Evidence for a role for trauma in psychosis

In the only study not supporting the relationship (Spataro et al. Reference Spataro, Mullen, Burgess, Wells and Moss2004) cases of abuse were drawn from official record data, in contrast to the other studies, all of which used retrospective self-report of abuse. Spataro and colleagues compared hospital admission rates in those with a history of sexual abuse with admission rates in a general population control sample and found no association between child sexual abuse and later hospital admission with a diagnosis of schizophrenia. However, a major limitation of this approach is that the majority of cases of abuse are never registered, and cases of abuse will therefore have been present in the control sample. In addition, the cases of abuse included in the study received some form of intervention, possibly mitigating the risk for later psychopathology.

Four population-based studies used a cross-sectional design to study the association between trauma and psychosis. These studies found significant associations between different types of self-reported trauma in childhood and psychotic disorder (Bebbington et al. Reference Bebbington, Bhugra, Brugha, Singleton, Farrell, Jenkins, Lewis and Meltzer2004; Shevlin et al. Reference Shevlin, Dorahy and Adamson2007b), hallucinations (Whitfield et al. Reference Whitfield, Dube, Felitti and Anda2005; Shevlin et al. Reference Shevlin, Dorahy and Adamson2007a) and delusions (Scott et al. Reference Scott, Chant, Andrews, Martin and McGrath2007). Prospective associations between self-reported trauma and incident psychotic symptoms (both hallucinations and delusions) were reported by Janssen et al. (Reference Janssen, Krabbendam, Bak, Hanssen, Vollebergh, Graaf and Van Os2004) and Spauwen et al. (Reference Spauwen, Krabbendam, Lieb, Wittchen and van Os2006). In both studies, baseline retrospective assessment of trauma was associated with the new development of psychotic symptoms over mean follow-up periods of 36 and 42 months respectively.

The risk-increasing effect of trauma appears to be related to interpersonal events in particular, but is not confined to sexual abuse. One study investigating multiple types of trauma reported the highest effect size for sexual abuse (Bebbington et al. Reference Bebbington, Bhugra, Brugha, Singleton, Farrell, Jenkins, Lewis and Meltzer2004), but this was not the case in the Spauwen et al. (Reference Spauwen, Krabbendam, Lieb, Wittchen and van Os2006) study.

Five studies reported dose–response relationships between number of traumatic events and risk for psychotic symptoms, in line with the hypothesized causal role for developmental trauma (Janssen et al. Reference Janssen, Krabbendam, Bak, Hanssen, Vollebergh, Graaf and Van Os2004; Whitfield et al. Reference Whitfield, Dube, Felitti and Anda2005; Spauwen et al. Reference Spauwen, Krabbendam, Lieb, Wittchen and van Os2006; Scott et al. Reference Scott, Chant, Andrews, Martin and McGrath2007; Shevlin et al. Reference Shevlin, Dorahy and Adamson2007a). Similarly, stronger effects were found for more severe trauma, as indicated by a diagnosis of post-traumatic stress disorder (PTSD; Scott et al. Reference Scott, Chant, Andrews, Martin and McGrath2007); or trauma meeting the DSM-IV A2 criterion of presence of intense fear, helplessness or horror (Spauwen et al. Reference Spauwen, Krabbendam, Lieb, Wittchen and van Os2006).

The consistency of the findings across these studies, all based on large samples and controlling for relevant demographic and mental health variables, has raised the possibility that developmental trauma may be an important environmental risk factor contributing interactively with genetic factors to the development of psychosis (van Os et al. Reference van Os, Krabbendam, Myin-Germeys and Delespaul2005). However, all studies in this area inevitably rely on observational data and a retrospective and relatively crude assessment of developmental trauma, making it difficult to prove causality. A strong argument in favour of causality is the fact that plausible psychological as well as biological mechanisms have been proposed that explain the relationship (discussed below).

