QUANTITATIVE EEG DURING STROOP TESTING UNMASKS PATHOLOGICAL AGING IN COGNITIVELY HEALTHY INDIVIDUALS

Abstract

Stroop interference has been shown to differentiate higher risk for Alzheimer's disease (AD) in cognitive healthy (CH) aging population, but no studies using detailed brain processing in this population have been described. We aim to explore quantitative electroencephalography (qEEG) during Stroop interference test to investigate the underlying neural oscillatory events of interference in pre-symptomatic AD. We employed qEEG to investigate brain activity during interference processing (Stroop) in a pre-symptomatic AD pilot study. Study participants, age 60–100 years, with normal cognition were recruited from the local community, consisting of two subgroups based on cerebrospinal fluid (CSF) proteins: cognitively healthy with normal amyloid/tau ratio (CH-NAT, n=20) or pathological amyloid/tau ratio (CH-PAT or pre-symptomatic AD, n=30). Behavioral performance and spectral power from 21 recording head sensors were collected during testing. EEG signal regularity (ir) or “noise” levels were measured by spectral entropy (SE). The relationships of alpha power, alpha SE, behavioral measures during Stroop, and spinal fluid amyloid levels were compared for CH-NAT versus CH-PAT. The behavioral performance was similar between the two groups during congruent or incongruent trials. The alpha power from occipital sensors were significantly different (p=0.034) between CH-NAT and CH-PAT groups during the congruent interference trials. The SE from CH-PAT brain occipital alpha signals were higher compared to CH-NATs (p=0.030). There are positive correlations between congruent trial alpha power and resting state heart rate (p<0.05, r>0.35) for frontal, central, left lateral, right lateral, and occipital regions in CH-PAT, but not CH-NAT group. The congruent trial alpha SE was negatively correlated with education (p<0.03, r<-0.60) over frontal and parietal regions in CH-NAT, but not for CH-PATs. Congruent trial alpha SE negatively correlated with fasting glucose (p<0.01, r<-0.45) over central and parietal regions and negatively correlated with CSF amyloid level (p<0.05, r<-0.35) for parietal, left lateral, and occipital regions in CH-PATs, but not in CH-NATs. These results suggest that neuroplasticity occurs during interference processing in the pre-symptomatic AD stage to likely compensate for compromised biochemistry.