Six Myths About Dissociative Identity Disorder (Part 2)


One of the most frequently repeated myths is that DID is iatrogenically created. Proponents of this view argue that various influences—including suggestibility, a tendency to fantasise, therapists who use leading questions and procedures, and media portrayals of DID—lead some vulnerable individuals to believe they have the disorder. Trauma researchers have repeatedly challenged this myth. Space limitations require that we provide only a brief overview of this claim.

A recent and thorough challenge to this myth comes from Dalenberg and colleagues. They conducted a review of almost 1500 studies to determine whether there was more empirical support for the trauma model of dissociation—that is, that antecedent trauma causes dissociation, including dissociative disorders—or for the fantasy model of dissociation. According to the latter (also known as the iatrogenic or sociocognitive model), highly suggestible individuals enact DID following exposure to social influences that cause them to believe that they have the disorder. Thus, according to the fantasy model proponents, DID is not a valid disorder; rather, it is iatrogenically induced in fantasy-prone individuals by therapists and other sources of influence.

Dalenberg and colleagues concluded from their review and a series of meta-analyses that little evidence supports the fantasy model of dissociation. Specifically, the effect sizes of the trauma-dissociation relationship were strong among individuals with dissociative disorders, and especially DID (i.e., .54 between child sexual abuse and dissociation, and .52 between physical abuse and dissociation). The correlations between trauma and dissociation were as strong in studies that used objectively verified abuse as in those relying on self-reported abuse. These findings strongly contradict the fantasy model hypothesis that DID individuals fantasise their abuse. Dissociation predicted only 1%–3% of the variance in suggestibility, thereby disproving the fantasy model’s notion that dissociative individuals are highly suggestible.

Despite the concerns of fantasy model theorists that DID is iatrogenically created, no study in any clinical population supports the fantasy model of dissociation. A single study conducted in a “normal” sample of college students showed that students could simulate DID. That study, by Spanos and colleagues, documents that students can engage in identity enactments when asked to behave as if they had DID. Nevertheless, the students did not begin to believe that they had DID, and they did not develop a wide range of severe, chronic, and disabling symptoms displayed by DID patients.

The study by Spanos and colleagues was limited by the lack of a DID control group. Several recent controlled studies have found that DID simulators can be reliably distinguished from DID patients on a variety of well-validated and frequently used psychological personality tests (e.g., Minnesota Multiphasic Personality Inventory–2), forensic measures (e.g., Structured Interview of Reported Symptoms), and neurophysiological measures, including brain imaging, blood pressure, and heart rate.

Two additional lines of research challenge the iatrogenesis theory of DID: first, prevalence research conducted in cultures where DID is not well known, and second, evidence of chronic childhood abuse and dissociation in childhood among adults diagnosed with DID. Three classic studies have been conducted in cultures where DID was virtually unknown when the research was conducted. Researchers using structured interviews found DID in patients in China, despite the absence of DID in the Chinese psychiatric diagnostic manual. The Chinese study and also two conducted in central-eastern Turkey in the 1990s where public information about DID was absent—contradict the iatrogenesis thesis. In one of the Turkish studies, a representative sample of women from the general population (n = 994) was evaluated in three stages: participants completed a self-report measure of dissociation; two groups of participants, with high versus low scores, were administered the DDIS by a researcher blind to scores; and the two groups were then given clinical examinations (also blind to scores). The researchers were able to identify four cases of DID, all of whom reported childhood abuse or neglect.

The second line of research challenging the iatrogenesis theory of DID documents the existence of dissociation and severe trauma in childhood records of adults with DID. Researchers have found documented evidence of dissociative symptoms in childhood and adolescence in individuals who were not assessed or treated for DID until later in life (thus reducing the risk that these symptoms could have been suggested). Numerous studies have also found documentation of severe child abuse in adult patients diagnosed with DID. For example, in their review of the clinical records of 12 convicted murderers diagnosed with DID, Lewis and colleagues found objective documentation of child abuse (e.g., child protection agency reports, police reports) in 11 of the 12, and long-standing, marked dissociation in all of them. Further, Lewis and colleagues noted that “contrary to the popular belief that probing questions will either instil false memories or encourage lying, especially in dissociative patients, of our 12 subjects, not one produced false memories or lied after inquiries regarding maltreatment.

