P 39. Evaluation of cognitive subtypes in Parkinson's disease – Correlations with qEEG power spectra

Abstract

Introduction. According to the Dual Syndrome Hypothesis cognitive impairment in Parkinson's disease (PD) patients can be divided into two subtypes: The frontal type versus the posterior type. The risk of progression to PD dementia was described to be higher in patients initially presenting with posterior deficits. Additionally, cognitive impairment in PD is related to slowing of EEG background activity. The aim of the present study was to investigate the correlation between different cognitive subtypes and qEEG measures in PD-patients. Methods. Fifty-nine PD-patients completed a comprehensive neuropsychological test battery covering five cognitive domains (Executive Function, Episodic Memory, Attention, Visuo-Spatial Ability, Language). Additionally, a 256-resting-state EEG was recorded. For statistics, Principal Component Analysis was calculated to group the neuropsychological measures into independent factors. Factor scores were further correlated with qEEG power spectra through Pearson“s correlation. Results. From screeplot two factors were derived. Factor 1 consisted of tests measuring constructional praxis, visuo-episodic memory as well as Stroop interference. Factor 2 consisted of tests assessing executive functions, attention, verbal-episodic memory and language. Pearson's correlation with qEEG measures and Factor 2 showed significant positive correlations with Global Alpha_2 Power (10–13 Hz; r = 0.35, p < 0.01) and Global Beta Power (13–30 Hz; r = 0.28; p < 0.05). There was a negative correlation between Factor 2 and Global Theta Power (4–8 Hz; r = −0.31, p < 0.05). No correlation with Factor 1 was found. Conclusion. From Principal Component analysis two factors were derived of our comprehensive neuropsychological test battery. In our sample, fronto-temporal associated neuropsychological measures were meaningfully correlated with qEEG power spectra. In a further step, correlation with connectivity measures of the qEEG power spectra could allow a closer insight into underlying network dynamics.