The role of cognitive biases in adolescent social anxiety

Social anxiety disorder (SAD) is one of the most common mental health disorders. It is characterized by persistent and impairing fear of negative evaluation and avoidance of social encounters. Age-of-onset data suggests that SAD symptoms are often first experienced at the adolescent juncture (Kessler et al., 2005a). Cognitive models of SAD derived from adult literature highlight the importance of systematic biases in information processing in maintaining anxious responses to social situations. There is little work to date on the role of these biases in youth populations. In this thesis, I extend the work on the role of cognitive biases in social anxiety to adolescent populations with a particular focus on the role of biases in the interpretation of social ambiguity. In a first study, a cross-lagged design, interpretation bias predicted anxiety levels six months later and vice versa. I further found a link between stress, anxiety and interpretation bias such that interpretation bias mediated the effects of stress. In a second study, I developed a novel tool to probe interpretation biases using complex social scenes. In a third study, I employed this new paradigm in conjunction with eye- tracking to investigate attentional deployment alongside the interpretative process in relation to social anxiety levels. While levels of social anxiety were not themselves related to viewing patterns, the relationship was moderated by attentional control such that youths with increased social anxiety levels and high levels of attentional control spent more time on social emotional cues. The final study explored evidence for social transmission of social anxiety and interpretation biases in peer relationships. Using a novel social network approach, I examined whether adolescents were similar in social anxiety level or in interpretation bias to their close friends. Gender emerged as a strong predictor of friendship ties: adolescents were more likely to nominate a peer of the same sex as a friend. However, I did not find evidence for a link between social anxiety or interpretation bias level and friendship ties in this peer network. Further investigations using this multi-method approach, incorporating cognitive, attentional and network methods, may be the way forward to gain insight in the complexities of interactions in this at-risk population. Understanding whether and how adolescent social anxiety is unique in its phenotype and risk correlates is crucial for the development of novel, developmentally suitable psychological interventions.