Although risk-taking has been studied from a developmental perspective, no study has examined how anxiety, age, risk-valence and social context interact to modulate decision-making in youths. This study probes this question using a risk-taking task, the Stunt Task, in clinically anxious children (n = 17, 10 F, age = 8.3–12.1 years), healthy children (n = 13, 4 F, age = 9.3–12.2 years), clinically anxious adolescents (n = 18, 6 F, age = 12.3–17.7 years), and healthy adolescents (n =14, 10 F, age = 12.5–17.3 years). Social context was manipulated: in one condition, participants were led to believe that a group of peers were observing and judging their performance (peer-judge), while, in the other condition, they were led to believe that peers were not observing them (control). Only anxious children showed an influence of social context on their risk-taking behavior. Specifically, anxious children bet significantly less and had slower reaction times (RT) during the peer-judge than control condition. However, across social conditions, risk-valence modulated RT differently in function of age and diagnosis. Anxious children were slower on the positive-valence risky trial, whereas anxious adolescents were slower on the negative-valence risky trials relative to their respective healthy peers. In conclusion, clinically anxious children were the only group that was sensitive (risk-averse) to the effect of a negative peer-judge context. The negative peer-judge context did not affect risky decision-making in adolescents, whether they were anxious or healthy. Future work using a stronger aversive social context might be more effective at influencing risky behavior in this age group.
