Attachment theory is almost unique among psychoanalytic theory in bridging the gap between general psychology and clinical psychodynamic theories. There has been a gulf between theories of the mind that have their roots in psychological research and clinical theories that focus on the significance of individual experience in determining life course, including psychopathology. Yet attachment theory has a home on both sides of the fault line [1].

Bowlby was among the first to posit that the human infant enters the world predisposed to participate in social interaction. Bowlby’s critical contribution was his unwavering focus on the infant’s need for an unbroken (secure) early attachment to the mother [2].

Attachment theory postulates that the affective bond that develops between the child and the caregiver has consequences for the child’s emerging self-concept and developing view of the social world. Based on ethological theory, John Bowlby conceptualized human motivation in terms of “behavioral systems,” and noted that attachment-related behavior in infancy such as clinging, crying, smiling, monitoring caregivers, and developing a preference for a few reliable caregivers or “attachment figures” is part of an evolution-based functional biological system that increases the likelihood of protection from dangers and predation, and comfort during times of stress. However, the fundamental survival gain of attachment lies not only in eliciting a protective caregiver response, but also in the experience of psychological containment of aversive affect states required for the development of a coherent and symbolizing self [1].

Bowlby proposed that through repeated transactions with their attachment figures, infants form mental representations or affective– cognitive schemata of the self and others and develop expectations about interpersonal relations, which he called “internal working models” [2]. The central feature of the internal working model concerns the expected availability of the attachment figure, not the physical proximity. A key feature of the evolving working model of the self is how acceptable or unacceptable the child feels in the eye of the attachment figure. A child whose internal working model of the caregiver is focused on rejection is expected to evolve a complementary working model of the self as unlovable, unworthy, and flawed. These models of the attachment figure and the self are transactional, interactive models representing self-others relationships [1]. Working models are thought to be initially encoded in procedural memory as expectations that help the infant feel secure.

In the 1970s the work of Ainsworth helped to refine the attachment concept. With her colleagues, Ainsworth developed a laboratory procedure called the Strange Situation, which was designed to assess the quality and organization of infant attachment and exploratory behavior in the context of incrementally increasing environmental stress [3]. The Strange Situation consists of a series of infant–caregiver separations and reunions, and the behavior that the infant manifests during the procedure serves as the basis for Ainsworth’s attachment classifications. They found that the mayority of the one-year-old children respond to the mother with proximity seeking and relief at reunion (securily attached infants), but about 25 percent respond with subtle signs of indifference (anxious avoidantly attached infants) and further 15 percent respond with proximity seeking but little relief at reunion (anxious resistantly attached infants). Later, a fourth category, disorganized–disoriented, was added [4].

The disorganized baby displays disorganized and/or disoriented behaviors in the parent’s presence, suggesting a temporary “collapse” of a behavioral strategy. Considering Bowlby’s statement that attachment influenced human relationships “from the cradle to the grave”, Main, Kaplan and Cassidy employed Ainsworth’s typology of attachment patterns in the development of the Adult Attachment Interview (AAI) [5]. The AAI is a semi structured interview designed to elicit thoughts, feelings, and memories about early attachment experiences, and to assess the individual’s state of mind with regard to early attachment relationships [6]. They found four major categories: Secure, dismissing, preoccupied and unresolved for trauma and loss. Interviews that do not fall into one of the above three categories are given a CC rating (can not classify). The first three categories parallel the parent–child attachment patterns originally identified in childhood –the secure, avoidant, and ambivalent- by Ainsworth [3]. The unresolved for trauma and loss category corresponds to the pattern of disorganizeddisoriented attachment later described in infants who had been subjected to maltreatment or to frightened or frightening behaviors on the part of parents. Since AAI was developed, several interviews and self-reports have been made to identify attachment styles. One of them is the CAMIR (Cartes: Modèles Individuels de Relation), the one we used in our study [7].

