Developmental Aspects of Affective Decision Making in ASD

Background: We previously presented data showing superior behavioral performance in ASD on an affective decision making task (the Iowa Gambling Task; IGT), in a sample of young, high-functioning ASD adolescents (ages 11-16) compared to age- and IQ-matched typical controls. Because there was no correlation in the ASD group between behavioral performance and anticipatory skin conductance response, we suggested that the ASD group used a hyper-rational strategy to achieve good results. However, we also noted significant correlations with age in both ASD and the typical (TYP) group and wondered how different developmental trajectories may influence the interaction of rational versus emotional strategies for completing the IGT. We have since collected IGT data for 50 additional children ages 8-11 to investigate age-related influences.

Objectives: To explore how participant age influences both strategy and performance during decision making on the IGT and how this influence may differ across ASD and typical development.

Methods: The Iowa Gambling Task presents two decks of cards that have big gains but also occasionally large losses, resulting in overall net loss; and two decks with smaller gains but also relatively smaller losses, resulting in overall net gain. We report behavioral data for 136 participants ages 8-16 (69 ASD, 67 TYP; mean age=12.6, Full Scale IQ mean=110). Skin conductance response data were available for a subsample of 88 (42 ASD) participants. We analyzed younger versus older age groups split at the cut of 12.5 years that was reported by Schumann et al. (2004) for changes in the trajectory of amygdala growth in ASD.

Results: Behavioral data of the number of “good” versus “bad” deck choices across 5 blocks of 20 trials, using a larger sample that also includes a younger age range, again shows significantly overall better performance in the ASD than TYP group. Older children outperform younger children across both groups, however the relative superiority of ASD performance persists even at younger ages, including a significant 3-way interaction of diagnosis x age x trial block of choices. Analysis of anticipatory SCR before making choices from bad decks versus good decks shows that the younger ASD group, but not the older ASD group, shows greater SCR response before choosing bad than good decks; there is no such differentiation at all in the TYP group.

Conclusions: These data may represent a functional example of Schumann et al.’s (2004) hypothesis of overactive amygdala function in younger children with ASD that slows in later adolescence. The ASD group is more successful than controls in both groups but may depend on different strategies (gist-based versus rational-based) across development. The failure of the TYP group to perform well at any age, using the task software published by the task authors, is surprising although recent work (see Smith, Xiao & Bechara, 2011) shows that typical adolescents at least have difficulty inhibiting impulsive reward-based choices. The mechanisms of decision making in ASD vis-à-vis typical development remain poorly understood and more research here is needed, in addition to work on the effects of age.