Borderline Personality Disorder and Couple Dysfunctions Sébastien Bouchard, MPS, and Stéphane Sabourin, PhD
Current Psychiatry Reports 2009, 11:55–62 Current Medicine Group LLC ISSN 1523-3812
http://www.springerlink.com/content/x588530017721153/fulltext.pdf[/color]
The central characteristics of borderline personality disorder (BPD) are bound to be associated with the development and maintenance of couple dysfunction. Although seven of the nine diagnostic criteria of BPD in the DSM-IV-TR refer directly to interpersonal functioning, very few empiric studies have addressed the exact nature of the relationship between BPD and couple functioning. This article examines recent studies describing couple outcomes—union formation and durability, partner choice, relationship satisfaction, intimate violence, attachment security, and sexual functioning—associated with BPD. The relationship between couple dynamics (including partner personality characteristics) and BPD symptomatology is probably bidirectional or reciprocal. The review concludes with an exploration of diagnostic and treatment implications.
IntroductionIn discussing the research agenda for DSM-V Axis II, some authors argue that future personality disorder criteria should explicitly refer to the ability to establish and maintain fulfilling, intimate, long-term relationships [1]. Couple quality and stability are associated with a diverse array of positive educational, occupational, social, and personal outcomes [2]. Thus, the proposition that the inability to pursue relational life tasks in adulthood is a defining feature of personality disorders appears promising [1]. In fact, the challenge to develop a durable, loving relationship is a daunting task for most people suffering from personality disorders [3••], particularly borderline personality disorder (BPD). Surprisingly, the research evidence supporting this hypothesis is sparse. This article reviews recent studies examining whether BPD is related to a clinically significant alteration of couple functioning. Specifically, we examine empiric data pertaining to six aspects of couple dysfunctions in BPD: 1) union formation rates and stability, 2) partner selection processes and partner psychosocial profile, 3) relationship satisfaction, 4) intimate violence, 5) attachment security, and 6) sexual functioning. Before assessing the research literature, we briefl y present relevant clinical data.
In clinically driven analyses of BPD, relationship dysfunctions are discussed from many perspectives. First, from a descriptive viewpoint, the central characteristics of BPD—uncontrolled anger, impulsivity, cycles of idealization and devaluation of self and others, hypersensitivity to rejection, and self-destructive behaviors—form a syndrome that is bound to manifest itself mainly in close relationships and represent an intrinsic aspect of this syndrome [4,5]. Second, BPD is a severe mental disorder characterized by high rejection sensitivity [6]. Thus, relationship difficulties can intensify or stabilize the expression of borderline symptoms. Inversely, couple satisfaction is also presented as a buffer—or as a protective factor— against the worsening of BPD symptoms. Paris [7,8] has reported a compelling series of case studies underlining the role of couple issues in the evolution of BPD. Third, when studying the natural course of BPD, couple events have been examined as a specific category of psychosocial outcomes, positive or negative, of the disorder. The remission of BPD symptoms potentially allows these patients to work through important couple life tasks (eg, choosing an adequate partner, strengthening engagement, solving daily problems, learning to tolerate individual differences in marriage, overcoming sexual inhibitions).
Finally, a perusal of recently developed manuals for the individual treatment of BPD—dialectical [9], psychoanalytic [10,11], and schema-focused [12]—clearly indicates that the association between BPD and couple functioning is pervasive but complex. Likewise, couple therapy practitioners taking diverse approaches increasingly report that BPD is overrepresented in distressed couples who seek help [4,13••,14]. Fruzzetti and Fruzzetti [4] estimate that by using subthreshold criteria of BPD (at least three criteria met), close to 50% of distressed couples seeking treatment have at least one member with borderline personality traits or the full syndrome of BPD. From a strictly clinical perspective, the BPD couple functioning equation is judged to be signifi cant; it is probably bidirectional and needs to take into account several vantage points involving elements of couple functioning relevant for scrutinizing BPD’s etiology, essential nature, and natural course.
