LAUSR

Abstract Background Recently, the efficacy of a novel virtual reality based cognitive behavior therapy (VR-CBT) for paranoia was demonstrated. Evidence is growing that the maintenance of psychosis may be influenced by affective processes. This study examined how treatment with VR-CBT influenced positive and negative affect states, and whether the interplay between mental states was affected. Methods The sample consisted of 91 patients with a psychotic disorder randomized either to 16-session individual VR-CBT or treatment as usual. The experience sampling method (ESM; a structured diary technique) was used to assess mental states at baseline, post-intervention and 6-month follow-up. Mixed model analyses were conducted to study treatment effects. Lagged associations between mental states were estimated at baseline and post-intervention, and were visualized with networks. Results VR-CBT, but not treatment as usual, resulted in reduced levels of paranoia and negative affect. At pre-intervention networks depicting the dynamic interplay between mental states over time had limited significant connections, with most stable connections being auto-relations. I.e., paranoia was best predicted by paranoia at the previous moment. The dynamic interplay between affective states did not change over time after VR-CBT. Discussion We found that VR-CBT specifically targets paranoia and there are indications that VR-CBT had an enduring effect on negative emotions. These beneficial treatment effects do not seem to transfer to positive affective states. Unexpectedly, we did not find evidence that negative mental states such as feeling down or lonely triggered paranoia in the next moment even at pre-intervention, and these temporal relations between mental states did not change over time in response to treatment.