Eye-Tracking and Borderline Personality Disorder: A Systematic Review

Background/Objectives: Borderline personality disorder (BPD) is a severe mental disorder characterised by emotion dysregulation, impulsivity and interpersonal hypersensitivity. Its prevalence ranges from 0.5% to 6.4%. Eye tracking and pupillometry provide objective indices of social attention and inhibitory control, but the BPD literature using these techniques has not been systematically reviewed. The aim of this work was to synthesise the empirical evidence on visuo-attentional and pupillary alterations in BPD. Methods: Following the PRISMA 2020 statement, Web of Science, Scopus and PubMed were searched up to 13 March 2026, with no date or language restrictions. Search terms combined borderline personality disorder and eye-tracking constructs. Two reviewers independently screened records with complete inter-rater agreement at the title-and-abstract stage (Cohen's κ = 1.00); two generative artificial-intelligence assistants (ChatGPT, NotebookLM) were additionally consulted as a non-systematic plausibility check and returned no eligible studies beyond the database search. Risk of bias was appraised with the framework appropriate to each study design (RoB 2 for randomised trials and Newcastle-Ottawa Scale logic for observational studies, with ROBINS-I held in reserve for non-randomised intervention designs). Results: Seventeen studies met the inclusion criteria, with sample sizes ranging from 19 to 164 participants and predominantly adult female samples. Designs included antisaccade and oculomotor tasks, free-viewing, dot-probe, affective priming and pharmacological challenge. Four findings recurred across studies. First, patients with BPD showed an early reflexive vigilance to the eye region of emotional and neutral faces, followed by reduced time on positive stimuli during longer presentations. Second, self-reported impulsivity was elevated, but laboratory inhibition was largely preserved; the deficits that did emerge were limited to preparatory control and were greater in patients with comorbid ADHD or under induced negative affect. Third, autonomic dysregulation was indexed by lower heart-rate variability and a larger baseline pupil size; in a single longitudinal study, pupillary reactivity was prospectively associated with subsequent symptom change. Finally, intranasal oxytocin reduced amygdala-driven vigilance. Conclusions: Eye-tracking and pupillometric measures appear to capture meaningful aspects of the BPD clinical picture. The two-stage profile of early vigilance followed by reduced sustained engagement is most parsimoniously described as a vigilance-avoidance pattern, which is compatible with, but not uniquely explained by, the hypersensitivity hypothesis of emotion dysregulation. Because thirteen of the seventeen studies recruited women only, these conclusions apply primarily to adult women with BPD. Methodological heterogeneity, the predominance of female samples and the scarcity of longitudinal data justify the need for standardised protocols, transdiagnostic comparisons and the inclusion of male and gender-diverse populations in future research.