4 93 Neural and affective responses to prolonged eye contact with parents and unfamiliar others in depressed and non-depressed adolescents Eight first-level SPM T-contrasts were specified, one for each condition (i.e., direct and averted gaze of parent, unfamiliar peer, unfamiliar adult, self). These T-contrast images were entered in a 2 × 4 full factorial ANOVA design with two within-subject factors (i.e., gaze direction and target). SPM F-maps were computed to assess main effects of gaze direction and target, and their interaction, followed up by post-hoc analyses between all conditions. Second, to examine group differences between DEP and HC adolescents, we entered previously described t-contrasts in a 2 × 2 × 4 full factorial ANOVA design with one between-subject factor (group) and two within-subject factors (gaze direction and target). SPM F-maps were computed to assess main effects of group, gaze direction and target, and their interactions (i.e., group × gaze direction, group × target, group × gaze direction × target), followed up by post-hoc analyses between all conditions. At the second-level, we first performed region of interest (ROI) analyses using independently defined functional ROIs (8-mm spheres MNI space) surrounding peak voxels of brain region based on previously found regions using the eye contact task in a sample of adults Wever et al. (2022) and Supplement S4.4). We used the MarsBar toolbox (Brett et al., 2002) to extract activity from five ROIs, i.e., left temporoparietal junction (TPJ), dorsomedial prefrontal cortex (dmPFC), bilateral inferior frontal gyrus (IFG), and right fusiform gyrus (FG). To assess the effects of group, gaze direction, target, and their interactions, we performed generalized linear mixed regression analyses and post-hoc tests in R. All ROI analyses were Bonferroni corrected for the number of tests (p <.05/5). Thereafter, to explore blood oxygenation level-dependent (BOLD)-responses in brain regions outside the ROIs, we performed complementary whole-brain analyses that were corrected for multiple comparisons with family-wise error cluster (FWE) correction at p <.05 (with a cluster-forming threshold of p <.001). To check whether results were not driven by differences in age, gender, handedness, and pubertal status of adolescents, we performed additional analyses to control for these variables (see Supplement S4.5 for details on associated questionnaires). RESULTS Adolescents’ general responses to prolonged eye contact in HC adolescents (aim 1) Affective responses To examine adolescents’ affective responses to eye contact in HC adolescents and how this may vary as a function of target, we performed a generalized linear mixed regression model with gaze direction and target, and their interaction on adolescents’ affect ratings (i.e., mood, connectedness, feelings about the targets).
RkJQdWJsaXNoZXIy MjY0ODMw