Managing treatment-resistant depression with comorbid personality disorders

Personality disorders are some of the most common comorbid psychiatric conditions in major depressive disorder (MDD). The combination of MDD and personality disorders provides a real challenge to clinicians, as treatment of patients with these comorbid conditions is hampered by significant deficits in psychosocial and interpersonal functioning, often leading to poorer outcomes and treatment resistance. This chapter explores the relationship between personality disorders and TRD by synthesizing relevant empirical data and their implications for clinical practice and discusses alternative, personality-derived theoretical models of TRD. We begin by reviewing the definition and criteria for personality disorders according to the Diagnostic and Statistical Manual of Mental Disorders, 5th Edition ( DSM-5 ; ), along with the Alternative DSM-5 Model for Personality Disorders included in Section III: Emerging Measures and Models of the manual. Then, we discuss the relationship between depression and personality dysfunction, including the challenges in assessing each of the conditions in comorbid presentations. We then turn our attention to reviewing the evidence of the influence of comorbid personality disorders on the functioning of patients with MDD and treatment outcome for MDD. The two sections that follow discuss the relationship between depression and specific personality disorders. In order to allow for sufficient depth of discussion, we chose to focus on two types of personality pathology, Borderline Personality Disorder (BPD) and Narcissistic Personality Disorder (NPD), because of their frequent comorbidity with TRD in clinical practice. An additional factor in our selection was the predominance of the diagnosis of BPD in female and NPD in male patients. In each of the two sections, we review the literature on effective biological, psychological, and combined treatments for the corresponding personality disorders and the influence of the personality pathology on the treatment outcome in cooccurring MDD. We provide case studies to illustrate each of the comorbid clinical presentations. We then discuss alternative, personality-derived models of TRD through the framework of the Section III of the DSM-5, the historical concept of Depressive Personality Disorder, and a novel empirically derived construct of malignant self-regard. Finally, we distill the implications for clinical practice in managing TRD with comorbid personality disorders in the “ Clinical summary ” section of this chapter.

Personality disorders in DSM-5

The DSM-5 describes personality disorders in both Section II (Diagnostic Criteria and Codes) and Section III (Emerging Measures and Models). Section II describes personality disorders and their criteria in a manner that is nearly identical to the DSM-IV-TR ( ), which describes 10 discrete personality disorders with specific criteria and instructions on how combinations of these criteria lead one to be diagnosed with the disorder. A personality disorder is “an enduring pattern of inner experience and behavior that deviates markedly from the expectations of the individual’s culture, is pervasive and inflexible, has an onset in adolescence or early adulthood, is stable over time, and leads to distress or impairment” (p. 645). The Section III description of personality disorders, known as the Alternative DSM-5 Model for Personality Disorders (AMPD), is a product of the DSM-5 Personality and Personality Disorders Work Group and represents a radical departure from the current diagnostic approach. This model was included because of considerable problems that have been leveled against the current categorical approach, including the “failure to ‘carve nature at its joints’, the lack of meaningful or well-validated boundary between normal and disordered personality, excessive heterogeneity within diagnostic categories, excessive diagnostic cooccurrence across categories, inadequate coverage of the full range of personality difficulties seen in clinical practice, questionable temporal stability of diagnoses, dissatisfaction among clinicians, and inadequate scientific foundation” (p. 45, ).

Because of these concerns, the AMPD was included in Section III of DSM-5, with the encouragement for the model to be further tested and validated. Here, “personality disorders are characterized by impairments in personality functioning and pathological personality traits” (p. 761). Personality functioning refers to the individual’s representations of both self and others, also known as interpersonal functioning. Within the self domain, individuals are assessed for the quality of their identity (i.e., experience of oneself as unique, separateness from others, self-esteem stability, and the sense of who one is as a person) and self-directedness (i.e., a sense of one’s ability to plan and act on personally meaningful short-term and long-term goals). Within the interpersonal domain, individuals are assessed for empathy (i.e., the ability to understand, appreciate, and tolerate other’s points of view and needs, as well as knowing how one’s actions affects others) and intimacy (i.e., the ability to have meaningful, in-depth relationships with others). Additionally, pathological personality traits are organized into five broad trait domains, each of which is composed of several facets. The domains and facets are as follows:

Negative affectivity (emotional lability, anxiousness, separation insecurity, submissiveness, hostility, perseveration, depressivity, suspiciousness, and restricted affectivity)
Detachment (withdrawal, intimacy avoidance, anhedonia, depressivity, restricted affectivity, suspiciousness)
Antagonism (manipulativeness, deceitfulness, grandiosity, attention-seeking, callousness, hostility)
Disinhibition (irresponsibility, impulsivity, distractibility, risk taking, rigid perfectionism, or lack thereof)
Psychoticism (unusual beliefs and experiences, eccentricity, cognitive and perceptual dysregulation)

It should be noted that certain facets can be part of differing trait domains (e.g., depressivity and restricted affectivity) and that the facets per domain vary in number. These trait domains were derived from extensive empirical investigation and are strongly grounded in empirical investigations of trait theory in personality psychology ( ). To be diagnosed with a personality disorder by way of the AMPD, an individual must have “moderate or greater impairment in personality (self/interpersonal) functioning” and “one or pathological personality traits”; p. 761). How the AMPD is related to TRD will be discussed below. However, suffice it to say that the AMPD offers a much wider range of personality pathology than the typical taxonomy and incorporates chronic affective states (e.g., depressivity, emotional lability, anxiousness, and hostility) into its assessment framework.

The relationship between personality disorders and depression

The combination of personality disorders and MDD is one of the most common comorbid psychiatric presentations ( ; ). Prevalence rates in community and clinical samples span a wide range from 20% to 85% ( ; ; ; ; ), with the average being approximately 50%. Prevalence data on comorbid personality disorders in TRD are virtually nonexistent, although one small study comparing rigorously-defined treatment-resistant depressed outpatients with their non-TRD counterparts did not find treatment resistance to be associated with increased rates of personality disturbance ( ).

From a conceptual standpoint, the development and nature of the association between depression and personality pathology remain poorly understood. These constructs are interrelated, but the relationship between them is likely complex and multifactorial ( ; ). summarizes relevant theoretical work by identifying three possible operative models in the relationship between depression and personality disorders: predisposition, complication, and coaggregation. Predisposition proposes that patients with certain personality traits or trait constellations, including neuroticism and low extraversion/detachment, are more susceptible to developing depressive disorders. Complication refers to pathoplastic effects of personality pathology on the expression or course of depression, such as increased severity or chronicity and poorer response to standard treatments. Conversely, the complication model can account for the development or exacerbation of personality pathology as a result of recurrent or persistent depressive episodes. This conceptual model is most applicable to elucidating the association between personality disorders and TRD and most relevant to the topic of this book chapter. Finally, coaggregation suggests that when a patient’s coexisting vulnerabilities for depression and personality pathology are simultaneously expressed, the personality disorder becomes a precursor to or an attenuated manifestation of depression.

review additional conceptual work to expand the discussion of the interrelatedness of depression and maladaptive personality traits/personality disorders. They highlight several additional explanations, including (a) depression and personality disorders have common causes, (b) the content and diagnostic criteria of the two conditions overlap, and (c) personality traits/disorders represent consequences or “scars” of depressive illness. The role of life stress and its shared association with depression and personality pathology deserves special mention here. observes that the relationship between personality traits/disorder and depression is intensified by adverse life events, which often serve as triggers or perpetuating factors in major depressive episodes. In addition, there is an interactive and reciprocal relationship between life stress and personality pathology, whereby individuals with maladaptive personality traits (e.g., antisocial, narcissistic, borderline) have more unstable life circumstances (e.g., work, financial, and legal problems) and interpersonal relations (e.g., marital discord, estrangement from family), which, in turn, increase their risk of being exposed to various stressors ( ).