Diagnostic issues

An often heard criticism of the population-based trauma studies is the focus on the outcome of individual psychotic symptoms in the general population rather than a diagnosis of clinical psychotic disorder (Morgan & Fisher, Reference Morgan and Fisher2007). The first issue here is whether the relationship between trauma and psychotic symptoms in the community can be extrapolated to the categorical diagnosis of clinical psychotic disorders, such as schizophrenia. Converging evidence that the psychotic experiences in the general population are not qualitatively distinct from the clinical disorder comes from studies showing psychopathological resemblance (Vollema & van den Bosch, Reference Vollema and van den Bosch1995; Gruzelier, Reference Gruzelier1996), shared demographic and clinical correlates (Van Os et al. Reference Van Os, Hanssen, Bijl and Vollebergh2001; Johns et al. Reference Johns, Cannon, Singleton, Murray, Farrell, Brugha, Bebbington, Jenkins and Meltzer2004), familial clustering (Krabbendam et al. Reference Krabbendam, Myin Germeys, Hanssen and van Os2005; Hanssen et al. Reference Hanssen, Krabbendam, Vollema, Delespaul and Van Os2006) and the presence of transitions over time between the subclinical and clinical expressions of psychosis (Chapman et al. Reference Chapman, Chapman, Kwapil, Eckblad and Zinser1994; Poulton et al. Reference Poulton, Caspi, Moffitt, Cannon, Murray and Harrington2000). This may indicate that the link with developmental trauma also pertains to clinical psychotic disorders. In fact, two of the seven population-based studies reported associations between trauma and psychotic disorder (Bebbington et al. Reference Bebbington, Bhugra, Brugha, Singleton, Farrell, Jenkins, Lewis and Meltzer2004, Shevlin et al. Reference Shevlin, Dorahy and Adamson2007b).

Second, the relationship between trauma and psychosis refers almost exclusively to the positive symptoms of psychosis, i.e. hallucinations and delusions (Read et al. Reference Read, van Os, Morrison and Ross2005). This observation is in agreement with the idea that there are at least two different dimensions in schizophrenia, each associated with their own pattern of risk factors, demographic associations and symptoms (Myin-Germeys & van Os, Reference Myin-Germeys and van Os2007). According to this view, developmental trauma is on the ‘affective pathway’ associated with positive symptoms, less cognitive impairment, and female sex. However, rates of positive psychotic symptoms are also elevated in individuals with PTSD (Butler et al. Reference Butler, Mueser, Sprock and Braff1996; Hamner et al. Reference Hamner, Frueh, Ulmer, Huber, Twomey, Tyson and Arana2000) and there is considerable diagnostic overlap between this disorder and schizophrenia (Seedat et al. Reference Seedat, Stein, Oosthuizen, Emsley and Stein2003; Mueser et al. Reference Mueser, Salyers, Rosenberg, Goodman, Essock, Osher, Swartz and Butterfield2004). This raises the question whether developmental trauma truly increases the risk for (positive psychotic symptoms in) schizophrenia. The focus on single symptoms is motivated by the heterogeneity of psychotic disorders such as schizophrenia, and by the lack of valid distinctions between diagnostic categories (Allardyce et al. Reference Allardyce, Gaebel, Zielasek and van Os2007). The single-symptom approach has highlighted the common processes involved in development and persistence of symptoms across diagnostic boundaries, for example, in psychosis and anxiety disorders (Garety et al. Reference Garety, Kuipers, Fowler, Freeman and Bebbington2001, Reference Garety, Bebbington, Fowler, Freeman and Kuipers2007; Freeman et al. Reference Freeman, Garety, Kuipers, Fowler and Bebbington2002). Along the same lines, a recent account of the relationship between trauma and psychosis argued that both PTSD and psychosis are characterized by intrusions and the subsequent interpretation of these intrusions, and suggested that it is the cultural acceptability of the intrusions and the subsequent interpretation that determines whether the experiences are labelled PTSD or psychosis (Morrison et al. Reference Morrison, Frame and Larkin2003). According to Read et al. (Reference Read, van Os, Morrison and Ross2005), we may ‘not need to separate abuse sequelae into seemingly discrete categories such as PTSD, dissociative disorder, schizophrenia, borderline personality disorder’ (p. 340), but rather see all abuse-related symptoms as part of a spectrum of reactions to aversive events. Development and persistence of these symptoms may be mediated by shared biological and psychological mechanisms.

