Mistrust in Borderline Personality Disorder

Today I’m presenting excerpts from an article available in full text. Kurt, Walker, Luyten & Fonagy (2025) published “Laboratory-Assessed Mistrust in Borderline Personality Disorder: A meta-analytic review” in Personality Disorders: Theory, Research, and Treatment. Here are highly edited excerpts:

Although a recent systematic review examined the relationship between mistrust and borderline personality disorder (BPD; Preti et al., 2023), it did not statistically quantify the strength of this association. To address this gap, we conducted a meta-analysis comparing laboratory-based assessments of mistrust between individuals with BPD and control groups. Our search yielded 1,717 reports, of which 26 met the inclusion criteria (N = 3,716). Based on 70 effect sizes from these reports, the meta-analysis indicated a small-to-moderate effect size (g = 0.44, 95% confidence interval = [0.27, 0.61], p < .001), demonstrating that individuals with BPD exhibited significantly higher levels of mistrust compared to controls. Subgroup analyses identified the mistrust paradigm as a statistically significant moderator. These results underscore the significant role of mistrust in BPD. 

Borderline personality disorder (BPD) is a multifaceted mental health condition characterized by an unstable self-image, turbulent interpersonal relationships, instability in affect, and impulsive behaviors. Prevalence estimates suggest that approximately 5.9% of the population is diagnosed with BPD, accounting for about 15%–20% of patients in both outpatient and inpatient settings. 

Individuals with BPD often experience severe interpersonal difficulties, including extreme shifts between affection and contempt and intense fears of real or perceived abandonment. Trust is essential for fostering interpersonal vulnerability, building stable relationships, and establishing genuine connections based on the expectation of positive intentions from others. In individuals with BPD, the propensity for interpersonal difficulties is partly attributed to a deep-rooted mistrust of others, contributing to unstable relationships and conflict in social situations. While mistrust is a universal human experience, it tends to be significantly more intense in people with BPD. 

Various theoretical frameworks explain the link between mistrust and BPD, consistently highlighting the perception of others as untrustworthy. Despite differing terminology, they converge on the role of temperament and early experiences in shaping these perceptions. Object relations theory emphasizes paranoid object relations, attachment theory highlights negative internal working models, schema theory addresses mistrust/abuse schemas, and cognitive theory examines mistrust beliefs. These theories suggest that in hostile early environments, mistrust may function as a protective mechanism, shaping interpersonal relationships into adulthood. Additionally, mentalizing theory posits that difficulties in understanding mental states contribute to mistrust. Deficits in mentalizing reduce epistemic trust—trust in communicated knowledge—hindering the assimilation of new information because of distorted perceptions of mental states. In BPD, this epistemic mistrust impairs the ability to recognize information as reliable, leading to rigid thought and behavior patterns. Contradictory findings in studies using laboratory-based tasks—some showing no relationship between mistrust and BPD, and others indicating that individuals with BPD are more trusting. These inconsistencies, particularly within laboratory-based paradigms, underscore the need for a meta-analysis to clarify the dynamics of mistrust measured through laboratory-based tasks.

Mistrust plays a significant role in the interpersonal difficulties observed in BPD. It also triggers core symptoms such as identity disturbance, mood instability, and impulsivity and is often linked to self-harm and suicidal behaviors following interpersonal conflict. Moreover, mistrust can undermine therapeutic alliances, leading to poorer treatment outcomes and higher dropout rates.

Interestingly, the discrepancy between self-reported and laboratory-assessed mistrust suggests that individuals with BPD may not consistently exhibit behavioral signs of mistrust, despite endorsing it in self-reports. Two studies in our meta-analysis observed mismatches between reported trust beliefs and actual trust behaviors. This points to the possibility that real-time trust decisions are less influenced by conscious beliefs and more by immediate context, with mistrust fluctuating rather than reflecting stable dispositions.

Another possible explanation for the smaller effect size is that laboratory-based measures in psychology often lack the rigorous psychometric validation typically applied to self-report instruments. For example, self-reported and behavioral measures of impulsivity show only minimal overlap. A similar pattern may explain the inconsistencies reported by Sharp et al. (2023) and Unoka et al. (2009), where self-reported mistrust did not align with laboratory-assessed mistrust. This suggests that the weaker effect sizes may reflect limitations in the measurement of behavioral mistrust rather than a genuinely weaker association.