Dimensions of Couple Functioning Associated With BPDUnion formation and durability A first generation of research showed that BPD is associated with a lower probability of being married [15,16], greater number of break-ups in significant relationships [17], shorter friendship duration, and absence of an intimate partner or confidante [18]. In a 15-year follow-up study involving an extensive series of BPD patients (n = 502), Stone [19] reported marriage rates below national rates in the United States at the time (52% for women, 29% for men). More recently, in their study on hospitalized BPD patients observed for 7 years, Links and Stockwell [20] found that borderline patients married at the same rate as the comparison group of former inpatients. However, they noted that patients remaining single were younger, more impulsive, and had more dissociative episodes than patients in an intimate relationship. Finally, in a 27-year follow-up of a cohort of 100 BPD patients diagnosed using the DSM-III criteria, Paris and Zweig-Frank [21] indicated that 67% of their participants had been married, whereas the rate of divorce reached 36%; nevertheless, only 42% of these BPD patients were presently involved in a stable relationship.
In their important study, Whisman et al. [3••] excluded BPD diagnoses but used a large sample size (n = 43,093 adult respondents) against which the occurrence, timing, and disruptions of marital unions in BPD eventually could be compared in future investigations. Their results showed that personality disorders were associated with decreased probability of marriage, higher rates of early marriage, and elevated risk of marital disruption. No comparable data exist with a representative sample of individuals diagnosed with BPD. However, an examination of recent naturalistic studies that analyzed the clinical course of BPD and randomized clinical trials looking at the efficacy of various treatments for BPD shows three interesting trends (Table 1)—two substantive and the other methodologic [22•,23–26].
First, a significant percentage of patients (30% to 45%) with BPD are involved in an intimate couple relationship. Second, the results of two large-scale, prospective studies on the longitudinal course and outcome of BPD [27,28] provide relevant data on the prognosis for couple stability in BPD. In the McLean Study of Adult Development (MSAD), the probability of BPD patients in remission being married or living with a partner increased from 15.4% to 38% over a 6-year period. For nonremitted BPD patients, this percentage remained stable at 15%, and the difference with remitted patients was signifi cant. In the Children in the Community Study [28], after controlling for Axis I disorders in adolescence, elevated borderline symptoms in adolescents predicted lower partner involvement 20 years later. We recently conducted a small-scale study of 35 couples in which the woman was diagnosed with BPD [22•]. Mean relationship duration was almost 6 years (SD = 8.8; range, 2 months–38 years). Most of these couples reported a chronic pattern of episodic relationship instability characterized by intermittent break-ups and reunions approximately once every 6.5 months. In addition, 28.6% of clinical couples reported having broken up definitively before the end of the study, which lasted about 18 months. Finally, in a longitudinal study of male batterers presenting with high borderline symptomatology, the rate of relationship instability (separation or divorce) reached 75% over a 3-year period [29].
Third, from a methodologic perspective, there is a striking lack of descriptive and relevant couple data in research reports investigating BPD (eg, union status and duration). Overall, recent studies suggest that couple formation and duration are problematic processes in BPD patients. However, the relationship picture is not as negative as clinicians tend to expect it to be. In a signifi cant proportion of cases, once formed and after a period of adaptation and increased conflicts, couples in which one member suffers from BPD can reach a sort of “instable stability.†These preliminary results are interesting, but to gather a more complete picture of relationship stability in BPD, future investigations should routinely include more systematic and descriptive measures of current and past cohabitating and marital unions.
Partner choice and partner well-being Concern has grown among clinicians and researchers that among BPD patients, partner choice may have a significant effect on symptom intensity and treatment prognosis [30,31]. In this context, the success of long-term couple relationships depends partially on the personality of the partners whom BPD patients select [30]. If empiric evidence confi rms this hypothesis, the systematic assessment of a partner’s personality and a couple’s dynamics could become an important addition to diagnostic protocols used with BPD patients.
