Voices from the storm: A critical review of quantitative studies of auditory verbal hallucinations and childhood sexual abuse

https://doi.org/10.1016/j.cpr.2011.05.004Get rights and content

Abstract

Although an association between hallucinations and sexual abuse has been documented, the relation between specifically auditory verbal hallucinations (AVHs) and childhood sexual abuse (CSA) is less clear. This study reviewed quantitative studies of AVHs and CSA. 36% of psychiatric patients with AVHs, and 22% of non-psychiatric patients with AVHs, reported CSA. At least 16% of the general population with auditory hallucinations also reported CSA. The majority of studies reviewed found that those with AVHs were more likely to be survivors of CSA than individuals without AVHs. 56% of psychiatric patients with CSA reported AVHs, and at least 21% of the general population with CSA reported auditory hallucinations. A majority of studies found survivors of CSA were more likely to report AVHs than individuals without CSA. Ability to impute a causal role for CSA was impaired by such studies' failures to control for potentially confounding variables. Yet, studies of AVH content showed links between the content of voices and the content of CSA in some voice-hearers. It is concluded that although a clear association between CSA and AVHs exists, there is not yet reliable quantitative evidence of a causal relation. Implications for mental health professionals and for future research, are discussed.

Research highlights

► 36% of psychiatric patients with AVHs report childhood sexual abuse (CSA). ► Individuals with CSA are more likely to report AVHs than those without CSA. ► The content of some, but not all, AVHs relate to CSA experiences. ► Studies of CSA and AVHs have not controlled for confounding variables. ► Studies controlling for confounders are needed to assess if CSA may cause AVHs.

Introduction

Recent decades have seen a number of shifts in the paradigm of hallucination research. First, a move has occurred away from studying such experiences within the context of specific psychiatric diagnoses, partly due to questions being raised as to the reliability and validity of certain diagnoses such as schizophrenia (e.g., Bentall, 2003, Boyle, 2002). This has resulted in the emergence of a complaint-oriented approach to psychopathology (Bentall, 2006, Frith, 1992), with researchers attempting to understand the causal mechanisms underlying the occurrence of hallucinations trans-diagnostically. As a result, a significant literature has been generated on the potential cognitive and neurocognitive mechanisms associated with hallucinations (e.g., Bentall, 1990, Jones, 2010, Seal et al., 2004). The meaning of hallucinations has also been reconsidered, with a movement away from viewing such experiences as meaningless symptoms of a mental illness (Romme & Escher, 1993), and towards a re-conceptualization of them as inherently meaningful events often having their origins in earlier stressful or traumatic events in the individual's life (Honig et al., 1998, Hornstein, 2009, Romme et al., 2009, Romme et al., 1992). Such work follows on findings such as that of Romme and Escher (1989) who found 70% of individuals with auditory hallucinations first began to hear voices following a traumatic/emotional event.

One traumatic event that has been found in many studies to be associated with the presence of hallucinations is childhood sexual abuse (CSA). In early studies Ellenson, 1985, Ellenson, 1986 found that many female incest survivors experienced hallucinations in a range of modalities including the visual (e.g., shadowy figures, movements in peripheral vision), the auditory (e.g., intruder sounds) and specifically the auditory verbal (e.g., voices giving commands, voices persecuting the voice hearer, or voices helping the voice-hearer). Following this work, an extensive body of work has documented a relation between hallucinations and CSA (Janssen et al., 2004, Read et al., 2005).