Alternatively, self-reported mistrust may be more reflective of real-world experiences than laboratory-based assessments, serving as a more direct measure of mistrust. Laboratory tasks might not prompt participants to deeply consider their trust beliefs in relation to personality traits, potentially overlooking a stronger link between mistrust and BPD because of their less explicit and more variable nature. For example, trust games might be perceived primarily as measures of cooperative behavior rather than trust itself.

In DSM-5, transient paranoid ideation is a symptom of BPD, often involving ideas of reference, persecution, and delusions. This differs from generalized, nondelusional interpersonal mistrust, which in BPD is more pervasive than transient. The tasks in this meta-analysis assess mistrust rather than paranoia, as they do not measure delusions or severe mistrust. Instead, they capture a more stable, reality-aligned form of interpersonal mistrust. Evidence from facial trustworthiness tasks and self-report studies suggests that mistrust in BPD is a persistent trait rather than a fleeting state. The Five-Factor Model of Personality Disorders also includes “distrust” in the BPD trait profile (Mullins-Sweatt et al., 2012), reinforcing this distinction. Recognizing this could refine diagnostic criteria by accounting for persistent, yet nondelusional, mistrust in BPD.

Furthermore, our findings support mentalizing-based explanations of mistrust (Fonagy et al., 2015). Specifically, individuals with BPD perceive neutral faces as more untrustworthy than controls, reflecting hypermentalizing tendencies, where mental state inferences extend beyond facial expressions alone. This suggests that overinterpreting social cues may contribute to heightened mistrust toward others.

Studies that identified the largest disparities in mistrust between BPD and control groups primarily employed trustworthiness appraisal tasks, supporting the notion that mistrust beliefs, rather than behaviors, are prominent in BPD. Furthermore, in one study showing large disparities between groups, the tendency to perceive an unattractive partner as untrustworthy may indicate a dysfunctional assessment of mistrust in individuals with BPD, as attractiveness may not reliably predict trustworthiness.

Upon analyzing the negative effect sizes, we observed that in tasks such as the social valuation task, judge-advisor task, and irony detection task, individuals with BPD demonstrated greater trust in communicated or social information compared to controls. This challenges the epistemic mistrust hypothesis in BPD and underscores the need for further investigation, as much of the discussion has remained theoretical. One possible explanation is that individuals with BPD may not have difficulty trusting information that lacks emotional salience or personal relevance. In other words, they may struggle more with affective-level learning than with cognitive-level learning, which requires further testing through experimental designs.

Additionally, negative effect sizes from some trust games support the view of potential dysfunction in assessing cues related to trustworthiness. For example, individuals with BPD were more trusting than controls when their game partner was perceived as attractive. Furthermore, the administration of acetaminophen led to increased trust in individuals with BPD, a phenomenon the authors explained as extreme shifts from an all-bad to an all-good representation. This helps explain the unexpected finding that individuals with BPD were more trusting than HCs.

Our analysis revealed that age did not significantly moderate the relationship between mistrust and BPD. While prior research indicates that both mistrust and the severity of BPD symptoms tend to decrease with age, this effect might be balanced by an increase in negative life events or the stabilization of one’s environment over time. These opposing factors could neutralize each other, explaining why age did not emerge as a significant moderator in our study.

Similarly, gender was not a significant moderator despite evidence from previous studies suggesting a higher prevalence of mistrust and BPD in women compared to men. This discrepancy might be because of women’s greater likelihood of seeking help and engaging in social support networks. Such behaviors could weaken the association between BPD and mistrust, which may account for the lack of a significant moderating effect of sex in our findings.

Education level also did not significantly moderate the effect size. Although higher educational attainment is generally associated with lower levels of mistrust and fewer BPD symptoms, the absence of a moderating effect might result from counterbalancing influences. Enhanced coping mechanisms associated with higher education could reduce mistrust, while increased critical thinking skills might lead to greater skepticism. These conflicting factors may cancel each other out, resulting in no significant overall effect.