The National Institute of Mental Health states that people with BPD generally have poor judgment in choosing partners [32]. Although evidence for assortative mating exists in many psychiatric disorders, including alcoholism, drug use, schizophrenia, antisocial disorder, and affective disorders [33,34], specific data for assortative mating in individuals diagnosed with BPD are scarce. However, some indirect evidence exists.
First, dysregulation of aggression is a hallmark of BPD, and externalizing problems has been shown to infl uence a partner’s choice among adolescents and adults. For example, aggressive girls are more likely to select aggressive men as intimate partners [35], and adult romantic pairs display substantial homophily of aggression and delinquency [36]. Second, as similarity in personality-related domains is generally related to marital quality and duration [36,37], one would expect to find higher rates of personality disorders in intimate partners of people with BPD, especially in well-established couples.
To our knowledge, only one study supports this contention [22•]. Using the Structured Clinical Interview for DSM-IV Axis II Personality Disorders, these researchers found a 55.9% rate of personality disorder in intimate partners of women with BPD, whereas the expected rate of personality disorder in the general population is 9% to 15.7% [37]. At this early stage, it is difficult to determine whether this high proportion of potentially dysfunctional pairings refl ects poor partner choice, attraction by shared developmental failures [38], or simply a limitation in the availability of adequate partners.
The well-being of spouses or family members living with BPD patients also has gained attention recently. Scheirs and Bok [39] administered the Symptom Check List (SCL-90) to 64 spouses and parents of BPD patients. The results showed a degree of psychological distress comparable with what is normally observed in families of schizophrenic, depressive, or post-traumatic stress disorder patients. This high distress was as elevated in parents as it was in intimate partners. Our own data on the romantic partners of patients with BPD suggest that their psychological distress is twice as high as what is found in men from nondistressed couples [22•]. Hoffman and colleagues [40] showed the following: 1) family members and partners generally had little knowledge about BPD; 2) the more knowledge they had, the more depressed, burdened, hostile, and psychologically distressed they were; 3) a high degree of agreement exists between patients and those close to them on the personality traits of the BPD patient; and 4) BPD patients and those close to them disagreed on the personality traits of family members (or partners) who were perceived by BPD patients as evidencing a higher level of neuroticism and lower levels of extraversion and openness.
Recent studies reviewed here suggest the validity of some clinical concerns about the quality of partner choice and the maladaptive processes at work in these families. The hypothesis that caregivers (spouses or parents) present significant mental health impairments that may be conceptualized as causes or consequences of BPD has received preliminary support. However, the research basis on which this hypothesis rests should be strengthened. Innovative psychoeducational programs for families with an individual diagnosed with BPD have the potential to shed light on this question and to provide relevant data [41]. There were also some methodologic weaknesses in these studies: small samples, samples mixing partner with relatives, and insufficient information to compare the psychosocial profi les of partners and parents with population norms.
Relationship satisfaction Poor relationship satisfaction and high relationship distress are expected in couples in which one member suffers from BPD. The results of early studies suggested that an Axis II diagnosis significantly decreased relationship satisfaction [42,43]. This negative effect was more important for couples in which one partner suffered from a personality disorder than for couples in which one partner suffered from an Axis I disorder.
To our knowledge, three studies have examined BPD and couple quality in clinical samples using a categorical DSM diagnosis. In the MSAD [27], self-reported relationship quality (being in a good or in an emotionally sustaining, close relationship involving at least weekly contact without elements of abuse or neglect) was lower in BPD patients (33.5%, n = 290) than in patients with other personality disorders (46.3%, n = 72). In addition, over a 6-year period, the evolution of relationship quality was more positive for remitted than nonremitted BPD patients. More specifi cally, the percentage of remitted patients who felt they were in a good relationship rose from 37.6% (n = 202) to 63% (n = 200), whereas for nonremitted BPD, this percentage increased from 26% (n = 88) to 43.8% (n = 64).