Assessment of comorbid personality and depressive disorders

The complexity of interrelatedness between depression and personality pathology makes the task of disentangling depressive symptoms from those of personality disorders exceptionally difficult in clinical settings. One challenge lies in the fact that the presence of depressive phenomenology may color the expression of personality and therefore complicate its assessment and vice versa. Furthermore, there is evidence to suggest that patients with MDD tend to overreport symptoms of personality pathology ( ; ), which makes the assessment of personality while the patient is acutely depressed problematic. While clinicians can reliably evaluate current depressive symptoms using patient self-report, the depressed patient’s description of their long-standing, habitual interpersonal patterns and self-perceptions are likely to be negatively influenced state effects of the illness, yielding a potentially biased assessment of personality functioning ( ; ). Conversely, personality pathology may color the quality of the patient’s experience of MDD and lead to overreporting of depressive symptoms. For example, one literature review found that patients with BPD tend to score higher on self-report measures of depressive symptoms, compared to clinician-administered scales ( ). In fact, patients with BPD without MDD may score as highly on self-rated symptom inventories as BPD patients with MDD and as highly as patients with MDD without BPD ( ). Finally, an initial, cross-sectional evaluation of an adult patient based primarily on self-report may not allow for accurate assessment of their personality traits/disorders, which are, by definition, longitudinal constructs with a developmental origin that typically manifest themselves by late adolescence ( ).

There are several possible strategies for addressing the challenge of assessment of comorbid depressive and personality disorders. One such strategy involves integration of collateral information in the evaluation process. Reports from collateral sources, such as long-term romantic partners or spouses, parents, and adult children can improve the accuracy of the assessment by providing additional data points to patient self-report, including longitudinal data ( ). However, this approach is not without its limitations, as it potentially introduces informant bias to the assessment process. In particular, familial risk of depression increases the probability that the informants themselves may be suffering from a mood disorder, which, in turn, can color their reports of the identified patient’s behavior or interpersonal patterns. In addition, the negative impact of the identified patient’s debilitating depressive illness or problematic personality traits on the informants’ quality of life and relationship with the proband ( ), may negatively bias their descriptions ( ; ).

Another strategy for navigating the challenge of accurately assessing personality disorders and depression in comorbid presentations involves incorporating clinician-administered semistructured interviews in the assessment process. Although clinicians typically rely on unstructured interviews in diagnosing personality disorders ( ), structured and semistructured interviews have shown greater interrater reliability ( ; ). Examples of commonly used measures for personality disorders include Structured Interview for DSM-IV Personality (SIDP-IV; ), Structured Clinical Interview for DSM-5 Personality Disorders (SCID-5-PD; ), Personality Inventory for DSM-5 (PID-5; ), and the DSM-5 Level of Personality Functioning Scale (LPFS; ). The assessment of depression can also be aided by incorporating structured interviews and self-administered inventories. proposed an evidence-based toolkit of measures with adequate psychometric properties for optimal depression assessment in clinical settings: (a) the Structural Clinical Interview for DSM (SCID), with most recent clinician version being SCID-5-CV ( ) to establish a formal mood disorder diagnosis; (b) the Mini-International Neuropsychiatric Interview (MINI; ) module on melancholic features to supplement the SCID; (c) the Seasonal Pattern Assessment Questionnaire (SPAQ; ) to assess for a seasonal component; (d) the Beck Depression Inventory-II (BDI-II; ) to assess depressive symptom severity and short-term change in symptoms; and (e) the Longitudinal Interval Follow-Up Evaluation (LIFE; ) to assess remission of MDD.

