The severity of dissociation and dissociative phenomena in non-affective psychosis: a systematic review and meta-analyses

Abstract

Non-affective psychosis is associated with elevated risk of suicide, self-harm, and reduced life expectancy. Symptoms of dissociation are common in non-affective psychosis, and dissociative experiences are known to be linked to multiple markers of elevated clinical risk, including sleep disturbance and negative affect. The present review represents the first comprehensive quantification of dissociative phenomena severity across multiple instruments in non-affective psychosis. Six databases (Embase, MEDLINE, PsycINFO, PubMed, Scopus, Web of Science) were searched from inception and systematically screened. Eligibility was established against ICD/DSM criteria or service-defined categorisation, including first-episode psychosis, brief limited intermittent psychotic symptoms, and at-risk of psychosis mental states. In total, data were extracted for 22 measures. Four random effects meta-analyses were conducted for measures with three or more independent datasets: the Dissociative Experiences Scale (DES) (k = 51; n = 3,878); the Cambridge Depersonalization Scale (CDS) (k = 6; n = 351); the Examination of Anomalous Self Experience scale, using the dichotomous scoring method (EASE-D) (k = 11; n = 340); and the Examination of Anomalous Self Experience scale, using the continuous scoring method (EASE-C) (k = 6; n = 235). DES data yielded a pooled mean of 21.28 (95% CI 19.19–23.38), consistent with moderately elevated dissociation. CDS data yielded a pooled mean of 36.38 (95% CI 26.89–45.87), also raised relative to community norms. EASE-D pooled means (M = 17.76 95% CI 15.29–20.24) and EASE-C pooled means (M = 74.86 95% CI 52.36–97.35) both indicated high and clinically relevant loads of basic self-disturbance. High heterogeneity (I2 > 85%) across measures highlights considerable variability in dissociative severity across samples, suggesting that levels of dissociation and dissociative experiences are not uniform in populations who experience non-affective psychosis. The present findings emphasise the importance of assessing dissociation in non-affective psychosis, and where necessary, intervening clinically. Nonetheless, the development of trauma-informed, dissociation-targeted clinical interventions is required to optimise approaches to support. Throughout the literature, there is a pressing need for more harmonised measurement efforts and further mechanistic research.