Cartes: Modèles Individuels de Relation (CAMIR)

CAMIR is a self-report questionnaire to access internal working models in adults. This questionnaire was created by Pierrehumbert and coauthors in Lausanne, Switzerland. Its objective is the evaluation of adult relational strategies, assuming the existence of a model of itself and others in interpersonal relationships [8]. The instrument tries to explore, on one hand, the person’s current estimation about infantile close relationships and, on the other hand, the characteristics of the current interpersonal exchange in their family system.

Theory of personality development by Otto Kernberg

Another theory that we focused on was the object relations theory M. Klein, M. Mahler, E. Jacobson and specifically Otto Kernberg’s contributions to the personality structure formation. According to his theory, during the first months of life multiple internal object relations based on prototype experiences with the primary caregiver are created. The nature of these experiences differs from one point to another in terms of emotional intensity. During times of quiescence the infant is connected with the surrounding environment in a kind of cognitive learning. Likewise, there are moments of high emotional intensity, usually related to the need or desire for pleasure or fear or desire to move away from pain. These periods of emotional high intensity involving the self in relation to an object, facilitate the internalization of primitive object relationships along the axis of the reward, that is, it undergoes either as idealized object (completely good) or aversive one (all bad). In other words, when the baby is under intense affects the experience of self and object acquires a unique importance to facilitate the settlement of affective memory structures [9].

During the course of infant development multiple experiences full of affection are internalized, in order that a part of the psyche is constructed with idealized images based on successful experiences on the one hand, and another segment is built on negative experiences, on the other, with devalued images of self and others.

In the normal development of the child, there is a gradual integration of these extreme representations of self and object during the first years of life. This integration results in internal representations of self and others that are more complex and realistic and recognize that people are a mixture of good and bad attributes and are capable of being nurturing and frustrating at different moments.

When the baby is not able to avoid pain or satisfy a need, it emits signals to the caregiver. The latter, reads these signs and responds, both in behavior and affectively. However, if the interactive system between mother and child is distorted by an insecure attachment, the baby experiences an inconsolable affect. One result of this process is that the normal integration of opposing affective experiences are not performed, and therefore, the motivational system remains decoupled generating a series of mental mechanisms to deal with the intense negative affect, based on splitting, such as proyective identification, idealization, devaluation and primitive forms of projection.

Another consequence of the lack of integration of good and bad aspects of self and others is identity diffusion. Without a complex and realistic self-image the identity is split and is neither consistent nor continuous in time.

Kemberg has theorized that personality “organization” falls into three broad classes, namely the neurotic, borderline, and psychotic levels of organization [10]. In Kernberg’s model, the level of personality organization is determined by an individual’s position on each of three separate components in a multidimensional model, namely, primitive psychological defenses, reality testing, and identity diffusion. In brief, primitive psychological defenses are those defensive propensities (and their behavioral referents) such as projection, denial, dissociation or splitting [10,11]. These defenses suggest more severe psychopathology and are distinguished from healthier variants of defensive operations such as reaction formation, isolation, undoing, suppression, and, of course, repression. Reality testing, in this model, “refers to the capacity to differentiate self from nonself, intrapsychic from external stimuli, and to maintain empathy with ordinary social criteria of reality” [11]. At its most extreme level of impairment, failure of reality testing is manifested by psychotic disorganization of thought and behavior as it appears in psychotic states. Finally, identity diffusion refers to those psychological and behavioral indicators that derive from a poorly integrated identity, particularly poorly integrated concepts of self and significant others.

Personality Disorders and their related behavioral and psychological referents, even those delineated within the DSM nomenclature, emanate from within the borderline level of personality organization. This level of organization is characterized by:

1. Broadly intact reality testing

2. Predominance of primitive psychological defenses.

3. Marked identity diffusion and it defines the underlying developmental matrix from within which all forms of Personality Disorders arise. The borderline level of organization, however, should not be confused with DSM-defined borderline personality disorder, which is but one disorder that can derive from borderline personality organization.

