Neural Correlates of Response Inhibition and Response Monitoring in Autism

Background: Several studies provide evidence for deficits in executive functions in autism, including on measures of response inhibition and response monitoring. These deficits may be particularly important in autism because they may lead to decreased ability to flexibly allocate resources to guide attention, thoughts and actions, and consequently may contribute to increased perseveration and repetitive behavior. Recently published functional magnetic resonance imaging (fMRI) findings reveal that, during correct response monitoring, adults with autism showed increased activation in rostral anterior cingulate cortex (rACC) compared to healthy controls, and that rACC activity was correlated with ratings of restricted and repetitive behavior (Thakkar et al 2008). A similar correlation between ACC function and repetitive behaviors was found with an executive function measure of shifting abilities using fMRI in adults with autism (Shafritz et al 2008).

Objectives: To examine the neural correlates of response monitoring and response selection/inhibition in children with high-functioning autism (HFA), and its correlation with measures of restricted/repetitive behavior extracted from the Autism Diagnostic Observation Schedule (ADOS) and the Autism Diagnostic Interview (ADI).

Methods: 13 children with HFA and 16 typically developing (TD) children completed a functional MRI scan while performing a Go/No-Go task. The neural correlates of correct response inhibition and commission errors (failure to inhibit responding to No-go stimuli) were examined using event related analysis to assess response inhibition and response monitoring, respectively. Groups were matched for commission errors, gender, age (8 to 12 years old) and non-verbal IQ (WISC IV Perceptual Reasoning Index).

Results: Contrast of commission (incorrect No-go) vs correct inhibition on No-go trials revealed that children with HFA showed increased activation in the medial prefrontal cortex (mPFC, BA10/9/32) and the left temporal gyrus (LTempG, BA20/21/22) compared to TD children. Subsequent analysis for each separate trial condition (commission error and correct inhibition trials) revealed abnormal activation in HFA in both regions and for both trial conditions. During correct inhibition, children with HFA showed significant deactivation in the mPFC and LTempG in contrast to TD children who showed no changes. During commission error, children with HFA showed significant activation in the mPFC and LTempG while TD children showed deactivation in these regions. Moreover, in children with HFA deactivation in the LTempG during correct inhibition was significantly correlated with measures of restricted/repetitive behavior (ADOS, p<0.009, ADI, p<0.06).

Conclusions:

Compared to the TD group, children with HFA show abnormal neural activity in the mPFC/rACC and LTempG during correct and incorrect response inhibition. In the present study, activation in the mPFC was not correlated with restricted and repetitive behavior scores. However, we found a significant correlation with deactivation in the LTempG. Currently, there is only limited data examining relationships between neural circuitry involved in response inhibition and response monitoring and repetitive behaviors in HFA. Further studies are warranted, however, our data and previous findings suggest that repetitive behaviors associated with autism might reflect dysfunction across a wider network of executive circuitry.

Shafritz KM, et al. (2008), Biol Psychiatry 63:974-980.

Thakkar KN, et al. (2008), Brain 131:2464-2478.