Does shame mediate the influence of trauma on psychosis? A systematic review and meta-analytic structural equation modelling approach

Meta-analytic evidence has linked shame separately to both potentially traumatic events (PTEs) and psychosis, but the influence of shame on the relationship between PTEs and psychosis has not yet been examined. This study used meta-analytic structural equation modelling (MASEM) to examine whether shame plays a mediatory role between PTEs and experiences of psychosis. A nested search was conducted within a previous systematic review on psychosis and shame to identify articles that contained a measure of PTEs. Included studies reported a quantitative association between psychosis and shame, and additionally a quantitative relationship between either i) PTEs and psychosis; or ii) PTEs and shame. Of the 40 articles initially included, 14 met criteria and 13 were included in the analyses. Overall, shame partially mediated the relationship between PTE's and psychosis, observed through a significant indirect effect (β = 0.15, 95 % CI: 0.11–0.19) and a reduction in the direct path that remained significant (β = 0.13, 95 % CI: 0.06–0.20). Indirect paths through shame between childhood PTEs and psychosis (β = 0.07, 95 % CI: 0.03–0.11), and between lifespan PTEs and psychosis (β = 0.09, 95 % CI: 0.03–0.11), were both small but significant. Both direct paths remained significant, suggesting that shame acts a partial mediator for both types of PTE. Shame is one path through which potentially traumatic events may influence the experience of psychosis and should be considered alongside other affective types in future modelling of psychosis. Qualitative research may aid further understanding of the mechanisms by which shame operates in this relationship.