Psychodynamic Approaches to Personality Disorders

Clinical context

The personality disorders have been somewhat marginalized by their relegation to a separate axis in the DSM system. Nevertheless, these disorders are ubiquitous in clinical practice, and they complicate the treatment of axis I disorders. Indeed, some clinicians attempt to focus their efforts on the obviously diagnosable axis I disorders only to find that their treatments are less than effective because they are not paying sufficient attention to the patient’s personality problems. Often clinicians obtain their first clue to the existence of a personality disorder when they experience a surprisingly entrenched emotional reaction to the patient. Many such reactions are negative, in keeping with the oft-repeated notion that personality disorders are more likely to create distress in others than in the patient. However, excessively idealized and protective reactions are also a potential pitfall in the treatment of personality disorders.

The psychodynamically informed psychiatrist is ideally suited to bringing a thoughtful diagnostic understanding and a carefully planned treatment approach to axis II conditions, because the interpersonal developments in the treatment relationship are at the heart of psychodynamic thinking (1). Other treatment strategies, such as dialectical behavior therapy (2), cognitive therapy (3), and pharmacotherapy (4), are also useful in treating personality disorders. In this article, however, I focus on a systematic psychodynamic approach that may be useful in both diagnostic understanding and treatment planning.

From a psychodynamic perspective, the achievement of a stable and positive sense of self and the establishment of mutually gratifying and enduring relationships are perhaps the two fundamental tasks of personality development (5). Indeed, they are probably the areas of major difficulty in most personality disorders that make them stand out in a clinical context. These two features are inextricably intertwined throughout the life cycle and lead to problems in multiple settings. In addition, there is now abundant evidence that some personality traits are heritable (6, 7). A set of defense mechanisms that are characteristic of particular personality disorders has also been traditionally associated with the psychodynamic perspective (1), and it must be taken into account as well. Vaillant and Vaillant (8) point out that defenses do not simply change the relationship between an emotional state and an idea, they also influence the relationship between self and object. Hence defenses are embedded in relatedness.

The psychodynamic clinician would thus view personality disorders as involving four major components: a biologically based temperament, a set of internalized object relations, an enduring sense of self, and a specific constellation of defense mechanisms. Some would also include the characteristic cognitive style that is intimately connected to the use of defense mechanisms (9).

Working from a psychodynamic perspective, a psychiatrist assumes that a series of internal object-relations units will be externalized in the process of diagnosis and treatment, and these will provide clues to the core features of the patient’s personality. Representations of self in relation to others connected by affect states become etched in neural networks in the course of development and become repetitive patterns of relatedness (10). If a little girl is repeatedly criticized and humiliated by her mother for reasons that are unclear, she is likely to develop an internal object-relations constellation involving an inadequate self, a humiliating, critical object, and an affect state of shame. If she comes into contact with a female clinician, she will be likely to recreate that pattern unconsciously in the work with that clinician.

The core idea from psychodynamic and psychoanalytic theory is that patients reenact their internal object relationships through the externalization of these childhood patterns. This mode of actualizing old relationships in the present is often referred to as projective identification (11, 12). In this model, patients behave in a characterologically driven way that exerts interpersonal pressure on the clinician to conform to what is being projected into him or her. In other words, patients with a personality disorder may behave in such an irritating fashion that the clinician becomes angry at the patient and thus contributes to the recreation of an experience from childhood involving an angry, critical parent and a child who is misbehaving. Even when the object relationship is a negative one, the patient may still recreate the unpleasant atmosphere because of its familiarity and predictability. Abused children often turn to the abusive parent to look for safety and protection because they have no alternative. Hence the pattern is established of seeking safety in the shadow of one who has abused them. This paradoxical situation may nevertheless be soothing and familiar to them and is associated with a sense of attachment and connection.

A psychodynamic approach to diagnosis is heavily influenced by what develops in the transference and countertransference embedded in the therapeutic relationship. Recent research suggests that countertransference phenomena can be measured in clinically sophisticated and psychometrically sound ways that reflect complex reactions in clinicians. Betan et al. (13), studying a national random sample of 181 psychiatrists and clinical psychologists, found that a factor analysis of a countertransference questionnaire yielded eight clinically perceptually coherent factors independent of the theoretical orientation that the clinicians espoused. The reaction patterns occurred in coherent ways with specific personality types, suggesting that these reactions are useful in diagnostic understanding. Narcissistic personality disorders, for example, elicited feelings of anger, resentment, and dread as well as feelings of being devalued and criticized. These feelings correlate well with clinical experience.