On the contrary, our subjects either denied or minimised their early experiences. We had to rely for the most part on objective records and on interviews with family and friends to discover that major abuse had occurred.” Notably, these inmates had already been sentenced; they were all unaware of having met diagnostic criteria for DID, and they made no effort to use the diagnosis or their trauma histories to benefit their legal cases.

Similarly, Swica and colleagues found documentation of early signs of dissociation in childhood records in all of the six men imprisoned for murder who were assessed and diagnosed with DID during participation in a research study. During their trials, the men were all unaware of having DID. And since their sentencing had already occurred, they had nothing to gain from DID being diagnosed while participating in the study. Their signs and symptoms of early dissociation included hearing voices (100%), having vivid imaginary companions (100%), amnesia (50%), and trance states (34%). Furthermore, evidence of severe childhood abuse has been found in medical, school, police, and child welfare records in 58%–100% of DID cases. These studies indicate that dissociative symptoms and a history of severe childhood trauma are present long before DID is suspected or diagnosed.

Perhaps the “iatrogenesis myth” exists because inappropriate therapeutic interventions can exacerbate symptoms if used with DID patients. The expert consensus DID treatment guidelines warn that inappropriate interventions may worsen DID symptoms, although few clinicians report using such interventions. No research evidence suggests that inappropriate treatment creates DID. The only study to date examining deterioration of symptoms among DID patients found that only a small minority (1.1%) worsened over more than one time-point in treatment and that deterioration was associated with revictimization or stressors in the patients’ lives rather than with the therapy they received. This rate of deterioration of symptoms compares favourably with those for other psychiatric disorders.


Some authors suggest that the symptoms of DID represent a severe or overly imaginative presentation of BPD. The research described below, however, indicates that while DID and BPD can frequently be diagnosed in the same individual, they appear to be discrete disorders.

One of the difficulties in differentiating BPD from DID has been the poor definition of the dissociation criterion of BPD in the DSM’s various editions. In DSM-5 this ninth criterion of BPD is “transient, stress-related paranoid ideation or severe dissociative symptoms.” The narrative text in DSM-5 defines dissociative symptoms in BPD (“e.g., depersonalization”) as “generally of insufficient severity or duration to warrant an additional diagnosis.” DSM-5 does not clarify that when additional types of dissociation are found in patients who meet the criteria for BPD—especially amnesia or identity alteration that is severe and not transient (i.e., amnesia or identity alteration that form an enduring feature of the patient’s presentation)—the additional diagnosis of a dissociative disorder should be considered, and that additional diagnostic assessment is recommended.

On the surface, BPD and DID appear to have similar psychological profiles and symptoms. Abrupt mood swings, identity disturbance, impulsive risk-taking behaviours, self-harm, and suicide attempts are common in both disorders. Indeed, early comparative studies found few differences in clinical comorbidity, history, or psychometric testing using the Minnesota Multiphasic Personality Inventory and the Millon Clinical Multiaxial Inventory. However, recent clinical observational studies, as well as systematic studies using structured interview data, have distinguished DID from BPD. Brand and Loewenstein review the clinical symptoms and psychosocial variables that distinguish DID from BPD: clinically, individuals with BPD show vacillating, less modulated emotions that shift according to external precipitants. Also, individuals with BPD can generally recall their actions across different emotions and do not feel that those actions are alien or so uncharacteristic as to be disavowed. By contrast, individuals with DID have amnesia for some of their experiences while they are in dissociated personality states, and they also experience a marked discontinuity in their sense of self or sense of agency. Thus, the dissociated activity and intrusion of personality states into the individual’s consciousness may be experienced as separate or different from the self that they identify with or feel they can control. Accordingly, using SCID-D structured interview data, Boon and Draijer demonstrated that amnesia, identity confusion, and identity alteration were significantly more severe in individuals with DID than in cluster B personality disorder patients, most of whom had BPD. However, DID and BPD patients did not differ on the severity of depersonalization and derealization. Both groups had experienced trauma, although the DID group had much more severe and earlier trauma exposure.

