Attachment disorganization in different clinical groups : what underpins unresolved attachment ?

This paper summarizes findings and clinical implications of research on attachment disorganization in diverse clinical groups. Disorganized/unresolved attachment is overrepresented in these groups compared to healthy control participants, but disorderspecific characteristics of this attachment pattern are still poorly understood. The focus of this study was to explore defensive processes in participants whose narratives were classified as disorganized/unresolved using the Adult Attachment Projective Picture System (AAP). Besides the predominance of disorganized attachment, clinical participants demonstrated more “segregated system material” especially in stories representing aloneness and more “Personal Experience material” compared to healthy individuals. Within the disorganized/ unresolved clinical individuals, BPD and PTSD patients showed the highest proportion of attachment disorganization and were less able to use other attachment-related defenses to maintain organized. Furthermore, PTSD patients were emotionally overwhelmed by the projective attachment scenes compared to the other clinical groups as indexed by an incapacity to complete sections of the AAP. BPD and addicted patients were characterized by a high degree of self-other boundary confusion. Depressive and schizophrenic patients showed a high overall defensive intensity to remain organized.

psychopathology (Westen et al., 2006).The attachment system is thought to influence the individual's interpersonal perception, expectation and behavior (Bretherton & Munholland, 2008).Several studies demonstrate the central role of attachment and attachment representations in the development of diverse clinical disorders (see review from Dozier et al., 2008).In particular, disruptions in early attachment relationships (for example as a result of experiencing loss) are associated with the emergence of psychopathology later in life.This influence is mainly based on a distorted internalization or fractured mental representation of relationship patterns that usually are elaborated, integrated and consolidated over the life span.In terms of attachment theory these representations are described in terms of disorganized/unresolved attachment status or states of mind (see George, West, & Pettem, 1999).Empirical investigations of disorganized/unresolved attachment are based on Main and Goldwyn's operational definition of lack of resolution (i.e., unresolved) using the Adult Attachment Interview (AAI, George, Kaplan, & Main, 1984/1985/1996;Main & Goldwyn, 1985/1988/1994).The Adult Attachment Projective Picture System (AAP, George et al., 1999;George & West, 2012) defines the dimension that is used to identify disorganized/ unresolved attachment following Bowlby's (1980) view of defensive exclusion of material related to pathological mourning.Unresolved attachment is associated with a particular form of defensive exclusion he termed segregated systems.Under severe conditions, Bowlby proposed, that memories and associated affects have to be segregated or blocked from conscious processing in order to prevent debilitating emotional dysregulation.Segregated systems, therefore, work to prevent overwhelming experiences and flooding as long as they can be contained.Here the discrimination to posttraumatic symptomatology becomes obvious: Uncontained or dysregulated segregated systems might overwhelm or emotionally flood the individual (breakdown of affect regulation) usually in attachment-related situations (not in others) and this temporarily lost regulatory capacity is recovered when the attachment system is deactivated by a specific stimulus, which could be the call upon an internally represented other or a coregulation via a physically present close person (Nolte et al., 2011).

