Effectiveness of CBT in Changing Attention Biases in Children with ASD and Comorbid Anxiety

Background:   Due to high rates of co-morbid anxiety and ASD, interventions have been developed to specifically target anxiety reduction. Although these interventions have proven effective in reducing parent or child-reported levels of anxiety, few studies have documented changes in cognitive processes. Anxious children without ASD tend to have an attentional bias to threat-related stimuli such as angry faces; changes in attention bias have been found with CBT (Waters et al., 2012). Attention bias has not been studied in anxious children with ASD although an object bias is typically found in children with ASD (Moore et al., 2012). 

Objectives:  The purpose of this study was to (1) Replicate the attentional bias towards objects compared to faces in children with ASD compared to those with typical development and (2) Examine whether attentional bias in children with ASD was changed after receiving a CBT-based anxiety intervention. 

Methods: Twenty-six high-functioning children with ASD (8-14 years) and 18 age-matched children with typical development (7-14 years) completed a dot-probe task to measure attentional bias. Each trial consisted of an emotional face (happy or angry), or an object paired with a neutral face. Participants were asked to respond to a target probe (e.g., an asterisk) that appeared immediately after the two stimuli. Faster reaction time (RT) to the target presented at the location of an emotional face or object compared to a neutral face was taken as evidence of an emotional bias (faster RT to happy or angry faces) or an object bias (faster RT to objects). Twelve children with ASD were diagnosed with comorbid anxiety and participated in a 14-week CBT anxiety-intervention program (Fighting Worries and Facing Your Fears; Reaven et al., 2009). The intervention focused on emotion identification, anxiety coping strategies, and graded exposure. Children completed the dot-probe task pre- and post-intervention to examine changes in attentional bias. 

Results:  Participants with typical development responded quicker to probes at the location of emotional faces over neutral faces (17ms) and neutral faces over objects (27ms), both results indicating a bias toward attending to faces, especially emotional faces. Participants with ASD (with and without comorbid anxiety) showed a preference for objects over neutral faces (17ms) but no preference for emotional faces over neutral faces (2ms). This interaction between diagnosis and stimulus type was a large effect, F(1, 42) = 21.80, p<.001.

 After intervention, children with ASD showed attentional bias scores that were similar to those with typical development including a preference for emotional faces over neutral faces (10ms) and a preference for neutral faces over objects (15ms). A time x stimulus type interaction was observed, F(1,11)=4.37, p=.06. 

Conclusions:   Children with ASD showed more bias towards objects compared to emotional faces than children with typical development suggesting that children with ASD were not attending to potentially important social cues in their environment. However, children who received the anxiety intervention showed increased attention to faces, especially emotional faces. These findings suggest that CBT can alter underlying cognitive processes in children with ASD and comorbid anxiety.