A number of studies have further attempted to examine if CSA is associated with hallucinations in specific modalities. Given that intrusive images and auditions, believed to underpin some visual and auditory hallucinations respectively (see Jones, 2010), are underpinned by different cognitive systems (Hagenaars, Brewin, van Minnen, Holmes, & Hoogduin, 2010) it is plausible that different modalities of hallucinations may be preferentially associated with specific causes. Indeed, research (e.g., Read, Agar, Argyle, & Aderhold, 2003) has led to the suggestion that there is a specific relation between CSA and auditory verbal hallucinations (or ‘hearing voices’), with Hammersley and Fox (2006) arguing that in studies of psychosis and childhood trauma the link between CSA and auditory hallucinations “is consistently the most reliable finding” (p. 152). Going beyond association, results from quantitative studies have been used to argue that “child abuse is a causal factor for… voices commenting and command hallucinations” (Read et al., 2005, p. 330), yet the causal role of specifically CSA has not been addressed. However, clinical experience, first-person testimonies from individuals who hear voices (e.g., Romme et al., 2009), and case studies (e.g., Heins et al., 1990, Bahn and Lee, 2007, Kaufman et al., 1997, Lysaker et al., 2007) are all suggestive of a causal role for CSA in the development of auditory verbal hallucinations (AVHs).

A number of theoretical models provide a basis for conceptualizing how CSA may play an etiological role in the development of AVHs. Memory-based models of AVHs propose they result from the unintentional activation of memories (Waters, Badcock, Michie, & Maybery, 2006). Consistent with this model, more severe AVHs (but not visual hallucinations) in people diagnosed with schizophrenia have been found to be associated with worse performance on temporal context memory tasks (Brébion, David, Jones, Ohlsen, & Pilowsky, 2007). This dovetails well with the context memory deficits that have been argued to stem from traumatic events such as CSA (Steel, Fowler, & Holmes, 2005). Steel and colleagues argue that traumatic events are not processed in the normal manner by the hippocampus. During highly traumatic events, information by-passes the hippocampus and is processed via the amygdala, in order to enable a faster release of stress hormones (LeDoux et al., 1988, Steel et al., 2005). Although this allows quicker processing, it does not allow the hippocampus to perform its normal role of integrating information within a spatial and temporal context (Brewin, 2001, Steel et al., 2005). Thus, potential context memory deficits resulting from CSA form a theoretical framework for understanding how such experiences may result in AVHs. If a review of the CSA and AVHs literature shows that the content of AVHs closely matches that of CSA experiences, then this would help support this memory-based model. However, if more subtle relations between AVHs and CSA are found then this would suggest a need to revise a simplistic memory-based model of AVHs.

In addition to context memory deficits, source-monitoring (Johnson, Hashtroudi, & Lindsay, 1993) provides another mechanism through which CSA may result in AVHs. Bentall (1990) has argued that the ability to discriminate between real, external events and imaginary, self-generated events is a skill (which has been given a number of names; reality-testing, source-monitoring, self-monitoring), in which individuals draw on a number of specific properties of a given cognitive event to make a judgment (not necessarily a conscious one) as to its origin. Bentall proposed that a failure in this skill could result in AVHs, with internal events being misattributed to an external source. While individual studies have found evidence both for and against this proposal (see Aleman & Larøi, 2008 for a review), a recent meta-analysis found deficits in self-monitoring (the ability to recognize self-generated thoughts or actions as one's own) to be more pronounced in people diagnosed with schizophrenia with AVHs than in such individuals without AVHs (Waters, Woodward, Allen, Aleman, & Sommer, in press). One specific property of cognitions that has been shown to encourage source-monitoring errors is the amount of cognitive effort associated their occurrence (Johnson et al., 1993), with spontaneously occurring cognitions (low cognitive effort) being more likely to be misattributed to a non-self source. Relating this to CSA, Hammersley et al. (2003) have noted that CSA is likely to produce intrusive memories, occurring automatically and with low cognitive effort, with such experiences potentially seeding AVHs. Allen, Coyne, and Console (1997) have further argued that trauma-induced dissociation may hamper reality-testing, leaving individuals “vulnerable to the nightmarish inner world” (p. 332).