Another line of recent research has allowed us to link neurobiological factors with the transference-countertransference constellation seen in personality disorders. A strong linkage exists between childhood trauma and borderline personality disorder (14). Traumatized borderline patients have been shown to have a hyperreactive hypothalamic-pituitary-adrenal (HPA) axis, leading to significantly enhanced ACTH and cortisol response compared with nonabused subjects (15).

Supplementing the findings on the hyperreactive HPA axis in traumatized borderline patients are other studies suggesting that these patients have reduced amygdalar volumes as well (16–17, 18). Reduced volume is also associated with hyperactivity in the amygdala compared with normal control subjects when exposed to standard presentations of faces (19). Of even greater importance, though, is the tendency for patients with borderline personality disorder to attribute negative attributes to neutral faces. Faces without expression were regarded as untrustworthy, threatening, and possibly malevolent. Hence a hyperactive amygdala may work synergistically with a hyperreactive HPA axis to create a familiar pattern of relatedness in borderline patients. In other words, the patient may be recreating a setting in which the clinician is perceived as a threatening or persecuting object and the patient as victimized self, and an affect of hypervigilant anxiety connects the two (20).

In addition to monitoring the characteristic object-relational themes, the psychodynamic clinician also looks for specific constellations of defenses that work in concert with relatedness patterns. For example, someone with obsessive-compulsive personality disorder would use defensive operations such as intellectualization, isolation of affect, and reaction formation (1). These defenses would be used in the setting of a clinical evaluation to tone down affect states and attempt to control the situation that potentially produces anxiety. The patient might, for example, talk about an emotional situation such as the death of a parent entirely in terms of data and facts rather than express any feelings about the event.

Treatment strategies and evidence

A growing body of research confirms that psychodynamic/psychoanalytic psychotherapy is efficacious in the treatment of personality disorders (21). Some axis II conditions have been more extensively studied than others. Randomized controlled trials of psychotherapy in the treatment of cluster C personality disorders (3) and borderline personality disorder (22, 23) demonstrate that psychodynamic therapy produces durable changes, but it may take anywhere from 10 to 18 months to work. Moreover, studies of these conditions show that patients continue to improve at follow-up points after the therapy is terminated, which suggests that patients internalize the therapeutic dialogue and continue to use what they’ve learned in therapy for further growth. In a longitudinal follow-along study (24), 40 patients with personality disorders had their defense mechanisms measured at the beginning of treatment and at regular intervals over the course of 3–5 years of long-term psychodynamic psychotherapy. Those with high initial scores on maladaptive and self-sacrificing defensive styles improved in terms of their defensive functioning, and this improvement was correlated with symptomatic improvement as well. In fact, changes in defensive style predicted symptomatic change.

An effective treatment follows directly from the diagnostic understanding of what is going on within the patient’s internal world. Patients with personality disorders tend to draw the psychotherapist into a “dance.” Through the process of projective identification, they unconsciously transform the therapist into a transference object (9). Therapists, in turn, may unconsciously enact old patterns with the patient before they realize the nature of the “dance.” For example, a patient with borderline personality disorder who experienced frequent childhood abandonment may behave in such a way that the therapist starts to dread seeing the patient. As a result, the therapist may consistently start sessions late or even occasionally forget a session. Only through observing that behavior would the therapist finally recognize that he or she has become the abandoning object to the patient’s victimized self.

Therapists must maintain a kind of free-floating responsiveness (25) to what is being evoked by the patient to recognize the recreated “dance.” One of the key components of technique, then, is for therapists to note these relatedness patterns and point them out to the patient as they occur in the therapeutic relationship. The next step is for the therapist to see how they are related to childhood wishes, fantasies, and traumatic experiences and how they create difficulties in adult life. Often the therapist’s observations of the patient’s behavior come as a surprise to the patient because the behavior is part of procedural memory and is totally outside the patient’s conscious awareness. Hence therapists must use tact in gradually introducing these observations to the patient, in a way that can be digested by the patient.

Psychodynamic therapists also recognize that the patient’s characteristic defense mechanisms become resistances in the treatment. Hence the patient who intellectualizes as a characteristic defense will begin intellectualizing in therapy as a resistance to facing emotionally painful issues. In short, defenses are intrapsychic while resistances are interpersonal (1). Psychodynamic therapists do not try to eradicate resistances. Rather, they expect resistances and use them to understand the patient. In the clinical setting, they are inextricably connected with patterns of object relations such that many of the ways the patient defends against emotions emerge as transference resistances, meaning a way of relating to the therapist that works against understanding. For example, the narcissistic patient may treat the therapist with contempt. The patient with histrionic personality disorder may attempt to seduce the therapist rather than benefit from understanding or insight. An avoidant patient may avoid the session altogether. These patterns of resistance are brought out in the open by the therapist in a gradual way and tied to the underlying fears they reflect through interpretation.

