Experimental Risk-Taking in Teens with an Autism Spectrum Disorder.

Background: Puberty is the hallmark of increased real-life risk-taking in typically developing (TD) teens. As teens with an autism spectrum disorder (ASD) develop non-typically, they might take fewer risks than their TD peers. One study investigated risk-taking in teens with and without ASD with an adjusted version of the Balloon Analogue Risk Task for youth (BART-Y; South, Dana, White, & Crowley, 2011), and found no group differences. This study needs replication with the original BART-Y.

Objectives: We will examine whether ASD teens show less risk-taking than TD teens on the BART-Y. In line with South et al.(2011), we will explore IQ and anxiety as predictors of experimental risk-taking. Additionally, we will include pubertal stage as a predictor, and explore whether parent-reported real-life risk-taking is positively associated with BART-Y performance.

Methods: 34 teens with a clinical DSM-IV-TR/DSM-5 ASD diagnosis and 34 age- and gender-matched TD teens (age:12-16, IQ>80) performed the BART-Y. They were assessed on ASD symptoms (SRS [and 3Di diagnostic interview for ASD group]), IQ (WISC-III-NL Block Design and Vocabulary), anxiety (SCARED-71, parent-report), pubertal stage (PDS, self-report), and risk-taking (CAMEL, parent-report).

Results: Groups did not differ on age, gender, and IQ. As intended, ASD teens had higher SRS scores than TD teens (p<.0001). Group significantly predicted BART-Y performance (ASD<TD;β=-5.54,p=.04) and explained 6.1% of its variance. However, Bayesian regression indicated only anecdotal evidence for the hypothesis that groups differed on the BART-Y, BF₀₁=1.47. All other predictors were non-significant. None of their Bayes factors showed substantial evidence for the hypothesis that these factors predicted BART-Y performance. Real-life risk-taking was higher in ASD as compared to TD teens (p=.008, BF₀₁=5.1). BART-Y performance did hardly correlate with CAMEL scores (r=.089. p=.473).

Conclusions: Although preliminary analyses indicated less risk-taking in ASD as compared to TD teens, Bayesian statistics showed that evidence for this group difference was only anecdotal. It is, therefore, more valid to conclude that teens with ASD do not differ from TD teens in experimental risk-taking, in line with South et al.(2011). IQ, anxiety, as well as pubertal stage did not impact experimental risk-taking. Surprisingly, parents reported that ASD teens take more risks in real-life than their TD peers. The next step is to determine whether this is also supported by self-report. Understanding the origin of the contrast between subjective and objective measures of risk-taking during puberty is important to fully grasp risk-taking in ASD teens in this crucial developmental stage.