The direction of the trust paradigm—whether neutral, mistrust enhancing, or trust enhancing—was not a significant moderator. All conditions showed comparable differences in mistrust between individuals with BPD and control groups. This suggests that individuals with BPD may maintain a consistent level of mistrust that is resistant to change, making it challenging to rebuild trust even in trust-enhancing situations. This finding implies that even trust-enhancing cues may not be effective in psychosocial interventions, confirming the hard to treat nature of this disorder.

The similar effects observed in neutral and mistrust-enhancing conditions imply that individuals with BPD inherently expect mistrustful outcomes, reacting to neutral situations as if they were already negative. When faced with mistrust-inducing conditions (e.g., noncooperative partners), their mistrust levels may not increase significantly because such outcomes align with their expectations. This observation aligns with findings that individuals with BPD in unpredictable environments do not readily adjust their behaviors based on advice, as they anticipate unpredictability as the norm. Therefore, situations that typically heighten mistrust do not substantially alter their mistrust levels because these outcomes are not surprising to them.

Interestingly, the comparable effect sizes across trust-enhancing, neutral, and mistrust-enhancing conditions diverge from literature documenting paradoxical trust behaviors in individuals with BPD. Previous studies have suggested that mistrust-enhancing conditions can lead to increased trust, while trust-enhancing conditions may result in greater mistrust among those with BPD. In contrast, our analysis indicates that regardless of the condition, individuals with BPD exhibit higher mistrust compared to controls.

We found no significant differences between self-report and interview methods for assessing BPD. This suggests that the method of BPD assessment does not substantially influence the relationship between BPD and mistrust. 

Our findings underscore the critical importance of addressing mistrust in the treatment of BPD. Neglecting to assess and prioritize mistrust can impede the development of interpersonal trust, thereby adversely affecting epistemic trust. When new social information is perceived as nonbenign, patients may reject therapeutic interventions, significantly hindering treatment progress. Therefore, focusing on interpersonal trust is essential in clinical practice.

Several therapeutic approaches have already integrated considerations of trust issues. Mentalization-based therapy emphasizes the significance of understanding both one’s own and others’ mental states as foundational for building epistemic trust. A lack of this understanding can lead to deficits in epistemic trust, which is crucial for integrating new thoughts and behaviors into existing knowledge frameworks and fostering adaptable thinking patterns. Mentalization-based therapy, therefore, prioritizes establishing trustworthy therapeutic relationships as an initial step toward cultivating epistemic trust. Similarly, Schema Therapy (ST) aims to reform maladaptive schemas related to interpersonal relationships. It focuses on nurturing trustworthy and stable connections through therapist–client interactions to address mistrust and abuse schemas. Transference-focused psychotherapy tackles the paranoid perceptions of others in BPD by striving to integrate polarized views, moving patients toward more balanced and realistic perceptions free from excessive trust or mistrust.

The smaller effect sizes observed in studies using laboratory-based and behavioral methods, compared to self-reports, suggest that individuals with BPD may be sensitive to immediate contexts, making them more receptive to interventions targeting mistrust. However, the absence of significant differences in mistrust between neutral, mistrust-enhancing, and trust-enhancing conditions contrasts with the idea that these individuals are prone to contextual influences on their mistrust levels. This finding suggests that creating a trusting therapeutic environment may be challenging, as the similar effect sizes across conditions imply that patients might not be highly receptive to trust-enhancing cues used by therapists. Therefore, our findings indicate that the complex and dynamic nature of how individuals with BPD respond to environmental influences on mistrust warrants further investigation.

In conclusion, this meta-analysis highlights a higher level of mistrust in individuals with BPD compared to control groups. This finding underscores the significance of mistrust as a critical concern within BPD. Our analysis supports the approach of psychosocial interventions aimed at readdressing mistrust in individuals with BPD.

I thought this was very helpful in highlighting (1) the problem of mistrust in persons with BPD, (2) the circumstances in which persons with BPD trust another are complex, and (3) the fact that this makes therapy challenging.

Next
Next

Studies of disordered eating