Cognitive inflexibility and frontal-cortical activation in pediatric obsessive-compulsive disorder

Abstract

Objective: Deficits in cognitive flexibility and response inhibition have been linked to perturbations in cortico-striatal-thalamic circuitry in adult obsessive-compulsive disorder (OCD). Although similar cognitive deficits have been identified in pediatric OCD, few neuroimaging studies have been conducted to examine its neural correlates in the developing brain. In this study, we tested hypotheses regarding group differences in the behavioral and neural correlates of cognitive flexibility in a pediatric OCD and a healthy comparison (HC) sample.

Method: In this functional magnetic resonance imaging (fMRI) study, a pediatric sample of 10- to 17-year-old subjects, 15 with OCD and 20 HC, completed a set-shifting task. The task, requiring an extradimensional shift to identify a target, examines cognitive flexibility. Within each block, the dimension (color or shape) that identified the target either alternated (i.e., mixed) or remained unchanged (i.e., repeated).

Results: Compared with the HC group, the OCD group tended to be slower to respond to trials within mixed blocks. Compared with the HC group, the OCD group exhibited less left inferior frontal gyrus/BA47 activation in the set-shifting contrast (i.e., HC > OCD, mixed versus repeated); only the HC group exhibited significant activation in this region. The correlation between set shifting-induced right caudate activation and shift cost (i.e., reaction time differential in response to mixed versus repeated trials) was significantly different between HC and OCD groups, in that we found a positive correlation in HC and a negative correlation in OCD.

Conclusions: In pediatric OCD, less fronto-striatal activation may explain previously identified deficits in shifting cognitive sets.

Figure 1

Set-shifting paradigm. Note: On each trial, three objects were presented for 500 ms with a 1500 ms interstimulus interval. One object had a unique attribute, either shape (square or circle) or color (gray or white). Subjects indicated the location of the unique object via button press. Unbeknownst to the subject, trials were presented in a blocked fashion, with 15 trials per block. In shape blocks, shape identified the unique object. In color blocks, color identified the unique object. In mixed blocks, the unique attribute alternated between shape and color.

Figure 2

Group differences in frontal activation and correlations between shift cost and caudate activation during set-shifting. Note: (A) The left inferior frontal gyrus activation in response to mixed versus single conditions was lower in the obsessive-compulsive disorder (OCD) group compared with the healthy comparison (HC) group. The group difference in activation [(−38, 32, −4), Z = 4.09, k = 31]is displayed with a peak threshold of p < .001. (B) Percent signal change values were extracted from the group difference in left inferior frontal gyrus activation for the mixed and repeated conditions relative to fixation and plotted. (C) The correlation between activation and the shift cost (i.e., the reaction time differential between mixed versus repeated conditions) showed a group difference in the right caudate [(8, 8, 2), Z = 3.58, k = 46]. The group difference in correlation is displayed with a peak threshold of p < .001. (D) Percent signal change values were extracted from the group difference in correlation found in the right caudate and plotted against shift cost for each group. The HC group showed a positive correlation (Pearson's r = 0.53, p < .02), and the OCD group showed a negative correlation (Pearson's r = −0.61, p < .02) between percent signal change in the caudate and shift cost.