What makes young people cut, scratch, carve or burn their skin, hit or punch themselves, or even bang their heads against a wall?
For years, psychologists theorized that such self-harming behaviors helped to regulate these sufferers' negative emotions. If a person is feeling bad, angry, upset, anxious or depressed and lacks a better way to express it, self-injury may fill that role.
Also known as non-suicidal self-injury, or NSSI, the condition is classified in the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (2013) as a "new disorder in need of further study," as well as a symptom of borderline personality disorder, which is marked by such tendencies as emotional instability, unstable relationships and chronic feelings of emptiness. (See companion article on who is more likely to self-injure.)
For treatment, practitioners often turn to dialectical behavior therapy because of its effectiveness in treating borderline personality disorder in general. But now, two researchers at Harvard University are taking a different tack on this inquiry. Joseph Franklin, PhD, a postdoctoral fellow in the laboratory of suicide researcher and psychologist Matthew Nock, PhD, wanted to step back from earlier assumptions about self-injury and think more broadly about what might be driving self-harming behavior.
Although researchers generally accepted that NSSI helped people deal with strong negative emotions, Franklin wanted to understand more about why that was the case. Around the same time, Jill Hooley, DPhil, who heads Harvard's experimental psychopathology and clinical psychology program, was finding that people who engaged in NSSI would endure experimentally induced pain for much longer than control participants, and she wondered why. What factors kept self-injurers from apparently acting in their best interests and avoiding pain? And could identifying those factors help explain why people were motivated to engage in NSSI in the first place?
"We were both interested in knowing why people hurt themselves," says Hooley, "but Joe was focused on the possible [emotional] benefits of NSSI and I was looking at the motivations behind it."
While over time their interests and investigations have converged, when they first discovered each other's work, "we were like, ‘Oh, this makes so much sense!'" Hooley adds. "This is how the various findings fit together. It was very exciting."
Why self-injury?
Franklin started his investigation with one of the central questions in the field: Why would people report feeling better after hurting themselves?
"It seemed very strange to me, and to a lot of other people," says Franklin. Most studies on the topic had relied on self-reports, "but I wanted to look at this experimentally and biologically to see if it was really true."
In a 2010 study in the Journal of Abnormal Psychology, Franklin and colleagues used a task that measured people's defensive eye-blink responses before and after they dipped their hands into ice-cold water. The results indicated that self-injurers do in fact feel better afterward, he found.
A second finding was more surprising.
"Everyone else reported feeling better, too," he says.
That is, healthy controls showed exactly the same degree of physiological defensiveness and subsequent physiological relief as those who engaged in self-injury. In a 2013 paper in Clinical Psychological Science, Franklin's team replicated the finding and also showed that most people had equivalent changes in positive emotions in response to shocking stimuli.
Franklin then turned to the pain literature to see if he could gain more understanding. There, he discovered something described by psychologists 70 years ago: a phenomenon called pain offset relief. According to this concept, virtually everyone experiences an unpleasant physical reaction to a painful stimulus. Removing the stimulus does not return the individual to their pre-stimulus state, however. Rather, it leads them into a short but intense state of euphoria.
Using a technique called pain offset relief conditioning, those scientists also found that if you paired the pain with a stimulus, over time, people would react more favorably to the pain because they had learned to associate it with pain relief. For example, when researchers shocked rats and then presented them with a pleasant odor, over time, the rats began seeking out the smell.
Testing this paradigm with various types of shocks and physiological measures, Franklin continued to find powerful pain-offset relief effects in all of his participants, self-injurers and controls alike. People who self-injure may unwittingly be tapping into this mechanism, Franklin surmises. The first time they hurt themselves, they experience unpleasant pain. But when they keep doing it and experience pain relief, they begin to associate cutting or other forms of self-injury with relief, and they return for more.
"That is contrary to what a lot of people assumed and what a lot of treatments focus on," Franklin says. "People, including myself, were thinking there was something unique about these folks" — that if they were injuring themselves repeatedly, they must perceive or experience pain differently from others.
"But that doesn't appear to be the case at all," he says. "It looks like it's this natural phenomenon that [people who self-injure] happen to be tapping into."
Harmonic convergence
The next research question may seem obvious: If anyone can experience pain relief by inflicting pain on themselves, why don't more of us do it? To ask the question a different way, why do self-injurers harm themselves instead of opting for healthier or more pleasant ways of relieving emotional pain, such as watching a movie, meeting a friend or going to a yoga class?
