People with borderline personality disorder’s (BPD’s) intimate relationships are associated with dysfunction, communication problems, and dissatisfaction [1, 2]. The Borderline Interpersonal-Affective Systems (BIAS) model suggests that BPD is maintained through transactions between people with BPD’s and their significant others’ (SOs) dysregulated emotions and communication, and SOs may also inadvertently reinforce destructive behavior in BPD. Including SOs in treatment may therefore optimize BPD interventions by targeting each member’s cognitive, emotional, and communication processes, and the transaction between them. Conjoint interventions may also target SOs’ mental health problems [1].

No manualized interventions target BPD, relationship problems, and SO mental health simultaneously. As a result, our team developed Sage [3], a manualized psychotherapy delivered conjointly to people with BPD and SOs to target BPD, relationship functioning, and SO mental health. Sage is outlined in detail elsewhere [3] but, in brief, is a 12-session intervention that targets the relational and emotional maintenance factors of BPD outlined by the BIAS model [4]. Phase 1 provides BPD psychoeducation and skills to mitigate safety concerns and relationship conflict. For example, these skills include learning how to monitor oneself for signs of escalating distress; effectively disengaging from the conflict (i.e., calling a “time out”); using strategies to decrease distress (e.g., paced breathing); and returning to the conversation when emotions are regulated and key areas of focus for the conflict are refined. Phase 2 teaches dyadic emotion regulation and effective communication skills. Phase 3 focuses on cognitions that influence emotion dysregulation and relationship dysfunction, dyadic strategies to challenge them, and relapse prevention. Most sessions are 75-min and weekly, although the first two sessions are 90-min within the same week where possible to support safety planning.

This manuscript describes a proof-of-concept case series of five individuals with BPD and their partners who received Sage from study investigators or supervised doctoral-level clinical psychology students. The purpose of this case series was to gather preliminary evidence regarding whether Sage is acceptable and can improve BPD symptoms (primary outcome), relationship conflict and SO mental health (secondary outcomes), and other relevant tertiary outcomes.

Method

Participants

Five adult intimate dyads participated, wherein one partner (1) met DSM-5 BPD criteria [5]; and (2) had elevated suicidal ideation (≥ 15 on the Beck Scale for Suicidal Ideation [6]) or chronic and recent suicidal or non-suicidal self-injury (2 + acts in the past five years with 1 + in the past eight weeks; e.g., [7]). Exclusion criteria included (1) severe, past-year intimate partner violence; (2) lack of English proficiency; (3) residing outside Ontario; and (4) 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.

Measures

All measures are described in Table 1. BPD, exclusion criteria, and comorbidities were assessed with diagnostic interviews. Assessors were calibrated quarterly against a gold-standard rater. Primary outcomes were participants with BPD’s BPD symptoms (self- and partner informant-report), suicidal/self-injurious behavior, and suicidal ideation (self- and partner informant-report). Secondary outcomes involved self-reports of conflict from both members, and partners’ reports of their emotion dysregulation, shame, depression, anxiety, positive affect, negative affect, and anger/hostility. Participants with BPD also provided informant-ratings of partners’ emotion dysregulation. Tertiary outcomes included participants with BPD’s self-reports of secondary outcomes, partners’ informant-report of their emotion dysregulation, and both members’ self-report of relationship satisfaction and intervention satisfaction. This case series was pre-registered (Clinicaltrials.gov Identifier: NCT04737252).

Table 1 Domains assessed, measures used, and measure citations
Full size table

Procedures

Study procedures received research ethics approval. After an online screening, prospective participants with BPD and SOs completed eligibility assessments. Couples were administered outcome assessments at baseline, after session 6,Footnote 1 and at the end of Sage.

Sage therapists who were not intervention developers learned the intervention from reading the manual and watching and discussing Sage intervention session recordings. Graduate students conducted co-therapy for their first case (e.g., two student therapists present instead of one; this occurred in one instance in the case series). Sage therapists met with study investigators for weekly group supervision which included review of session recordings.

Data analytic strategy

Jacobson and Truax’s [22] Reliable Change (RC) indices were calculated for each outcome to classify responses as improved, worsened, or unchanged.Footnote 2 Reliability and standard deviation estimates used to calculate RC for each measure were obtained from studies using samples of people with BPD or related problems (e.g., inpatient psychiatric samples).

Results

See Table 2 for sample demographic and clinical characteristics. See Table 3 for means, standard deviations, and RC thresholds for each outcome. See Table 4 for reliable improvement, worsening, or no change outcomes for each measure by couple. Four of five couples completed the intervention, and one dropped out of the intervention after session 10 without providing post-intervention data. This couple is excluded from RC analyses. Primary and secondary outcomes are described for each case below.

