Communication Between Individuals With Borderline Personality Disorder and Their Partners

ABSTRACT

Borderline personality disorder (BPD) is characterized by frequent and intense conflict in intimate relationships. Emerging theoretical perspectives have suggested that communication from both individuals with BPD and their partners during times of conflict may exacerbate BPD pathology and therefore reflect an important target for intervention. Communication samples between individuals with BPD and their partners (N = 18 couples) who were seeking conjoint therapy were coded using the Rapid Marital Interactions Coding System‐2. No differences were found in the amount of positive, constructive or hostile communication between individuals with BPD and their partners, and both partners used significantly more constructive communication than positive or hostile, though they were misaligned in their use of constructive communication. These findings challenge the notion that individuals with BPD communicate in a more hostile or conflictual way than their partners, despite the stigma they face in healthcare and community settings.

1. Introduction

Borderline personality disorder (BPD) is a debilitating and often lethal disorder characterized by disturbances in close relationships, fraught with frequent and intense conflict (American Psychiatric Association ²⁰²²; Navarro‐Gómez et al. ²⁰¹⁷). Research has demonstrated that individuals with BPD are most at risk for dangerous or lethal behaviours, such as self‐harm, during times of conflict in intimate relationships, potentially due to an inability to engage with and express emotions in a safe and effective way (Brodsky et al. 2006; Sauer et al. ²⁰¹⁴). As such, understanding how individuals with BPD and their partners communicate during moments of conflict may highlight potential intervention pathways that could optimize BPD recovery and mitigate some of its life‐threatening or destructive outcomes (Fitzpatrick et al. 2021).

1.1. Conflictual Intimate Relationships

Extensive research demonstrates that individuals with BPD experience frequent and higher intensity relationship conflict compared with healthy controls and other clinical populations (e.g., Bouchard et al. ²⁰⁰⁹; de Montigny‐Malenfant et al. ²⁰¹³; Jackson et al. ²⁰¹⁵; Lazarus et al. ²⁰²⁰; Navarro‐Gómez et al. ²⁰¹⁷). Several studies also report that while people with BPD have more intimate relationships than healthy controls, their relationships are typically shorter‐term (see Navarro‐Gómez et al. ²⁰¹⁷, for review). Women with BPD experience significantly higher rates of intimate partner violence than women without BPD (Bouchard et al. 2009), and BPD symptom severity is associated with a greater likelihood to both perpetrate and experience intimate partner violence (Brownridge and Tyler 2023; Jackson et al. ²⁰¹⁵). Overall, the extant evidence suggests that intimate relationships of those with BPD are generally volatile, conflictual and less satisfying than those of other groups.

1.2. The Borderline Interpersonal Affective Systems (BIAS) Model

The Borderline Interpersonal Affective Systems (BIAS) model posits that relationship disturbances (e.g., interpersonal threat sensitivity, negative appraisal biases) and emotion dysregulation (e.g., difficulties with emotion processes and problems regulating emotions) are at the core of BPD, and that reciprocal transactions between these features are the basis of the aetiology and maintenance of the disorder (Fitzpatrick et al. 2021). Importantly, The BIAS model suggests that these core features lead people with BPD to have communication problems with their partners, who, due to their own difficulties with emotions and thoughts, also engage in problematic communication in return.

When exposed to interpersonal situations, including even calm conversations about conflictual topics, individuals with BPD may negatively appraise and perceive them as interpersonally threatening, exacerbating emotion dysregulation that in turn may obstruct their ability to communicate effectively (e.g., using overly hostile or negative communications, or inability to engage with emotionally neutral problem discussion). Preliminary research indicates that in a non‐clinical sample, BPD symptoms correlated with greater use of negative communication behaviours (e.g., devaluation of partner and denial of responsibility), but were uncorrelated with the use of positive communication behaviours (e.g., problem solving and warmth; Lavner et al. ²⁰¹⁵). When asked to discuss a medium‐intensity conflictual topic, both individuals with BPD and their partners exhibited more attempts to control the conversation than healthy control couples (de Montigny‐Malenfant et al. ²⁰¹³), and women with BPD exhibited more criticism, blame, threat, non‐verbal hostility, negative mind‐reading and escalation than their partners without BPD (Miano et al. 2017). These communication difficulties may amplify conflict in the moment that it occurs, increasing interpersonal threat sensitivity and emotion dysregulation over time which, in turn, further deteriorates one's ability to communicate effectively in the future.