The impact of personality disorders on functioning and treatment outcome in patients with MDD

Very few studies have examined the role of personality disorders on psychosocial functioning of patients with MDD. The best available evidence comes from the Collaborative Longitudinal Personality Disorders Study (CLPS; ), a large prospective multisite study with over 600 participants recruited at Brown, Columbia, Harvard, and Yale University medical schools. In a cross-section analysis of data obtained at 36-month follow-up, compared the psychosocial functioning and well-being of patients with comorbid MDD and schizotypal, borderline, avoidant, or obsessive-compulsive personality disorder to those with MDD. They utilized the Medical Outcomes Study Short-Form Health Survey (SF-36; ), a widely used self-report inventory that assesses health-related quality of life. They found that personality disorder significantly contributed to functional impairment and diminished sense of well-being in at least three areas measured by the SF-36: emotional role limitations (i.e., problems with daily activities or work), social functioning (i.e., interpersonal relationships with relatives, friends, or neighbors), and general health perceptions (i.e., attitudes toward one’s health, both present and anticipated, and one’s appraisal of their tendency to get sick). These results were consistent with the findings from the baseline data from CLPS ( ; ; ; ; ), which showed that patients with comorbid MDD and personality disorders experienced greater impairment in employment and interpersonal relationships, compared to those suffering from MDD alone.

In a longitudinal, naturalistic follow-along study using data from the first 2 years of the CLPS cohort, compared three distinct groups of patients: (a) those with baseline MDD and personality disorder diagnoses, whose personality pathology remained stable and did not remit during the study period; (b) those with the same baseline comorbid presentation, but whose personality disorders remitted over time; and (c) those with MDD who did not meet the criteria for a personality disorder diagnosis at baseline. Using the LIFE psychosocial scales, which assess employment, interactions with friends, spouse/partner, and parents, recreation, global social adjustment, and the DSM-IV Axis V Global Assessment Functioning (GAF), they found that the presence of a persistent personality disorder predicted impairment in psychosocial functioning over and above the debilitating effects of comorbid MDD. They also found that the patients whose personality disorders remitted were more likely to experience a remission of MDD and improvement in social and occupational functioning. concluded that the burden of personality disorders in depressed patients extended beyond suffering from symptoms to psychosocial functioning and that the clinical course of comorbid personality disorders directly influenced outcome of MDD.

Numerous studies have examined the pathoplastic effects of personality pathology on the course and treatment outcome of depression. There is considerable evidence that patents with MDD suffering from a comorbid personality disorder have an earlier age of illness onset, experience more severe symptoms and longer duration of depressive episodes and are at higher risk of relapse after remission ( ; ; ). The literature on the impact of personality pathology on depression treatment outcome has been more mixed and has been the subject of a long-standing debate in psychiatry going back nearly 70 years ( ). In order to synthesize the findings from scores of relevant original studies, several narrative and systematic literature reviews and metaanalyses have been published in the last two decades. In a narrative review of over 50 studies, found that the prognostic significance of personality pathology for treatment outcome in patients with MDD depended on study design, with the best-designed studies reporting the least amount of impact. A systematic literature review and metaanalysis of six randomized controlled trials (RCT) of pharmacotherapy for depressed patients with or without personality disorders also found no difference between the two groups ( ). On the other hand, a more inclusive systematic review and metaanalysis of 34 studies, incorporating cohort, case-control, and RCT methodologies reported poorer outcomes for the comorbid group ( ). This finding supported the conclusions of several earlier narrative reviews that found comorbid personality disorders to predict poorer outcome of depression treatment, including pharmacotherapy and electroconvulsive therapy (ECT; ; ; ; ).

