CHPR: Please begin by telling us what’s meant by nonsuicidal self-injury.

Dr. Dixon-Gordon: Nonsuicidal self-injury (NSSI) refers to self-injury that results in tissue damage, done on purpose for reasons not socially sanctioned—so, not tattooing, not piercing—and without any suicidal intent. Typical behaviors include cutting or scratching or hitting oneself in a way that results in a bruise, head banging, and self-strangulation. But if someone engages in self-injury and thinks “This could kill me, and I kind of hope it does,” then that would be suicidal behavior, not NSSI.

CHPR: On our inpatient unit, sometimes these self-injurious behaviors, like self-strangulation with bedsheets, are done directly in front of staff. Does attention-seeking explain at least some self-injurious behaviors? 

Dr. Dixon-Gordon: It’s a misconception to think people engage in self-injurious behaviors for attention. Most of the time, it’s not manipulative behavior. In fact, I avoid the term “manipulative” because frankly, we are all manipulating our environment all the time to get our needs met. These people are often just not very good at it. They are not influencing others effectively to get the support they need, and that’s why they end up in this painful spot of self-harm being their best tool to self-soothe. 

CHPR: Right. I think many people in the mental health field use the term “manipulative” as a way of expressing countertransference reactions of frustration and exhaustion when working with these patients. 

Dr. Dixon-Gordon: That’s understandable. Here we are trying to help people get better, so it can be frustrating to work with individuals who deliberately hurt themselves. It’s important to keep in mind that, most of the time, people report they harm themselves to feel better. 

CHPR: How do self-injurious behaviors help them feel better? 

Dr. Dixon-Gordon: There’s some evidence that the opioid system plays a role in helping individuals feel better from self-injury. Interestingly, people with borderline personality disorder (BPD) are not only more likely to engage in self-injury, but they are also more likely to gravitate toward opiates rather than other drugs (Sansone RA et al, Int J Psychiatry Med 2008;38:217–226; New AS and Stanley B, Am J Psychiatry 2010;167(8):882–885). Of course, there are lots of ways of managing distress. If their other coping strategies aren’t working, individuals can use substances, they can binge eat, they can binge watch television, they can engage in risky sex. But for some people, self-injury is ego syntonic. For example, if somebody is highly self-critical or experiences a lot of shame, then it becomes more understandable why they would want to cause themselves harm.

CHPR: What groups are at greatest risk of NSSI?

Dr. Dixon-Gordon: Individuals with BPD are at high risk. After all, self-harm is one of the criteria for BPD, and 75%–80% of people with BPD engage in self-injury (Goodman M et al, Pers Ment Health 2017;11:157–163). But there is a lot of self-harm across a range of psychiatric diagnoses, like depression and PTSD, as well as in the context of no psychiatric diagnosis (Klonsky ED et al, Am J Psychiatry 2003;160(8):150–158). That’s especially the case for adolescent emerging adult populations.

CHPR: How common is NSSI among adolescents? 

Dr. Dixon-Gordon: They are perhaps the highest risk category. In general, about 17% of adolescents report some lifetime engagement in NSSI, but that range is up to around 40% in some samples depending on how NSSI is measured (Swannell SV et al, Suicide Life Threat Behav 2014;44(3):273–303). And we see a decrease in NSSI over people’s lifespans, with NSSI occurring on average in about 13% of young adults and about 5% of adults (Swannell et al, 2014). But prevalence rates of self-injurious behaviors range widely depending on how you ask the question.

CHPR: How do you ask questions about NSSI? 

Dr. Dixon-Gordon: Researchers often ask about NSSI through questionnaires or brief interviews. A question might sound like “Have you ever purposely hurt yourself without wanting to die, like cutting or burning yourself?” The more different strategies you mention, the more people are likely to respond “Oh, yes, I have done that.” How you define self-harm really matters, like whether you differentiate self-injury without any suicidal intent from suicidal behavior. Or if someone reports they hit themselves and it hurt, but there was no bruise, does that count as NSSI? Self-report versus interview-based methods are going to differ. We do know there’s been an increase in self-injury over time in the general population, particularly during the pandemic (Zetterqvist M et al, Psychiatry Res 2021;305:114208). And there are regional effects, as self-injury seems to be somewhat of a contagious behavior (Jarvi S et al, Arch Suicide Res 2013;17(1):1–19).

CHPR: Wow. Can you say more about that?

Dr. Dixon-Gordon: A lot of people report their first instance of self-injury as “Oh, I heard of a friend who did this,” or “I saw a show where they did this, and I thought maybe that could work for me.” There does seem to be some clustering, and NSSI seems to come and go in various regions.

CHPR: Are there gender differences?

Dr. Dixon-Gordon: Yes. Self-injury in women is the more prototypical skin cutting, while among males, it more often involves hitting oneself (Gratz KM et al, Personal Disord 2012;3(1):39–54). 

CHPR: What are the differences between NSSI and suicide attempts?