Hill et al. [44] compared romantic relationship dysfunction in patients with BPD (n = 46) with those with avoidant personality disorder (n = 27) and individuals with no personality disorder (n = 25). The main findings revealed that compared with other patient groups, BPD was related to more romantic relationship dysfunction. BPD patients also evidenced elevated avoidant romantic relationship. To our knowledge, only one study has directly assessed couple satisfaction in BPD patients and their partners. Bouchard et al. [22•] first showed the absence of within-partner difference on couple distress in these unions. In addition, both partners of these BPD couples reported lowered dyadic satisfaction than community couples. These differences were statistically strong (d = 0.80 for the difference between men, d = 1.26 for women). When comparing these satisfaction scores with population norms [45], the results indicated that 49% of women with BPD and 40% of their male partners were clinically distressed. This is an important finding because rates of couple distress in population samples vary from 20% to 30% [46,47].
Borderline symptoms and traits were related to relationship quality or adjustment in three recent studies. First, in a report based on a subsample of older adolescents from the Children in the Community Study (n = 200) observed over a 10-year period (from age 17 to 27 years), Chen et al. [48] observed that after controlling for Axis I disorders and other symptoms of personality disorders, self-reported borderline symptoms were associated with sustained elevations in partner confl ict throughout the 10-year period. These findings suggest that borderline traits have a negative impact on relationship satisfaction and that this effect may be lasting and even increase over time. Second, in a 4-year longitudinal study of 142 older adolescent girls, Dailey and colleagues [49] found that the presence of borderline personality traits predicted increases in partner dissatisfaction, romantic confl icts, and unwanted pregnancy. These associations remained significant after controlling for depression. Third, in a community sample of 82 married couples, South et al. [50] used multilevel modeling to study the relationship of self- and partner-reported symptoms of BPD to marital satisfaction. Results showed that after controlling for other personality disorder symptoms, self-reported marital satisfaction was predicted by higher levels of spouse-reported BPD symptomatology.
Taken together, BPD generally seems to have a negative impact on couple satisfaction, but we believe that closer scrutiny should be applied to the significant proportion of couples who still report being satisfied with their relationship despite the presence of BPD in one of the partners. These particular couples may have developed particular coping skills from which other couples like them could benefit. Also, the specific relationship expectations of patients with BPD need to be examined, as the extent to which their intimate partner has adjusted to them probably has a signifi cant infl uence on couple satisfaction.
Intimate violence People with BPD are at risk of being abusive and becoming victims of abuse in intimate relationships. Using data from the MSAD, Zanarini et al. [51] reported that the prevalence of physical/sexual abuse experienced as adults was higher in BPD patients (45.5%) than in patients with another Axis II disorder (16.1%). Unfortunately, this study did not report to what extent these abusive experiences took place in an intimate relationship. In the study by Bouchard et al. [22•], the percentage of women with BPD who reported being psychologically violent with their partner was 88.6%, whereas the rate of physical violence was 54.3%. Physical violence consisted mainly of minor assaults, and a minority of couples exhibited mutual violence.
Using a 3-year prospective design, Holtzworth-Munroe et al. [29] studied a group of 102 maritally violent men. As expected, a reliable subgroup of male batterers had elevated self-reported borderline symptoms (16% of the sample). In the prospective study by Dailey et al. [49], borderline symptoms observed at the fi rst assessment period were related to subsequent experiences of physical abuse by a romantic partner. However, this prospective association disappeared after researchers controlled for initial depression and other personality disorder symptoms. They did observe that higher levels of verbal aggression were related to higher levels of borderline pathology as reported by each spouse. Finally, in a cross-sectional study of undergraduate students, borderline traits were associated with relationship (eg, gossiping, rumor spreading, flirting in front of one’s romantic partner) and overt aggression [52].
