LAUSR

The borderline construct has remained enigmatic across generations of clinicians and research traditions since its first appearance in psychiatric literature in the 20th century.1 This construct has traveled from early conceptions of the individual on the “border” of neurosis and psychosis to its current home in a DSM cluster of personality disorders characterized by an unstable negative affect and behavior that is considered unacceptable in polite society. Yet, one essential element of borderline personality disorder (BPD) that survives these changes in our descriptions and understanding of this condition is that of relational dysfunction. Certainly the early-life adversity, mood instability, and various forms of behavior that antagonize a person’s well-being, security, and occasionally survival are prevalent and somewhat distinguishing features of the condition. Nonetheless, perhaps the single clinical hallmark of BPD is the disturbance in social relations, often declaring that the patient has something other than uncomplicated depression, bipolar disorder, or posttraumatic stress. This phenomenology has been characterized as “oscillations of attachment,”2 and social attachment in BPD has been incorporated into interesting work from the perspectives of developmental cognitive neuroscience3 and even neuroeconomics.4 Nevertheless, we still have an incomplete understanding of how these disturbances work in the brain, a problem that should interest anyone other than those who are still under the sway of Descartes. A growing empirical literature describes altered function in limbic and frontal brain regions when patients with BPD process social/emotional perceptual or mnemonic information.5 These are interesting and suggestive studies, yet they take the traditional route in studying complex social-emotional phenomena with ethologically impoverished probes and patients isolated from an actual social interaction. However, a recent study in JAMA Psychiatry by Bilek and colleagues6 shows us a path forward. These investigators used a novel functional magnetic resonance imaging (fMRI) neuroimaging method called hyperscanning. This is an effective method that was introduced by Montague and colleagues7 in which a closed-circuit video connects 2 participants in realtime visual contact while each are undergoing fMRI in separate scanners. This enables investigators to cross the chasm between static 2-dimensional grayscale images of middle-aged men with 1970s sideburns and the brain function that presumably mediates the patient’s spiraling chaos in response to provocations from the other end of the telephone line. Three groups were studied, including 23 symptomatic (“current”) patients with BPD by DSM-IV criteria, 17 patients with remitted BPD (who had not met BPD criteria for at least 2 years before the study), and 20 healthy nonpatient controls (HC). Each participant (all women) was paired with a healthy participant who was not in the comparison group, and all pairs were previously unknown to each other. All but 5 participants were untreated with psychotropic medication at the study. Each pair was linked by fMRI hyperscanning during a joint visual attention task performance. This task involved 1 participant (the “sender”) sharing the information about the target location by shifting her gaze toward that location. The partner (the “receiver”) had to follow the sender’s eye-gaze direction while viewing nontargets. A trial was considered successful only if both participants brought their gaze to the correct target location. A control condition required each participant to perform the task individually while the live video stream remained active. Participants received real-time feedback on their trial performance, and task roles were switched between participants after every 20 trials of each type. The imaging analysis involved an initial contrast of a taskrelated blood oxygenation level–dependent signal change between interacting pairs vs noninteracting (randomly selected) pairs of participants, using an independent components analysis, to find those components that were driving blood oxygenation level–dependent signal variation that could be attributed to the state of social interaction. Then the degree of coupling between these components was evaluated between the different groups of interacting pairs (eg, current BPD-HC vs HC-HC and remitted BPD-HC vs HC-HC). The results indicated that the group with BPD exhibited reduced coupling with their paired HC participants in right temporal-parietal junction (rTPJ) during task performance, compared with HC-HC dyads, while neither simple task-related TPJ activity nor cortical volumes were altered in the group with BPD. The group with remitted BPD showed coupling that was not significantly different from HC-HC dyads. All groups were comparable in task performance. To my knowledge, this finding reveals for the first time the altered brain coupling of individuals with BPD during realtime social interactions. Joint attention is an important cognitive process that emerges during infancy with infantparent bonding, forming an important basis for proper social Related article Opinion