Test-Retest Reliability Of Behavior Measurements In Sustained Attention And Their Neuroelectrophysiological Mechanism

Sustained attention refers to focusing on a certain task for a period of time and reflects the stability of attention.Reaction time variability is considered as a critical cognitive index of sustained attention,it refers to the variations in an individual's reaction time during engagement in a continuous reaction time task.It is usually measured by the standard deviation of reaction time and the coefficient of variation of reaction time to indicate the overall fluctuation of sustained attention.It is worth noting that the fluctuation in sustained attention is periodical,which suggested by many researches.And the amplitude of low frequency fluctuation of reaction time as a new measure of sustained attention has attracted widespread attention from researchers.Although there are many research reports,the test-retest reliability of these behavioral measurements and the neurophysiological mechanism behind these measurements are still unclear.Therefore,two experiments are carried out to explore these problems in this study.In experiment 1,we examined the test-retest reliability of the behavioral measurements of sustained attention.41 healthy college students participated in two-choice reaction time task,twice over a mean interval of 7 days.The conventional measurements of reaction time variability including the standard deviation of reaction time and the coefficient of variation of reaction time were assessed first.Then the amplitude of low frequency fluctuation of reaction time was further calculated for sub-frequency bands including <0.01 Hz,0.01–0.027 Hz,0.027–0.073 Hz,and 0.073–0.167 Hz.The test-retest reliability of these measurements was evaluated through intra-class correlation.In experiment 2,an EEG experiment method was used to investigate the association between reaction time variability and EEG activities during sustained attention task.70 healthy college students participated in the experiment with the two-choice response time task.Behavioral measurements,i.e.,standard deviation,coefficient of variation and amplitude of low frequency fluctuation of individual data of reaction time,and EEG measurements,including the average power and the trial-by-trial variations of EEG rhythmic activities(Delta: 1-4 Hz,Theta: 4-8 Hz,Alpha: 8-14 Hz,Beta: 14-30 Hz and Gamma: 30-50 Hz)were calculated,and the EEG-behavior correlations were further assessed.Our results showed that the standard deviation of reaction time and the coefficient of variation of reaction time had high levels of test-retest reliability(ICC: 0.82-0.85).For the amplitude of low frequency fluctuation of reaction time,it showed relatively high levels of test-retest reliability in each frequency band(ICC: 0.65-0.85).There was no significant correlation between the standard deviation and the coefficient of variation of reaction time and the measurements of EEG rhythmic activities,while the amplitude of low frequency fluctuation of reaction time showed frequency-dependent correlations with the trial-by-trial variations of Theta and Delta.The trial-by-trial variation of Theta had correlation with the amplitude of low frequency fluctuation of reaction time in 0.01-0.027 Hz,and the trial-by-trial variation of Delta had correlation with the amplitude of low frequency fluctuation of reaction time in 0.073-0.167 Hz.These findings revealed good test-retest reliability for the behavior measurements of sustained attention and provided a methodological basis for future study of sustained attention.Moreover,the correlation between the trial-by-trial variation of EEG rhythmic activity and the amplitude of low frequency fluctuation of reaction time provided evidence for the neurophysiological underpinning of the periodic fluctuation of sustained attention.