The BIAS model further proposes that partners of those with BPD may also engage in unhelpful communication patterns that ultimately contribute to the maintenance of BPD symptoms. Over time, partners may develop negative beliefs and patterns of cognition about their relationship, increasing their own sensitivity to emotional distress in response to the emotions of, or communications from, the individual with BPD. Consequently, they may communicate with their partner with BPD ineffectively or problematically, for example, by attempting to avoid conflict altogether and/or withdrawing from their partners in order to mitigate conflict and distress (Fitzpatrick et al. 2021). Partners and individuals with BPD may also struggle with constructive problem discussion, where the partner attempts to problem solve when empathy or emotional support would be more beneficial. Thus, both partners' ineffective communication behaviours may contribute to conflict and thus the eventual exacerbation of relationship disturbances and emotion dysregulation which transact to maintain BPD.

1.3. Analysing Relational Communication Through Observational Coding

Two studies (de Montigny‐Malenfant et al. ²⁰¹³; Miano et al. ²⁰¹⁷) have explored communication between individuals with BPD and their partners using macrocoding systems, whereby each partner is given an overall code for their communication style throughout the sample. Although macrocoding provides valuable information about the general tone of communication, it does not account for momentary variability throughout the course of the conversation. By contrast, microcoding observational measures are applied to communication samples in much smaller increments, offering more granularity and specificity than broader coding systems. One such system is the Rapid Marital Interactions Coding System‐2 (RMICS2; Heyman et al. ²⁰¹⁵), which is applied to 5‐s increments to assess communication patterns during conversation between two people. Research applying the RMICS2 demonstrates that more hostility (e.g., criticism, blaming) is correlated with lower relationship satisfaction and more distress (Heyman 2001; Heyman et al. ²⁰²²), while more positivity (e.g., warmth, taking responsibility) is correlated with higher satisfaction (Hiew et al. 2016; Williamson et al. ²⁰¹²). To summarize, using microcoding to explore communication in individuals with BPD may offer an important advancement for characterizing the amounts of hostility, positivity and constructive problem solving in communication between partners, rather than just the overall tone of their conversations. Importantly, this would offer precision with respect to elucidating which elements of communication during conflict, if any, require targeting in BPD interventions.

1.4. Present Study

The BIAS model highlights communication behaviour as a potential maintaining feature of the relationship disturbances which characterize BPD and suggests that couple communication may be a fruitful target for conjoint therapy (Fitzpatrick et al. 2021). As such, this study aimed to explore communication behaviours of individuals with BPD and their partners. The primary aim of this study was to assess for any differences in communication behaviour between individuals with BPD and their partners. We predicted that higher hostility, lower positivity and lower problem solving would be evident in individuals with BPD compared with their partners.

2. Method

2.1. Participants

Twenty‐two couples were recruited for a case series and uncontrolled trial testing Sage, a novel conjoint therapy for individuals with BPD and their partners (Fitzpatrick et al. 2023, 2024). The full study protocol is available at Clinicaltrials.gov (Identifier: NCT04737252). One partner in each dyad met DSM‐5 criteria for BPD (American Psychiatric Association ²⁰²²), based on diagnostic assessments by trained assessors who were calibrated quarterly against a gold‐standard assessor. Participants were excluded from the study on the basis of (1) severe, past‐year intimate partner violence, (2) lack of English proficiency, (3) residing outside Ontario and (4) comorbidities including clinically significant psychosis not explained by BPD, bipolar I disorder with past‐month mania or a past‐year hospitalization for mania, severe current substance use disorder or major cognitive, intellectual and/or medical impairment. Further demographic information is available in Table 1. All elements were completed virtually over the videoconferencing platform Zoom (including assessment, communication sample and intervention).

