A new approach to the assessment of structural personality pathology: Theory-driven profile interpretation of the Dutch Short Form of the MMPI

https://doi.org/10.1016/j.newideapsych.2007.03.002Get rights and content

Abstract

This article presents a new approach to the assessment of structural personality pathology: theory-driven profile interpretation of the Dutch Short Form of the MMPI (DSFM). The assessment method integrates results of a self-report personality questionnaire with psychodynamic theoretical concepts to yield a working hypothesis about underlying structural features of personality. As an illustration, results of a validation study with psychotic phenomena in projective drawings as the external criterion are described. Results show that the DSFM profiles, that are presumed to measure psychotic vulnerability predict psychotic phenomena in projective drawings, after statistically controlling for the effect of age, the single scales that are being used to construct the profiles and their statistical interaction.

Results suggest that theory-driven profile interpretation of the DSFM may be considered a time and cost efficient method to derive a working hypothesis about structural vulnerability.

Introduction

Whether personality pathology should be diagnosed in terms of categories or in terms of dimensions has, for many years, been a focal issue within the field of the assessment of personality disorders (Widiger & Frances, 1985; Widiger & Simonsen, 2005a). A major problem with categorical systems such as DSM-IV (APA, 1994) is that personality disorders represent a heterogeneous set of symptoms and that all that is required for diagnosis is the presence of a specific number of symptoms from a defined set. The result is that the symptom profile of patients diagnosed with the same disorder may differ considerably. In addition, there appears to be high comorbidity among the personality disorder diagnoses and, finally, the distinction between normal and abnormal personality is rather arbitrary (Morey et al., 2002; Widiger & Simonsen, 2005b).

There is growing recognition that a dimensional approach to assessing personality pathology offers notable advantages (Verheul, 2005; Widiger & Simonsen, 2005b). Using a dimensional approach, patients are rated on various trait dimensions, creating a specific trait profile (Clark, McEwen, Collard, & Hickok, 1993; Widiger & Simonsen, 2005b). In a dimensional approach, there is a fluid transition between “normal” and “abnormal”, because patients are rated on a continuum.

The usefulness of combining traits in personality assessment into profiles is currently being acknowledged in the literature. As suggested by Morey et al. (2002), the interaction between dimensions is considered more important for assessing personality pathology, rather than the single dimensions. Miller (2003) also points to the importance of combining dimensions in order to understand the association between personality and psychopathology. In his review of research on the influence of personality on the development, course and behavioural expression of PTSD, he shows that a combination of high negative emotionality combined with low constraint predicts externalizing forms of post-traumatic stress reactions, whereas a combination of high negative emotionality with low positive emotionality (which relates to high constraint) predicts internalizing types of reactions.

Finally, within the assessment of personality pathology, descriptive (DSM-IV) and structural-dynamic diagnosis need to be differentiated. As early as 1970, Kernberg proposed a psychoanalytic classification of character pathology and has, since then, extensively described his structural-dynamic vision on personality pathology (Kernberg (1975), Kernberg (1976), Kernberg (1984), Kernberg (1994), Kernberg (1996); Kernberg and Caligor, 2005). The core concept in Kernberg's model is that of “personality organization”, a relative stable structure that consists of various inner representations of early relationships of the self with significant others (in terms of psychodynamic theory: “objects”), including the affective quality of these relationships. During early developmental stages, various affect laden “units” of self and object representations constitute the basic building blocks of personality structure. During these early stages of development, self and object representations are not yet differentiated: boundaries between self and others are absent and therefore it is impossible to distinguish between the inner and outer world. Later in development, self and object representations are differentiated and boundaries between self and others are established. The next developmental task is to integrate positive and negative affects, resulting in a synthesis of contradictory self and object representations and consequently in an integrated concept of self and others. In general, the better these developmental tasks have been accomplished, the stronger the personality structure (Kernberg, 1994).

On the basis of this developmental model, Kernberg describes three levels of personality organization: the neurotic, the borderline, and the psychotic. Within the borderline domain Kernberg (1975) distinguishes several different types of patients. The three personality organizations are distinguished on the basis of three dimensions: identity integration versus identity diffusion, defence mechanisms (predominantly mature/neurotic versus predominantly primitive), and reality testing (absent, frail or good). In patients with a neurotic organization, self and object are clearly differentiated and positive and negative affects related to self and others are integrated. So the patient has a well-integrated identity and is able to tolerate ambivalence. The patient mainly uses neurotic defences, centring around repression.