STUDIES ON ATTACHMENT AND PSYCHOPATHOLOGY
The relationship between disorganized/unresolved attachment and psychopathology is consistent with Bowlby's original predictions regarding psychiatric instability as a potential response to the death of attachment figures (Bowlby, 1980).
For PTSD and abused patients (n = 271 in total), the review of Bakermans-Kranenburg & van IJzendoorn (2009) reported 68% of unresolved individuals, while only 14% were classified as secure.Studies investigating the association between attachment and trauma revealed that attachment disorganization (unresolved trauma) was related to high levels of distress and PTSD symptoms in the aftermath of a traumatic event in adulthood (e.g.rape, stillbirth, war veterans) (Harari et al. 2009;Hughes, Turton, Hopper, McGauley, & Fonagy, 2004;Stovall-McClough & Cloitre, 2006).
The findings regarding the association between states of mind of attachment and depression have been rather inconsistent with some studies reporting depression being associated with preoccupied attachment (e. g.Rosenstein & Horowitz, 1996;Fonagy et al., 1996;West & George, 2002) but others reporting certain entities of depression (episodically depressed and dysthymic) as linked with dismissing states of mind (deactivating attachment-related themes) (see Dozier et al., 2008).In our own study using the Adult Attachment Projective Picture System (AAP, George & West, 2001, 2012) we found a moderate amount of unresolved loss (54%) in chronically depressed (CD) outpatients compared to healthy controls (Buchheim et a l., 2012).This is in line with the results of Fonagy et al. (1996) who reported a high percentage of unresolved loss (72%) in depressed inpatients.However, Bakermans-Kranenburg and van IJzendoorn (2009) state in their meta-analysis comprising various populations assessed by the Adult Attachment Interview that in depressed individuals dismissing as well as preoccupied representations are equally present.What might be unique for depressed subjects is the relative absence of unresolved loss or trauma compared to other clinical groups.
De Rick and Vanheule (2007) examined patients with alcoholism and found an overrepresentation of insecure attachment styles using self-report measures.Caspers, Yucuis, Troutman and Spinks (2006) investigated the association between attachment representations (as assessed by the AAI) and addicted behavior in a non-clinical sample.Insecure participants showed more addicted behavior than secure ones and insecure-dismissing individuals showed less compliance for treatment compared to preoccupied ones.However, studies using representational attachment measures such as the AAI or AAP in patients with alcoholism to date are rare (Arnold, 2013).Dozier et al. (1999) reported that schizophrenic patients show more insecure dismissing attachment representations compared to patients with an affective disorder when assessed with the AAI.This finding was replicated by Tyrrell et al. (1999) showing that 89% of schizophrenic patients were classified as dismissing (AAI data).When including the "unresolved" category (four way analysis), 44% of the patients were classified as unresolved.
In sum and in line with the recent meta-analysis reporting on the distribution of attachment representations in clinical and non-clinical groups, we can conclude that clinical participants show insecure and disorganized/unresolved attachment representations at a much higher rate compared to healthy controls.Bakermans-Kranenburg and van IJzendoorn (2009) suggest in their meta-analysis that in future studies more differentiated scales or dimensional approaches should be applied in order to study the range of and subtleties underpinning attachment disorganization in different clinical disorders.
Before presenting our results we describe an established, valid measure of adult attachment, the Adult Attachment Projective Picture System (AAP, George & West, 2001;George et al., 1999;George & West, 2012), that we have used in our research to unravel the nuances of disorganized/unresolved attachment.We provide an overview of the measure with special emphasis on the representational elements and attachment-related defensive processes (for more details see George & West, 2012):