TABLE 1.

Demographic information.

	BPD (n = 18)	Partners (n = 18)
Gender
Cisgender male	1	16
Cisgender female	14	1
Transgender female	1	0
Nonbinary	2	1
Race
White/Caucasian/European origin	13	13
Aboriginal Canadian/First Nations/Métis/Inuit	1	0
South Asian	0	2
East or Southeast Asian	0	1
Non‐Mexican Latinx/Hispanic	2	0
Biracial/multiracial	2	1
Employment status
Employed	10	16
Unemployed	7	2
Prefer not to answer	1	0
Sexual orientation
Asexual	0	1
Bisexual	5	1
Gay or Lesbian	0	0
Heterosexual	11	16
Pansexual	1	0
Marital status
Married	6	6
Remarried	1	1
Never married	11	10
Unknown	0	1
Age (mean, standard deviation)	30.94 (8.33)	31.06 (8.33)

Abbreviation: BPD, individuals with borderline personality disorder.

2.2. Communication Sample

The present study includes 18 communication samples, after four couples from the clinical trial were excluded from the present analyses because their communication samples were incomplete or unusable due to technical recording issues. Communication samples were collected at the beginning of the first session of Sage. Each couple was asked to collaboratively choose a moderately distressing topic in their relationship (i.e., not too ‘hot’ or ‘cold’) and communicate about it for approximately 10 min without clinician intervention while being videorecorded. Sage clinicians turned off their own video cameras, and participants were asked to turn to each other and communicate as naturally as possible.

2.3. Measures

Communication samples were coded using the RMICS2 (Heyman et al. 2015), which involves assigning a code to each partner's communication behaviour (verbal and paraverbal) for every increment of 5 s. Codes are organized into a continuum ranging as follows: high hostility, low hostility, constructive problem discussion (CPD), low positivity and high positivity. Hostile communication includes paraverbal indicators of negative affect (e.g., tone, rolling eyes, scowling) and content including blame, criticism, negative mindreads and negative appraisals of the relationship. By contrast, positive communication includes positive affect, statements which convey acceptance, warmth, absolution or self‐disclosure (Heyman et al. 2015). High and low designations were determined by the intensity of the interactions and the extent to which they facilitated connection or discord (e.g., criticism of behaviour is low hostility, while criticism of global character is high hostility). Dysphoric affect was coded if participants exhibited negative emotionality without hostile intent or meaning, and Other was coded for off‐topic or irrelevant remarks (see Chakravarthula et al. ²⁰¹⁹ for in‐depth discussion of codes with examples).

2.4. Coders

Coding was completed by authors C.L. and A.a.d., both Caucasian, female, doctoral‐level graduate students, respectively aged 27 and 24 at time of coding. Coders completed a two‐day didactic RMICS2 training and independently coded 14 training samples until reliable with each other and a gold‐standard coder (based on Kappa ≥ 0.06 and Guilford's G ≥ 0.07). Twenty percent of communication samples were double coded to ensure interrater reliability. Kappa values ranged from 0.68 to 0.90, and Guilford's G ranged from 0.8 to 0.88.

2.5. Analyses

For each communication sample, the proportion of time spent by each participant using each communication code was calculated. Low hostility/high hostility and low positivity/high positivity were collapsed into ‘hostility’ and ‘positivity’, respectively, while dysphoric affect and other were excluded from all analyses due to low frequency in our sample. Preliminary analyses included linear regressions to assess correlations of positivity and CPD within couples (hostility was excluded due to low frequency and violation of the assumptions of statistical tests). A within‐subjects ANOVA test on trimmed means was used to assess the differences among hostility, positivity and CPD between individuals with BPD and their partners. BPD status was a within‐subjects variable, and couple was the subject unit. Yuen‐Welch t‐tests on trimmed means were used to estimate effect sizes among data cells. Trimmed methods were chosen because of their robustness against non‐normality and heteroskedasticity (Ng and Cribbie 2019; Wilcox et al. ²⁰⁰⁰).

