Adolescent paranoia: Prevalence, structure, and causal mechanisms

Poster B55, Tuesday, October%209, 11:30%20am%20-%201:00%20pm, Essex%20Ballroom

Jessica C. Bird1, Robin Evans1, Waite Felicity1, Bao S. Loe2, Daniel Freeman1; 1University of Oxford, 2University of Cambridge

Paranoia is heightened during adolescence, but little is known about the content of paranoid thoughts at this age. Paranoia likely builds upon normative adolescent social concerns and may be shaped by the psychological and social context of adolescence. In this study we examine the prevalence, structure, and the probabilistic causal mechanisms of adolescent paranoia using a Bayesian approach with Directed Acyclic Graphs (DAGs). A cross-sectional UK school cohort of 801 adolescents (11-15 years) completed measures of paranoia and hypothesised affective, cognitive, and social factors. Paranoid thoughts were common, followed a continuous distribution, and were hierarchically structured. Fifteen percent of participants scored within an estimated clinical range for paranoia. A second-order model of items demonstrated excellent fit comprising a paranoia factor and sub-factors of social fears, conspiracy, and physical threat. All variables predicted paranoia in a linear regression, but a DAG sampled from the posterior distribution identified independent relationships between paranoia and negative affect, peer difficulties, bullying, and engagement with social media (i.e. safety behaviours, self-comparison, and affective reactivity). The direction of some relationships was uncertain, however, a directed effect was indicated for negative affect causing paranoia, and paranoia causing peer difficulties and social media safety behaviours. This study supports a continuum of paranoid thoughts in young people, rooted within the context of adolescence. Although causal conclusions are limited in cross-sectional data, this study demonstrates clear interactions between paranoia and affective and social processes in adolescence. Interventions targeting these mechanisms are needed to test their causal role.

Topic Area: Other

Back to Poster Schedule