In recent years, efforts have been made in terms of instruments to operationalize this general model of Personality Disorder diagnosis and classification. One of these instruments is the Inventory of Personality Organization (IPO), developed by Kernberg and colleagues [12,13]. The IPO is a 155-item self-report instrument that assesses the three major dimensions relevant to Kemberg’s model of personality organization (identity diffusion, primitive psychological defenses, and reality testing) as well as several other supplementary scales of interpersonal phenomena. The three primary clinical scales consist of 57 items on the IPO; the remaining items concern secondary scales and are not addressed in this study. The IPO is intended to aid in the assessment of the behaviors and psychological features reflective of identity diffusion, primitive defenses, and reality testing in both clinical and nonclinical populations. Prior research Foelsch and coauthors has demonstrated the internal consistency and test-retest reliability of the IPO in nonclinical community and clinical samples as well as basic criterion validity relationships with an established measure of PD in a clinical sample [13]. To the same extent, the internal consistency and reliability of the IPO has been proved in clinical and nonclinical population in Santiago, Chile [14].

In recent decades, psychopathology researchers and theorists have begun to understand fundamental aspects of personality pathology, specifically borderline personality disorder, such as unstable, intense interpersonal relationships, feelings of emptiness, bursts of rage, chronic fears of abandonment and intolerance for aloneness, and lack of a stable sense of self as stemming from impairments in the underlying attachment organization [15]. Our research is an attempt to establish this link between personality pathology and impairments of attachment.

Although both theories- kernberg theory and attachment theoryhave many points of divergence, such as the conceptualization of the internal world, there are fewer important points of contact to consider. As Fonagy posits, Kernberg’s formulation of borderline pathology, translated into attachment theory language might be the activation of poorly structured, highly distorted unstable internal working models with loose assignments of object and subject [16].

Since both theories that we presented, consider as crucial the responsiveness of the caregiver for the normal development of the personality, we hypothesize that the unavailability of an attachment figure that meets the needs of the child:

• Does not create complex and integrated subjective experiences of self and others, with its good and bad aspects, through which primitive defensive mechanisms are maintained, (avoiding “bad” aspects threatening the “good” ones) and also does not allow the identity integration, whereby the experience of cohesion and coherence of self is lacking.

• “Internal working models” that are internalized and displayed in situations of close relationships are either preoccupied (exhibiting more anxiety in relationships) or dismissing (which exhibits greater avoidance).

Hence, our working hypothesis is that people having insecure attachment, either the dimension of anxiety or avoidance, have a greater tendency to use primitive defensive mechanisms and to have identity diffusion. Moreover, there is abundant evidence of the relationship between attachment styles and psychopathology reported in several research papers and pointing to major problems in insecure attachment styles thus, we expect to confirm that there are differences in scores on the OQ-45.2, between different types of attachment evaluated through CAMIR, insecure styles tending to have higher scores on scales of the OQ-45.2 [17]. It is also expected that people with higher scores on IPO also have greater psychopathology, reflected in higher scores on the OQ-45.2

Participants

Participants were two samples, one of them clinical and the other nonclinical. The clinical sample consisted of 27 inpatient women recruited from a Psychiatry Service of a General Hospital in Santiago, Chile. The nonclinical sample was 24 women who were attending health checks at their corresponding primary care center, in the same urban area (Providencia, Santiago). Regarding the clinical sample, the inclusion criterion was women with the clinical DSM IV diagnosis of Borderline personality disorder. Patients were excluded if they were psychotic or did not receive approval for participation from the attending psychiatrist because of severely unstable conditions. The inclusion criteria for the community sample was women who were attending a regular check-up without a reason for mental health consultation or who were accompanying someone and who also did not have a current mental health problem.

Procedures

All procedures were reviewed and approved by the hospital staff and the Psychiatry Department Director. During intake interviews, patients were assessed by their attending psychiatry resident to determine DSM-IV-TR diagnosis and the appropriateness of their participation in the study, based on clinical status. Eligible patients interested in participating scheduled an appointment with a research assistant. After the study was explained, the patient signed a consent form. Medical students and Psychiatry residents, who were previously trained, acted as research assistants. They administered both, the CAMIR and the IPO at different times, separated by roughly a week. During the first interview they also collected epidemiological information and registered the OQ- 45.2, which is applied in every intake interview by the hospital staff.

