital regions compared to PD-MCI and PDD, which did not differ. Alpha dominant frequency was directly associated with global cognitive ability across all regions (all βs>0.39, p<0.0001). In the posterior and occipital regions, power of alpha frequency relative to the aperiodic component was associated with global cognition and was higher in HC compared to PD-N and PD-MCI. We also found a negative association between the periodic offset and global cognition across all regions. However, this negative association disappeared after adjusting for L-dopa equivalent daily dose in the PD cohort. Aperiodic components did not differ between groups.

Conclusion: We found the dominant frequency of alpha to be strongly associated with global cognitive ability in PD, which was most prominent in the posterior and occipital regions. A longitudinal follow-up is needed to examine whether these EEG features predict progression to dementia in PD.

« Back to 2022 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/eeg-power-spectra-parametrization-and-cognitive-ability-in-parkinsons-disease/