Attentional Biases in Currently Depressed Children: An Eye-Tracking Study of Biases in Sustained Attention to Emotional Stimuli

Abstract

Cognitive theories state attentional biases contribute to the development and maintenance of depression. Like depressed adults, there is growing evidence for the presence of attentional biases to sad stimuli in depressed youth. Although the direction of this bias among children remains unclear, preliminary evidence indicates attentional avoidance of sad stimuli in children. This is the first known sudy to use eye-tracking to investigate the exact nature of attention biases among depressed children. To assess sustained attention, the current study used eye-tracking and a passive viewing task in which children viewed a series of four facial expressions (angry, happy, sad, neutral) presented simultatiously for 20 s on a computer screen. The current study compared the attentional allocation of currently depressed children (n = 19; M age = 11.21) to a group of never depressed children (n = 22; M age = 10.82). Consitent with earlier research with children, we found that children with current major or minor depression, compared to children with no history of depression, exhibited attentional avoidance of sad facial stimuil as well as some evidence for preferential attention to happy faces. This study provides additional evidence that although depressed children demonstrate mood congruent attentional biases like that observed depressed adults, the nature of these biases may reflect attentional avoidance of sad stimuli, rather than preferential attention.

Notes

¹Although we present here the results with Epoch treated as a fixed, categorical variable, we also examined Epoch as a continuous (linear or nonlinear) influence in these models. In all instances, interactions with depression group were nonsignificant.

²In this study, only three of the participants included in the MDD group met criteria for minor depressive disorder. As noted in the Method section, these participants had four rather than the five required symptoms for as DSM MDD diagnosis. To determine whether the results presented were unduly affected by the inclusion of these three children, we reconducted each of our analysis focusing only on children meeting criteria for MDD. The Group × Emotion interaction remained significant, F(3, 312.77) = 8.25, p < .001. As before, the group difference in attention to sad faces was significant, F(1, 34.37) = 4.57, p < .04, though the group difference in attention to happy faces was reduced to nonsignificant with this smaller sample size, F(1, 32.78) = 3.19, p = .08.

³The specific results for the nonsignificant group difference in attention to happy faces with the inclusion of each covariate are child age, F(1, 33.41) = 3.35, p = .08, r_{effect size} = .30; gender, F(1, 33.06) = 3.57, p = .06, r_{effect size} = .31; race, F(1, 33.50) = 2.08, p = .16, r_{effect size} = .24; and current anxiety, F(1, 33.59) = 2.41, p = .13, r_{effect size} = .26.