Lynn et al., (2012). Dissociation and dissociative disorders:Challenging conventional wisdom

The current (fourth) edition of the Diagnostic and StatisticalManual of Mental Disorders (DSM-IV-TR) defines dissociation as “adisruption in the usually integrated functions of consciousness,memory, identity, or perception of the environment” (AmericanPsychiatric Association, 2000, p. 519). Many psychologists andpsychiatrists view dissociation as a coping mechanism designed todeal with overpowering stress (Dell & O’Neil, 2009). Onewell-known form of dissociation is depersonalization, in whichindividuals feel disconnected from themselves; they may feel likean automaton or feel as if they are watching themselves from adistance. Another is derealization, in which individuals feeldisconnected from reality; they may feel as though they are in adream or that things seem to be moving in slow motion. StevenSpielberg’s 1998 film, Saving Private Ryan, vividly depicts anepisode of derealization (spoiler alert): After being shot, CaptainJohn Miller (portrayed by Tom Hanks) witnesses the events aroundhim unfolding as if in a silent, slow-motion movie. Certain formsof dissociation are widespread in the general population; forexample, most estimates suggest that nearly 50% of individuals haveexperienced depersonalization at some point in their lives(Aderibigbe, Bloch, & Walker, 2001). When mild andintermittent, such symptoms are rarely of clinical concern.Nevertheless, in some cases, dissociation may take the form ofgrossly impairing dissociative disorders. These puzzling conditionsinclude dissociative identity disorder (DID), formerly known asmultiple personality disorder, dissociative fugue, anddepersonalization disorder. In the best known dissociativedisorder, DID, individuals supposedly develop multiple coexistingpersonalities, known as “alters.” In dissociative fugue,individuals purportedly suddenly forget their past, travel fromhome or work (fugue has the same root as fugitive), and adopt a newidentity; in depersonalization disorder, individuals experiencefrequent bouts of depersonalization, derealization, or both.Dissociation also features prominently in other psychologicalconditions not formally classified as dissociative disorders, suchas panic disorder, borderline and schizotypal personalitydisorders, and posttraumatic stress disorder. The origins ofdissociation are poorly understood. Nevertheless, the clinicalliterature on dissociation has been marked by three widely acceptedassumptions associated with what is often referred to as theposttraumatic model. Specifically, it has long been assumed thatchronic dissociation is (a) a coping mechanism to deal with intensestressors, especially childhood sexual and physical trauma; (b)accompanied by cognitive deficits that interfere with theprocessing of emotionally laden information; and (c) marked by anavoidant informationprocessing style characterized by a tendency toforget painful memories. The coping mechanism outlined in (a) istypically assumed to play a key causal role in dissociativedisorders. For example, many authors have argued that DID reflectsindividuals’ attempts to “compartmentalize” and obtainpsychological distance from traumatic experiences such as childabuse (Dell & O’Neil, 2009). In this article, we review recentresearch that calls these widespread assumptions into question andproposes novel and scientifically supported approaches forconceptualizing dissociation and dissociative disorders.

The Posttraumatic Model The posttraumatic model (Bremner, 2010;Gleaves, 1996) is ostensibly supported by very high rates—sometimesexceeding 90%—of reported histories of childhood trauma, mostcommonly child sexual abuse, among patients with DID and perhapsother dissociative disorders (Gleaves, 1996; Simeon, Guralnik,Schmeidler, Sirof, & Knutelska, 2001). Nevertheless, a numberof authors (e.g., Giesbrecht, Lynn, Lilienfeld, & Merckelbach,2008, 2010; Kihlstrom, 2005; Merckelbach & Muris, 2001; Piper& Merskey, 2004; Spanos, 1994, 1996) have questioned theoft-cited link between child abuse/ maltreatment and dissociationfor several reasons. First, in most studies (e.g., Ross & Ness,2010), objective corroboration of abuse is lacking. Second, theoverwhelming majority of studies of self-reported trauma anddissociation are based on cross-sectional designs that do notpermit causal inferences; in these designs, individuals aretypically assessed for DID or other dissociative disorders andasked to recollect whether they had been abused or neglected inchildhood. Prospective studies that circumvent the pitfalls of suchretrospective reporting often fail to substantiate a link betweenchildhood abuse and dissociation in adulthood (Giesbrecht et al.,2008; but see Bremner, 2010). Third, researchers have rarelycontrolled for overlapping conditions or symptoms, such as those ofanxiety, eating, and