Instruments

CAMIR The CaMir is an auto-questionnaire aimed at measuring attachment cognitions. It investigates participant’s evaluations of past and present experiences (respectively with family of origin and with current attachment figures), their personal interpretations of parent’s attitudes during childhood and the impact these had, and their conceptions of family functioning. Hence, the items are defined to cover 4 levels of reality: The past, the present, the mood status (current elaboration of parental implication in upbringing) and generalizations (representations about parenthood and a child’s emotional needs). In each of these levels, the items explore relational strategies: Primary strategy: the person values social support and relational security (secure prototype). Secondary strategy: the individual values independence at the expense of relational support (dismissing prototype) or, on the contrary, the individual values interpersonal implication at the expense of autonomy (preoccupied prototype). [18]. Because of its Q-sort format and the requirement of a forced distribution of items, the tendency to give socially desirable responses is reduced (although not eliminated). Measuring proximity of participants’ scores on all of the items to those of prototypes allows a finer assessment than that obtained with self-reports based on a few or sometimes even single items. In addition, continuous scales provide more analytic power than categories. [19].

Attachment cognitions are investigated in two different ways: the first consists of determining participants’ overall attachment style, while the second focuses on cognitions regarding specific aspects of attachment (e.g., parental attitudes, experiences, personal reactions in determined circumstances). Hence, two approaches are proposed to code participants’ responses: computing proximity scores (Spearman correlations) between the participants’ answers with those of prototypes to determine attachment style, and computing mean scores among the items included in a scale to investigate specific aspects of attachment. Prototypes and scales have been constructed independently.

The 72 items, grouped in the following scales, constituted the final instrument:

a) Parental interference

b) Over involvement

c) Restrictive parenting

d) Parental support

e) Open communication

f) Positive evaluation of Childhood

g) Lack of parental concern

h) Self-sufficiency

i) Spite against parents

j) Childhood trauma

k) Lack of memory

l) Inconsistent and resigned parental attitudes

m) Traditional family values.

Scales ABC approximate to the notion of preoccupation, scales DEF are relatives to autonomy, scales GHI evoke dismissing, scales JK relate with no resolution and scales LM with familiar system structure. The 72 items are printed on cards. Participants are asked to sort them into three and then five piles ranging from most characteristic to least characteristic. Final sorts are forced into a bell-shaped distribution with respectively, 12, 15, 18, 15, and 12 items in each pile.

Inventory of Personality Organization (IPO)

The IPO is a 155-item self-report instrument that assesses the three major dimensions relevant to Kemberg’s model of personality organization (identity diffusion, primitive psychological defenses, and reality testing) as well as several other supplementary scales of interpersonal phenomena. We used this instrument in its 2001 version, which has been validated in our country, showing adequate validity and reliability [14], but we focused on the three primary clinical scales consisting of 57 items on the IPO. The rest of the items were not applied. The score of structural scales IPO were classified into three levels according to previous investigations (<40: Low score; 40-60: Medium score; >60 high score), where a person having higher score has more primitive defenses, identity diffusion and reality testing impairments [20].

The Outcome Questionnaire-45.2

The OQ-45 is an instrument developed to examine patient progress [21]. The OQ-45 is a 45-item patient self-report instrument designed to assess experience of psychological distress, interpersonal functioning, and contentment with social role functioning [22]. The 45 items are assessed with a 5-point Likert scale (0=never, 1=rarely, 2=sometimes, 3=frequently, 4 = almost always), with nine of the items reverse scored to limit the likelihood of response bias. The sum of all items gives a total distress (TD) score, ranging from zero to 180, with higher scores being indicative of greater levels of psychological distress.

The psychometric properties of the OQ-45 have been studied extensively in many countries, including Chile and the instrument has been found to be reliable and valid [23].

Analysis

The analysis scheme considered the description of the sociodemographic characteristics of the sample and the distribution and reliability of scales evaluated in the study. The comparison of these characteristics between the two samples (patients and nonpatients) was performed using Fisher’s exact test for variables binary or categorical response and by t test for continuous variables.