In attempt to bring clarity to this debate, the authors of several previous reviews came together as the Personality Disorder and Depression Outcome Group (PDDOG) and published the most comprehensive systematic literature review and metaanalysis on this topic to date, encompassing a total of 58 studies ( ). Of the included studies, 35 were prospective case series and 17 RCTs. In the metaanalysis, showed that, irrespective of treatment modality, a diagnosis of personality disorder in patients with MDD was associated with a poorer treatment outcome, with an overall odds ratio of 2.16 (1.83–2.56). The authors also noted that, in the included studies, the treatment was limited to depressive disorders and did not address coexisting personality pathology, highlighting a potential reason for treatment resistance in MDD. They concluded that assessing personality status and treating cooccurring personality disorders were essential for optimizing treatment response in depressed patients. Furthermore, in the most up-to-date narrative review of recent literature on this topic, highlights additional evidence to support the conclusion that untreated personality pathology can negatively affect the clinical course and treatment outcome in MDD. These findings are consistent with the clinical experience of the authors of this book chapter and many other clinicians working with patients with chronic and treatment-resistant depression.

Borderline personality disorder and depression

Clinical presentation

Borderline personality disorder (BPD) is a serious mental illness characterized by “a pervasive pattern of instability of interpersonal relationships, self-image, and affects, and marked impulsivity that begins by early adulthood and is present in a variety of contexts" ( , p. 663). Individuals with BPD frequently experience extreme fear of abandonment and make frantic and often self-sabotaging efforts to avoid real or perceived rejection, such as responding with inappropriate anger, suicidal threats, or parasuicidal gestures. This dynamic is produced by an intense fear of being alone and an overreliance on others for a sense of personal identity and worth. BPD patients struggle with individuation and often experience their identity as “fused” with that of other people. At the same time, their interpersonal relationships are marked by instability and intensity, often resulting in significant fluctuations from idealization to debasement or derision, depending on their perception of the other individual’s readiness to meet their needs or prove their loyalty. Their descriptions of other people may lack complexity and are often extreme: for example, “the best human being ever” or “a total loser.” These individuals’ self-image and sense of self are also highly unstable, and they are prone to intense shame, guilt and self-condemnation. They experience these feelings as so intolerable, that they often resort to radical and ultimately counterproductive mitigating measures, such as self-mutilation and suicidal behaviors. Individuals with BPD also display significant impulsivity, including behaviors that prove to be self-damaging (e.g., unprotected sex, excessive spending, substance abuse, unsafe driving, etc.). Finally, BPD is characterized by affective volatility, frequently resulting in depressive states marked by intense albeit transitory dysphoria, several hours to a few days in duration, anger and hostility. These fluctuations stem from BPD-associated mood reactivity, which, in turn, is triggered by a real or imagined abandonment or anticipated rejection. In some cases, the affective dysregulation can be accompanied by transient paranoid ideation or dissociative experiences.

Due to its characteristic instability of emotions and predominance of negative affect, BPD can often mimic a depressive disorder. Both BPD and MDD are marked by dysphoric affect and increased risk for suicidal behavior, which makes the task of determining whether the depressive symptoms should be attributed to the cooccurring depression or to the personality disorder itself, more complicated. Furthermore, individuals with this type of personality pathology appear to have a heightened subjective experience of depression, so much so that patients with BPD without MDD often score as highly on standardized self-rated depression measures as BPD patients with MDD and as highly as patients with MDD without BPD ( ). In addition, patients with comorbid MDD and BPD tend to endorse greater severity of symptoms on self-report measures as compared to clinician-rated severity ( ). Finally, the experience of depression in individuals with MDD is often qualitatively different from that in BPD patients, with the former being marked by melancholy, anhedonia, and apathy, accompanied by neurovegetative symptoms, and the latter by a sense of boredom, alienation, emptiness and inner “badness,” as well as anger and hostility ( ; ; ).

BPD comorbidity and its impact on the course and treatment outcome in MDD

BPD is a relatively common mental illness, with prevalence rates of approximately 1.7% in the general population and 15%–28% in patients receiving psychiatric care in inpatient and outpatient settings ( ). BPD is also by far the most-studied personality disorder, and its interface with depression has received a great deal of attention in professional literature. The lifetime prevalence of cooccurrence of depression with this type of personality pathology is extremely high, with recent studies suggesting that 61%–85% of individuals with BPD eventually meet the criteria for MDD ( ; ).