What is known about the biological impact of CSA on the developing brain also suggests that such experiences may lead to neurological changes that predispose individuals to AVHs. The relatively new discipline of developmental traumatology (Crozier, Van Voorhees, Hooper, & De Bellis, 2011) has begun to trace the neurobiological impact of chronic trauma on the developing child. This considers how measureable aspects of traumatic experiences (such as the type, age of onset, and duration of child maltreatment) and other biopsychosocial factors (e.g., the child's temperament, social support for the child and family) relate to specific neurobiological changes in the brain. While such studies have typically focused on the effects of childhood abuse per se, rather than the differential effects of specific types of abuse (e.g., sexual, physical, emotional), the results of such studies are consistent with abuse causing structural changes that may lead to AVHs. For example, changes in anterior cingulate cortex volume (Cohen et al., 2006, Kitayama et al., 2006), superior temporal gyrus volume (De Bellis et al., 2002, Tomoda et al., 2011), and fractional anisotropy in the arcuate fasciculus1 (Choi, Jeong, Rohan, Polcari, & Teicher, 2009) have all been found to be associated with childhood abuse. All these areas have been found by both functional and structural neuroimaging (Allen, Laroi, McGuire, & Aleman, 2008) and electrophysiological (Heinks-Maldonado, Mathalon, Houde, Gray, & Ford, 2007) studies to be related to AVHs, and have been interpreted as relating to abnormal functioning in the normal cognitive mechanisms that allow us to recognize internally generated cognitions as self-produced (e.g., Feinberg, 1978, Jones and Fernyhough, 2007, Seal et al., 2004). In addition to these findings, Read, Perry, Moskowitz, and Connolly (2001) have also noted similarities between abused individuals and people diagnosed with schizophrenia in terms of hyperresponsivity of the hypothalamic–pituitary–adrenal axis stress system (see Shenk, Noll, Putnam, & Trickett, 2010, in relation to CSA specifically), and hyperactivation of the dopaminergic system. Although much work remains to be done on developmental traumatology, at present it appears plausible that traumatic events (such as CSA) may be the cause of many of the gray and white matter changes, as well as hormonal and neurotransmitter abnormalities found in those with AVHs.

Despite the confluence of good theoretical reasons for predicting a role of CSA in the etiology of AVHs with suggestions of a relationship derived from clinical experience, first-person testimonies from voice-hearers, and case studies, there has been no critical review of the empirical studies in this important area. The empirical establishment of a relation between CSA and AVHs would be important for a number of reasons. First, if an association does exist then this would have major implications for voice-hearers, mental health professionals, families, and researchers. This would be the case even if the relation were not a causal one, as it would highlight the need to address the increased probability of the existence of CSA in individuals presenting with AVHs, and vice versa. Second, if evidence exists for a causal relation, this would have implications for our understanding of the mechanisms underlying AVHs, the design of interventions, and ultimately for our conceptualization of the meaning of AVHs themselves. Third, studies of any potential relation between the content of AVHs and CSA events could help inform theoretical models of the relation between them, particularly regarding the role of memory processes.

This study aimed to perform a critical review of studies reporting on AVHs and CSA to assess the current state of knowledge in this field. Three specific aims were defined: 1) To establish prevalence of CSA in individuals with AVHs, as well as the prevalence of AVHs in those who have undergone CSA, 2) To examine evidence that CSA may play a causal role in AVHs, including an examination of the relation between the content of AVHs and its relation (if any) to earlier CSA experiences, and 3) To examine implications of these findings for future research directions and clinical practice.