To analyze the association between secure attachment style with the score scales of OQ45.2 and IPO, simple linear regression was used and to evaluate the correlation between personality structure and quality of mental health the Pearson correlation coefficient was estimated.

All analyses were performed using Stata 13.1 software and the significance level considered was 0.05.

Table 1 describes the main demographic characteristics and medical history of both samples, clinical and non-clinical. In general, samples are similar in age, marital status and educational level, however, they differ as to the employment situation, where a higher proportion in the clinical group is inactive or unemployed (55.6% vs. 8.3%), and the number of children, where a higher proportion in the clinical sample having 2 or more children (59.2% vs. 20.8%). Regarding the medical and psychiatric history, no women in the control group have indicated that background, in contrast to what was observed in the other group, where 48.2% had medical history and 88.9% one or more psychiatric hospitalizations. Consistent with these results, the questionnaire score OQ-45.2 was significantly higher among patients in the case group, indicating greater presence of psychiatric pathology.

Table 1: Socio-demographic characteristics and medical history in patients with borderline personality disorder (cases) and a community sample (controls).

Table 2 shows mean, standard deviation and reliability of both CAMIR and IPO. The average score for the three scales of the IPO was on average 22 points higher among cases.

Table 2: Mean, SD and reliability of subscales CAMIR and IPO.

The reliability of the subscales of both instruments was good, except for five subscales of CAMIR: Parental Interference, self sufficiency, lack of parental concern, lack of memory, inconsistent and resigned parental attitudes and traditional family values.

The distribution between cases and controls regarding attachment styles and scores of IPO are presented in (Table 3). The distribution of attachment styles in the total sample was: secure 25,4%, dismissing 23,5% and preoccupied 50,9%. The secure attachment style was lower in the clinical group (14.8% vs. 37.5%). Regarding dismissing prototype, there is a significant difference between the groups, with a higher percentage in the case group (37.0% vs. 8.3%). For the preoccupied style, there is a higher percentage in the control group than in the clinical group (48.2% vs 54.8%), although the difference is not significant.

Table 3: Attachment Style and personality structure in BPD patients and controls.

With respect to the IPO, significant differences in the distribution of the three structural scales between cases and control was observed, where the case group had a higher proportion of people in the categories of high score compared to the control group (Table 3).

Finally, the association between secure prototype (vs. the other two) and scales of IPO and OQ (Table 4) was analyzed. No significant association between attachment style and the three primary scales of the IPO was observed, although the score on these three scales was lower among those with a secure attachment style. It was also observed that those with secure attachment prototype had on average 42 points lower on the OQ-45.2 than women with no secure prototype, a difference that was statistically significant.

Table 4: Association between secure attachment prototype and scores of IPO and OQ45.2.

Table 5 shows the assocciation between OQ and IPO. The results show a direct association between both tests, ie a worse personality structure, lower quality of mental health.

Table 5: Pearson correlation of OQ45.2 score and IPO scales.

Our study compared two samples with clinical and sociodemographic differences as Table 1 shows. The clinical group had a diagnosis of Borderline Personality Disorder, and were unemployed or inactive at a higher proportion compared to the control group and also had a higher number of children.

14.8% of secure attachment in the clinical group was observed, which are those with high scores in IPO, ie patients with personality pathology. Levy, after a thorough review of the literature, pointed out that secure attachment in BPD group is extremely low, especially compared to other groups, which is the case [15]. Across interview measures, secure attachment ranges from 0 to 30%, usually around 6 to 8%. In contrast, in the control group, there is a percentage of 37.5% of secure prototype, which is lower than what has been found in the general population around the world (59%) [24]. However, two studies in Chile had observed that the percentage of secure attachment in this country is lower than what has been found worldwide. Garrido et al found 51.4% of secure attachment prototype using the CAMIR test and Spencer, meanwhile, found 37% of secure attachment style in a sample of 549 subjects, using the ECR (experiences in close relationships). In the same study, the autor made a comment about the higher proportion of preoccupied and fearful attachment style and less secure attachment in Chilean population compared to previous studies evaluating attachment styles in different cultures. However, the autor posits that the intercultural study [25].