In patients with a borderline organization the differentiation of self from object has occurred to a sufficient degree to allow for the establishment of some ego-boundaries (Kernberg, 1994). Therefore, they have good, although vulnerable, reality testing. However, the developmental task of integrating positive and negative affects about self and others has failed and therefore these patients suffer from identity diffusion and mainly use primitive defences such as splitting. Kernberg (1994) considers a pathological predominance of aggression in these patients as an important explanation of the failure of these patients to integrate positive and negative affects.

Patients with a psychotic organization suffer from the most severe character pathology because both the developmental task of differentiating self from others as well as the task of integrating positive and negative affects have failed. As a result, ego boundaries are absent, reality testing is lost and they suffer from identity diffusion.

According to Kernberg, most of the DSM IV axis II disorders are organized at the borderline level (Kernberg & Caligor, 2005).

A Dutch Short Form of the MMPI (DSFM) has been developed in the Netherlands (Luteijn & Kok, 1985).The reason for constructing the DSFM was to create a test that would contain the core components of the MMPI, relatively free of what is now described as the common psychopathology factor “demoralization” (Sellbom, Ben-Porath, & Graham, 2006).

The DSFM measures five dimensions: Negativism, Somatization, Shyness, Psychopathology and Extraversion. The results of a study among psychiatric patients by Achterveld and Snellen (1999) show that four of the five dimensions display conceptual overlap with the PSY-5 (Harkness, McNulty, & Ben-Porath, 1995), a dimensional model of abnormal personality. The DSFM subscale Negativism is positively correlated with the PSY-5 subscale Negative Emotionality (0.42; p<0.01), the subscale Shyness shows a negative correlation with Aggressiveness (−0.52; p<0.01) and Disconstraint (−0.40; p<0.01) and a positive correlation with low positive emotionality (0.39; p<0.01). The DSFM subscale Psychopathology appears to be rather strongly associated with the PSY-5 subscale Psychoticism (0.62; p<0.01). Finally, the DSFM subscale Extraversion is positively associated with Disconstraint (0.39; p<0.01) and Aggressiveness (0.36; p<0.01) and negatively associated with Low Positive Emotionality (−0.68; p<0.01). Eurelings-Bontekoe and Snellen (2003) have attempted to integrate the dimensional assessment of personality pathology using the DSFM with the structural-dynamic theory of personality pathology as described by Kernberg. They developed a theory-driven profile interpretation of the DSFM. The profile interpretation implies that the dimensions of the DSFM are not interpreted in isolation, but rather that the scores on the dimensions are combined a priori in a theory-driven way into specific profiles, that are subsequently interpreted by using psychodynamic concepts. As noted, combining dimensions into profiles is not new, and resembles the traditional strategy used in the interpretation of the MMPI where combinations of scales are classified into code types (e.g., Friedman, Lewak, Nichols, & Webb, 2001). However, the use of psychodynamic theoretical concepts to construct profiles a priori and to interpret the meaning of these a priori defined trait-profiles does constitute a new assessment approach. In fact, psychological constructs are operationalized as a set of scores combined on the basis of theory (Petot & Djuric Jocic, 2005; Ganellen, 1996). The idea is that the gap between dimensional self-reported (overt) features of personality pathology and structural (covert) indices of personality can only be bridged by applying theory in the interpretation process. For example: individual MMPI scales 2 (depression), 4 (psychopathic deviation), and 7 (psychasthenia) do not express the dependence of subjects, but when combined in the 2–4–7 profile, they may indicate strong, unsatisfied needs for attention and support, and conflicts around dependency (Petot & Djuric Jocic, 2005). Such a diagnostic approach is more complex, but similar to the process of clinical data integration (Petot & Djuric Jocic, 2005). The main difference between the DSFM profile interpretation and the approach of Friedman and of most other MMPI textbook authors, is the great emphasis on descriptive diagnosis in traditional MMPI profile interpretation versus the emphasis on psychodynamic theory to interpret profiles in the DSFM profile approach.

