Background: Previous studies suggested that childhood trauma is an important etiologic factor for the development of borderline personality disorder (BPD). Moreover, insecure attachment and maladaptive emotion regulation (ER) might be related to childhood trauma and BPD. This study was aimed to explore the relationships among childhood trauma, insecure attachment, maladaptive ER, and BPD features. Methods: A cohort of 637 patients with psychological disorders completed a series of psychometric instruments such as the Personality Diagnostic Questionnaire-4+ (PDQ-4+), the 23-Item Borderline Symptom List, the Childhood Trauma Questionnaire, the Attachment Style Questionnaire, and the Cognitive Emotion Regulation Questionnaire. The path analyses were conducted to investigate the experience-driven model that whether insecure attachment and maladaptive ER could mediate the relationship between childhood trauma and BPD features. The random forest regression was performed to select variables that contribute significantly to BPD features, which variables would be incorporated into the datadriven model to further confirm the experience-driven model. Results: Both the experience-driven model and the data-driven model verified that there were three significant mediation pathways (childhood trauma → insecure attachment/maladaptive ER → BPD features, childhood trauma → insecure attachment → maladaptive ER → BPD features; all p < .05), and the most weighted mediation pathway by which childhood trauma influencing the BPD features was through insecure attachment and then through maladaptive ER (weighted 53.16%). Conclusion: The influence of childhood trauma on BPD features was mainly mediated by the combination of insecure attachment and maladaptive emotion regulation.

In this study, combining degree centrality (DC) and fractional amplitude of low frequency fluctuation (fALFF) analyses of resting state (rs)-functional magnetic resonance imaging (fMRI) data, we aimed to explore functional connectivity density, local brain spontaneous activity, and their coupling strengths in borderline personality disorder (BPD). Forty-three BPD patients and 39 demographically-matched controls underwent rs-fMRI after completing a series of psychological tests. Two-sample t-tests were performed to compare DC and fALFF between these two groups. Across-voxel correlation analysis was conducted to assess DC-fALFF coupling strengths in each group. Imaging parameters and psychological variables were correlated by Pearson correlation analysis in the BPD group. Altered DC and fALFF values in the BPD group, compared with the control group, were distributed mainly in default mode network (DMN), and DC-fALFF coupling strengths were decreased in the left middle temporal gyrus (MTG) and right precuneus in the BPD group. Additionally, insecure attachment scores correlated positively with left precuneus DC and negatively with fALFF of the right posterior cingulate cortex (PCC) in the BPD group. These altered DC and fALFF findings indicate that the BPD patients had disturbed functional connectivity density and local spontaneous activity in the DMN compared with control subjects. Their decreased connectivity-amplitude coupling suggests that the left MTG and right precuneus may be functional impairment hubs in BPD. Disturbed rs function in the left precuneus and right PCC might underlie insecure attachment in BPD.