Specificity of Risk Averse Decision-Making to the Autism Spectrum

Background:   

Our previous studies of decision-making in youth with Autism Spectrum Disorder (ASD) have shown evidence for risk aversion during several decision-making tasks (e.g., Johnson et al., 2006). Specifically, mathematical modeling analyses have revealed that people with ASD attend to loss and avoid potentially “risky” behaviour significantly more than their typically developing peers.

Objectives:   

To examine the specificity of these findings to ASD, we examined risk-aversion in clinical groups that are characterized by symptoms frequently observed in ASD, namely anxiety and attention problems. We hypothesized that a risk-averse decision-making style would also be present in a group of participants diagnosed with an Anxiety Disorder, but not in youths with Attention Deficit/Hyperactivity Disorder.

Methods:  

Current analyses included 28 youths with ASD (24M/4F), 16 with an Anxiety Disorder (11M/5F), 10 with ADHD (6M/4F), and 83 typically developing controls (40M/43F). Data collection for the ADHD group is ongoing. The ADHD and Anxiety groups (mean = 11.2 and 11.3 years, respectively) were significantly younger than the ASD group (mean = 13.1) and thus, we created two control groups. One matched the ASD group (N = 36) and the other matched the Anxiety and ADHD groups (N= 54), on age, IQ, and sex ratio. All participants completed a battery of decision-making tasks including the Iowa Gambling Task (IGT) and Balloon Analogue Risk Task (BART). In addition to traditional group comparisons of task performance, we utilized formal cognitive models to assess IGT and BART performance. The ASD group and their controls did not differ with regard to anxiety severity (Screen for Child Anxiety Related Disorders). The Anxiety group had a significantly higher anxiety score than all other groups.  The ASD and ADHD groups were higher than their control groups and the Anxiety group on the Conners Rating Scale-Revised. The ASD group far exceeded all other groups on the Social Responsiveness Scale, a measure of autism symptom severity.

Results:  

Despite reporting minimal anxiety, the ASD group showed significantly fewer risky choices on the BART compared to the controls, whereas the ADHD and Anxiety groups did not differ from their control group. Mathematical modeling results for the BART indicated group differences on two parameters: 1) the choice consistency parameter, with the ASD group exhibiting more consistent (i.e., less exploratory) responses than the ADHD and control groups; and 2) the sensitivity to rewards parameter, with the ASD group showing lower sensitivity to rewards relative to control and ADHD participants. In contrast to our main hypothesis performance of the Anxiety group did not differ from controls on any measure. Currently, there are no group differences on the IGT.

Conclusions:

Overall, these findings indicate that risk-averse decision-making may be specific to youth with ASD. Although our ASD group had a low level of anxiety, they showed more risk-averse behaviour than a group of youth with Anxiety Disorders. As expected, youth with ADHD showed a pattern of risky decision-making. Results will be discussed in the context of the preference for sameness and aversion to change often associated with the autism spectrum.