BPD and DID can also be differentiated on the Rorschach inkblot test. Sixty-seven DID patients, compared to 40 BPD patients, showed greater self-reflective capacity, introspection, ability to modulate emotion, social interest, accurate perception, logical thinking, and ability to see others as potentially collaborative. A pilot Rorschach study found that compared to BPD patients, DID patients had more traumatic intrusions, greater internalisation, and a tendency to engage in complex contemplation about the significance of events. The DID group consistently used a thinking-based problem-solving approach, rather than the vacillating approach characterised by shifting back and forth between emotion-based and thinking-based coping that has been documented among the BPD patients. These personality differences likely enable DID patients to develop a therapeutic relationship more easily than many BPD patients.

With regard to the frequent comorbidity between DID and BPD, studies assessing for both disorders have found that approximately 25% of BPD patients endorse symptoms suggesting possible dissociated personality states (e.g., disremembered actions, finding objects that they do not remember acquiring) and that 10%–24% of patients who meet criteria for BPD also meet criteria for DID. Likewise, a national random sample of experienced U.S. clinicians found that 11% of patients treated in the community for BPD met criteria for comorbid DID, and structured interview studies have found that 31%–73% of DID subjects meet criteria for comorbid BPD. Thus, about 30% or more of patients with DID do not meet full diagnostic criteria for BPD. In blind comparisons between non-BPD controls and college students who were interviewed for all dissociative disorders after screening positive for BPD, BPD comorbid with the dissociative disorder was more common than was BPD alone (n = 58 vs n = 22, respectively). It is important to note that despite its prevalence in patients with DID, BPD is not the most common personality disorder that is comorbid with DID. More common among individuals with DID are avoidant (76%–96%) and self-defeating (a proposed category in the appendix of DSM-III-R; 68%–94%) personality disorders, followed by BPD (53%–89%).

When the comorbidity between BPD and DID is evaluated specifically, the patients with comorbid BPD and DID appear to be more severely impaired than individuals with either disorder alone, for example, the participants who had both disorders reported the highest level of amnesia and had the most severe overall dissociation scores. Similarly, individuals who meet criteria for both disorders have more psychiatric comorbidity and trauma exposure than individuals who meet criteria for only one, and they also report higher scores of dissociative amnesia.

In the future, the neurobiology of BPD and DID might assist in their comparison. Preliminary imaging research in BPD suggests the prefrontal cortex may fail to inhibit excessive amygdala activation. By contrast, two patterns of activation that correspond to different personality states have been found in DID patients: neutral states are associated with overmodulation of affect and show corticolimbic inhibition, whereas trauma-related states are associated with under-modulation of affect and activation of the amygdala on positron emission tomography. Similarly, recent fMRI studies DID found that the neutral states demonstrate emotional under-activation and that the trauma-related states demonstrate emotional overactivation. Perhaps BPD might be thought of as resembling the trauma-related state of DID with amygdala activation, whereas the dissociative pattern found in the neutral state in DID appears to be different from what is found in BPD. Additional research comparing these disorders is needed to further explore the early findings of neurobiological similarities and differences.

What remains open for debate is whether a personality disorder diagnosis may be given to DID patients, because attribution of a clinical phenomenon to a personality disorder is not indicated if it is related to another disorder—in this instance, DID. Hence, the DSM-5 criteria for BPD may be insufficient to diagnose a personality disorder because DID is not excluded. In this regard, some DID researchers have concluded that unmanaged trauma symptoms—including dissociation—may account for the high comorbidity of BPD in DID patients. For example, one study found that only a small group of DID patients still met BPD criteria after their trauma symptoms were stabilised. Resolution of this debate may hinge on whether patients diagnosed with BPD are conceptualised as having a severe personality disorder rather than a trauma-based disorder that involves dissociation as a central symptom.

To be studied is the possibility that several overlapping etiological pathways—including trauma, attachment disruption, and genetics may contribute to the overlap in symptomatology between BPD and DID. To clarify which variables increase the risk for one or both developmental outcomes, research that carefully screens for both DID and BPD is needed. The apparent phenomenological overlap between the two psychopathologies does not create an insurmountable obstacle for research, because distinct influences may be parsed out via statistical analysis. Screening for both disorders would prevent BPD and DID from constituting mutually confounding factors in research specifically about one or the other.

