The Weakened Relationship Between Prestimulus Alpha Oscillations and Response Time in Older Adults With Mild Cognitive Impairment

Front Hum Neurosci. 2020 Mar 12:14:48. doi: 10.3389/fnhum.2020.00048. eCollection 2020.

Abstract

Background: Prestimulus alpha oscillations associated with preparatory attention have an impact on response time (RT). However, little is known about whether there is a deficit in the relationship between prestimulus alpha oscillations and RT in older adults with mild cognitive impairment (MCI). Method: We collected electroencephalography (EEG) data from 28 older adults with MCI and 28 demographically matched healthy controls (HCs) when they were performing an Eriksen flanker task. For each participant, single-trial prestimulus alpha power was calculated for combinations of congruency (congruent vs. incongruent) and response speed (fast vs. slow). Result: Statistical analysis indicated that prestimulus alpha power was significantly lower for fast trials than slow trials in HCs but not in older adults with MCI. The Fisher's z scores of the within-subject correlation coefficients between single-trial prestimulus alpha power and RT were significantly larger in HCs than in older adults with MCI. In addition, machine learning analyses indicated that prestimulus alpha power and its correlation with RT could serve as features to distinguish older adults with MCI from HCs and to predict performance on some neuropsychological tests. Conclusion: The reduced correlation between prestimulus alpha activity and RT suggests that older adults with MCI experience impaired preparatory attention.

Keywords: machine learning; mild cognitive impairment; preparatory attention; prestimulus alpha oscillations; response time.