Dissociation as a causal pathway from sexual abuse to positive symptoms in the spectrum of psychotic disorders

Childhood adversities have been recognized as a known risk factor in numerous mental disorders [1, 2]. As recently recommended, childhood maltreatment such as sexual abuse, physical abuse, emotional abuse, and neglect, might be accompanied by an increase in psychosis risk [3‐7]. In this respect, a recent meta-analysis of 29 studies showed that the severity of the positive and negative symptoms was associated with childhood sexual abuse and neglect, respectively. However, no association was observed between positive symptoms severity and childhood neglect and between negative symptoms severity and childhood sexual abuse [7].

In recent decades, the long-term negative effects of childhood maltreatment on the social and cognitive functioning of psychotic patients have led to significant researches into the understanding of possible underlying psychological mechanisms [8‐13]. In this regard, numerous psychological models have been proposed to explain the association between childhood maltreatment and psychosis. For instance, one of the most important of these models has shown that childhood maltreatment may result in psychosis via a pathway of posttraumatic stress disorder-related symptoms, including intrusive memories and dissociation [12, 13]. In this conceptual model, “flash-backs could be interpreted as being externally generated, which leads to hallucinatory experiences and hampers reality testing” [12].

The posttraumatic dissociation role in psychosis was already highlighted when Janet [12] characterized hysterical psychosis by its stress-related and dissociative nature. Although dissociation has not been recognized as a diagnostic criterion for any form of psychosis, Moskowitz et al. [14] and Ross [15] (among contemporary authors) suggested that a certain dissociative type of psychosis could potentially respond to trauma-focused psychotherapy. In this regard, accumulative evidence from an updated meta-analysis stated that dissociation—characterized by disruptions to the integrative functioning of several core mental domains— was related to positive psychotic symptoms (especially hallucinations) and less robustly associated with negative psychotic experiences. These findings proposed that certain psychotic symptoms might be better conceptualized as dissociation in nature and support the development of interventions targeting dissociative experiences in treating psychotic symptoms [16].

Although researchers have paid much attention so far to the causal pathways from certain forms of childhood maltreatment to specific psychotic symptoms, they have remained highly unexplained among patients with schizophrenia spectrum and other psychotic disorders [7]. However, a meta-analysis conducted by Alameda et al. [12] showed that dissociative experiences might play a mediating role in the relationship between childhood maltreatment and the development of psychotic symptoms. Nevertheless, several methodological and conceptual problems have been consistently identified in studies on childhood maltreatment, dissociation, and psychosis, which need to be addressed. Firstly, these studies adopted the Dissociative Experiences Scale (DES) to assess dissociation. Although the DES is a well-validated measure of dissociation, its mere use in psychotic patients is controversial due to item content overlap. For instance, item 27 of the DES is designed to get information about voice-hearing experiences directly. Reality distortions and perceptual disturbances are also reflected by eight items within the DES, which might not be able to discriminate between hallucinatory and dissociative experiences [17]. Besides, dissociation appears to be state-dependent among patients with schizophrenia spectrum disorders [18]. Therefore, the diagnostic interviews by applying Structured Clinical Interview for DSM-IV Dissociative Disorders (SCID-D) should be incorporated alongside the DES to avoid measurement artifacts [17, 18]. Given the relatively significant prevalence of undiagnosed dissociative disorders and borderline personality disorder among psychotic patients, the second problem is that these comorbidities emerge when exploring the relationship between psychotic symptoms and dissociation. Hence, it is crucial to evaluate these comorbidities through Structured Clinical Interview for DSM-5 Personality Disorders (SCID-5-PD) and SCID-D [16, 19‐21]. Thirdly, it is required to pay attention to different dissociation subtypes. Although recent evidence suggested that absorption may be more associated with psychotic-like experiences, the evidence synthesis based solely on the bivariate relationships between dissociation and psychotic experiences cannot confidently prove whether symptom-specific associations exist between psychotic symptoms and dissociation or between psychotic symptoms and specific dissociative subtypes. In this respect, multivariate analyses might be more practical to answer such questions [16]. Fourthly, it is debated whether neglect and abuse constructs can be statistically differentiated in psychosis. If separate neglect and childhood abuse constructs are validated, studies should apply statistical procedures (e.g., hierarchical regression) so that the effect of abuse is partialled out from neglect impact, and vice versa [22]. Finally, investigations might need to address the measurement of any differences in strength of association between childhood maltreatment and delusions/hallucinations across the psychosis continuum from delusions/hallucinations in nonclinical samples through first-episode psychosis to chronic disorder [7].