MEASURING ADULT ATTACHMENT REPRESENTATIONS WITH THE ADULT ATTACHMENT PROJECTIVE PICTURE SYSTEM
The Adult Attachment Projective Picture System (AAP, George & West, 2001;George et al., 1999, George & West, 2012) is a set of eight drawings: a neutral scene and seven attachment scenes.The AAP stimuli are line drawings indicating a range of theoretically derived attachment events (e.g.illness, separation, solitude, death, and threat).The interviewee is asked to describe the events that comprise a "story" for each picture, what is happening, what led up to the scene, what the depicted characters are thinking or feeling, and what might happen next.The individual's responses to the stimulus set provide a composite view of representation of self and attachment figures in situations that activate the attachment system.The AAP classification system designates four main adult attachment categories consistent with classifications used with the AAI (secure, dismissing, preoccupied, unresolved).Classifications are based on the analysis of verbatim transcripts of the narratives provided in response to the seven attachment scenarios.The coding system evaluates "alone" stories (in which only one character is portrayed) for evidence of markers for agency and flexible integration of attachment including representations of seeking or receiving care from attachment figures, engaging in constructive behavior to solve the situation, and the desire and ability to seek connections within intimate relationships with others (including friendships and romantic relationships).Dyadic stories are evaluated for evidence of the "goal-corrected partnership" described by Bowlby (1969) as an integrated attachment-caregiving relationship that is characterized by care or mutual enjoyment.All responses are evaluated for attachment-based defensive processes (deactivation, cognitive disconnection, segregated systems) and the presence or absence of autobiographical personal experience (inclusion of memories related to personal life-experience while responding to the AAP stimuli) material in the narrative (George & West, 2012).
Defensive processes.In the volume Loss, Bowlby (1980) discussed at length his view of defense, representation, affect regulation, and the relationship between defensive processes and psychiatric symptoms.His unique approach was to consider defensive processes, a core psychoanalytic concept, from an information processing perspective.He viewed defense as a set of automatic unconscious attentional processes that select, exclude and transform behavior, thoughts and emotions in order to prevent psychological breakdown (see also Hesse & Main, 2006).Bowlby (1980) delineated three qualities of defense: deactivation, cognitive disconnection and segregated systems.The AAP is the only assessment tool that provides descriptions and valid operationalisations of these defenses of adult attachment "in action" (George et al., 1999).
Deactivation.Deactivating defensive processes are defined as attempts to shift attention away from attachment events, individuals or feelings.Deactivation works so as to deflect and prevent the individual from becoming consciously aware of attachment distress.In the AAP system, deactivation is coded for story themes that emphasize the importance of rules, social scripts, power, achievement, authority, distance, or romance.Deactivating defenses produce evaluations of individuals as not deserving care and with attachment needs being rejected or neutralized.
Cognitive disconnection.The associated processes disconnect the elements of attachment from their source.Thus, they undermine consistency and the capability of holding a unitary view of events and associated emotions as well as individuals in mind.This results in confusion and representational shifts that make it difficult to turn away from attachment distress.Cognitive disconnection is evidenced in the AAP by uncertainty, anxiety, and the heightening of emotional involvement.The individual is unable to make decisions about or final evaluations of characters or events, sometimes to the extent of being inextricably caught between opposing themes (e.g. it is nighttime or daytime; the girl is happy or sad).Disconnection is associated with heightened emotional intensity (e.g.anger, frustration) and characterized by coping with arousal through withdrawal, withholding or attempts to gloss over attachment difficulties with the implied hope that these will simply disappear.
George and Solomon conceive deactivation and cognitive disconnection as adaptive forms of defense (George & Solomon, 2008).Adaptive hereby describes that these processes successfully divert or splinter attention from attachment distress and event-related memories and feelings to keep attachment organized.These forms are the primary defensive strategies associated with organized attachment classifications (i.e.secure, dismissing or preoccupied).Segregated Systems.Disorganized attachment is associated with Bowlby's (1980) third form of defense, segregated systems (George & Solomon, 2008;George et al., 1999).A segregated system is defined as the result of an intensified defensive process that locks away the painful "package" of memories and affects associated with a particular threatening attachment relationship from conscious awareness (from working memory in terms of information processing).
Bowlby's approach in defining segregated systems helps to clarify and to understand signs and symptoms of rigid control, frozen constriction, frightened hypervigilance, over-sensitivity and dysregulation.The AAP narratives are coded first for evidence of segregated systems and then evaluated as to whether or not the segregated material is re-organized or contained.Evidence of segregated systems in the AAP includes any feature of an individual's response that indicates helplessness, fear, being out of control, isolated, unprotected or abandoned.In a next step, the responses are evaluated with regard to whether or not the individual is able to contain or re-organize representational dysregulation, a capacity referred to as "resolution" following the nomenclature in the field (Dozier et al., 2008).This re-organization occurs when there is evidence of agency of self or functional assistance from others.The failure to re-organize or contain segregated systems indicates that attachment remains dysregulated leading to the classification of unresolved (U).

RESEARCH QUESTIONS
In our study assessing several clinical groups we first assume that disorganized/unresolved attachment is predominant in those groups compared to healthy individuals.
In a second step we explore patterns of attachment disorganization in diverse disorders more specifically by comparing the degree and intensity of defensive processes (deactivation, cognitive disconnection and segregated systems marker).
Moreover, we include an analysis of the amount of self-other boundary confusion by the evaluation of Personal Experience (PE) material in order to potentially identify and differentiate disorder specific characteristics of attachment disorganization.

Method
Participants.The overall sample consisted of n = 218 adults.72.9% suffered from a psychiatric disorder compared to 27.1% healthy individuals.Clinical samples were diagnosed with addiction and substance abuse (25.7%),PTSD (8.3%), depression (13.3%),Borderline Personality Disorder, BPD (15.6%) or Schizophrenia (10.1%).All patients were diagnosed along DSM-IV criteria by independent and trained clinicians using the SCIDinterview (Wittchen et al. 1997).We only included patients with a clear primary major diagnosis such as e.g.BPD, PTSD or major depression.Except for the schizophrenic patients, all other patients were not medicated.The schizophrenic patients were medicated but showed no severe cognitive impairments.They were able to cooperate in the projective test and did not show formal thought disorder.Patients with addiction and substance abuse (n=55) and patients with schizophrenia (n=22) were recruited from a psychiatric hospital and treated in an inpatient setting.Patients with Borderline Personality Disorder (n=34), depressed patients (n=29) and PTSD patients (n=18) were treated in an outpatient setting.All patients were administered the AAP before the beginning of their treatment.The control group was recruited via advertisements in daily newspapers.All control subjects were screened for possible psychopathology and showed no clinically relevant symptoms in the SCID-Interview.Clinical groups and control group differed in gender distribution with significantly more female patients than controls (Chi 2 = 4.21, p = .03).Moreover, patients were significantly younger (M = 30.13,SD = 8.95) than healthy individuals (M = 39.48,SD = 12.38) (t(216) = 4.63, p <.01).

