Statistical Nonparametric fMRI Maps in the Analysis of Response Inhibition in Abstinent Individuals with History of Alcohol Use Disorder

Abstract

Inhibitory impairments may persist after abstinence in individuals with alcohol use disorder (AUD). Using traditional statistical parametric mapping (SPM) fMRI analysis, which requires data to satisfy parametric assumptions often difficult to satisfy in biophysical system as brain, studies have reported equivocal findings on brain areas responsible for response inhibition, and activation abnormalities during inhibition found in AUD persist after abstinence. Research is warranted using newer analysis approaches. fMRI scans were acquired during a Go/NoGo task from 30 abstinent male AUD and 30 healthy control participants with the objectives being (1) to characterize neuronal substrates associated with response inhibition using a rigorous nonparametric permutation-based fMRI analysis and (2) to determine whether these regions were differentially activated between abstinent AUD and control participants. A blood oxygen level dependent contrast analysis showed significant activation in several right cortical regions and deactivation in some left cortical regions during successful inhibition. The largest source of variance in activation level was due to group differences. The findings provide evidence of cortical substrates employed during response inhibition. The largest variance was explained by lower activation in inhibition as well as ventral attentional cortical networks in abstinent individuals with AUD, which were not found to be associated with length of abstinence, age, or impulsiveness.

Keywords: abstinent AUD; common cognitive control network mechanisms; executive functions; functional neuroimaging; reactive stopping; response inhibition; right hemisphere inhibition network.

Figure 1

Schematic diagram illustrating the Go/NoGo task.

Figure 2

Thirteen voxel clusters where (a) 7 regions in the right hemisphere and bilateral anterior insula were activated, (b) 3 regions in the default-mode network were deactivated, and (c) 2 regions in the primary sensory areas were deactivated during successful response inhibition. All figures are in radiological orientations.

Figure 3

Boxplots of the 1st principal component scores in the AUD and control groups. The scores were significantly different (p < 0.02) between groups, as determined by a nonparametric Mann–Whitney U test.

Figure 4

Right hemispheric cortical regions where positive BOLD response in the AUD group was signif-icantly lower relative to in the control group: (a) medially presupplementary motor area, (b) bi-lateral anterior insula, inferior and middle frontal cortices, and (c) posteriorly superior parietal lobule, angular gyri, and middle temporal cortices.