Our goal here is to more precisely investigate the causal pathways from childhood maltreatment to psychotic symptoms among psychotic patients by overcoming the methodological problems discussed above. In this respect, we replicated the study by Holowka et al. [23] on childhood maltreatment and dissociation among chronic psychotic patients, while extending the design by including first-episode psychotic patients, and community controls. Additionally, we hypothesized that, in patients with schizophrenia spectrum and other psychotic disorders, the particular forms of dissociation may mediate the association between five major types of childhood abuse and psychotic symptoms, by controlling for gender [8, 24].

To the best of our knowledge, this is the first study on the relationship between childhood maltreatment, dissociative experiences, and psychotic symptoms in Iranian patients with schizophrenia spectrum and other psychotic disorders. The findings of the present study can be divided into six major parts. As the first part, the results of this study indicated that only the mean scores of sexual abuse, emotional abuse, and physical abuse (not mean scores of emotional neglect and physical neglect) were significantly higher in psychotic patients than the community controls. Despite that, no significant difference was observed between first-episode psychotic patients and chronic psychotic patients. These findings were consistent with a priori hypothesis of the association between childhood abuse and psychotic symptoms, proposed in previous studies [39‐41]. For example, Daalman et al. [41] observed higher rates of childhood sexual and emotional abuse in both groups of hallucinated objects, irrespective of their disease status. These findings have introduced childhood maltreatment as a specific risk factor in the development of psychotic symptoms.

As the second main finding of the present study, the mean scores of dissociative experiences in the three study groups were significantly different, and the results obtained from the post hoc analysis confirmed higher mean scores of dissociative experiences in chronic psychotic patients. Previously, Braehler et al. [42] investigated first-episode psychotic patients, chronic psychotic patients, and community controls, realizing that chronic psychotic patients experienced higher levels of dissociative symptoms. The greater dissociation in chronic psychotic patients might be attributed to traumatic experiences after childhood; the issue disregarded in the present study. Psychiatric patients, compared to the general population, are more likely to experience additional traumatic events (e.g., assault) in adulthood [43]. Meanwhile, chronic patients might be at a higher risk of re-traumatization due to more coercive admissions and hospitalization [44]. Accordingly, if dissociative symptoms are considered the result of a set of childhood and adulthood traumatic events, chronic psychotic patients are expected to be at a greater risk of more severe dissociation due to experiencing multiple traumas [42]. Nevertheless, further studies are required to investigate potential cumulative effects of adulthood trauma on dissociation. Another probable reason for higher levels of dissociation in the patients with schizophrenia spectrum and other psychotic disorders (particularly chronic psychotic patients) is the etiological and phenomenological overlap between dissociative and psychotic symptoms [19]. In this regard, previous studies identified a subgroup of schizophrenic patients with high levels of childhood maltreatment and dissociation who met diagnostic criteria for borderline personality disorder or dissociative disorders [20]. Childhood adversities have been shown to be higher in schizophrenia patients with comorbid borderline personality disorder than those without borderline personality disorder [21]. Nonetheless, in our study, only four chronic psychotic patients (including 3 patients with comorbid borderline personality disorder and one patient with comorbid dissociative disorders), and one first-episode psychotic patient (with comorbid borderline personality disorder) made this explanation improbable. Finally, in recent years, it has been suggested that schizophrenia is best understood as a disorder of consciousness and self-experience (disturbed ipseity) that involved two key aspects of “hyper-reflexivity” (i.e., forms of exaggerating and alienating self-consciousness) and “diminished self-affection” (i.e., a diminished sense of existing as a subject of awareness or agent of action) [45]. This hypothesis may explain the apparent heterogeneity of psychotic symptoms among patients with schizophrenia spectrum and other psychotic disorders.