Attachment security Preoccupied and fearful attachment styles are overrepresented in BPD populations [53]. Likewise, rejection anxiety—but not avoidance of proximity—generally is prevalent in BPD patients [54]. A growing body of research also suggests that the insecure attachment status of patients with BPD is closely associated with their relationship difficulties [55]. However, only a few of these studies investigated the association of BPD with couple quality.
People suffering from BPD generally represent an exacting and highly unpredictable attachment figure for their romantic partner. Also, for these couples, with an insecure attachment style generally observed on both sides of the dyad, the possibility of successful coregulation of negative affect may be almost out of reach. For example, in the Bouchard et al. [22•] sample, both partners evidenced insecure attachment representations in 68.6% of all cases. These couples seem bound to continue feeling insecure with each other and have to resort to hyperactivating and deactivating strategies to regulate affects [56]. Still, as reported previously, a significant percentage of couples in which one member has BPD seem satisfied with the union and prove to be somewhat stable. Research revealing that couples in which both partners are insecurely attached generally can remain longer in an unsatisfying relationship compared with securely attached couples [57] or suggesting that similarity in attachment characteristics—not attachment security per se—is strongly associated with relationship satisfaction [58] helps us make sense of these somewhat contradictory intuitive observations. Male batterers presenting with borderline symptoms sampled in the study by Holtzworth-Munroe et al. [29] also evidenced fear of abandonment and preoccupied/fearful attachment. These attachment difficulties were stable over a 3-year period.
Two studies directly evaluated the attachment styles of partners of women with the full syndrome of BPD [22•,59]. Goldstein [59] used a sample mainly composed of people with subclinical BPD (up to four BPD criteria; n = 36) and included only four couples in which one member had fi ve or more BPD criteria. In their sample (n = 40 couples), the number of BPD symptoms correlated moderately with a partner report of feeling insecure in the relationship and worrying about losing a partner’s affection. The results of the study by Bouchard et al. [22•] strongly suggest that insecurity of attachment is not only a defining feature of BPD (100% of their sample of women with BPD had an insecure attachment status). Also, a rate of insecure attachment representation of 68.6% was found in partners of women with BPD (vs 34.3% in the control group). For these men, both dimensions of attachment—avoidance of intimacy and rejection anxiety—were statistically higher than what was observed in men from community couples.
Sexual functioning Theoretical and clinical analyses frequently underlined how BPD is associated with problematic sexual behaviors. Neeleman [60] recently reported only six empiric studies addressing the sexual functioning of patients with BPD. However, most were published before 1995. Neeleman [60] concluded that individuals with BPD tend to have significant problems with regard to intimate and sexual relationships. These problems seem to be related to heightened sexual impulsivity, reduced sexual satisfaction, increased sexual boredom, greater preoccupation with sex, avoidance of sex, and a wide range of sexual complaints. In addition, evidence indicates that gender identity disorder and ambivalence about sexual orientation occur more frequently in people with BPD.
In the past 5 years, three relevant studies have been published. First, Zanarini et al. [61], in an analysis of the MSAD data, assessed three patterns of sexual diffi culties: avoidance of sex for fear of experiencing negative symptoms, symptoms experienced after consensual sexual relations, and general sexual difficulties. The prevalence of these problematic patterns was high in BPD patients, reaching 41%, 34.1%, and 47.6%, respectively, for each of the problematic behavioral patterns and 5.6%, 9.7%, and 12.5%, respectively, for patients with other personality disorders. Female BPD patients reported more sexual difficulties (65.2%) than men with BPD (45.9%). Finally, over a 6-year period, these sexual diffi culties significantly declined, more so for men than women. Second, Sansone et al. [62] surveyed the sexual histories of 76 female outpatients diagnosed using the Personality Diagnostic Questionnaire-4 and a screening interview. Results revealed that compared with women who were not diagnosed with BPD, women with BPD reported an earlier onset of sexual intercourse and were more likely to have experienced date rape.