Measures
Attachment Representation.All participants were administered the Adult Attachment Projective Picture System (AAP, George et al., 1999;George & West, 2012) as the first measure within an assessment battery.Interviews were administered by independent and trained psychologists, within a standardized one-to-one setting.The AAP is a reliable and valid measure of attachment representations for adults.The AAP coding system evaluates content and process elements of the responses to attachment stimuli (see above and George & West, 2001; 2012 for a comprehensive description of the AAP coding and classification system).The pictures depict theory-derived attachment events and are administered as follows: #2 "Child at Window"; #3 "Departure"; #4 "Bench"; #5 "Bed"; #6 "Ambulance"; #7 "Cemetery"; #8 "Corner".There are four "monadic" (an individual alone) and three "dyadic" (two or more individuals) scenes.Transcripts are classified into one of the four standard adult attachment categories: secure (F), insecure-dismissing (Ds), insecure-preoccupied (E) and unresolved (U).Organized attachment (F, Ds, E) is characterized by the capacity to contain dysregulated fear.Disorganized/unresolved attachment is identified by attachment dysregulation produced by the failure of integrated and functional representations of the self to contain or organize segregated systems.The defensive processes (deactivation, Ds and cognitive disconnection, E) have been described above.We counted the occurrences of markers of each defensive process in the stories: sum of Ds and sum of E.Moreover, we developed a score of defensive intensity, which consists of the sum of all defensive markers (E + Ds).
Personal Experience (PE) is defined as the inclusion of memories related to personal life-experience while responding to the AAP stimuli.During the AAP task, individuals are asked to tell a hypothetical story about the characters portrayed in the pictures; individuals are never asked to describe their own experience as an element of their response.Personal Experience is a representational marker for high attachment anxiety and preoccupation with one's own distress that blurs and potentially dissolves the capacity to maintain self-and other boundaries under stress (George & West, 2012).Personal Experience can be found in any of the classification groups as the experience of heightened distress and self-and other boundary blurring is a sign of high anxiety and is not a classification criterion per se.
In the present study two independent certified and reliable judges coded 20% of the AAPs of the total sample (n = 44).There was a 92% agreement in the four attachment categories (kappa = .88).Personal Experience was coded as either present or absent per story.100% agreement for coding Personal Experience was achieved in our study by two independent certified judges.

Results
Attachment distribution.First we analyzed the distribution of attachment classifications in our sample with the assumption that insecure as well as disorganized/unresolved attachment was predominant in the clinical groups compared to healthy controls.Overall, the distribution was significantly different comparing clinical and healthy participants (Chi 2 = 82.96,p <.001).As expected we found significantly more attachment insecurity in the clinical groups (Chi 2 =65.33, p<.001) with a predominance of the disorganized/ unresolved classification (Chi 2 =18.72, p<.001).More than half of our clinical participants showed attachment disorganization (51.6%), and almost all (96.2%) of these individuals were characterized by an insecure internal working model of attachment (see table 1).As shown in table 1 the differences in distribution within the various clinical groups were also significant (Chi 2 =30.11, p<.01).PTSD and BPD showed the highest percentage of attachment disorganization (U = Unresolved) in combination with preoccupied (E) states of mind.
Depressive and addicted patients were mainly characterized by a moderate percentage of attachment disorganization.Addicted patients showed both dismissing (Ds) and preoccupied (E) attachment states of mind and depressive patients showed a higher amount of dismissing (Ds) states of mind.Patients with schizophrenia demonstrated the lowest rate of attachment disorganization and showed both dismissing (Ds) and preoccupied (E) attachment states of mind.

