Attention Biases in Anxious Youth with Co-Occurring ASD Symptoms

Background:  Children with anxiety disorders typically show numerous processing biases that favor negative or threatening information and these processes are believed to be responsible for the etiology and/or maintenance of these disorders.  Children with autism spectrum disorders (ASD) frequently show elevated levels of anxiety, and examining the cognitive mechanisms in children with ASD symptoms and co-occuring anxiety may prove useful in better understanding the co-occurrence of these disorders.

Objectives:  We explored if children with elevated ASD traits demonstrate an attentional bias towards threat similar to that of typically developing, low-ASD children.  We used a standard dot-probe task that assessed attentional biases towards emotional (threatening or happy) vs neutral faces.

Methods: Fifty-nine typically developing children with a primary anxiety disorder diagnosis (mean age = 11.52) completed a dot-probe task (500 ms presentation). ASD symptoms were assessed with the Social Responsiveness Scale-Parent Version (SRS-P).  Anxiety disorders were diagnosed using the Anxiety Disorders Interview Schedule for Children, a semi-structured interview for diagnosing anxiety and other disorders.  Continuous measures of anxiety were assessed using the Multidimensional Anxiety Scale for Children, administered to parents and children.

Results: Of the 59 participants, 29 had elevated ASD symptoms (as determined by cut-off scores of 60 or higher, indicating moderate to severe symptom severity).  Children in the elevated ASD group were found to have higher parent-reported anxiety t(57) = -2.76, p = .01 and child-reported anxiety, t(57) = -3.05, p = .00.  Children in the elevated ASD group had more cases of a primary diagnosis of social phobia, χ² (1, N=59) = 9.05, and this was included as a control factor in regression analyses.  At the group level, neither group (low or elevated ASD) showed a significant bias towards threatening or happy faces.  However, regression analyses using parent-reported anxiety, ASD symptoms, and their interaction significantly predicted bias scores for happy faces, R²=.22, F(4,55)=3.61 p<.01.  Both SRS-P scores (β= .86, p < .05) and the interaction between SRS-P scores and parent-reported anxiety (β=-1.90, p< .05) significantly predicted biases towards happy faces.  Specifically, those with elevated SRS-P symptoms and high parent-reported anxiety showed an attention bias toward happy faces.  In contrast, those with high ASD symptoms and low parent-reported anxiety showed an attention bias away from happy faces.

Conclusions: In the present sample, no overall differences were found in the attentional biases of children with and without elevated ASD symptoms. However, a significant interaction between parent-reported anxiety and SRS-P symptoms was observed, such that children with high ASD symptoms and high anxiety were more likely to attend to happy faces than children with low SRS scores or low anxiety. Considering the majority of studies examining children and adults does not find biases towards happy faces, these results suggest that happy faces were particularly favored in attentional processing, perhaps due to being perceived as threatening, due to the difficulties in social processing seen in children with elevated ASD symptoms.  These results suggest that ASD symptoms play a role in the relationship between attentional biases and anxiety.