In light of the frequent comorbidity encountered in both research and clinical practice, a number of studies have explored the possibility that the two conditions have shared causes, with some authors going as far as conceptualizing BPD as a variant of a mood disorder ( ; ). A recent review by has in fact revealed an overlap in neurobiology of BPD and MDD, including reduced serotonin neurotransmission, amygdala hyperreactivity to emotional stimuli, and volume abnormalities in the anterior cingulate cortex and hippocampus. At the same time, the review also identified significant differences in brain region involvement, neurohormonal activation, heritability, symptomatology, prognosis, and pharmacotherapy response. The above similarities notwithstanding, the consensus of expert opinion remains that BPD and MDD are separate and distinct disorders ( ; ; ; ; ; ; ) that may share certain biological, temperamental, and early environmental risk factors ( ; ).

There has been a long-standing consensus of expert opinion that antidepressant medication is not as effective in ameliorating the symptoms of major depression cooccurring with BPD as in MDD without personality pathology ( ; ). In a recent systematic literature review synthesizing the findings of three longitudinal studies ( ; ; ) with a total sample of nearly 2400 patients, reported that the presence of comorbid BPD in fact slightly worsens the outcome of MDD. Cooccurring BPD is also associated with earlier onset of depression, longer duration of Major Depressive Episodes (MDE), increased rate of persistence of depressive symptoms, and lower rate of remission ( ; ). These findings are consistent with the results of an earlier review of the impact of personality pathology on treatment outcome in major depression ( ) that reported that high neuroticism scores, which are characteristic of BPD ( ), were associated with poor prognosis, particularly in long-term outcomes. Furthermore, in a 6-year prospective study, found that high neuroticism scores also predicted earlier relapse and shorter remission of major depression, suggesting that that depressed patients with cooccurring BPD respond more poorly to pharmacological treatment. In addition, in a prospective epidemiological study using a large nationally-representative sample from the National Epidemiologic Survey on Alcoholism and Related Conditions, observed that BPD accounted for over half of cases of persistent depression at the 3-year follow-up. BPD robustly predicted persistence of MDD, even after controlling for age at onset of depression, the number of previous episodes, duration of the current episode, family history, treatment, and other personality disorders. Finally, in another longitudinal study from CLPS that followed a sample of BPD patients with and without cooccurring MDD for 3 years, found that the rate of remission of BPD was not affected by the presence of comorbid MDD; however, the remission rate of MDD was significantly reduced by the presence of coexisting BPD. When considered together, the above evidence identifies BPD as a key risk factor for TRD.

It should be noted here that the prognostic significance of comorbid BPD extends to somatic treatments for major depression, although the available evidence is more limited. A narrative review by suggested that depressed patients receiving ECT had poorer outcomes on some measures in the presence of coexisting BPD. This finding was supported by more recent original studies ( ; ). For example, in a prospective longitudinal study, compared the outcome of ECT in three different groups of depressed patients: (a) those with BPD, (b) those with PDs other than BPD, and (c) those without a PD and found that the BPD group showed the least amount of symptomatic improvement following treatment completion. In addition, in a recent retrospective chart review of patients with MDD receiving ECT and transcranial magnetic stimulation (TMS), found that the presence of coexisting borderline personality traits modestly predicted poorer response to ECT. Interestingly, the authors also found that borderline personality traits had no impact on the outcome of TMS. The available data does indicate that MDD could be successfully treated with ECT in patients with cooccurring BPD, leading some researchers to speculate that “the depressed patient appears to have two distinct disorders, one with is responsive to ECT and the other which is not” ( , p. 91).

Treatment

In light of the negative prognostic impact of BPD on cooccurring major depression, a proper initial treatment of BPD can improve the chances of achieving remission in MDD. In fact, found that improvements in BPD were followed by improvements in MDD, but not vice versa. This section will provide an overview of treatments for BPD and review the available limited data on treatments for the combination of MDD and BPD.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here