Mentioned is not part of any Latin American sample, so it is not possible to say that the distribution of attachment found corresponds to the Latin countries [26]. According to the Spencer study, what is noticeable in our study is the high percentage of preoccupied attachment style in the control group, which means that not all people having insecure attachment develop pathology.

It is interesting that the clinical group (people with personality pathology) had higher scores on IPO, which means that the instrument demonstrates its capacity to discriminate this kind of pathology, as has been found in previous researches. [13]

Secure attachment is low in people with personality pathology. That is, having secure attachment would protect against Personality Pathology but it is not the only factor. Moreover, having insecure attachment does not mean that a person will develop pathology. As the study shows, a considerable percentage of people in the control group showed preoccupied attachment style. As Cassidy et al posit, early attachment is not expected to be perfectly predictive of later outcomes [27]. Furthermore, attachment insecurity per se is not psychopathology nor does it guarantee pathological outcomes. Instead, insecurity in infancy and early childhood is thought to be a risk factor for later psychopathology if subsequent development occurs in the context of other risk factors (e.g., poverty, parental psychopathology, abuse). Security is a protective factor that may buffer against emotional problems when later risks are present [28]. Another fact that indicates the same point is the lower score of OQ45.2 in the secure group compared to the non-secure, as Table 4 showed.

With respect to the association between personality structure and secure vs insecure attachment style (Table 4), the hypothesis that women with secure attachment would have low scores in IPO and those with insecure attachment would have higher rates in IPO, could not be proved. What we found in the literature regarding this, was that all studies found an inverse relationship between scores on borderline dimensions and secure attachment [15]. Although the association in our study was not statistically significant, there was a tendency in that direction (the securely attached group had an average of 7 points lower on the IPO scales versus the group with no secure attachment), therefore, we believe that the small sample size does not permit the association to be statistically significant.

Higher OQ scores were also observed in patients with IPO high scores, this being related to what has already been shown in the literature about the ability of both tests to discriminate pathology. The IPO, personality pathology and OQ, mental health in general. [13,22]. Moreover, confirm the fact that people with personality pathology (the clinical group) had psychological distress and problems in interpersonal and role functioning.

In relation to the instrument used to assess attachment in this study, CAMIR, was used due to its easy applicability and as it had validation studies in Chilean population with adequate reliability and validity. However, we could not find in the literature any study with this instrument in which personality pathology is also measured. Although we believe that the present study is an initial approach to what is going on since attachment style conforms toward the development of personality pathology, more research is needed on mechanisms, or mediators, that help to explain how insecurity, or a particular form of insecure attachment, leads in some cases to psychopathology. Also it is important to consider, that insecurity is associated with other disorders beyond personality pathology, such as posttraumatic stress disorder and mood disorders [16,17]. What is interesting is Levy’s proposal about the idea that the two primary types of insecure attachment- preoccupied and avoidant- can occur at several developmental levels bringing about an array of possibilities in a continuum from people without pathology, to those with PTSD or depression, to those with healthier forms of personality pathology, and finally at the lowest developmental level, to those with BPD or antisocial PD. [15]. To him, differences in the content and structure of mental representations-or internal working models- distinguishes between greater and lesser adaptive forms of avoidant and preoccupied attachment, thereby bringing a fuller developmental perspective to the study of attachment patterns. One should consider several mediators for these various forms of adaptation at different levels of analysis: neurological, hormonal, cognitive, behavioral, and social-interactional. Mediators may include difficulties in emotion regulation and deficits in social skills, for example [27].

Also, we need to know whether attachment status per se is the issue or whether, for example, poor parenting, or the recent concept of epistemic trust/distrust intervene more in the development of psychopathology. All these questions need further research [29].

To Luzmenia Garrido who gave to me the CAMIR with its application instruccions and analysis. To the Psychiatry Department of Del Salvador Hospital and his Chief Director, Dr Ramon Florenzano, who gave to us the Chilean validation of IPO