As the third main finding of the present study, there was no association between gender, child maltreatment, dissociation, and the severity of psychotic symptoms among patients with schizophrenia spectrum and other psychotic disorders. However, a recent review article revealed that women might be at a greater risk of sexual abuse than men [24]. This may be explained by cultural norms (such as shame, taboos and modesty, virginity, status of females, etc.) affecting the likelihood of whether child sexual abuse is discovered by an adult or disclosed by the child. For example, where cultural norms favor males over females, a girl’s report of sexual abuse by a boy or a man may be discounted [46].

As the fourth main result of the present study, the summary of regression analysis results revealed that for the patients with schizophrenia spectrum and other psychotic disorders, positive symptoms were related to sexual abuse, dissociative amnesia, absorption, and physical abuse, negative symptoms were associated with physical abuse, and general psychopathology symptoms were related to absorption and physical abuse. These findings were consistent with studies conducted by Sheffield et al. [39], Bendall et al. [47], and Read et al. [48], whereas they were inconsistent with results obtained by Bell et al. [49]. However, contrary to Sheffield et al. [39], we observed no relationship between emotional abuse and psychotic symptoms. This suggests that further studies are required to decide whether emotional abuse can account for psychotic symptoms or not. In addition, our study illustrated that physical abuse could be associated with psychotic symptoms even in the absence of sexual abuse and dissociation as well. This finding argues that some of the psychotic patients might have adapted differently to childhood adversities. In this respect, observing a relationship between childhood maltreatment and negative symptoms, Vogel et al. [50] concluded that negative symptoms (i.e., a constant state of down-regulation of emotion and social engagement) could be an alternative adaptive response to childhood adversities.

As the fifth finding of this study, dissociation mediated the relationship between sexual abuse and positive symptoms, which agreed with the similar results in clinical groups and preliminary research with nonclinical participants [51‐54]. Overall, this finding supports the information-processing theory proposed by Holmes et al. [55] who argued that peri-traumatic dissociation resulted in poorly encoded autobiographical representations by disruption of information processing, which might be later re-experienced as traumatic intrusions (e.g., hallucinations). Further, recent studies have suggested that weakened cognitive inhibition might represent the prevailing cognitive concomitant of dissociation [56]. In line with this theory, recent experimental evidence has highlighted the importance of inhibitory processes in the explanation of auditory hallucinations [57, 58]. Nevertheless, further research is required to determine whether such processes are able to explain the observed relationship between dissociation and positive symptoms.

As another interesting finding of our study, the regression analysis revealed that dissociative symptoms were related to psychotic symptoms in some cases independent of exposure to childhood maltreatment, suggesting that dissociation did not develop as a result of childhood adversities in some patients. These findings led to the following question: “Do various psychotic symptoms have distinct etiologies?” In answer to this question, recent evidence has suggested the presence of substantial overlaps in socio-environmental and biological risks across specific symptoms and diagnostic categories; accordingly, a move has been taken toward transdiagnostic therapies [59‐62]. Nevertheless, many risk and resilience factors might exist for each symptom, and the relative significance of each factor is likely to differ from person to person. This highlights the necessity to create individualized formulations to better understand the development of distressing symptoms from a psychological therapy perspective [16].

