Perceived threat mediates the relationship between psychosis proneness and aggressive behavior.

Psychotic symptoms are associated with aggressive tendencies, but this relationship is both complex and imperfect. In contrast to psychotic disorders, little is known about aggressive behavior and sub-clinical psychotic symptoms (e.g., "psychosis proneness"), which are relatively common in the general population. Threat/control-override (TCO), which is the propensity to overestimate the likelihood that an outside agent will (1) inflict harm (threat) or (2) control one's behaviors (control-override), has been associated with aggression in both psychiatric and community samples. The purpose of this study was to determine if psychosis proneness is related to aggression, and if one or both aspects of TCO mediate this relationship. We hypothesized that the propensity to overestimate threat would mediate this relationship, but control-override would not. Sixty men and sixty women (mean age=20.00 years, sd=3.00) with no history of psychotic disorder completed measures assessing psychosis proneness, threat control/override, aggressive history, aggressive ideation, and aggressive behavior. Three structural equation models were tested: (1) Threat and control-override modeled as separate mediating variables, (2) TCO as a unitary mediating latent construct, and (3) TCO considered as part of a psychosis proneness latent variable. Results indicated that psychosis proneness is positively related to aggression and that the best model fit was obtained when threat and control-override were modeled as separate variables, with mediation through threat alone. The utility of TCO for explaining the relation between psychosis spectrum symptoms and aggression is discussed.