Social Learning and Preferences in Adolescents With Conduct Problems and Varying Levels of Callous-Unemotional Traits

Learning to successfully navigate the social world, in particular when to trust others and how to work together with them, is crucial to well-adjusted social development. This is especially the case during adolescence, when individuals are undergoing significant biological and social changes. Adolescents with conduct problems (CP) tend to have difficulties in social relationships and to display aggressive behaviors as well as reduced cooperation with others. This pattern appears to be particularly pronounced in adolescents with CP and high callous-unemotional traits (CP/HCU). However, very little is currently known about the mechanisms that might drive reduced cooperative behaviors in adolescent CP, and whether these differ for individuals with high vs low levels of CU traits.

We used a series of economic games to assess how adolescents with CP/HCU (n = 46), CP, and lower levels of CU traits (CP/LCU) (n = 46), and typically developing adolescents (TD) (n = 59) interacted with social (human) and non-social (computer) partners that varied in their degree of cooperation (trustworthy vs untrustworthy and friendly vs unfriendly), and whether this related to group differences in social preferences (aversion to inequality) and prior beliefs.

Adolescents with CP (both HCU and LCU) had more difficulty than TD adolescents in differentiating between trustworthy and untrustworthy social environments in our task. Adolescents with CP/LCU also had more difficulty coordinating with friendly and unfriendly social partners to produce rewarding outcomes than TD adolescents. Surprisingly, we saw no relationship between participants' inequality aversion or prior beliefs and social learning in our games.

These findings indicate that, under controlled experimental conditions, adolescents with CP have more difficulty learning to differentiate between social environments that vary in cooperation, particularly adolescents with CP/LCU. These findings were not explained by inequality aversion or prior beliefs. Our findings also raise important questions regarding methods used to understand the mechanisms underlying social behaviors in adolescents with CP.