T114. THE CAUSAL DYNAMICS OF PARANOIA IN PATIENTS WITH SCHIZOPHRENIA: A THEORY DRIVEN NETWORK ANALYSIS

BACKGROUND: A cognitive account identifies six key psychological maintenance factors for persecutory delusions. However, a complex system of causation is likely where these factors interact in their influence on paranoid ideas. We set out to evaluate the causal dynamics of paranoia with theory-driven network approaches. METHODS: 1809 patients with non-affective psychosis attending UK mental health services completed assessments of paranoia, hallucinations, insomnia, self-esteem, worry, anxious avoidance, analytic reasoning, and psychological well-being. To assess causal patterns, we estimated, first, an undirected partial correlation network and then, second, adopted a Bayesian approach with Directed Acyclic Graphs to discover the directed causal pathways best supported by the data. RESULTS: The networks showed that with all other variables controlled, paranoia had direct causal interactions with hallucinations, negative self-beliefs, insomnia, worry, and avoidance. Hallucinations and negative self-beliefs were most directly linked to paranoia, whereas indirect paths had prominent influences on the causal effects for insomnia, worry, and avoidance. The direction of these interactions was uncertain, but negative self-beliefs and insomnia were more likely to influence paranoia than vice versa. Self-report reasoning was likely unrelated to paranoia once other factors were controlled. Causal factors were highly interconnected, with insomnia, negative self-beliefs, avoidance, and worry most directly linked to other variables. Most interactions were likely reciprocal, except for hallucinations which were unlikely to influence other variables and significantly caused by insomnia and avoidance. DISCUSSION: The findings are consistent with a complex system of interacting causation in the maintenance of paranoia. The patterns observed support the cognitive model of persecutory delusions, highlighting multiple pathways of causal interaction between paranoia and theoretically important factors. Interventions directly targeting these factors are likely to lead to multiple benefits, alleviating paranoia both directly and indirectly through connections with other causally related symptoms.