Retest Reliability of the N2 Event-Related Potential Component and Conflict Processing Behavioral Task in Children with High Functioning ASD

Background:

Evidence indicates good retest-reliability of electrophysiological measures of performance monitoring in the anterior cingulate cortex (ACC) over time in healthy adolescents and young adults and moderate retest reliability of the conflict monitoring N2 component in a large group of healthy individuals (Segalowitz et al., 2010; Clayson & Larson, 2013). However, to our knowledge, no investigations have explored the retest reliability of this component in children, particularly those with autism spectrum disorders (ASD), which is important considering its recent uses as neural developmental measure of executive control (EC) (Espinet et al., 2011). In addition, we will investigate the retest reliability of a modified version of the Change Task as a behavioral measure of EC within the same subjects. Ultimately, it is critical to examine the retest reliability of both the N2 component and the Change Task as the psychometric properties of EC measures are often poorly understood but represent an important domain in the exploration of intervention and biomarkers of treatment response.

Objectives:

To determine the retest reliability of the N2 event-related potential component and a conflict processing behavioral task in children with ASD.

Methods:

The test-retest reliability was evaluated for the conflict monitoring N2 event-related potential (ERP) and for an adapted version of the Change Task, which measures inhibitory control, shifting and monitoring behaviors. ERPs were obtained from 10 children diagnosed with ASD between the ages of seven and eleven, while they completed a modified version of the Eriksen Flanker Task (Rueda et al., 2004). The N2 ERP was calculated from the mean amplitude in electrodes Fz and those directly adjacent (EGI GSN: 19, 4). All participants completed the Flanker task at two time points, around ten-weeks apart and eight of these participants completed the change task at two time points, also around ten-weeks apart.

Results:

Results showed that the differential N2 amplitude between congruent and incongruent trials was not reliable over the two time points, while evidence showed that N2 amplitude in congruent trials was reliable over the 10-week period (ICC of .88, p < .05). In terms of behavior, the modified change task showed strong 10-week test-retest stability in measures of inhibitory control (ICC of .74, p < .05), shifting (ICC of .85, p < .05), and monitoring behavior (ICC of .98, p < .05).

Conclusions:

Although we found that the differential N2 amplitude was not stable over the 10-week time period, we found the general N2 amplitude in congruent trials was. In addition, the Change Task showed retest reliability over the 10-week period within these same subjects. This suggests that that behaviorally these subjects did not exhibit improvements in EC as measured by the Change task, despite changes in the differential N2. This could suggest that the N2 measure isn’t stable over time, or that the changes in neural EC perhaps, precipitate changes in behavior. Implications of this finding might be limited by our small sample size, which we anticipate will increase by at least a factor of two over the next few months.