Section snippets

Method

Papers reporting on the quantitative relation between AVHs and CSA were searched for using the MEDLINE (via Ovid Web of Science: 1950–2010) and Scopus (1966–2010) databases. The search string employed was (abuse OR trauma OR assault) AND (hallucination* OR psychosis). In addition, citation reports were utilized, and reference lists examined for other papers that might be pertinent to the review. Inclusion criteria for the study were that the study should quantitatively examine the relation

Prevalence studies

Seven studies were identified that reported on either the prevalence of AVHs in individuals who had received psychiatric diagnoses and who had experienced CSA, or on the prevalence of CSA in those with psychiatric diagnoses and AVHs. These studies are summarized in Table 1. Psychiatric diagnoses given to individuals in these studies included bipolar disorder, depression, dissociative disorder, post-traumatic stress disorder, schizophrenia, and substance abuse. A weighted mean showed that 36% of

Discussion

The studies reviewed in this paper showed a clear association between AVHs and CSA. First, in terms of CSA rates in those with AVHs, 36% of psychiatric patients with AVHs were found to report CSA. 22% of healthy voice-hearers (individuals who hear voices but who have not had contact with psychiatric services) also reported CSA. In the general population, 7% of those who reported auditory hallucinations (an unknown, but likely significant, subset of whom would have experienced specifically AVHs)

References (97)

  • J.C. Crozier et al.

    Effects of abuse and neglect on brain development

    Child Abuse & Neglect

    (2011)
  • M.D. De Bellis et al.

    Superior temporal gyrus volumes in maltreated children and adolescents with PTSD

    Biological Psychiatry

    (2002)
  • M. De Pradier et al.

    Influence of the serotonin transporter gene polymorphism, cannabis and childhood sexual abuse on phenotype of bipolar disorder: A preliminary study

    European Psychiatry: The Journal of the Association of European Psychiatrists

    (2010)
  • D.L. Dill et al.

    The reliability of abuse history report: A comparison of two inquiry formats

    Comprehensive Psychiatry

    (1991)
  • D.C. Goff et al.

    Self-reports of child abuse in chronically psychotic patients

    Psychiatry Research

    (1991)
  • L.C. Johns et al.

    The continuity of psychotic experiences in the general population

    Clinical Psychology Review

    (2001)
  • S.R. Jones et al.

    Thought as action: Inner speech, self-monitoring, and auditory verbal hallucinations

    Consciousness and Cognition

    (2007)
  • J. Kaufman et al.

    Case study: Trauma-related hallucinations

    Journal of the American Academy of Child and Adolescent Psychiatry

    (1997)
  • N. Kitayama et al.

    Smaller volume of anterior cingulate cortex in abuse-related posttraumatic stress disorder

    Journal of Affective Disorders

    (2006)
  • S. Moritz et al.

    Differences and similarities in the sensory and cognitive signatures of voice-hearing, intrusions and thoughts

    Schizophrenia Research

    (2008)
  • P. Mundy et al.

    The prevalence of psychotic symptoms in homeless adolescents

    Journal of the American Academy of Child and Adolescent Psychiatry

    (1990)
  • L. Offen et al.

    Is reported childhood sexual abuse associated with the psychopathological characteristics of patients who experience auditory hallucinations?

    Child Abuse & Neglect

    (2003)
  • C.E. Shenk et al.

    A prospective examination of the role of childhood sexual abuse and physiological asymmetry in the development of psychopathology

    Child Abuse & Neglect

    (2010)
  • M. Shevlin et al.

    Childhood traumas and hallucinations: An analysis of the National Comorbidity Survey

    Journal of Psychiatric Research

    (2007)
  • A. Tomoda et al.

    Exposure to parental verbal abuse is associated with increased gray matter volume in superior temporal gyrus

    NeuroImage

    (2011)
  • C.L. Whitfield et al.

    Adverse childhood experiences and hallucinations

    Child Abuse & Neglect

    (2005)
  • A. Aleman et al.

    Hallucinations: The science of idiosyncratic perception

    (2008)
  • S.L. Anderson et al.

    Preliminary evidence for sensitive periods in the effect of childhood sexual abuse on regional brain development

    The Journal of Neuropsychiatry and Clinical Neurosciences

    (2008)
  • N.C. Andreasen

    Scale for the assessment of positive symptoms

    (1984)
  • E.M. Andrew et al.