personality disorders, raising the possibilitythat the correlates of abuse are not specific to dissociativedisorders. Fourth, the reported high levels of child abuse amongDID patients may be attributable to selection and referral biases(Pope & Hudson, 1995); for example, individuals withdissociative disorders may be especially likely to enter treatmentif they are struggling with problems stemming from early abuse.Fifth, correlations between abuse and psychopathology decreasesubstantially or disappear when participants’ perception of familypathology is controlled statistically (Nash, Hulsey, Sexton,Harralson, & Lambert, 1993), which could mean that thisassociation is due to global familial maladjustment rather thanabuse itself. These five points of contention suggest ample reasonsto be skeptical of the claim that child abuse plays a central ordirect causal role in DID—although, as we will suggest later, itmay be one element of the complex etiological network thatcontributes to this condition. The Sociocognitive Model In contrastto the posttraumatic model, the sociocognitive model (Spanos, 1994;see also Aldridge-Morris, 1989; Lilienfeld et al., 1999; McHugh,1993; Sarbin, 1995) proposes that DID is a consequence of sociallearning and expectancies. This model holds that DID results frominadvertent therapist cueing (e.g., suggestive questioningregarding the existence of possible alters, hypnosis for memoryrecovery, sodium amytal), media influences (e.g., television andfilm portrayals of DID), and sociocultural expectations regardingthe presumed clinical features of DID. In aggregate, thesociocognitive model posits that these influences can leadpredisposed individuals to become convinced that indwellingentities— alters—account for their dramatic mood swings, identitychanges, impulsive actions, and other puzzling behaviors (seebelow). Over time, especially when abetted by suggestivetherapeutic procedures, efforts to recover memories, and apropensity to fantasize, they may come to attribute distinctivememories and personality traits to one or more imaginary alters. Anumber of findings (e.g., Lilienfeld & Lynn, 2003; Lilienfeldet al., 1999; Piper, 1997; Spanos, 1994) are consistent with thesociocognitive model and present serious challenges to theposttraumatic model. For example, the number of patients with DID,along with the number of alters per DID patient, increaseddramatically from the 1970s to the 1990s (Elzinga, van Dyck, &Spinhoven, 1998), although the number of alters at the time ofinitial diagnosis appears to have remained constant (North, Ryall,Ricci, & Wetzel, 1993). In addition, the massive increase inreported cases of DID followed closely upon the release in themid-1970s of the bestselling book (turned into a widely viewedtelevision film in 1976), Sybil (Schreiber, 1973), which told thestory of a young woman with 16 personalities who reported a historyof severe child abuse at the hands of her mother (see Nathan, 2011;Rieber, 2006, for evidence that many details of the Sybil story areinaccurate). Manifestations of DID symptoms also vary acrosscultures. For example, in India, the transition period as theindividual shifts between alter personalities is typically precededby sleep, a presentation that reflects common media portrayals ofDID in that country (North et al., 1993). Moreover, mainstreamtreatment techniques for DID often reinforce patients’ displays ofmultiplicity (e.g., asking questions like, “Is there another partof you with whom I have not spoken?”), reify alters as distinctpersonalities (e.g., calling different alters by different names),and encourage patients to establish contact and dialogue withpresumed alters. Interestingly, many or most DID patients show fewor no clear-cut signs of this condition (e.g., alters) prior topsychotherapy (Kluft, 1984), raising the specter that alters aregenerated by treatment. Indeed, the number of alters per DIDindividual tends to increase substantially over the course ofDID-oriented psychotherapy (Piper, 1997). Curiously,psychotherapists who use hypnosis tend to have more DID patients intheir caseloads than do psychotherapists who do not use hypnosis(Powell & Gee, 1999), and most DID diagnoses derive from asmall number of therapy specialists in DID (Mai, 1995), againsuggesting that alters may be created rather than discovered intherapy. These sources of evidence do not imply that DID cantypically be created in vacuo by iatrogenic (therapist-induced) orsociocultural influences. Sociocognitive theorists acknowledge thatiatrogenic and sociocultural influences typically operate against abackdrop of preexisting psychopathology. Indeed, the sociocognitivemodel is consistent with findings that many or most patients withDID, and to a lesser extent other dissociative disorders, meetcriteria for borderline personality disorder, a condition marked byextremely