It is hypothesized that specific profiles of DSFM dimensions refer to underlying, structural personality organizations, which cannot be assessed either by studying personality dimensions in isolation, or by merely describing their combinations, but by defining and interpreting several combinations in a theory-driven way. This method of interpretation uses Kernberg's tripartite model of personality pathology (Kernberg (1970), Kernberg (1984); Kernberg & Caligor, 2005) as the theoretical frame of reference, using concepts such as integrative capacity, impulse control, and anxiety tolerance.

The primary aim of this paper is to present the theory-driven profile interpretation, illustrated with an empirical study on the validity of the model, addressing the issue of whether the theory-driven profile interpretation indeed measures structural features of personality pathology.

The theory-driven profile interpretation was developed on the basis of the integration of almost 20 years’ clinical experience and systematic clinical observations of both inpatients and outpatients in both mental health care and forensic settings. Hypotheses about structural diagnoses are derived from specific combinations of raw scores on DSFM subscales Negativism (Negative affectivity), Psychopathology (Psychoticism) and Shyness (Constraint). Cut-offs for the different subscales are based on an iterative process of refinement over the years, and are currently the subject of empirical investigation.

In addition, it should be noted that the DSFM subscales Somatization and Extraversion are not included in the model. Although these two scales are always used in the profile interpretation of the individual patient to refine the clinical picture, they are not used for structural diagnosis: the DSFM subscale Somatization is considered to be an important general affect regulator, independent of structural pathology. The subscale Extraversion is considered to be a (normal) temperamental trait and not a marker of structural personality pathology. This is in agreement with Kernberg and Caligor (2005) who include in their psychoanalytic model of nosology the dimension Introversion/Extraversion as a temperamental disposition influencing the type of personality disorder (internalizing or externalizing) but not the severity of structural personality pathology. The DSFM Negativism subscale is hypothesized to measure the level of negative emotionality (inner tension and subjectively experienced anger). The DSFM subscale Shyness (in PSY-5 terms: low Aggressiveness, low Disconstraint and Low Positive Emotionality) is assumed to reflect an individual's inhibitory capacity (impulse control). A high level of shyness renders the individual overly adapted, controlled, prudent, inhibited and constrained; a low level of shyness may give rise to impulsivity and acting out. The DSFM Psychopathology subscale (in PSY-5 terms: Psychoticism) is assumed to measure severe psychopathology such as paranoid anxiety, perceptual aberrations and magical ideation. This scale is assumed to reflect anxiety tolerance or the propensity to paranoid and dissociative reactions during periods of high emotional stress. In times of relative emotional quiescence, this propensity need not be manifested.

The hypothesis is that, on the basis of combinations of these three dimensions, patients with a neurotic organization can be distinguished from the borderline and the psychotic organized patients, and that several subtypes can be distinguished within the borderline organization. The interpretation scheme is presented in Table 1.

Patients with a neurotic personality organization (NPO) have, according to Kernberg (1976), in general a moderate to good anxiety tolerance. However, these patients are characterized by a severe and punitive superego. They are over controlled and rigidly adhere to social norms and values and are overly constrained, rigid and inhibited. Aggression and other conflictive impulses are repressed. They tend to accept and comply rather than protest. Based on this theoretical notion, they are thought to combine above average to very high scores on DSFM Shyness with low to average scores on Psychopathology. Associated with their tendency to repress anger, scores on DSFM Negativism are expected to be low. However, high scores are equally possible in the case of a high level of experienced tension and negative affectivity, for instance when there is state-pathology such as depression. Therefore, Negativism is not included in this structural diagnosis.

Note in Table 1 that the higher the score on Psychopathology, the higher the score on Shyness needs to be to still be able to speak of sufficient control. However, what both neurotic profiles have in common is that they combine a relatively low level of Severe Psychopathology with overcontrol (a relatively high level of Shyness).

A characteristic of patients with a psychotic personality organization (PPO) is their fusion experiences due to absent boundaries between self and object representation. They long for symbiosis and for a mystical experience of oneness, goodness, and love. The reverse of this situation is the experience of engulfment into an aggressive relationship, without being able to differentiate between engulfer and engulfed (Kernberg, 1975). Both symbiotic fusion as well as aggression is avoided as long as possible, mainly by keeping distance from others and by efforts to control aggression as long as possible, which clinically manifests itself in obsessive compulsiveness, problems with concentrating, ruminating, and indecisiveness. Therefore, they often strongly resemble neurotic organized patients on the descriptive level.

