Frontal Asymmetry and Reward-Based Decision Making in Children with High Functioning Autism Spectrum Disorder

Friday, May 12, 2017: 5:00 PM-6:30 PM
Golden Gate Ballroom (Marriott Marquis Hotel)
R. Gilbert, A. M. Zhou, J. Donehey, J. Buirkle and S. Faja, Boston Children's Hospital, Boston, MA

Children with autism spectrum disorder (ASD) have impairments in motivation and reward-based decision making compared to typically developing (TD) children (Dawson et al. 2005; Kohls et al. 2012; Faja et al. 2013). In a study of decision-making and goal directed behavior, researchers suggested that anxiety and other influences such as personality and temperament may be moderators of risk-taking behavior (South et al. 2011). Among TD children, temperament has been related to frontal asymmetry, but studies of frontal asymmetry in ASD have been less conclusive. Most resting state EEG studies of ASD have shown greater frontal asymmetry across alpha frequency in comparison to a TD sample (Wang et al. 2013) and within a high functioning ASD sample, individuals with greater right frontal asymmetry exhibited greater social impairment and better visual analytic skills than individuals with greater left frontal asymmetry (Sutton et al. 2005). Studying resting EEG asymmetry may offer insight on the neural underpinnings of motivational approach or avoidance in goal-directed decision making.


To explore the relationship between reward-based decision-making, anxiety and internalizing/externalizing behaviors, and frontal asymmetry among children with ASD.


Participants included 48 children diagnosed with ASD and 31 IQ and age-matched TD children. Participants completed the Hungry Donkey Gambling Task, which offers insight on long-term and short-term reward-based decision-making. Anxiety, internalizing, and externalizing behaviors were reported by parents using the Child Behavior Checklist (CBCL). EEG data was recorded from a subset of 28 subjects with ASD during eyes open and eyes closed resting conditions. Frontal asymmetry between F3 and F4 was calculated from the mean alpha power found during the eyes-closed condition, as this was found to be a better measure of baseline arousal (Barry et al. 2009).


In comparison to an age and IQ-matched sample of TD individuals, we found no differences in risk taking on the Hungry Donkey gambling task with our ASD sample. Looking at individual differences within the ASD sample, we also did not find significant correlations between anxiety or internalizing/externalizing behaviors reported in the CBCL and performance on the Hungry Donkey Task or with frontal asymmetry. However, we did find a significant correlation between reward-based decision making on the Hungry Donkey task in the final two blocks and frontal asymmetry, r(27)= -0.42, p<.05.


While there were no significant group differences in performance on the Hungry Donkey gambling task between groups, the correlation between risk taking and frontal asymmetry within the ASD sample poses interesting questions about frontal asymmetry as a biomarker of approach/avoidance behavior in both a social and nonsocial context. Further exploration is needed to understand this relationship, the heterogeneity of ASD, as well as group differences between ASD and TD populations. We plan on recruiting and collecting resting state EEG data from approximately 30 TD participants to compare to compare to our ASD sample. We also anticipate recruiting an additional 20 ASD participants over the coming months, which will increase statistical power across our measures.