unstable behaviors, such as unpredictable shifts in mood,impulsive actions, and self-mutilation (Lilienfeld et al., 1999).Individuals with this disorder are understandably seeking anexplanation for their bewildering behaviors. The presence of hiddenalters may be one such explanation, and it may assume particularplausibility when suggested by psychotherapists or sensationalmedia portrayals. Cognitive Mechanisms of Dissociation Much of theliterature on cognitive mechanisms of dissociation is moreconsistent with the sociocognitive model than with theposttraumatic model. For example, researchers have found littleevidence for inter-identity amnesia among patients with DID usingobjective measures of memory (e.g., eventrelated potentials orbehavioral tasks; Allen & Movius, 2000; Huntjens et al., 2006).In such studies, investigators present certain forms of informationto one alter and see whether it is accessible to another alter. Inmost cases, it is, demonstrating that alters are notpsychologically distinct entities. Contradicting the claim thatindividuals with heightened dissociation are defending against theimpact of threat-related information and therefore exhibit sloweror impaired processing of such information, patients with DID andother “high dissociators” display better memory for to-be-forgottensexual words in directed-forgetting tasks (Elzinga, de Beurs,Sergeant, van Dyck, & Phaf, 2000). This finding is strikinglydiscrepant with the presumed coping function of dissociation.Studies of cognitive inhibition in highly dissociative clinical andnonclinical samples typically find a breakdown in such inhibition,challenging the widespread idea that amnesia (i.e., extremeinhibition) is a core feature of dissociation (Giesbrecht et al.,2008, 2010). The extant evidence therefore questions the widespreadassumption that dissociation is related to avoidant informationprocessing and suggests that apparent gaps in memory ininteridentity amnesia, or dissociative amnesia more generally,could reflect intentional failures to report information. Moreover,the literature indicates that dissociation is marked by apropensity toward false memories, possibly mediated by heightenedlevels of suggestibility, fantasy proneness, and cognitive failures(e.g., lapses in attention). Indeed, at least 10 studies fromdiverse laboratories have confirmed a link between dissociation andfantasy proneness. In addition, heightened levels of fantasyproneness are associated with the tendency to overreportautobiographical memories and the false recall of aversive memorymaterial (Giesbrecht et al., 2010). Accordingly, the relationbetween dissociation and fantasy proneness may explain whyindividuals with high levels of dissociation are especially proneto develop false memories of emotional childhood events. Thisexplanation dovetails with data revealing links betweendissociative symptoms and hypnotizability (Frischholz, Lipman,Braun, & Sachs, 1992) and high scores on the GudjonssonSuggestibility Scale (Merckelbach, Muris, Rassin, &Horselenberg, 2000). Similarly, dissociation increases the numberof commission memory errors (e.g., confabulations/false positives,problems discriminating perception from imagery) but not omissionmemory errors, which are presumably associated with dissociativeamnesia (Holmes et al., 2005). These findings, together withresearch demonstrating a link between dissociation and cognitivefailures, point to an association between a heightened risk ofconfabulation and pseudomemories. They also raise questionsregarding the accuracy of retrospective reports of traumaticexperiences. Still, these findings do not exclude some role fortrauma in dissociation. Suggestibility, cognitive failures, andfantasy proneness could contribute to an overestimation of agenuine, although perhaps modest, link between dissociation andtrauma. Alternatively, early trauma might predispose individuals todevelop high levels of fantasy proneness, absorption (the tendencyto become immersed in sensory or imaginative experiences; Tellegen& Atkinson, 1974), or related traits. In turn, such traits mayrender individuals susceptible to the iatrogenic and culturalinfluences posited by the sociocognitive model, thereby increasingthe likelihood of DID. Sleep, Memory, and Dissociation A recenttheory connecting sleep, memory problems, and dissociation mayprovide a conceptual bridge between the posttraumatic model and thesociocognitive model. In a review of 23 studies, van der Kloet,Merckelbach, Giesbrecht, and Lynn (2011) concluded that data fromclinical and nonclinical samples provide strong support for a linkbetween dissociative experiences and a labile sleep–wake cycle.This link, they contend, is evident across a range of sleep-relatedphenomena, including waking dreams, nightmares, and hypnagogic(occurring while falling asleep) and hypnopompic (occurring