Mechanism of risk

At the biological level, trauma may contribute to the risk of psychosis by impacting on the developing brain, leading to neurodevelopmental abnormalities (Read et al. Reference Read, Perry, Moskowitz and Connolly2001; Nemeroff, Reference Nemeroff2004). Specifically, persistent exposure to stressors may lead to chronically heightened stress-induced glucocorticoid release, which may lead to permanent changes in the HPA axis and to structural abnormalities in the hippocampus (Bremner, Reference Bremner1999; Heim et al. Reference Heim, Newport, Heit, Graham, Wilcox, Bonsall, Miller and Nemeroff2000; Teicher et al. Reference Teicher, Andersen, Polcari, Anderson, Navalta and Kim2003). Dysregulation of the HPA axis may contribute to the dopaminergic abnormalities that are generally thought to be involved in psychosis (Walker & Diforio, Reference Walker and Diforio1997). This means that trauma may contribute to the oversensitivity to later stress, compatible with the view that vulnerability to psychosis is not determined exclusively by genetic factors, but may be moulded in an interaction with a variety of environmental exposures.

At the psychological level, exposure to early adverse life-events may create a cognitive vulnerability, characterized by a tendency to perceive the self as powerless and others as malevolent, which in combination with an externalizing attribution style may lead to paranoid interpretation of anomalous experiences (Bentall et al. Reference Bentall, Corcoran, Howard, Blackwood and Kinderman2001; Garety et al. Reference Garety, Kuipers, Fowler, Freeman and Bebbington2001, Reference Garety, Bebbington, Fowler, Freeman and Kuipers2007; Birchwood, Reference Birchwood2003). The common assertion in these models is that it is the interpretation of anomalous experiences which determines whether an individual will develop psychosis, and that exposure to trauma influences such interpretations. Empirical support for this claim comes from the study by Bak et al. (Reference Bak, Krabbendam, Janssen, de Graaf, Vollebergh and van Os2005) who showed that in individuals with anomalous psychotic-like experiences, prior exposure to trauma in childhood and adolescence was associated with less subjective control over these experiences and greater level of psychological distress. The recent study by Gracie et al. (Reference Gracie, Freeman, Green, Garety, Kuipers, Hardy, Ray, Dunn, Bebbington and Fowler2007) in a student sample also pointed to the key mediating role of negative beliefs in the relationship between developmental trauma and predisposition to psychosis, particularly paranoid ideation, and to a lesser extent hallucinatory experiences. Predisposition to hallucinations was also weakly but significantly associated with PTSD re-experiencing symptoms, suggesting that there may also be a direct link between trauma and hallucinations.

The study by Andrew and colleagues in this issue (Andrew et al. Reference Andrew, Gray and Snowden2008) explored the link between developmental trauma on the one hand and beliefs about psychotic experiences and associated distress on the other in greater detail, further corroborating the predictions of the cognitive models of psychosis. A particularly interesting feature of this study was the inclusion of a group of non-psychiatric voice-hearers (i.e. mediums or psychics), who have predominantly positive beliefs about voices, as well as a group of psychiatric voice-hearers, who have predominantly negative beliefs about voices. The study not only assessed the prevalence of developmental trauma but also the extent to which the trauma continued to affect the individuals in terms of current symptoms of intrusions, avoidance and hyperarousal. Current trauma symptoms were found to be a significant predictor of beliefs about malevolence, benevolence and omnipotence of the voices, suggesting that the extent to which the psychological effects of the trauma persist is a particularly important factor in determining beliefs about voices.