Childhood Adversity is Associated with Altered Exploration and Reward-Driven Learning in Adolescents

• In Event: Early childhood adversity

Abstract

Study objective
Childhood adversity is often associated with problem behaviors later in life. This study investigates explore-exploit decision-making as a cognitive mechanism mediating this association. In uncertain environments, making decisions poses conflicts between staying with a known option (exploitation) versus trying a new, unknown alternative (exploration). Children and adolescents gather information and learn about their environments through sampling and exploration. It is possible that childhood maltreatment impairs social learning and leads to maladaptive behaviors through discouraging sampling and exploration.

This study examines whether childhood adversity is associated with reduced exploration. Further, it tests one possible explanation of this reduced exploration – inaction bias. Past research shows that acute stress suppresses adults’ learning to act (Go) in a Go No-Go task. It is hypothesized that chronic early life stress, such as childhood adversity, may similarly bias children to refrain from initiating actions. Specifically, we test the effect of parental unpredictability, a characteristic of adverse parenting that has been shown to alter cognitive processing differently from parental harshness.

Methods
This ongoing study (pre-registered) has thus far recruited 87 10-13-year-old adolescents; data collection and analyses will be completed before the conference. Participants completed a Horizon task (decision-making; Wilson et al., 2014) and a Go No-Go task (action learning; de Berker et al., 2016) in counterbalancing orders. In the Horizon task, participants played a two-armed bandit task and made decisions to tradeoff between rewards and information across 80 trials. In the Go No-Go task, participants used reward information (win/lose points) to learn how to respond to different stimuli across 120 trials. In addition, participants reported parental harshness and unpredictability using Parental Environment Questionnaire Conflict Subscale (Miller & Hauser, 1989) and Questionnaire of Unpredictability in Childhood (Glynn et al., 2018), respectively. We assessed working memory and socio-economic status as control variables.

Preliminary Results
Horizon task: Children who reported more unpredictable parenting explored less (β = -0.05, p = 0.005). Specifically, children with unpredictable, but not harsh, parents were more indifferent to the future utility of exploration (OR =0.72, 95% CI (0.58-0.89), p = 0.002; Figure 1), suggesting a less future-oriented decision-making strategy.

Go No-Go task: Children with unpredictable, but not harsh, parents, showed selectively impaired performance in No-Go trials. They had significantly more Go responses in No-Go trials even after repeated feedback (OR =1.69, 95% CI (1.22-2.34), p = 0.002; Figure 2).

Conclusions
Children with unpredictable and harsh parents explored less. Specifically, children with unpredictable, but not harsh, parents, used less strategic exploration, suggesting that parental unpredictability is associated with a present-oriented decision-making strategy. Also, children with unpredictable, but not harsh, parents were less successful in withholding their actions in reward-driven instrumental learning. This finding suggests potentially divergent effects of parental unpredictability and harshness, such that exposure to unpredictability may selectively impair future-oriented exploration and inhibitory control. This altered explore-exploit decision-making and impulsive actions may result from an adaptation to the demands of an unpredictable parenting environment. Yet they implicate compromised learning abilities and problem behaviors that we often see among individuals who experienced severe childhood adversity.

Authors

• Yuyan Xu, University of Wisconsin - Madison
  Presenting Author

• Seth David Pollak, University of Wisconsin - Madison