Abstract

Background

Although no pharmacological treatments have received regulatory approval for borderline personality disorder (BPD), antipsychotic medications—particularly haloperidol, olanzapine, and risperidone—are frequently prescribed off-label in clinical practice. This study aims to quantitatively compare the efficacy of these three agents in the management of BPD. Moreover, considering the heterogeneous presentations of BPD among patients, the efficacy of these agents is also quantitatively compared in controlling individual symptom domains, including depression, anxiety, interpersonal challenges, cognitive impairment, impulsivity, and hostility. Due to the lack of head-tohead trials, indirect comparisons were conducted through a meta-analysis of effect sizes derived from open-label studies.

Results

A total of 21 studies met the inclusion criteria and were included in the analysis. In the overall assessment, risperidone demonstrated the greatest effect size (−1.99 [−2.34,−1.64], p<0.001), followed by haloperidol (−0.98 [−1.25,−0.72], p<0.001) and olanzapine (−0.85 [−1.24,−0.46], p<0.001). Comparative analyses of these agents across six core symptom domains are presented in the graphical abstract. Subgroup analyses indicated that olanzapine may exert greater efficacy in female patients and individuals under the age of 30, with the most favorable outcomes observed when treatment duration approaches 8–9 weeks. Overall, the certainty of evidence was low to moderate, primarily due to the limited number of available studies and heterogeneity across findings. 

Conclusions 

Risperidone emerged as the most effective agent for the overall management of BPD. Risperidone also appears to be the optimal choice for targeting anxiety and impulsivity. However, haloperidol may be preferable for patients presenting predominantly with depressive symptoms, interpersonal difficulties, and hostility. Clinicians should interpret these indirect comparison results with caution, since the certainty of the evidence is generally low to moderate. While this study used an innovative approach for symptom-specific pharmacotherapy of BPD, further high-quality, large-scale trials are needed to strengthen the certainty of evidence. This study obtained the best possible statistical conclusions despite the very limited prior investigations.

Keywords Borderline personality disorder, Antipsychotic, Olanzapine, Risperidone, Haloperidol, Meta-analysis