Emotional body language in borderline personality disorder : an eye tracking study

Abstract

The research on the identification of emotions in individuals with Borderline Personality Disorder (BDP) indicated that the challenges in personal relationships observed in BPD patients could be attributed to a modified interpretation of social cues. Studies that employed a facial emotion recognition task revealed that BPD patients report enhanced identification of subtle, ambiguous signs of emotions, coupled with the tendency to misattribute negative valence to neutral or low intensity emotional cues (negativity bias). The present study focuses on recognizing emotions conveyed by body postures, considering that the human bodies are essential to correctly comprehend others’ emotions and intentions. To better understand the underlying mechanism of a biased perception of emotions we used an eye-tracking technique to investigate the visual exploration patterns of emotional body postures, while participants were instructed to execute two distinct tasks. The study involved a total of 26 patients with Borderline Personality Disorder (BPD) and 26 Healthy Controls (HC), participants in each group being matched for age and gender. Participants completed two tasks that requested them to evaluate static body postures that were either emotional (Happiness, Anger, Surprise) or non-emotional (Neutral). During the Detection Task, participants were tasked with identifying whether the body posture presented conveyed any emotions. Following that, in the Valence-estimation Task, they were required to rate the valence of the postures utilizing a visual analogue scale. In both tasks we measured the performance through accuracy and reaction times. Then, we evaluated the visual exploration pattern with different parameters: the latency and the durationof the first fixation in the Detection Task and the number of fixations on salient regions (Head, Right Hand and Left Hand) in the Valence-estimation Task.In the Detection Task, the accuracy in identifying emotional or neutral stimuli was similar between BPD patients and HC; even though HC responded faster when they correctly identified both emotional and neutral postures as compared to BPD patients. Conversely, in the Valence-estimation task, BPD patients rated neutral postures as more negative and angry postures as less negative, as compared to HC. Considering the eye-tracking results, we observed that individuals with BPD showed delayed first fixation to the Left Hand of neutral bodily positions as compared to HC, indicating that BPD patients were slower in directing their attention towards significant areas of neutral stimuli. Additionally, BPD patients exhibited a longer duration of first fixation for neutral stimuli, regardless of the body's region, compared to HC. In the Valence-estimation task, BPD patients displayed a lower number of fixations to the Hands as compared with HC, irrespective of the postures' valence. This preliminary study indicated that BPD patients showed a different visual exploration pattern of body postures as compared with HC. Patients with BPD exhibited slower reflexive eye movements towards significant cues of neutral stimuli indicating potential difficulties in focusing the involuntary attention on salient regions normally useful to discriminate threatening signals. Moreover, once the attention is directed to neutral stimuli the focus of attention is captured for longer time, indicating that neutral stimuli in general seem to be more attention-grabbing for BPD patients as compared to HC. Furthermore, in the Valence estimation phase of the emotion recognition process, patients with BPD confirmed their tendency to misinterpret neutral cues as more negative (negativity bias) and they showed a low level of interest (i.e. lower number of fixations) to the hands of all the postures. This could lead to speculate that the biased later stages of emotional information processing in BPD patients might be related with difficulties in focusing visual attention on important source of emotional information.