As the latest finding of this study, the results revealed that absorption mediated the relationship between sexual abuse and positive symptoms, agreeing with the results obtained by Cole et al. [51] and Perona-Garcelán et al. [54]. Absorption is a form of intensively focused attention wherein an individual becomes immersed in their mental imagery so that these events seem to happen in reality, just like what takes place in a hallucinatory experience [51]. In addition, the confusion between reality and imagination due to the disability to determine the veracity of memories can also lead to a fixed, false, and idiosyncratic belief that is perceived as delusion [51, 54]. We also found out that dissociative amnesia had a mediating role in the relationship between sexual abuse and positive symptoms. Previously, Kennerley [63] following Holmes et al. [55], emphasized distinct functions of different types of dissociation. Indeed, tuning in (absorption) can cause the re-living of intrusive peri-traumatic information in the forms of flash-backs and hallucinations. Tuning out (dissociative amnesia) might make a person unable to access traumatic information in the memory [51]. Accordingly, absorption and dissociative amnesia are expected to have positive and negative mediating roles, respectively, exactly consistent with what was observed in our study. An unexpected finding was about the lack of mediating role of depersonalization/derealization in the relationship between sexual abuse and positive symptoms; consistent with the study by Cole et al. [51] and inconsistent with the results obtained by Perona-Garcelán et al. [53, 54]. This might be due to negligible detrimental effects of non-pathological depersonalization/derealization on processing information related to adverse events in clinical groups [53].

The present study suffered from some methodological limitations. First, the findings could not be generalized to various cases since the sample size was small and participants were selected from a single geographic region. Second, cross-sectional studies mostly fail to specify a definite reason behind a correlation. This restriction might avoid a deep understanding of the essence of the causal relationship between childhood maltreatment, dissociative experiences, and psychotic symptoms. As the third limitation, this study used self-report scales that can only identify the emotions of patients through the assessment and are not able to reflect their real emotions. Hence, it is suggested that future studies should focus on methodological limitations, such as sole reliance on self-report scales due to memory bias and demand characteristics, lack of empirical data, and disregarding ethnic differences. It is worth noting that side effects of antipsychotics (e.g., memory problems, affective flattening, and detachment) may also be mistaken for dissociation during the evaluation of psychotic patients; thus, they should be considered in future studies [16].

Despite the above limitations, our study improved psychopathological comprehension of psychotic symptoms in patients with schizophrenia spectrum and other psychotic disorders. To the best of our knowledge, the present work is the first study that equally assesses and compares the effect of five major types of childhood abuse on various types of psychotic symptoms in Iranian patients with schizophrenia spectrum and other psychotic disorders. This systematic evaluation provided a better opportunity to observe childhood maltreatment as a risk factor in psychosis and allowed us to find specific relationships between sexual abuse, dissociative experiences, and positive symptoms. Understanding such internalized representations can be essential to develop therapeutic interventions and preventive approaches in victims of child maltreatment [64].

The present study aimed to assess the mediating role of dissociation in the relationship between childhood maltreatment and psychotic symptoms among patients with schizophrenia spectrum and other psychotic disorders. The study results showed that the mean scores of sexual abuse, emotional abuse, physical abuse, and dissociative experiences were higher in psychotic patients (particularly chronic psychotic patients) than community controls. However, in the three study groups, no significant gender-based difference was observed between the mean scores of dissociative experiences and various types of childhood maltreatment. In addition to confirming a relationship between physical abuse and psychotic symptoms (even in the absence of sexual abuse), this study implied the mediating role of absorption and dissociative amnesia in the relationship between sexual abuse and positive symptoms. These findings emphasized the importance of assessing the history of childhood maltreatment (particularly sexual and physical abuse history) and dissociative experiences in psychotic patients (regardless of their gender) to ensure the proposal of the most suitable and effective therapeutic interventions for this group of patients. Nonetheless, further studies using prospective or longitudinal designs are required to better understand the differential contribution of various forms of childhood maltreatment and their potential interactions.