3. Results

Partners' use of positivity was found to significantly predict individuals with BPDs' use of positivity: adjusted R ² = 0.50, F(1, 16) = 18.24, p < 0.001, β = 0.97. For every 10% increase in partner's positive responses, individuals with BPD's positive responses are estimated to increase by 9.7%. By contrast, the regression model exploring use of CPD by individuals with BPD and their partners did not reach statistical significance: adjusted R ² = 0.07, F(1,16) = 2.32, p = 0.15, β = 0.37 (Figure 2).

FIGURE 2.

Correlation of communication codes within couples. BPD, borderline personality disorder.

A within‐subjects ANOVA test on trimmed means found a significant main effect of code, F(2, 34) = 28.93, p < 0.001; no main effect of BPD status F(1, 17) = 0.217, p = 0.647; and no significant interaction between status and code, F(2, 34) = 0.547, p = 0.583 (Figure 1). Yuen‐Welch tests reveal a significantly higher proportion of CPD than hostility, with an estimated mean difference of 0.33, 95% CI: [0.23, 0.43], t(37.8) = 6.52, p = 0.000. An explanatory measure of effect size was calculated using trimmed means: ξ = 0.86, 95% CI: [0.74, 0.98], α = 0.05. The proportion of CPD was also significantly higher than positivity, with an estimated mean difference of 0.42 (95% CI: [0.33, 0.52]), t(30.63) = 9.34, p = 0.00, ξ = 0.92, 95% CI: [0.80, 0.99], α = 0.05.

FIGURE 1.

Proportion of 5‐s increments coded as hostile, constructive or positive communication for each participant. BPD, borderline personality disorder.

4. Discussion

Importantly, ours is the first study to use an observational microcoding, rather than global coding system, to assess communication between individuals with BPD and their partners. Contrary to our initial hypothesis, we found no difference in the amounts of positive, hostile or constructive communication used by individuals with BPD and their partners (Figure 1). While previous studies have reported that women with BPD used more negative communication than their partners (de Montigny‐Malenfant et al. ²⁰¹³; Miano et al. ²⁰¹⁷), this was not observed in our sample. Though microcoding systems like the RMICS2 have the disadvantage of being substantially more time and labour‐intensive, they allow for more precise and arguably more objective analyses of communication. These systems may be less sensitive to rater bias, and less likely to capture conscious or subconscious preference for one member of a couple than global rating systems (Julien et al. 1989; VanLear and Canary ²⁰¹⁵).

The finding that individuals with BPD are using various modes of communication at similar rates to their partners is especially informative given the level of stigma and discrimination they encounter, both socially and within the healthcare system (see Stiles et al. ²⁰²³, for review). Individuals with BPD may be disproportionately blamed for their relationship discord by clinicians and close others (Kyratsous and Sanati 2017). This stigma may even extend to the self—individuals with BPD often overidentify with harmfully stereotypical traits of BPD, seeing themselves as more ‘difficult’ or ‘unreasonable’ once diagnosed (Ociskova et al. 2017). In this study, we did not observe evidence that those with BPD communicate in a more conflictual or less effective way than their partners, suggesting that relationship problems in this population are truly dyadic, rather than being wholly caused by the person with BPD.

Additionally, we found a strong correlation between levels of positivity within couples, with nearly a one‐to‐one relationship in the proportions of total communication time each partner spent using positive communication (Figure 2). Our data suggest that positivity is readily received and reciprocated in individuals with BPD and their partners. By contrast, CPD was not significantly correlated within couples in our sample. The estimated slope of CPD in our sample was notably shallower than that of positivity: An increase of 10% more CPD by partners was associated with an increase of only 3.7% in individuals with BPD, compared with a respective 10% and 9.7% for positivity. As such, we see strong evidence that positivity from one partner begets positivity in the other, but the same is not necessarily true for CPD. Though we found no difference in the amount of CPD between individuals with BPD and their partners overall, our data suggest that individuals with BPD are not strongly reciprocating CPD from partners in our sample.