Table 2 Demographic and current comorbid information
Full size table
Table 3 Pre, mid-, and post-intervention means and standard deviations for study measures for people with BPD and their partners
Full size table
Table 4 Reliable improvement, worsening, or no change results by dyad
Full size table

Participant with BPD #1 exhibited pre- to post-Sage improvement in BPD severity, suicidal ideation, conflict, emotion dysregulation, anxiety, emotional reactivity, and relationship satisfaction. They exhibited no change in depression, shame, positive emotion, negative emotion, and anger. Partner #1 exhibited improvement in conflict and informant-reported emotion dysregulation, worsening in depression, shame, and positive emotion, and no change in self-reported emotion dysregulation, anxiety, emotional reactivity, negative emotion, anger, and relationship satisfaction.

Participant with BPD #2 exhibited pre- to post-Sage improvement in BPD severity, suicidal ideation, conflict, self-reported emotion dysregulation, depression, anxiety, shame, positive emotion, negative emotion, and anger. They exhibited no change in informant-reported emotion dysregulation, emotional reactivity, or relationship satisfaction. Partner #2 exhibited improvement in conflict, emotion dysregulation, emotional reactivity, and negative emotion, and no change in depression, anxiety, shame, positive emotion, anger, and relationship satisfaction.

Participant with BPD #3 showed improvement in BPD severity, suicidal ideation, shame, and emotional reactivity from pre- to post-Sage. They exhibited no change in conflict, emotion dysregulation, depression, anxiety, negative emotion, and anger, and worsening in positive emotion and relationship satisfaction. Partner #3 exhibited improvement in self-reported emotion dysregulation and positive emotion, no change in conflict, informant-reported emotion dysregulation, depression, anxiety, shame, emotional reactivity, negative emotion, and anger, and worsening in relationship satisfaction.

Participant with BPD #4 exhibited no change in informant-reported BPD severity, suicidal ideation, and emotion dysregulation, but they self-reported worsening in each of these domains along with conflict and relationship satisfaction. Finally, they exhibited no change in depression, anxiety, shame, emotional reactivity, positive emotion, negative emotion, and anger. Partner #4 exhibited worsening in conflict, self-reported emotion dysregulation, shame, and relationship satisfaction, and no change in informant-reported emotion dysregulation, depression, anxiety, emotional reactivity, and positive emotion, negative emotion, and anger.

Across couples, intervention satisfaction was high for participants with BPD and partners (Mean = 29.75 out of 32 for both; SD = 2.87 and 2.50, respectively). RC in the frequency of suicidal/self-injurious acts in the past month was not computed because this index does not have Cronbach alphas. However, on average across the four participants with post-intervention data, the frequency of suicidal and self-injurious behaviors in the past-month decreased from baseline (Mean = 3.25, SD = 5.25) to post-Sage (Mean = 1.25; SD = 1.5).

Discussion

Results provide preliminary evidence that Sage is a promising brief conjoint intervention for participants with BPD and partners. Couples found the intervention highly acceptable, with four out of five completing it and the fifth coming close to completion. Three of four participants with BPD and their partners agreed that there were improvements in BPD severity and suicidal ideation, and the average frequency of suicidal/self-injurious behaviors across participants showed a reduction from pre- to post-Sage. Moreover, three of four participants with BPD improved in other mental health symptoms.

Although three of four couples exhibited largely positive outcomes, one couple exhibited no change or poor outcomes which accounted for almost all instances of worsening. Sage may not have been beneficial or possibly even iatrogenic for this couple. This couple may have experienced a considerable stressor during the post-assessment period, resulting in the post-assessment capturing acute but temporary relational distress. Alternatively, Sage may have raised awareness of significant relationship issues for this couple, increasing their post-assessment distress. Further testing is needed to understand who may and may not be good candidates for Sage, or whether these outcomes would be sustained at a follow-up.

Secondary and tertiary outcomes were generally positive but less consistent than primary outcomes. Conflict improved in half of the couples, did not change in one couple, and worsened in the couple discussed above. Similarly, half of partners reported improvements in mental health outcomes. Ceiling and floor effects may explain the more limited improvement in these domains. Average relationship satisfaction remained well above the clinical threshold (M = 104; [20]) across all assessments, and partner baseline mental health problems were relatively low, which may have limited the detection of change. However, it is also possible that the benefits of partner involvement in Sage are specific to BPD outcomes.

We are unable to identify meaningful patterns in outcomes in the absence of a control group, a larger sample, and follow up assessment. Sample diversity was also limited mainly to white, heterosexual couples with female-identifying participants with BPD who were, on average, young and early in their relationship. Greater demographic variability is needed in future work. However, our preliminary findings are encouraging and provide proof-of-concept that Sage may have a positive impact on symptoms of BPD as well as some partner and relationship outcomes.