Unresolved attachment in clinical and non-clinical groups.
Comparing only the disorganized/unresolved participants in clinical and non-clinical groups, all 92 individuals classified unresolved (12.0%healthy, 25.0% addiction, 16.3% PTSD, 12.0% depression, 29.3% borderline and 5.4% schizophrenia) were included.An unresolved classification in clinical samples was characterized by a significantly higher amount of unresolved material (segregated system markers) especially in monadic AAP stories (Mann Whitney, Z = 3.37, p <.001), but not in dyadic stories (Z = 0.28, p = .82)(see table 2).In addition to the higher degree of unresolved material in general (see figure 1a) stories with material of Personal Experience (PE) were more prevalent in unresolved patients than in healthy controls with unresolved states of mind (Chi 2 = 8.61, p <.01) (see Figure 1b).Comparing healthy and clinical participants with an unresolved state of mind with respect to intensity and quality of normative defensive processes (E, Ds) we did not find any significant differences (Z = 0.91, p = .53).
Unresolved attachment in diverse clinical groups.We therefore compared only clinical participants classified as unresolved with respect to their use of defensive processes.Using a Kruskal-Wallis Test analysis we found significant differences between the clinical groups in the amount of stories with markers of deactivation (F (,216, 4) =11.80, p <.01), with markers of cognitive disconnection (F (216, 4) = 17.12, p <.01) and overall defensive intensity (F (216, 4) =14.22,p <.01).
Applying a U-Test post hoc analysis of the intensity and quality of defenses, we found the following characteristics (see figure 2): Depressive patients showed the highest rate of overall defensive intensity and of deactivation compared to patients with PTSD (Z = 2.71, p <.01), BPD (Z = 1.92, p <.01), addiction (Z = 3.81, p <.001) and schizophrenia (Z = 3.21, p <.001).Cognitive Disconnection was equally high in patients with schizophrenia and depression and significantly lower in patients with PTSD (Z = 3.15, p <.001), BPD (Z = 1.61, p <.01) and addiction (Z = 2.85, p <.01).When analyzing the number of AAP stories including material of Personal Experience (PE) and of segregated systems (Unresolved, U) we found the following characteristics (see Figure 3): PTSD patients (U = 3.11, p <.001) compared to those with schizophrenia and BPD (U = 4.73, p <.001) demonstrated the highest amount of unresolved material in their narratives.
Regarding Personal Experience in the stories, BPD patients (M = .85,SD = 1.26) and especially addicted patients (M = 2.13, SD = 2.30) had difficulties in self -other boundaries by showing the highest amount of Personal Experiencerelated material compared to the other clinical groups.