The lack of association of CPD within couples could arise through a number of non‐exclusive mechanisms. Given cognitive models of interpersonal threat sensitivity in BPD (e.g., Fitzpatrick et al. ²⁰²¹), individuals with BPD may be less likely to respond in kind to their partners' attempts to discuss relationship problems because they perceive even neutral, non‐critical problem discussion to be threatening. Alternatively, this finding could suggest that partners are responding to more emotional communication (i.e., positivity or hostility) with practical solutions (i.e., CPD), which may feel invalidating to those with BPD. In either case, our findings suggest a misalignment in communication, where one partner is trying to problem‐solve and the other is not. These data are in alignment with couples therapy modalities which include psychoeducation about the importance of distinguishing among two different purposes of communication: sharing one's experiences or solving problems (Fitzpatrick et al. 2023; Monson and Fredman ²⁰¹²). The Sage therapy model posits that partners of individuals with BPD often jump to solution‐oriented communication in an attempt to ease distress, which can feel invalidating and ultimately lead to heightened conflict and emotion dysregulation (Fitzpatrick et al. 2023). Thus, couples learn to ‘channel check’ before engaging in constructive problem discussion in order to ensure that they understand the other partner's needs from the interaction (i.e., simply being heard and understood versus receiving solutions or advice; Monson and Fredman ²⁰¹²).

4.1. Limitations and Future Directions

This study has several limitations to note. Primarily, our small clinical sample was not sufficiently powered to properly explore the less frequently used communication categories, including hostility and dysphoric affect. Further, we were unable to fully address the possible impact of covariates such as gender, age, ethnicity and sexual orientation due to the limited diversity of our small sample. Future investigations could also benefit from the inclusion of control couples without BPD in order to understand whether and how the communication of couples wherein one partner who has BPD differs from those who do not.

We also acknowledge that there is an inherent sense of artificiality when a couple is asked to spontaneously argue in a laboratory setting, potentially priming couples to be on their ‘best behaviour’ and resulting in less hostility and more positivity than would occur organically. Though this is a reality of all laboratory communication studies, it nevertheless hampers ecological validity. One option for future studies would be to record several communication samples over time and average the results, to capture potential changes in communication as couples become more comfortable or to capture naturalistic communication in couples' daily lives rather than in the laboratory. Importantly, the generally low levels of hostility in our sample could be indicative of conflict‐avoidant strategies that couples have adopted over time (Fitzpatrick et al. 2021), and it is possible that hostility may increase throughout the course of therapy as distressing topics are discussed more openly. As such, we intend to explore whether there are changes to communication throughout and following treatment with Sage, in order to better understand how communication relates to psychological wellbeing in this population.

4.2. Clinical Implications and Conclusions

Despite these limitations, our results clearly outline novel findings about the way individuals with BPD communicate with their intimate partners. We do not see any evidence that individuals with BPD use more or less hostile, positive or constructive communication behaviours, suggesting that individuals with BPD are not necessarily communicating in a more harmful or conflictual way than their partners. Given the complexity of this population and variability in methods and findings among studies, we acknowledge the necessity for future research to further contextualize these findings.

Our research highlights the difficulty couples have with engaging in constructive communication in a productive and meaningful way—we see that members of a couple are often misaligned in this regard and thus may benefit from psychoeducation about how to balance solution‐ and empathy‐focused communication. Taken together, these data reinforce the rationale for treating BPD in a conjoint manner, where both partners are empowered and held responsible for improving the relationship.

Disclosure

Other than approving it for funding, the funders had no role in the design of the study, data collection, data analysis or interpretation of findings.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.