    The relationship between trauma and beliefs about hearing voices: A study of psychiatric and non-psychiatric voice hearers

    Psychological Medicine

    (2008)
  • C. Anketell et al.

    An exploratory analysis of voice hearing in chronic PTSD: Potential associated mechanisms

    Journal of Trauma & Dissociation

    (2010)
  • G.H. Bahn et al.

    Psychotic symptoms 10 years after a single episode of childhood sexual abuse

    Primary Care Companion to the Journal of Clinical Psychiatry

    (2007)
  • V. Beavan et al.

    Hearing voices and listening to what they say: The importance of voice content in understanding and working with distressing voices

    Journal of Nervous and Mental Disease

    (2010)
  • R. Bentall

    The illusion of reality: A review and integration of psychological research on hallucinations

    Psychological Bulletin

    (1990)
  • R.P. Bentall

    Madness Explained: Psychosis and Human Nature

    (2003)
  • M. Boyle

    Schizophrenia: A scientific delusion?

    (2002)
  • C.M. Bulik et al.

    Features of childhood sexual abuse and the development of psychiatric and substance use disorders

    British Journal of Psychiatry

    (2001)
  • P. Chadwick et al.

    The omnipotence of voices II: The beliefs about voices questionnaire

    The British Journal of Psychiatry

    (1995)
  • D. Cicchetti et al.

    Equifinality and multifinality in developmental psychopathology

    Development and Psychopathology

    (1996)
  • J.A. Cohen et al.

    Trauma-focused cognitive behavioral therapy for children and adolescents: An empirical update

    Journal of Interpersonal Violence

    (2000)
  • M.A. Conway et al.

    The construction of autobiographical memories in the self-memory system

    Psychological Review

    (2000)
  • G. Dodgson et al.

    Avoiding false negatives: Are some auditory hallucinations an evolved design flaw?

    Behavioural and Cognitive Psychotherapy

    (2009)
  • M.J. Dorahy et al.

    Auditory hallucinations in dissociative identity disorder and schizophrenia with and without a childhood trauma history

    The Journal of Nervous and Mental Disease

    (2009)
  • G.E. Ellenson

    Detecting a history of incest: A predictive syndrome

    Social Casework: The Journal of Contemporary Social Work

    (1985)
  • G.S. Ellenson

    Disturbances of perception in adult female incest survivors

    Social Casework: Journal of Contemporary Social Work

    (1986)
  • I. Feinberg

    Efference copy and corollary discharge: Implications for thinking and its disorders

    Schizophrenia Bulletin

    (1978)
  • C. Feiring et al.

    Trying to understand why horrible things happen: Attribution, shame, and symptom development following sexual abuse

    Child Maltreatment

    (2002)
  • M.B. First et al.

    Structured Clinical Interview for DSM-IV (SCID)

    (1996)
  • Cited by (38)

    • Linguistic dual tasking reduces emotionality, vividness and credibility of voice memories in voice-hearing individuals: Results from a controlled trial

      2020, Schizophrenia Research
      Citation Excerpt :

      More specifically, several models state that some AVH may be considered to be decontextualized memories that are experienced as sensory perceptions (Hardy, 2017; Steel et al., 2005; Waters et al., 2012, 2006). This is supported by strong associations between experiencing childhood trauma and AVH (Bentall et al., 2012; McCarthy-Jones, 2011), objective observation of similarities in content and theme between traumatic experiences and AVH (Hardy et al., 2005), and the subjective experience of people reporting hallucinations that appear to be replays of actual past events (McCarthy-Jones et al., 2014). It is established that predominantly emotionality of voices and negative voice-related beliefs (e.g. believing the negative comments that the voices make) are an important cause of voice-related distress (Daalman et al., 2011; Hacker et al., 2008), and decreasing or modifying these beliefs in psychological therapies, such as cognitive therapy (e.g. Birchwood et al., 2014) and imaginary rescripting (Paulik et al., 2018), is proven to be effective in decreasing voice-hearing distress.

