Borderline personality disorder (BPD) is one of the most damaging mental illnesses. By itself, this severe mental illness accounts for up to 10 percent of patients in psychiatric care and 20 percent of those who have to be hospitalized. The defining characteristic of BPD is a pervasive instability in the patient’s life, especially when it comes to interpersonal relationships. BPD patients also have difficulty controlling their impulses and regulating their emotions. Close relationships of patients are often tumultuous and compromised by highly unpredictable behavior that can leave others baffled, angry and frightened. This behavior exerts a tremendous toll not only on those afflicted with the illness, but also on their social network and the health care system. (A well known, if dramatized, example is the character Glenn Close played in the movie Fatal Attraction.) Surprisingly, despite the importance of this disorder, little is yet known about what brain mechanisms might underlie it. In a recent paper in the journal Science, Brooks King-Casas and colleagues at Baylor College of Medicine provide an illuminating look into the brain in people suffering from BPD.

In this study, patients and healthy controls played a game in which money is exchanged between an investor, who decides how much to invest, and a trustee, who decides how much of the investment, which is tripled during the transfer, to repay. For example, if the investor decides to invest $10, then the trustee has $30 to divide ($10 x 3). Although this game is at first glance about money, it is really about the development of trust. If both players cooperate, both benefit from the exchange, much more so than if the investor keeps most of the money for him- or herself.

This trade requires a degree of trust between the players that is built up through repeated fair offers, however. An investor who does not trust the other player will not put in much money. This small offering is exactly what happened at the end of games with BPD trustees, indicating that they had difficulties establishing and maintaining cooperative relationships. In contrast, players without BPD (they were the healthy “control” subjects) still had high levels of at the end of the game. They accomplished such amounts through a “coaxing” strategy, in which wary investors transferring small amounts of money were encouraged by generous returns, which signal trustworthiness. The study found that healthy players used this strategy twice as often as BPD subjects.

Investing in Trust

To find out why patients behaved this way, the researchers used neuroimaging (fMRI) to study brain activation of trustees confronted with a small investment, which usually signals a lack of trust on the investor side. Although most healthy trustees would respond to such a move with the coaxing strategy—is a demonstration of their own trustworthiness—patients with BPD did not. Furthermore, the fMRI scans revealed a crucial difference between BPD subjects and healthy players. In healthy subjects, a region of the brain called the anterior insula seemed to neurally represent the investment level, so that small investments corresponded to large activation and vice versa. In BPD patients, however, no such correlation existed. As expected from previous work, this same brain area also represented the amount of money subjects were about to repay to the investor, so that a large activation in the insula predicted a small payment from the trustee. Strikingly, however, this correlation was now the case in both subjects with BPD and healthy controls. In other words, although healthy controls had insula activations both to “distrustful” offers from investors and “stingy” repayments, subjects with BPD represented only their own actions. Their impairment seemed to selectively concern the representation of the other player.

The anterior insula has long been associated with the representation of uncomfortable bodily sensations, such as pain. In addition, many studies have since shown that this area also strongly reacts to adverse or uncomfortable occurrences in social interactions, such as unfairness, excessive risk, frustration or impending loss of social status. This body of work suggests that the anterior insula tracks information about the intentions and behavior of others and colors them with a feeling of discomfort. If true, then one reason BPD subjects may be impaired in maintaining cooperation is because they lack the “gut feeling” (corresponding to the anterior insula signal) that there is a problem with the relationship. Because they can’t detect the breakdown of trust, they are less likely to trust others at all.

A Network Problem?

This exciting finding prompts many new questions. The first is, What causes this abnormal brain activity? Most research indicates that BPD commonly arises from a combination of genetic predisposition and severe early child trauma. Not everyone traumatized as a child develops BPD, but it could be that a combination of risk genes makes the impact of trauma on the developing brain more severe and enduring. It would be of high interest to determine whether such genetic variants, which are beginning to be identified, compromise the anterior insula structure and function. Because no brain region operates in isolation, it will also be relevant to fully characterize the brain network of which the insula is a part.

Beyond the domain of BPD, the innovative approach used in this study of a personality disorder can also be applied to other severe mental illnesses in which social dysfunction is a prominent source of disability, such as schizophrenia or autism. Such an advance should be warmly welcomed by patients, their therapists and researchers.


Mind Matters is edited by Jonah Lehrer, the science writer behind the blog The Frontal Cortex  and the book Proust Was a Neuroscientist.