On the basis of this, patients with a psychotic personality organization are thought to combine a moderate to poor anxiety tolerance with some but insufficient control. Based on clinical experience, we have distinguished between two subtypes: one latent type with a moderate anxiety tolerance and one manifest type with a low anxiety tolerance.

Therefore, patients with a presumed psychotic personality organization are expected to be characterized by a combination of average scores on DSFM Psychopathology and average scores on Shyness or by a combination of high scores on Psychopathology and average to above average scores on Shyness. Also here, the higher the Psychopathology score, the higher the Shyness score, but what both psychotic profiles have in common is that control is weak, relative to the anxiety tolerance: patients are neither undercontrolled (acting out and a lack of social inhibition is presumed to be less likely among these structurally very vulnerable patients) nor overcontrolled. Therefore, Shyness scores are neither below average, nor are the Shyness scores high or very high.

Negativism is average or lower, based upon the theoretical notion that patients with a psychotic organization are afraid of their aggression and try to control their aggression as long as possible (Kernberg (1976), Kernberg (1996)). In accordance with this theoretical notion, Diguer et al. (2001) found patients with psychotic personality organization produce the least negative responses using The Core Conflictual Relationship Theme (CCRT) method.

Patients with a borderline organization comprise a heterogeneous and thus a broad group (Grinker, Werble, & Drye, 1968; Kernberg, 1975). On the basis of the integration of systematic clinical observations and Kernberg's and Grinker's descriptions of the various subtypes of borderline organized patients, we have distinguished between several borderline profiles.

  • 1.

    The high level (overcontrolled) borderline organized group (neurotic borders according to Grinker et al., 1968) suffers from low anxiety tolerance and is at the same time overly inhibited. We therefore presume that patients with a high level borderline personality organization combine high to very high scores on Psychopathology because of their low anxiety tolerance, with high to very high scores on Shyness, reflecting their strong inhibition. With respect to Negativism, the same consideration applies as described for the neurotic organized patients: associated with their tendency to overcontrol negative affects, scores on DSFM Negativism are expected to be low. However, high scores are equally possible in the case of a high level of experienced tension and negative affectivity, for instance when there is state-pathology such as depression. Therefore, Negativism is not included in this structural diagnosis.

  • 2.

    The low level (undercontrolled) borderline organized patients, referred to by Grinker et al. as core borders, have a moderate to low anxiety tolerance and are in general impulsive and chaotic. They are therefore presumed to combine average to high scores on Psychopathology, reflecting their moderate to poor anxiety tolerance, with below average scores on Shyness, reflecting their lack of inhibition. Scores on Negativism are expected to be generally low, associated with strong externalizing tendencies, but can be high in the case of external stress. Therefore, again, Negativism is not included in this structural diagnosis.

  • 3.

    It is well known that patients with narcissistic pathology present themselves on self-report measures as healthy, stable and sociable (Miller, Pilkonis, & Clifton, 2005). This may reflect the inability of narcissistic patients to view themselves realistically and their tendency to deny vulnerabilities on self-reports (Hilsenroth, Handler, & Blais, 1996). In fact they present “illusionary mental health” (Shedler, Mayman, & Manis, 1993). Grinker et al. (1968) refer to these patients as the “as-if” group. The subgroup of borderline organized patients with narcissistic personality pathology is therefore presumed to score both low on Psychopathology as well as below average and lower on Shyness. Considerations with respect to Negativism are the same as described for the low level borderline organized patients: scores on Negativism are expected to be generally low, associated with strong externalizing tendencies, but can be high in the case of external stress.

  • 4.