    • Prevalence of auditory pseudohallucinations in adult survivors of physical and sexual trauma with chronic post-traumatic stress disorder (PTSD)

      2018, Behaviour Research and Therapy
      Citation Excerpt :

      For those experiencing civilian PTSD this raises the question of whether particular types of trauma exposure or trauma history are more or less likely to be associated with the experience of pseudo-hallucinations, as we know that dissociation is differentially associated with particular profiles of trauma exposure (Briere, 2006). The experience of childhood sexual abuse, has been established as a predictor of pseudohallucinations in samples both with and without psychosis (Hammersley & Fox, 2006; McCarthy-Jones, 2011; Read, McGregor, Coggan, & Thomas, 2006; Wearne et al., 2017), although the properties of the voices in these populations do not appear to differ between those with and without CSA (e.g. Offen, Waller, & Thomas, 2003). For this reason, the present study focused on a civilian sample presenting with PTSD following sexual assault, abuse or violence either in childhood or adulthood.

    • The relationship between dissociation and voices: A systematic literature review and meta-analysis

      2015, Clinical Psychology Review
      Citation Excerpt :

      When compared to other symptoms of psychosis, voice-hearing in particular seems to be associated with dissociative experiences (e.g., Kilcommons & Morrison, 2005). These seemingly strong associations, and the body of literature suggesting that dissociation may represent a post-traumatic sequelae or a regulatory response to trauma (for a review Dalenberg et al., 2012), has led researchers to examine dissociative processes as possible explanatory or mediating mechanisms of the robust and well-replicated relationship between adverse life experiences and the risk to experience psychotic symptoms, including AVHs (e.g., Bentall et al., 2012; Bentall et al., 2014; McCarthy-Jones, 2011; Shelvin, Dorahy, & Adamson, 2007; Varese et al., 2012). For example, Moskowitz and Corstens (2008) proposed that AVHs might be regarded as trauma-induced dissociative experiences.

    • Preliminary evidence of an interaction between the FOXP2 gene and childhood emotional abuse predicting likelihood of auditory verbal hallucinations in schizophrenia

      2014, Journal of Psychiatric Research
      Citation Excerpt :

      Alternatively, FOXP2 may be especially vulnerable to environmental modulation, such that individual variation among FOXP2 SNPs would make some people particularly susceptible to the presence (and/or absence) of stressors that likely interact with all genotypes (i.e., the interactions are not restricted to a perceived “risk” genotype; Belsky et al., 2009). A candidate environmental stressor for such an interaction is one already shown to be associated with AVHs, namely child abuse, with substantial evidence already demonstrating a relationship between AVHs and a range of forms of child abuse, including sexual (Bentall et al., 2012; McCarthy-Jones, 2011), physical (Bentall et al., 2012; Read et al., 2003) and emotional abuse (Daalman et al., 2012), as well as neglect (Varese et al., 2012). Since not all child abuse results in AVHs, variations in FOXP2 genotypes may be one factor in determining who goes onto develop AVHs following such abuse in childhood.

    • Childhood sexual abuse increases risk of auditory hallucinations in psychotic disorders

      2013, Comprehensive Psychiatry
      Citation Excerpt :

      While these studies did not systematically assess the influence of multiple types of abuse, they provide strong evidence of a link between the experience of childhood sexual abuse and the development of psychotic symptoms later in life. Studies have also examined the relationship between childhood abuse and the experience of specific positive symptoms, many of which point to an association between sexual abuse and hallucinations (for review see McCarthy-Jones, 2011 [37]). Shevlin et al. [7] found that neglect, rape, and molestation all significantly increased an individual’s risk for visual hallucinations, while childhood rape and molestation increased the risk of experiencing auditory hallucinations.

    View all citing articles on Scopus
    View full text