As Franklin was pondering this question in light of possible benefits to the self-injurer, Hooley and her team wondered about psychological explanations. She already knew that self-injurers would endure physical pain for longer, but why? Was this increased pain endurance linked to some of the psychological factors that are commonly associated with self-injury — depression, hopelessness or dissociation, for instance?
Somewhat surprisingly, her team found no significant associations. So having interviewed all of her self-injuring research participants in detail, she returned to her notes in search of clues.
That's when one factor stood out: How often they spontaneously described themselves as being "bad," "defective" or "deserving of punishment."
"It was as if harming themselves or experiencing pain was somehow congruent with their highly negative self-image," she explains.
To test this possibility, her team developed a measure that specifically assesses self-beliefs about being "bad" and deserving criticism. This time, they found an answer: The higher a person's score on negative self-beliefs, the longer they were willing or able to endure pain.
Given his conversations with Hooley, Franklin was thinking along similar lines. When he asked himself why people would undertake this behavior, he looked at it in context of the fact that most people probably like themselves and therefore don't want to hurt themselves. In ongoing, still unpublished work, he asked participants to rate words like "me," "myself" and "I" on a 10-point scale ranging from most unpleasant to most pleasant. Most people rated themselves between a seven and eight, but self-injurers gave themselves only a two or a three.
Likewise, Franklin reasoned that most people would not be overly fond of stimuli that depict blood, wounds, knives or equivalent images. But he surmised that people who self-injure might feel differently, partly because his findings suggested they would associate such images with pain relief. A 2014 study in Clinical Psychological Science shows this is the case: People who had engaged in NSSI over the past year or who had 10 or more lifetime episodes of self-cutting were much less likely to report aversion to these kinds of stimuli than non-injuring controls.
Meanwhile, Hooley has recently completed a neuroimaging study looking at how people process such stimuli.
"We're predicting that images of self-injury will activate reward-processing areas in the brains of people who engage in NSSI," she says, "but not in non-self-injuring controls."
Treatment targets
Now, based on some of these findings, the researchers and their colleagues are exploring new ways to treat self-injury.
In keeping with her discovery that self-worth is an important intervention variable, Hooley is taking a cognitive route. Clinicians might be able to lift self-injurers' inclination to do bad things to themselves, she thinks, by helping them change what are often deeply engrained negative self-views.
In a 2014 paper in Clinical Psychological Science opens in new window, Hooley and then-graduate student Sarah A. St. Germain report on the results of a five-minute intervention they developed that seeks to change beliefs about self-worth, but in ways that are believable, grounded in reality and relatively subtle.
"You can't do this by simply telling people who self-injure that they should think more positively about themselves," she comments.
The team asked people with NSSI and controls to choose which positive characteristics on a list best described themselves, and then to elaborate on some of those characteristics in detail using examples. Before and after the intervention, the researchers tested people's pain endurance by recording how long they kept their fingers clamped in a pressure device. (To control for noncognitive effects of good mood, the researchers also included a "happy music" condition with NSSI and healthy controls.)
After the intervention, self-injurers in the cognitive condition kept their fingers in the device for only about half the time they did initially. And the more their sense of self-worth increased, the less willing they were to stay in the painful situation. Meanwhile, the pain endurance of those in all other conditions remained the same. She has not yet looked at effect duration.
"The natural and adaptive response is to say, ‘I'm done with this.' But people who engage in self-injury don't necessarily see pain as something to escape from," Hooley explains. Instead, experiencing pain validates their sense of being a bad or damaged person.
Increasing their sense of self-worth may undermine this propensity, she adds.
"The more valuable that people feel, the less willing they are to endure a bad situation," she says. "Conversely, the worse people feel about themselves, the more inclined they will be to try [painful] methods of mood regulation that most other people would not even consider." Franklin, meanwhile, is using his findings to test a behavioral intervention using novel counter-conditioning techniques. The intervention targets self-injurers' propensity to pair pain images with relief, and to harbor negative associations to self-related words. The study is under review, but preliminary evidence suggests that changing people's feelings about self-related words and NSSI-related images can be effective treatments, Franklin says.
Studies also show that associating pain with relief and negative thoughts with self-related words are good predictors of future self-injury, suggesting such tasks could be used as early screening tools for NSSI, report Franklin and colleagues in a 2014 article in the Journal of Abnormal Psychology . opens in new window
These efforts are a step toward effectively treating people who self-injure, and Hooley and Franklin hope that psychologists who create and study interventions will add their own insights to the mix.
"We hope that our research is improving our knowledge about how self-injury works and providing novel treatment targets," Franklin says. "Our goal is for researchers and clinicians to use their creativity to attack these targets in new ways."
