BackgroundFear of abandonment and aloneness play a key role in the clinical understanding interpersonal and attachment-specific problems in patients with borderline personality disorder (BPD) and has been investigated in previous functional Magnet Resonance Imaging (fMRI) studies. The aim of the present study was to examine how different aspects of attachment representations are processed in BPD, by using for the first time an fMRI attachment paradigm including personalized core sentences from the participants’ own attachment stories. We hypothesized that BPD patients would show increased functional involvement of limbic brain regions associated with fear and pain (e.g., the amygdala and the anterior cingulate cortex) when presented personalized attachment relevant stimuli representing loneliness compared to healthy controls (HC).MethodsWe examined the attachment classifications of 26 female BPD patients and 26 female HC using the Adult Attachment Projective Picture System (AAP). We used an fMRI-adapted attachment paradigm to investigate the neural correlates of attachment. All participants were presented three personalized (vs. neutral) sentences extracted from their AAP attachment narrative, combined with standardized AAP pictures representing being alone (monadic) or in interactive (dyadic) attachment situations.ResultsAs expected, the classification of unresolved attachment was significantly greater in BPD compared to HC. BPD patients showed increased fMRI-activation in brain areas associated with fear, pain, and hyperarousal than HC when presented with personalized attachment-relevant alone stimuli. In particular, pictures with monadic attachment situations induced greater anterior medial cingulate cortex, anterior insula, amygdala, thalamus and superior temporal gyrus activation in the patient group.ConclusionThe results point to increased fMRI-activation in areas processing emotional distress and painful experiences in BPD patients. In particular, the emotional cascade reflecting attachment distress was evoked by combining monadic pictures, representing abandonment and aloneness, with the patients’ personalized narrative material. Our results confirmed and replicated previous results that illustrate once again the high relevance of aloneness and feelings of abandonment for BPD in the context of attachment trauma. Moreover, our results support the hypothesis of hypermentalization in response to attachment distress as a core feature of social-cognitive impairment in BPD associated with common treatment implications across different therapeutic orientations.