Dr. Dixon-Gordon: NSSI is more commonly associated with an immediate improvement in mood. Also, the methods used in NSSI are typically different than those used in suicide attempts. Someone engaging in NSSI might cut themselves, whereas suicidal behavior usually involves medication overdose or other more potentially lethal means. And yet, there is a huge association between NSSI and suicide risk. Adolescents with either NSSI histories or suicidal behavior histories are twice as likely to engage in the other behavior (Groschwitz RC et al, Psychiatry Res 2015;228(3):454–461). Self-injury without suicidal intent is perhaps the most robust risk factor for future suicide attempts other than a past suicide attempt.

CHPR: So, we must take NSSI incidents very seriously, almost at the level of a past suicide attempt.

Dr. Dixon-Gordon: And that’s tricky because a lot of our patients engage in this behavior all the time. For us to treat that as an acute risk is difficult.

CHPR: How do you approach these patients? What treatment strategies can we use?

Dr. Dixon-Gordon: There are several treatments that have shown efficacy for reducing nonsuicidal self-injurious behaviors. A few pharmacotherapies have been examined and show some promise for NSSI. For instance, one case-controlled study looked at the efficacy of naltrexone (50 mg daily), which targets the opioid system, and found it reduced instances of NSSI (Sonne S et al, J Nerv Ment Dis 1996;184(3):192–195). Selective serotonin reuptake inhibitors have also shown promise, with or without cognitive behavior therapy (Turner BJ et al, Can J Psychiatry 2014;59(11):576–585). But pharmacotherapeutic strategies for NSSI are still very poorly studied. The most well studied and efficacious are dialectical behavior therapy (DBT), especially for patients with BPD and self-injury, and emotion regulation group therapy.

CHPR: Please review the key points of DBT.

Dr. Dixon-Gordon: DBT is a comprehensive cognitive behavior treatment that incorporates acceptance-based principles and mindfulness practice. It includes individual therapy, group therapy, between-session phone coaching, and a consultation team for therapists. One element of DBT that I think makes it really useful for reducing self-injury is that we directly target self-injury. We treat it as a potentially life-threatening behavior because we know it is one of the most robust risk factors for suicidal behavior. So, when patients start DBT, we talk about how they’ve used self-harm to feel better, and how, paradoxically, NSSI has been what kept them alive because it was their one coping tool. It was their life raft in the ocean. At the same time, I tell them they’re not going to get anywhere by floating around in the middle of the ocean on a life raft. They must give that up and get on the boat. So, we track NSSI on a diary card, as well as patients’ urges to engage in self-injury. We do a deep dive to try to understand the behavior’s antecedents, the consequences, the factors maintaining this behavior, alternative behaviors they could do to achieve the same function, and how to block anything that’s potentially reinforcing.

CHPR: You also mentioned emotion regulation group therapy. Can you describe that?

Dr. Dixon-Gordon: That draws on many of the same principles as DBT, but it’s a 14-week group-based treatment. It’s quite a bit more efficient, and there’s some research showing its effectiveness, but not yet as much as for DBT (Sahlin H et al, BMJ Open 2017;7(10):e016220).

CHPR: Do you think something like that could work on an inpatient unit if the patient stays two to three weeks?

Dr. Dixon-Gordon: I don’t see why not. You have the whole milieu to work with you in monitoring and reinforcing the behavior, so there are some real advantages to being with people every day and being able to support them in developing adaptive coping strategies while on an inpatient unit.

CHPR: On our inpatient unit, we use a rewards-based token economy system where patients can get stars on their star card for demonstrating healthy behaviors, and they lose their star card for the day if they engage in assaultive or self-injurious behavior. For our patient population, this program has been very helpful in reducing self-injurious behaviors.

Dr. Dixon-Gordon: I think that is a phenomenal system. There is a bit of a risk that you might get a “what the hell” effect if someone has self-harmed once, because they might feel “Oh well, I’ll just keep self-harming for the rest of today because I’ve lost my star card for the day anyway.” But I do think that having immediate reinforcement of alternative behaviors is very important. So, you might ask “What strategies can you use when you feel really bad that don’t involve hurting yourself?” And then you give stars when they engage in those alternative behaviors.

CHPR: That’s a good suggestion.

Dr. Dixon-Gordon: And it’s great that you have a rewards-based system. Punishment-based approaches don’t usually reduce the behavior; they just reduce the publicity of the behavior. So, people might end up harming themselves in secret.

CHPR: Right. Is there anything else we should know before we wrap up?

Dr. Dixon-Gordon: I always remind providers that people resort to self-injurious behavior because they’re in a lot of pain and don’t have other tools in their toolbox. It’s important to treat them in a nonjudgmental way and try to understand the factors that keep that behavior going. It can be really stressful for providers to work with these patients, especially since it may seem like they are contributing to their own pain, and they are often high utilizers of EDs and inpatient units. It’s hard to watch when someone you’re trying to help self-harms, but it helps me personally to remember that they are doing the best they can to survive their pain. A big part of what makes DBT effective is that it provides support groups for the therapists and helps them to work on holding onto a nonjudgmental stance toward themselves and their patients. 

CHPR: Thank you for your time, Dr. Dixon-Gordon.