UC Berkeley

Adolescence is a period in development characterized by a greater tendency to engage in risky behaviors. The onset of adolescence is marked by puberty, which involves a dramatic rise in sex steroids. Existing neurobiological models have proposed that the rise in sex steroids during puberty influences the development of the adolescent brain, in particular the brain regions involved in the processing of socio-emotional information. According to these models, adolescents engage in more risk taking compared to children and adults because they process rewards differently and are more sensitive to their social environment. In a separate line of research, it has been hypothesized that changes in sleep during adolescence also contribute to the greater tendency to take risks. In this dissertation, we explored the role of pubertal hormones, social information, and sleep in adolescent risky decisions. To measure risk taking, we designed a child-friendly probabilistic decision-making game called the Jackpot task. In this task, participants could choose to take a risk or play it safe based on explicitly provided information about the risk level and stakes involved in their decision, and the type of cumulative performance feedback (social or monetary) they received. This task was administered in two independent samples of adolescents while they were lying in an MRI scanner, which allowed for examination of the reward-related brain processes associated with their risky choices. Pubertal hormone levels were measured based on saliva provided by the participants. Sleep was measured using a self-report questionnaire that was administered across five days. Participants also filled out various measures to capture individual differences in personality. While we did not find evidence for a peak in reward-related brain activation during adolescence, we did find large individual differences among adolescents in their behavior as well as their neural responses to rewards and social information. Reward-related brain responses associated with risk taking corresponded with higher levels of testosterone and stronger self-reported approach tendencies. The influence of social feedback, in the context of risky decisions, was stronger in girls with higher levels of estradiol and girls who reported greater susceptibility to peer influence. Adolescent girls with a preference for later bedtimes and with more irregular sleep patterns tended to make more risky decisions. These findings provide insight into some of the factors that contribute to adolescent risk taking and highlight the importance of using an interdisciplinary approach to investigate adolescent behavior.