The benefit of accurately diagnosing (1) BPD without DID, (2) DID without BPD, and (3) comorbid DID BPD is that treatment can be individualised to meet patients’ needs. A diagnosis of BPD without DID can lead clinicians to use empirically supported treatment for BPD. By contrast, the treatment of DID is different from the treatment of BPD and comprises three phases: stabilisation, trauma processing, and integration (discussed below). Given the severity of illness found in individuals with comorbid BPD/DID, clinicians should emphasise skills acquisition and stabilisation of trauma-related symptoms in an extended stabilisation phase. Early detection of comorbid DID and BPD alerts the therapist to avoid trauma-processing work until the stabilisation phase is complete. The trauma-processing phase should be approached cautiously in highly dissociative individuals, and only after they have developed the capacity both to contain intrusive trauma material and to use grounding techniques to manage dissociation.

In summary, DID and BPD appear to be separate, albeit frequently comorbid and overlapping, disorders that can be differentiated on validated structured and semi-structured interviews, as well as on the Rorschach test. While the symptoms of DID and BPD overlap, preliminary indications are that the neurobiology of each is different. It is also possible that differences between DID and BPD may emerge regarding the respective etiological roles of trauma, attachment disruption, and genetics.


Some critics claim that DID treatment is harmful. This claim is inconsistent with empirical literature that documents improvements in the symptoms and functioning of DID patients when trauma treatment consistent with the expert consensus guidelines is provided.

Before reviewing the empirical literature, we will present an overview of the DID treatment model. The first DID treatment guidelines were developed in 1994, with revisions in 1997, 2005, and 2011. The current standard of care for DID treatment is described in the International Society for the Study of Trauma and Dissociation’s Treatment Guidelines for Dissociative Identity Disorder in Adults. The DID experts who wrote the guidelines recommend tri-phasic, trauma-focused psychotherapy. In the first stage, clinicians focus on safety issues, symptom stabilisation, and the establishment of a therapeutic alliance. Failure to stabilise the patient or a premature focus on a detailed exploration of traumatic memories usually results in deterioration in functioning and a diminished sense of safety. In the second stage of treatment, following the ability to regulate affect and manage their symptoms, patients begin processing, grieving, and resolving trauma. In the third and final stage of treatment, patients integrate dissociated self-states and become more socially engaged.

Early case series and inpatient treatment studies demonstrate that treatment for DID is helpful, rather than harmful, across a wide range of clinical outcome measures. A meta-analysis of eight treatment outcome studies for any dissociative disorder yielded moderate to strong within-patient effect sizes for dissociative disorder treatment. While the authors noted methodological weaknesses, current treatment studies show improved methodology over the earlier studies. One of the largest prospective treatment studies is the Treatment of Patients with Dissociative Disorders (TOP DD) study, conducted by Brand and colleagues. The TOP DD study used a naturalistic design to collect data from 230 DID patients (as well as 50 patients with dissociative disorder not otherwise specified) and their treating clinicians. Patient and clinician reports indicate that over 30 months of treatment, patients showed decreases in dissociative, posttraumatic, and depressive symptomatology, as well as decreases in hospitalisations, self-harm, drug use, and physical pain. Clinicians reported that patient functioning increased significantly over time, as did their social, volunteer, and academic involvement. Secondary analyses also demonstrated that patients with a stronger therapeutic alliance evidenced significantly greater decreases in dissociative, PTSD, and general distress symptoms.

Crucial to a discussion of whether DID treatment is harmful is the importance of dissociation-focused therapy. A study of consecutive admissions to a Norwegian inpatient trauma program found that dissociation does not substantially improve if amnesia and dissociated self-states are not directly addressed. The study, by Jepsen and colleagues, compared two groups of women who had experienced childhood sexual abuse—one without, and one with, a dissociative disorder (DID or dissociative disorder not otherwise specified). None of the dissociative disorder patients had been diagnosed or treated for a dissociative disorder, and dissociative disorder was not the focus of the inpatient treatment. Thus, the methods of this study reduce the possibility of therapist suggestion. Although both groups had some dissociative symptoms, the dissociative disorder group was more severely symptomatic. Both groups showed improvements in symptoms, although the effect sizes for change in dissociation were smaller for the dissociative disorder group than for the non–dissociative disorder group (d = .25 and .69, respectively). As a result of these findings, the hospital developed a specialised treatment program, currently being evaluated, for dissociative disorder patients (Jepsen E, personal communication, June 2013).

