Brain Activity During Antisaccades to Faces in Adolescence

Abstract

Cognitive control and social perception both change during adolescence, but little is known of the interaction of these 2 processes. We aimed to characterize developmental changes in brain activity related to the influence of a social stimulus on cognitive control and more specifically on inhibitory control. Children (age 8-11, n = 19), adolescents (age 12-17, n = 20), and adults (age 24-40, n = 19) performed an antisaccade task with either faces or cars as visual stimuli, during functional magnetic resonance brain imaging. We replicate the finding of the engagement of the core oculomotor and face perception brain regions in all age-groups, with increased involvement of frontoparietal oculomotor regions and fusiform face regions with age. The antisaccade-related activity was modulated by stimulus category significantly only in adolescents. This interaction was observed mainly in occipitotemporal regions as well as in supplementary motor cortex and postcentral gyrus. These results indicate a special treatment of social stimuli during adolescence.

Keywords: adolescence; cognitive control; development; face perception; neuroimaging.

Figure 1

Schematic representation of the antisaccade task. Participants are instructed to look toward the appearing visual stimulus (i.e., produce a prosaccades, PS) when the visual cue following the fixation cross is a green dot. They are instructed to look to the opposite side of the appearing visual stimulus (i.e., produce an antisaccade, AS) when the visual cue following the fixation cross is a red dot.

Figure 2

Antisaccade error rate as a function of stimulus category (face and car) in children, adolescents, and adults. Bars indicate standard errors of the mean (SEM).

Figure 3

Group-averaged Zmax in all ROIs in children, adolescents, and adults. Bars indicate SEM. (A) For the contrast antisaccades > prosaccades. (B) For the contrast prosaccades to face > prosaccade to car. Significant differences between groups (P < 0.05) are indicated by ^*.

Figure 4

(A) Whole brain univariate results for the production of correct antisaccades (anti vs. pro) in children, adolescents, and adults. (B) Whole brain univariate results for viewing faces (proFace vs. proCar) in children, adolescents, and adults. (C) Whole brain univariate results for the task × stimulus interaction in adolescents. The contrasts are displayed at a threshold using clusters determined by Z > 2.3 and a (corrected) cluster significance threshold of P = 0.05.

Figure 5

ROI-based MVPA: mean percentage decoding accuracy in children, adolescents, and adults. SVM classification accuracy for saccade category (task) detection was averaged per age-group per region. Statistical significance was assessed with 1-sample t-test against 50% chance and corrected for multiple comparisons using the FDR method. Above chance statistical significance (P < 0.05) is represented by ^*. Significant differences between groups of P < 0.05 are represented by ^*. (A) Saccade type (task) decoding. (B) Stimulus type (category) decoding.