DISCUSSION
As expected, all clinical groups in our study showed significantly more attachment insecurity and disorganized/unresolved attachment than healthy individuals.This is in line with the majority of studies comparing clinical and non-clinical samples (Bakermans-Kranenburg & van IJzendoorn 2009).In our sample with n=218 adult participants we were able to replicate this robust finding using a relatively new measure of attachment representations: the Adult Attachment Projective Picture System (George et al., 1999;George & West, 2012).As shown in the large meta-analysis of Bakermans-Kranenburg and van IJzendoorn (2009) 15% of non-clinical subjects or healthy controls were classified as unresolved/disorganized.Our results (15.3%) confirm this finding.Unresolved patterns of attachment were mainly due to unresolved experiences of loss.As stated in the introduction this differs from posttraumatic symptomatology.In addition to this replication the focus of the present study was the exploration of potentially specific patterns of attachment disorganization comparing patients and healthy individuals (classified unresolved in attachment) in general and within the clinical groups in particular.We found that PTSD and BPD patients were characterized by the highest amount of attachment disorganization.This result is in line with data from the meta-analysis by Bakermans-Kranenburg and van IJzendoorn (2009) demonstrating that individuals who have been exposed to abuse and who were suffering from posttraumatic stress were always classified as unresolved.It is stated that unresolved attachment is a promising marker for dissociative disorders such as PTSD which sheds light on the etiology and underlying mechanisms involved.On the other hand depressed and addicted patients in our study showed a rather moderate proportion of attachment disorganization whereas schizophrenic patients showed the lowest degree.
When comparing only individuals classified as unresolved in attachment, clinical participants demonstrated more segregated system defenses, especially in monadic AAP stories representing aloneness compared to healthy individuals classified as disorganized/unresolved (see Table 2).The coding system of the AAP evaluates monadic stories for evidence of agency and flexible integration of attachment, including representations of seeking or receiving care from attachment figures, engaging in constructive behavior and the desire and ability to seek connections in intimate relationships with others.From an attachment perspective representations or experiences of being alone activate the attachment system by potentially inducing threat (Bowlby, 1980).Our findings reveal that the disorganized clinical participants were less able to contain and regulate their fears when confronted with stimuli representing aloneness.Thus, disorganization in clinical samples may be based on a reduced capacity to be alone (which requires access to the individual's own internal, representational world).
Moreover, the amount and the number of stories with Personal Experience (PE)-material were significantly higher in disorganized/unresolved clinical individuals than in healthy controls with the same attachment classification.This implies that only clinical groups showed a lack of capacity to maintain self-and other boundaries under the particular stress that was induced by completing the AAP stories.
There is clear evidence that attachment disorganization is not only a posttraumatic phenomenon.We consider attachment disorganization not as specific psychic disorder per se but as a risk factor for disorder-unspecific psychopathological development underpinning multifinality.A dysregulated attachment system is not necessarily related to a reduced overall internal capacity to regulate emotions.It is rather based on individually delineated and scaled triggers of attachment distress that activate attachment in a way and intensity that leads to dysregulation.More support for this notion is provided when comparing healthy and clinical participants with respect to the intensity and quality of "normative" defensive processes (E, Ds) where no significant group differences were found.
In order to identify possible patterns of attachment disorganization underpinning specific psychopathologies we analyzed patients classified as unresolved with respect to the amount of segregated material, their use and intensity of defensive processes as well as the amount of self-other boundary confusion.We found the following characteristics: Depressed patients showed the highest amount of overall defensive intensity and of deactivation compared to all other clinical groups.This finding suggests that depressed patients are able to maintain organized regulatory levels with high effort using defensive mechanisms before becoming disorganized.In this respect Dozier et al.'s (1999Dozier et al.'s ( , 2008) ) suggestion that mood disorders show a combination of dismissing and preoccupied features rather than having a high degree of unresolved attachment was confirmed.This is in contrast to patients suffering from severe abuse and/or posttraumatic stress.We may conclude that the intensity of defenses in unresolved patients can be interpreted as a sign of their efforts to remain organized until their affect regulation strategies break down.From this perspective, PTSD as well as BPD patients in our sample appeared less deficient compared to depressive and even schizophrenic patients.Indeed, PTSD and BPD patients demonstrated the highest amount of unresolved material in their narratives.An additional characteristic for (disorganized) PTSD patients is the high amount of "constriction".Constriction is a mental state that is the result of extreme fear (George & Solomon, 2008) and appears to be a "desperate" mental strategy that prevents segregated attachment material from becoming activated and, thus, blocks painful attachment material from flooding consciousness.In some cases, constriction is revealed by mental freezing; here the individual refuses to or cannot tell a story about a picture and in some situations the individual may return the picture to the interviewer or ask to move on to the next.This could mean that PTSD patients are unable to continue their narratives because they were emotionally overwhelmed or flooded by the attachment scenario without even being able to use organized defenses strategies as a regulatory attempt.Mental constriction is considered an indication of failed integration or failed containment of segregated systems.
Furthermore, the analysis of clinical participants classified as unresolved with respect to the amount of Personal Experience also showed specific qualities: BPD patients and especially addicted patients had severe difficulties in keeping self -other boundaries by showing the highest amount of Personal Experience material compared to other clinical groups.In our sample these two clinical groups included elements from their own abuse or maltreatment experiences or addiction history with alcohol in their narratives (see also Arnold, 2013).Therefore, these two patient groups seem incapable of remaining in the hypothetical realm of an AAP story but rather activate and slip into their own biographical material.
In addressing the significant age differences between clinical and healthy individuals in the current sample it has to be noted that the distribution of attachment classifications has been shown to be clearly independent of age in adults (see George & West, 2012).
In sum, we can conclude that BPD and PTSD patients showed the highest proportion of attachment disorganization and were less able to use other attachment-related defenses to maintain organized.PTSD patients were significantly more "blocked" to continue their narratives because they were emotionally overwhelmed by the AAP stimuli compared to the other clinical groups.Moreover, BPD and addicted patients showed a high amount of self-other boundary confusion.Depressive patients and schizophrenic patients showed a high defensive intensity; while depressive patients used both, deactivation and cognitive disconnection as defense significantly more than any other group, schizophrenic patients relied especially more upon cognitive disconnection to remain organized.These preliminary findings may serve as a useful framework to generate and test further hypotheses in order to differentiate the disorganized/ unresolved categories in different psychiatric disorders.

Figure 1a .
Figure 1a.Number of unresolved stories in healthy and clinical groups Figure 1b.Amount of Personal Experience in healthy and clinical groups

Figure 2 .
Figure 2. Number of stories including Ds and E-markers in unresolved clinical groups Note.Defensive intensity (E + Ds), E = amount of E-markers, Ds = amount of Ds-markers.

Figure 3 .
Figure 3. Number of stories including Personal Experience and Unresolved material

Table 1 .
Distribution of attachment classifications (4-way) in healthy and clinical groups

Table 2 .
Distribution of "unresolved stories" in monadic and dyadic picture