PTSD as a mediator of the relationship between trauma and psychotic experiences, Daniela Strelchuk et al, 2020

PTSD as a mediator of the relationship between trauma and psychotic experiences

Daniela Strelchuk, Gemma Hammerton, Nicola Wiles, Jazz Croft, Katrina Turner, Jonathan Heron and Stanley Zammit

Psychological Medicine, 2020, 1–9.

doi : 10.1017/S0033291720004821



Background : Traumatic experiences are associated with a higher risk of psychotic illnesses, but little is known about potentially modifiable mechanisms underlying this relationship. This study aims to examine whether post-traumatic stress disorder (PTSD) symptoms mediate the relationship between trauma and psychotic experiences (PEs).

Methods : We used data from the Avon Longitudinal Study of Parents and Children to examine whether: PTSD symptoms mediate the relationships between (a) childhood trauma and adolescent PEs (study of adolescent PEs; n = 2952), and (b) childhood/adolescent trauma and PEs in early adulthood (study of adult PEs; n = 2492). We examined associations between variables using logistic regression, and mediation using the parametric g-computation formula.

Results : Exposure to trauma was associated with increased odds of PEs (adolescent PEs: ORadjusted 1.48, 95% CI 1.23–1.78; adult PEs: ORadjusted 1.57, 95% CI 1.25–1.98) and PTSD symptoms (adolescent PTSD: ORadjusted 1.59, 95% CI 1.31–1.93; adult PTSD: ORadjusted 1.50, 95% CI 1.36–1.65). The association between PTSD symptoms and PE was stronger in adolescence (ORadjusted 4.63, 95% CI 2.34–9.17) than in adulthood (ORadjusted 1.62, 95% CI 0.80–3.25). There was some evidence that PTSD symptoms mediated the relationship between childhood trauma and adolescent PEs (proportion mediated 14%), though evidence of mediation was weaker for adult PEs (proportion mediated 8%).

Conclusions : These findings are consistent with the hypothesis that PTSD symptoms partly mediate the association between trauma exposure and PEs. Targeting PTSD symptoms might help prevent the onset of psychotic outcomes.

Key words : Longitudinal; mediation; psychotic experiences; PTSD; trauma



Psychotic disorders are some of the most disabling illnesses worldwide (Kyu et al., 2018; Rehm & Shield, 2019). However, current treatment approaches have limited efficacy and a better understanding of the aetiology of psychotic disorders is needed to guide the development of novel targets for intervention. There is robust evidence that psychosis exists on a continuum (Linscott & Van Os, 2013). Trauma exposure is associated with the development of both psychotic experiences (PEs) in the general population and psychotic disorders in clinical samples, with meta-analyses describing moderate to large associations between childhood adversity and psychosis across this spectrum (Matheson, Shepherd, Pinchbeck, Laurens, & Carr, 2013; Varese, Barkus, & Bentall, 2012). However, less is known about the modifiable pathways through which trauma might contribute to the development of psychosis. A clear understanding of these mechanisms is important because it may allow the identification of targets for interventions to prevent psychosis in people with a history of trauma.

The clearest trauma-related psychopathological outcome is post-traumatic stress disorder (PTSD). PTSD disorder occurs in approximately 10–40% of individuals following exposure to a severe traumatic event (Kessler, Sonnega, Bromet, Hughes, & Nelson, 1995), though symptoms at a sub-diagnostic threshold level are equally common (McLaughlin et al., 2015). Individuals with PTSD have a substantially higher risk of developing schizophrenia and other psychotic disorders (Okkels, Trabjerg, Arendt, & Pedersen, 2017), and PTSD is present, but usually undetected, in approximately 29% of patients in secondary care with a psychotic disorder on screening (Zammit et al., 2018).

It has been proposed that psychotic symptoms could manifest as a more extreme representation of PTSD-like trauma-related psychopathology (Morrison, Frame, & Larkin, 2003). A recent multifactorial model of the role of PTSD in psychosis suggests that unhelpful emotion regulation strategies and high fragmentation of traumatic memories can give rise to intrusions which, depending on their interpretation and coping mechanisms, lead to PEs (Hardy, 2017). For example, the re-experiencing symptoms of PTSD can take the form of hallucinations if intrusions are not recognised as trauma-related and are attributed externally instead (Steel, Fowler, & Holmes, 2005). In addition, dissociation, which is common in PTSD and is associated with childhood trauma in people with psychosis (Rafiq, Campodonico, & Varese, 2018), can deprive the individual from internal and external anchors, increasing vulnerability to impaired reality testing (Allen, Coyne, & Console, 1997) and hallucinations
(Moskowitz & Corstens, 2008).

Despite the high prevalence of PTSD in psychosis and plausible explanations for an association, only a few studies have examined the role of PTSD as a mediator in the relationship between trauma and psychosis (Choi et al., 2015; Hardy et al., 2016; McCarthy-Jones, 2018; Murphy, Murphy, & Shevlin, 2015; Peach, Alvarez-Jimenez, Cropper, Sun, & Bendall, 2019; Powers, Fani, Cross, Ressler, & Bradley, 2016; Soosay et al., 2012). Whilst these studies found some evidence for mediation, they were all cross-sectional and therefore unable to determine the direction of relationships, which could have led to biased mediation estimates. Furthermore, concerns about selection bias and confounding in these studies (Williams, Bucci, Berry, & Varese, 2018) highlight that longitudinal studies that can more adequately address these methodological limitations are required to determine whether symptoms of PTSD lie on the causal pathway between trauma and psychosis.