Two papers in this issue of the Journal provide important new findings regarding the prevalence and neurobiology of dissociative disorders. This form of psychopathology has been a stepchild in American psychiatry for centuries, included uncomfortably at best in the family of mental disorders. Pierre Janet’s dissociationist model of psychopathology (1) was influential in Europe but was eclipsed in the United States by Freud’s mental topography emphasizing repression (2). Janet used the term desaggregation mentale, which is poorly translated by the word "dissociation." The English term merely implies separation, whereas the French indicates a kind of forced separation of elements that would normally aggregate, which is actually a better description.

Memory processing depends on the creation of associations, all part of encoding, storage, and retrieval. To be successfully retrieved, episodic memories must be encoded in a temporal context and stored with salient cues that will reliably trigger retrieval. Traumatic experiences have predictable effects on this process. They are typically sudden, create discontinuities with prior experience, involve arousal of intense affect (including fear, anger, and sadness), and may create conflicting patterns of association (e.g., a parent previously viewed as loving and protective is seen in this context as sexually abusive or a threat to life) (3). Thus, the encoding of traumatic experiences tends to be distinct from that of more ordinary events, and the associated implications of such memories for one’s view of oneself (safe, worthwhile, and loved versus in danger, worthless, and hated) can create, especially in children, conflicting networks of information. Just as in depression information is selectively retrieved that tends to perpetuate the dysphoria (I am worthless, disliked, incapable) despite the presence of memories that would contradict this self-evaluation, an inconsistent and at times terrifying environment may create selective networks of association that preclude a more balanced view of the world (sometimes dangerous, sometimes safe) or of the self (good versus deserving of punishment). Processing traumatic memory stores—which convey starkly different associations regarding experience, implications for the self, and emotional arousal—would be difficult under the best of circumstances. Add to that the clear evidence of smaller hippocampal and amygdala volume among those with dissociative disorders (presented in this issue by Vermetten and colleagues) and the ability to encode, store, and retrieve memories and manage associated affect would be sorely constrained (4). The hippocampus is a context generator, helping us to put information into perspective. It has been shown, for example, to buffer the effects of stressful input on HPA activation (5). Dissociation in response to script-driven imagery is associated with decreased activity in the parahippocampal gyrus (6). Limitations on hippocampal size and function hinder memory processing and the ability to comprehend context, especially in the light of contradictory memory encoding and storage.

Do clinicians "remember" to make the diagnosis when it occurs? In another study featured in this issue, Foote and colleagues carefully assessed 231 consecutive admissions to an inner city mental health clinic and interviewed 82 of those willing to cooperate with the study. Patients were not selected on the basis of a screening measure for dissociative symptoms, and the clinic itself had no particular reputation for interest in the disorder. Twenty-four (29% of the subjects) met DSM-IV criteria for a dissociative disorder (dissociative amnesia [N=8], dissociative disorder not otherwise specified [N=7], dissociative identity disorder [N=5], and depersonalization disorder [N=4]). This is a surprisingly high figure, and it suggests that dissociative disorders may be underdiagnosed and undertreated. Indeed, only 5% of this group had previously been correctly diagnosed. As the authors note, there is no definitive psychopharmacological treatment for these disorders, and some treatments, such as antipsychotics, may worsen rather than improve symptoms by reducing cognitive control, increasing depersonalization, and blunting affective response while ending the search for other treatment. Furthermore, Foote and colleagues provide more evidence linking both physical and sexual abuse to dissociative symptoms, with odds ratios of 5.86 for physical abuse and 7.87 for sexual abuse. Similarly, all of the dissociative identity disorder subjects in the Vermetten et al. study also met diagnostic criteria for PTSD. These strong but not absolute associations between trauma and dissociation suggest that a stress-diathesis model linking traumatic experience with vulnerability to dissociation may account for an even greater amount of the psychopathology associated with traumatic dissociation (7).

One of the serious problems in the diagnosis and treatment of dissociative disorders is a tendency to confound observation of the symptoms with belief in their content. A clinician can note that a patient experiences him/herself as having more than one identity or personality state without believing that there really are four people in that body. The DSM-IV work group on dissociative disorders addressed this problem in two ways. The best known was changing the name of multiple personality disorder to dissociative identity disorder to place the correct emphasis on the failure to integrate aspects of identity, memory, and consciousness rather than the apparent proliferation of "personalities." Indeed, the problem is not having more than one personality, it is having less than one. The components of such a personality structure are often quite limited, associated with one primary affect or segment of experiences. This fragmentation complicates the patient’s ability to respond to complex life circumstances and form meaningful relationships. The other change involved one word: "presence" instead of "existence" of more than one identity or personality state. The term "presence" was drawn from the description of delusions in schizophrenia. One can describe a patient’s delusions without believing them. Similarly, one can note disrupted retrieval of memories associated with dissociative fragmentation in identity, memory, and consciousness without seeing the world in the same way the patient does.

These and other studies provide compelling evidence regarding the nature of dissociative disorders, their etiology, and now their neuropathology. Why, then, do dissociative disorders continue to remain underdiagnosed, undertreated, and, frankly, insufficiently respected? They are at once dramatic and puzzling disorders, with symptoms as extreme as one might see in bipolar disorder, with stark changes in mood and behavior, yet allowing for (at least potentially) greater control by the patient. This presentation tends to provoke all-or-none thinking: the patient must be "faking" since he or she can, under some circumstances, suppress the antisocial "alter" and behave. Dissociative identity disorder has been called a "disease of hiddenness" (8). Patients try to put on a good appearance despite chaotic internal lives, in part to try to get by, in part to ward off further anticipated abuse. Thus, they will tend to hide rather than reveal their symptoms, expecting (and often experiencing) disbelief when their symptoms do emerge. Furthermore, sexual and physical abuse and its aftermath are disturbing; they arouse strong affect in observers as well as survivors, and sometimes necessitate legal action or protection from ongoing threat. The clinician is burdened with applying the integrative understanding to the situation that the patient is incapable of utilizing, which can be taxing.

These patients are difficult to treat. There are no quick fixes, although many patients do respond to long-term psychotherapy (8). Harry Stack Sullivan referred to psychotherapy as "participant observation" (9). The therapist needs to interact directly with all elements of the patient’s emotional world. One has to participate in a real enough relationship with the patient so that one comprehends the patient’s world, but the therapist always has to be able to step back and observe while intervening. This involves what one of my patients called applying "Krazy Glue" to fragmented elements of her identity, working through traumatic memories, helping the patient navigate current relationships with family and others, and avoiding further traumatization. The therapist needs to recognize that the patient is fragmented. Efforts to reify each fragment into a "personality" are not helpful. Hypnosis can be useful in teaching patients about the dissociative nature of their symptoms by helping them to gain control over transitions among personality states, with the goal of improving internal communication and integrating disparate aspects of their identity.

We do our patients a disservice if we fail to correctly diagnose and treat them. It appears from these and other studies that as a profession we have dissociated dissociation and would do well to integrate and remember this new information.