    The subgroup of borderline organized patients with a heightened psychotic vulnerability, the so-called “psychotic borders” (Grinker et al., 1968) are presumed to score high on Psychopathology; their anxiety tolerance is hypothesized to be insufficient. In addition, these patients are, just like the psychotic organized patients, presumed to exert some but insufficient control over their impulses, hence acting out is less likely. Therefore their scores on Shyness are presumed to be average or above average, hence neither below average nor (very) high. The level of Negativism subsequently distinguishes the group of the psychotic borderline organized patients from the patients with the profile belonging to the manifest psychotic organization: negativism is high among the first group and relatively low among the latter. This profile is based on the theoretical notion that borderline organized patients with a psychotic vulnerability, suffer from a high level of negative affectivity but that they will try to maintain control over the acting out of aggression as long as possible (Kernberg, 1984).

  • 5.

    In addition to these profiles we have defined two profiles pertaining to the so-called “infantile personality” as described by Kernberg (1975), Kernberg (1976), Kernberg (1984), Kernberg (1996). According to Kernberg, these types of patients may show a combination of both psychotic, narcissistic/borderline and neurotic features (Kernberg (1975), Kernberg (1976), Kernberg (1984), Kernberg (1996); Zetzel, 1968) or in other words signs of personality disorders from both the A, the B and the C clusters of DSM-IV (APA, 1994). In concordance with Kernberg's idea and on the basis of clinical observation, two types are distinguished: a type with a good anxiety tolerance and one with a moderate anxiety tolerance. Common to both types is a certain capacity to control, which is, however, too weak to qualify them as neurotic and at the same time too strong to qualify them as either narcissistic (in the case of the first type with good anxiety tolerance), or as psychotic (second type with moderate anxiety tolerance). Therefore the first group is situated between the narcissistic and the neurotic profiles, the second group is situated between the psychotic and the neurotic profiles. Again, a high level of Negativism is not considered to be a structural feature.

In summary, the profiles cluster into three subgroups: a group of patients with a presumed neurotic organization, a group with a presumed borderline organization, and a group with a presumed psychotic organization.

It may be useful to illustrate our method with an example. When a patient scores low on Psychopathology and above average or higher on Shyness, the structural hypothesis is that this patient has a neurotic personality organization. Subsequently the patient is described in terms of this personality organization: it is hypothesized that this patient has good reality testing, a well-integrated identity and sense of self, that he is able to engage in complex and differentiated personal interactions and that he uses mainly mature defences. Anxieties centre around guilt and loss of love from others (Gabbard, 2005). Hence Kernberg's descriptions of patients with an NPO are used to describe the patient. None of these descriptions can be derived from the meaning of the single scales, nor from the mere description of the combination of Psychopathology and Shyness. On the basis of theory, a specific combination of these two variables has been defined that is subsequently interpreted using theory.

The study described here uses signs of psychotic phenomena in projective drawings. Projective drawings are likely to reveal psychotic vulnerability, because the test is highly unstructured. It is hypothesized that patients with a Psychotic Personality Organization profile will show the highest mean score. In addition, since the claim of the theory-driven profile interpretation is that it measures structural personality features that cannot be assessed by merely describing the single dimensions or their statistical interaction, the hypothesis is that the theory-based a priori defined profiles have incremental predictive power beyond and above the single DSFM scales that are used to construct the profiles and their statistical interaction.

Section snippets

Participants

The sample consisted of 100 outpatients (45 women and 55 men). Their age ranged from 18 to 74, with a mean age of 37 (SD 13.34). Educational level was low (a maximum of ten years of education) in 35, middle (approximately 12 years of education) in 55 and high (a minimum of 18 years of education) in 10 patients.

Procedure

The sample was drawn from a large group of outpatients that had undergone a standard psychological assessment procedure. Part of this standard procedure is that all patients are informed

Interrater reliability

The interrater reliability of ratings of psychotic phenomena was evaluated via a one-way random effects model ICC (1) (Shrout & Fleiss, 1979). This ICC model was chosen as it is considered to be the most conservative and generalizable calculation of ICC (see Eudell-Simmons, Stein, DeFife, & Hilsenroth, 2005). ICC (1) was 0.68, which is good (Cicchetti, 1994). As both raters scored all drawings, mean scores across the two raters were calculated and used in the analysis. Therefore, the

Psychotic drawings and personality organization

This paper describes a new approach to the structural assessment of personality pathology, integrating results of self-reported DSFM personality dimensions with psychodynamic theory. In this “theory-driven profile interpretation of the DSFM” scales are combined in a theory-driven way and specific combination of scales (profiles) are interpreted using psychodynamic theory, more in particular Kernberg's tripartite model of structural personality pathology. The idea is that scores on single

References (50)

  • E.H.M. Eurelings-Bontekoe et al.