Large, diverse samples, standardised assessments, and longitudinal designs with lengthy follow-ups were utilised in the studies by Brand and colleagues and Jepsen and colleagues. However, neither study used untreated control groups or randomisation. Additionally, Brand and colleagues’ TOP DD study had a high attrition rate over 30 months (approximately 50%), whereas Jepsen and colleagues had an impressive 3% patient attrition rate during a 12-month follow-up.

DID experts uniformly support the importance of recognising and working with dissociated self-states. Clinicians in the TOP DD study reported frequently working with self- states. While it is not possible to conclude that working with self-states caused the decline in symptoms, these improvements occurred during treatment that involved specific work with dissociated self-states. This finding of consistent improvement is another line of research that challenges the conjecture that working with self-states harms DID patients.

Brand and colleagues reviewed the evidence used to support claims of the alleged harmfulness of DID treatment. They did not find a single peer-reviewed study showing that treatment consistent with DID expert consensus guidelines harms patients. Those who argue that DID treatment is harmful to cite little of the actual DID treatment literature; instead, they cite theoretical and opinion pieces. In their review—from 2014—Brand and colleagues concluded that claims about the alleged harmfulness of DID treatment are based on non-peer-reviewed publications, misrepresentations of the data, autobiographical accounts written by patients, and misunderstandings about DID treatment and the phenomenology of DID.

In short, claims about the harmfulness of DID treatment lack empirical support. Rather, the evidence that treatment results in remediation of dissociation are sufficiently strong that critics have recently conceded that increases in dissociative symptoms do not result from DID psychotherapy. To the same effect, in a 2014 article in Psychological Bulletin, Dalenberg and colleagues responded to critics, noting that treatment was consistent with the expert consensus guidelines benefits and stabilises patients.


As we have shown, current research indicates that while approximately 1% of the general population suffers from DID, the disorder remains undertreated and underrecognized. The average DID patient spends years in the mental health system before being correctly diagnosed. These patients have high rates of suicidal and self-destructive behaviour, experience significant disability, and often require expensive and restrictive treatments such as inpatient and partial hospitalisation. Studies of treatment costs for DID show dramatic reductions in overall cost of treatment, along with reductions in utilisation of more restrictive levels of care, after the correct diagnosis of DID is made, and appropriate treatment is initiated.

Delay in recognition, and adequate treatment of DID likely prolongs the suffering and disability of DID patients. Younger DID patients appear to respond more rapidly to treatment than do older adults, which suggests that years of misdirected treatment exact a high personal cost from patients. Needless to say, if clinicians do not recognise the disorder, they cannot provide treatment consistent with expert guidelines for DID.

The myths we have dispelled also have substantial economic costs for the health care system and, more broadly, for society. For example, the myths may deter clinicians and researchers from seeking training in the assessment and treatment of DID, thereby compounding the problems of misunderstanding, lack of recognition, and inappropriate treatment, as we have discussed. The misconception that DID is a rare or iatrogenic disorder may lead to the conclusion that this disorder is one on which resources should not be expended (whereas we have shown the opposite to be the case). In combination, these myths may discourage scholars from pursuing research about DID and also inhibit funding for such research, which exacerbates, in turn, the lack of understanding about, and the currently inadequate clinical services for, DID.


An enduring interest in DID is apparent in the solid and expanding research base about the disorder. DID is a legitimate and distinct psychiatric disorder that is recognisable worldwide and can be reliably identified in multiple settings by appropriately trained researchers and clinicians. The research shows that DID is a trauma-based disorder that generally responds well to treatment consistent with DID treatment guidelines.

Our findings have some clinical and research implications. Clinicians who accept as facts the myths explored above are unlikely to carefully assess for dissociation. Accurate diagnoses are critical for appropriate treatment planning. If DID is not targeted in treatment, it does not appear to resolve. The myths we have highlighted may also impede research about DID. The cost of ignorance about DID is high not only for individual patients, but for the whole support system in which they live (e.g., loved ones, health systems, and society). Empirically derived knowledge about DID has replaced outdated myths, and for this reason, vigorous dissemination of the knowledge base about this complex disorder is warranted.

Credits: Bethany L. Brand, PhD, Vedat Sar, MD, Pam Stavropoulos, PhD, Christa Krüger, MB BCh, MMed (Psych), MD, Marilyn Korzekwa, MD, Alfonso Martínez-Taboas, PhD, and Warwick Middleton, MB BS, FRANZCP, MD for Harvard Review of Psychiatry July 2016.