Third, a recent study on a sample of 35 women diagnosed with BPD reported that the percentage who had more than 30 sexual partners across their lifetime was about 10 times higher than that of control women [63]. The clinical observations that often report high frequency of sexual partners in this population thus seem to hold, but we believe that a more refi ned interpretation is needed. In this sample, once they were involved in a more intimate relationship, women with BPD did not differ from other women on number of sexual thoughts, frequency of masturbation, and number of voluntary sexual contacts over the past year [63]. These results strongly suggest that the relationship status of women with BPD highly infl uences the type and frequency of sexual behaviors.
Bouchard et al. [63] also noted that among women with BPD involved in an intimate relationship, subjective experiences of sexuality differed from those of other women for attitudes such as ambivalence, negative attitudes, and feeling pressured. More specifically, the feeling of being pressured by a partner was uniquely associated with meeting criteria for BPD, even after controlling for the effect of childhood sexual abuse. Further analyses revealed that rejection anxiety played a mediational role in this association. Thus, attachment status in patients with BPD seems to interfere not only with general interpersonal functioning but also in the more intimate area of subjective sexual experience. This meaningful aspect of the impact of BPD on intimate relationships is seldom considered in clinical practice and research.
Clinical Implications Assessment and diagnostic implications Actual treatment facilities and future treatment effi cacy research should integrate a short evaluation of the couple’s variables that are suspected to negatively infl uence treatment outcome when ignored. A measure for both partners of dyadic adjustment, intimate violence (physical and psychological), style of communication, and attachment style may be most informative for treatment orientation and treatment process. Also, the taboo of feminine intimate violence must be overcome, and this dimension of the interpersonal functioning of some women with BPD must be evaluated and integrated into treatment plans. In some cases, an investigation of the more intimate sphere of activities and attitudes toward sexuality and sexual history is most important to the diagnostic process. Finally, we strongly believe that clinicians should be aware of the tendency to conclude prematurely that patients with BPD systematically distort their descriptions of others. Asking for a precise description of the personality characteristics of their intimate partners and, when possible, directly assessing the presence of personality pathology in these partners can sometimes open our eyes to neglected factors in some treatment-resistant patients in psychotherapy.
Treatment implications Disturbances between romantic partners are important to understand insofar as they probably help to generate and maintain personality disorders in these individuals [30]. Treatment resources should consider offering more specifi c psychoeducation to partners of individuals with BPD. In the hope of preventing the escalation and chronicization of BPD symptomatology, we as clinicians should try to better educate young patients with BPD as to how to be aware of themselves in their attraction to certain types of love partners and in the way they cope with rising intimate confl icts. Furthermore, that a high prevalence of personality pathology probably exists among the romantic partners of women with BPD suggests that merely educating the partner about BPD and how to behave with such a person may fail to address some clinically important behavioral patterns that could be part of the non-BPD partner. As many authors believe [13••,22•,30,31], an effective approach to treating women with BPD should strive to integrate, when possible, a couple approach, as the intimate partner can be viewed as a vicarious victim or an unrecognized contributor to the manifestations of BPD.
Conclusions This paper reviewed the available theorization and data on the impact of BPD on intimate relationships. The literature strongly suggests that BPD has a negative impact on important aspects of couple functioning. Future research should look more closely at the impact of BPD on intimate relationships and the impact of a couple’s dynamic on manifestations and outcome of this severe psychopathology. Ultimately, a better integration of personality disorder theorization with couple research will help to reduce the negative impact of BPD on relationship satisfaction of both partners and eventually contribute to enhancing the often limited efficacy of available psychosocial treatments for BPD.
Corresponding author SébastienBouchard, MPS Centre de Traitement Le Faubourg Saint-Jean, Programme-ClientÂle des Troubles SévÂres de la Personnalité du Centre Hospitalier Robert-Giffard, 175 Rue Saint-Jean, 3e étage, Québec City, Québec G1R 1N4, Canada. E-mail: sebastien_bouchard@ssss.gouv.qc.ca
Poll