In cases of severe borderline personality disorder, it is common to combine the use of medication with psychodynamic psychotherapy, as suggested by the American Psychiatric Association practice guideline on the disorder (26). While a variety of psychopharmacologic agents appear to be useful for patients with borderline personality disorder, the selective serotonin reuptake inhibitors (SSRIs) have garnered supportive data in four double-blind placebo-controlled trials (15, 27–29). SSRIs may facilitate therapy by reducing the intense anger, dysphoria, and hypervigilant anxiety connected with the patient’s internal object world (20). Rinne et al. (30) noted that fluvoxamine is associated with a significant reduction of ACTH and cortisol response in borderline patients, particularly in those who sustained childhood abuse. It may well be that toning down the hyperreactivity of the HPA axis with an SSRI moderates the patient’s specific object-relations unit of victimized, hypervigilant self linked to a potentially malevolent object. The diminution of the hypervigilant anxiety promotes greater reflective capacity, allowing the patient to consider what is going on in the therapeutic relationship more effectively (20).

Questions and controversies

To a large extent, research on personality disorders is still in its infancy. Part of the marginalization of these disorders is related to the fact that funding for research is still catching up with the long-standing tradition of supporting investigations of axis I disorders and their treatments. Hence we have much to learn.

One of the ongoing controversies has to do with how we can incorporate the methods used by clinicians into the standard diagnostic frameworks of the DSM system and empirical investigations (31). Asking a series of questions from standardized assessment instruments often promotes defensiveness in patients with personality disorders, and the information generated lacks the clinical usefulness of immediate interactions within the therapeutic setting. The translation of transference and countertransference data into systematic diagnostic criteria constitutes a formidable challenge for the field.

A second controversy involves the question of how much therapy is enough. While we generally regard personality disorders as relatively refractory to brief treatments, we cannot be sure what duration of therapy is really needed. There are two confounding factors in attempts to answer this question: 1) the extensive comorbidity of axis I syndromes with personality disorders, which makes the course of different patients highly variable; and 2) the evidence that some personality disorders remit without extensive treatment when followed longitudinally (32, 33). More recent research (34) suggests that more traitlike and attitudinal criteria are relatively fixed, while more behavioral features and reactive symptoms tend to be intermittent. More research is needed to identify those features that are likely to be durable and to determine how to tailor the treatment accordingly.

Yet another controversy lies in deciding which disorders are treatable with dynamic psychotherapy, with or without the use of medication, and which are not. Randomized controlled trials are badly needed for conditions such as antisocial personality disorder (traditionally thought to be untreatable), histrionic personality disorder, narcissistic personality disorder, and the relatively uncommon (in clinical practice) cluster A disorders.

Summary and recommendations

One of the implications of this article is that psychodynamic clinicians must consider their own reactions as part of the diagnostic process. Systematic application of countertransference understanding is one way that psychiatrists distinguish themselves from other medical specialists. In both diagnosis and treatment, we can benefit from viewing ourselves as part of a two-person “dance” that the patient recreates in multiple settings and that leads to much of the patient’s distress, not to mention the distress of those who live and work with the patient.

There is a natural tendency to want to rid oneself of distressing countertransference feelings, whether they be erotic, contemptuous, sadistic, or disorganizing. However, clinicians may benefit from postponement of this impulse to slough off unpleasant feelings. Often by steeping oneself in the transference-countertransference experience, the clinician gains a great deal of information about the internal world of the patient that cannot be gleaned from other sources. Hence, in dynamic psychotherapy of patients with personality disorders, it is useful to immerse oneself in the experience for a period of weeks or months rather than trying to make the experience go away through interpretation.

Making observations about these patients from an outside perspective is often a jolt to the patient’s self-esteem. The experience might be compared with seeing oneself on a videotape. One’s first reaction is “I can’t possibly look like that or sound that way.” Because the patient’s recreated internal object relationship is largely unconscious, it may appear as something completely alien when pointed out to the patient. Hence one needs to develop a solid therapeutic alliance through empathic validation and sympathetic understanding before presenting interpretations of transference to the patient (35).

Finally, in the treatment of severe personality disorders, such as borderline personality disorder, the psychiatrist is attempting to treat both biology and psychodynamics. It is ill advised in treating personality disorders to try to separate brain and mind as though they are independent entities. Both medication and psychotherapy may be necessary for maximal gains, and the dynamic psychiatrist must be bilingual in the language of mind and the language of brain—which are equally important—in treating these patients (20). In treatment, one must avoid sacrificing optimal treatment because of a preference for one theory over another or one view of psychiatry over another.