whileawakening) hallucinations. Supporting this hypothesis, studies ofthe association between dissociative experiences and sleepdisturbances have generally yielded modest correlations (in therange of .30 to .55), implying that unusual sleep experiences anddissociation are moderately related constructs (see also Watson,2001). Nevertheless, these studies typically relied oncross-sectional designs. To address this limitation, Giesbrecht,Smeets, Leppink, Jelicic, and Merckelbach (2007) deprived 25healthy volunteers of one night of sleep and found that sleep lossengenders a substantial increase in dissociative symptoms. Theyalso found that this increase could not be accounted for by moodchanges or response bias. van der Kloet, Giesbrecht, Lynn,Merckelbach, and de Zutter (in press) later conducted alongitudinal investigation of sleep experiences and dissociativesymptoms among 266 patients who were evaluated on arrival and atdischarge 6 to 8 weeks later. Sleep hygiene was a core treatmentcomponent. Prior to treatment, 24% of participants met the clinicalcut-off for dissociative disorders (i.e., Dissociative ExperiencesScale > 30; Bernstein-Carlson & Putnam, 1993); at follow-up,this number dropped to 12%. Although sleep improvements wereassociated with a reduction in global psychopathology (e.g.,anxiety, depression), this reduction did not account fully for thespecific effect of treatment on dissociation. The fact that asleep-hygiene intervention reduces dissociative symptomsindependent of generalized psychopathology bears noteworthyclinical implications. It also suggests that researchers may wishto revisit the treatment of dissociative disorders. Surprisingly,this clinically important area has received minimal investigation:For example, Brand, Classen, McNary, and Zaveri (2009) reportedthat only eight nonpharmacological studies, none of which was awell-controlled randomized trial, have examined treatment outcomesfor DID. van der Kloet et al.’s (in press) findings suggest anintriguing interpretation of the link between dissociative symptomsand deviant sleep phenomena (see also Watson, 2001). According totheir working model, individuals with a labile sleep– wake cycleexperience intrusions of sleep phenomena (e.g., dreamlikeexperiences) into waking consciousness, in turn fosteringdissociative symptoms. This labile sleep–wake cycle may stem inpart from a genetic propensity (Lang, Paris, Zweig-Frank, &Livesley, 1998), distressing trauma-related memories, or otherunknown causal influences. van der Kloet et al.’s model furtherproposes that disruptions of the sleep– wake cycle degrade memoryand attentional control, thereby accounting for, or at leastcontributing to, the cognitive deficits of highly dissociativeindividuals. Accordingly, the sleep-dissociation perspective mayexplain (a) how aversive events disrupt the sleep–wake cycle andincrease vulnerability to dissociative symptoms, and (b) whydissociation, trauma, fantasy proneness, and cognitive failuresoverlap. Thus, this perspective is commensurate with thepossibility that trauma engenders sleep disturbances that in turnplay a pivotal role in the genesis of dissociation and suggeststhat competing theoretical perspectives may be amenable tointegration. The SCM holds that patients become convinced that theypossess multiple selves as a by-product of suggestive media,sociocultural, and psychotherapeutic influences. Their sensitivityto suggestive influences may arise from increased salience ofdistressing memories (some of which may stem in part from trauma)and susceptibility to memory errors and a propensity to fantasizeand experience difficulties in distinguishing fantasy from reality,brought about at least in part by sleep disruptions. The data wehave summarized have received only scant attention in the clinicalliterature. Nevertheless, they have the potential to reshape theconceptualization and operationalization of dissociative disordersin the upcoming edition of the Diagnostic and Statistical Manual ofMental Disorders (DSMV, publication scheduled in 2013). Inparticular, they suggest that sleep disturbances, as well associocultural and psychotherapeutic influences, merit greaterattention in the conceptualization and perhaps classification ofdissociative disorders (Lynn et al., in press). From thisperspective, the hypothesis that dissociative disorders can betriggered by (a) a labile sleep cycle that impairs cognitivefunctioning, combined with (b) highly suggestive psychotherapeutictechniques, warrants empirical investigation. More broadly, thedata reviewed point to fruitful directions for our thinking andresearch regarding dissociation and dissociative disorders in yearsto come.

Respond to whether you think DID (Dissociative IdentityDisorder) “exists” and what you mean by that. In your opinion, doyou think it exists? What do you think DID is and what causes itand why?