    DSM-III-R and ICD-10 personality disorders and personality dimensions as assessed by the Dutch Short Form of the MMPI: Preliminary results

    Personality and Individual Differences

    (1995)
  • W. Achterveld et al.

    Vergelijking van de PSY-5 en de NVM [Comparison of PSY-5 and Dutch Short Form of the MMPI]

    Diagnostiekwijzer

    (1999)
  • DSM-IV. Diagnostic and statistical manual of mental disorders

    (1994)
  • D.V. Cicchetti

    Guidelines, criteria, and rules of thumb for evaluating normed and standardized assessment instruments in psychology

    Psychological Assessment

    (1994)
  • L.A. Clark et al.

    Symptoms and traits of personality disorder: Two new methods for their assessment

    Psychological Assessment

    (1993)
  • J. Cohen

    Statistical power analysis for the behavioral sciences

    (1977)
  • L. Diguer et al.

    The core conflictual relationship theme of psychotic, borderline, and neurotic personality organizations

    Psychotherapy Research

    (2001)
  • L. Diguer et al.

    Personality Organization Diagnostic Form (PODF). Manual for scoring, 2006 edition

    (2006)
  • E.M. Eudell-Simmons et al.

    Reliability and validity of the Social Cognitions and Object Relations Scale (SCORS) in the assessment of dream narratives

    Journal of Personality Assessment

    (2005)
  • E.H.M. Eurelings-Bontekoe et al.

    Dynamische Persoonlijkheidsdiagnostiek met behulp van projectieve technieken: Thematische Apperceptie Test (TAT), Tekeningen en Zinnen Aanvul Test (ZAT) [Dynamic personality assessment using projective techniques: Thematic Apperception Test (TAT), projective drawings and Sentence Completion Test]

  • E.H.M. Eurelings-Bontekoe et al.

    Dynamische persoonlijkheidsdiagnostiek. [Dynamic personality assessment]

    (2003)
  • E.H.M. Eurelings-Bontekoe et al.

    Een aanzet tot constructvalidering van de profielinterpretatie van de Nederlandse Verkorte MMPI (NVM). Een onderzoek met behulp van de TAT [A first attempt to validate the theory driven profile interpretation of the DSFM using TAT-narratives]

    Tijdschrift voor Psychotherapie

    (2005)
  • Exner, J. E. (1991). The Rorschach: A comprehensive system, Vol. 2: Interpretation (2nd ed.). New York:...
  • Exner, J. E. (1993). The Rorschach: A comprehensive system, Vol. 1: Basic foundations (3rd ed.) New York:...
  • J.E. Exner

    A primer for Rorschach interpretation

    (2000)
  • Exner, J. E. (2001). A Rorschach workbook for the comprehensive system (5th ed.). Asheville, NC: Rorschach...
  • A.F. Friedman et al.

    Psychological assessment with the MMPI-2

    (2001)
  • G.O. Gabbard

    Psychodynamic psychiatry in clinical practice

    (2005)
  • R.J. Ganellen

    Integrating the Rorschach and MMPI-2 in personality assessment

    (1996)
  • R.R. Grinker et al.

    The borderline syndrome: A behavioral study of ego functions

    (1968)
  • G. Groth-Marnatt

    Handbook of psychological assessment

    (1990)
  • A.R. Harkness et al.

    The Personality Psychopathology-Five (PSY-5): Constructs and MMPI-2 scales

    Psychological Assessment

    (1995)
  • M.J. Hilsenroth et al.

    Assessment of narcissistic personality disorder: A multi-method view

    Clinical Psychology Review

    (1996)
  • S. Hoermann et al.

    The construct of effortful control. An approach to Borderline Personality Disorder Heterogeneity

    Psychopathology

    (2005)
  • T. Ingenhoven et al.

    Nederlandse vertaling van de Inventory of Personality Organization. [Dutch translation of the Inventory of Personality Organization]

    (2004)
  • Cited by (15)

    View all citing articles on Scopus
    View full text