Prevention of self harm in adolescents

BMJ 2011; 342 doi: http://dx.doi.org/10.1136/bmj.d592 (Published 07 April 2011) Cite this as: BMJ 2011;342:d592
  1. David Brent, academic chief, child and adolescent psychiatry
  1. 1Western Psychiatric Institute and Clinic, University of Pittsburgh School of Medicine. Pittsburgh, PA 15213, USA
  1. brentda{at}upmc.edu

Interventions should be tailored to individual risk factors and the social context

The linked randomised trial (Assessment of Treatment in Suicidal Teenagers; ASSIST; doi:10.1136/bmj.d682) is the largest clinical trial so far to target self harm in adolescents, including non-suicidal self injury and suicide attempts.1 Despite a promising pilot study and excellent trial management, when added to routine care the intervention of developmental group therapy did not significantly reduce the occurrence, frequency, or severity of self harm. Possible explanations for this lack of effect are that self harm is too heterogeneous for any one intervention to be effective, the effects of routine care overwhelmed those of the experimental treatment, or the treatment targeted individual rather than contextual factors.

Each group received nine or 10 sessions of routine care, much more than either condition received in the pilot study. As the authors note, the increased amount of treatment and possible improvement in routine treatment over the past decade may have overwhelmed the effect of developmental group therapy, which was essentially compared with minimal treatment in the original study.

ASSIST included adolescents who engaged in non-suicidal self injury and suicidal behaviour. Although non-suicidal self injury and suicide attempts often occur in the same individual and share some common risk factors, their motivations, reinforcers, and neurobiology are distinct. Non-suicidal self injury is most commonly used as a mood regulation strategy and is associated with higher pain thresholds, lower opioid activity, and supersensitivity to the µ opioid receptor.2 3 Non-suicidal self injury, which is thought to relieve negative affect through the release of endogenous opioids, is highly reinforcing, and this raises the question of how useful opioid antagonists might be in preventing it. One treatment may therefore not be effective for both types of disorder. For example, dialectic behavioural therapy, one of the treatments on which developmental group therapy is based, decreased suicidal behaviour but not non-suicidal self injury when compared with expert community care.4

Developmental group therapy targets the full range of problems that adolescents with either form of self harm might have, including depression, substance use, conduct problems, abuse, and peer and parental conflict in an average of only 10 sessions.1 It is possible that participants in ASSIST may not have received enough of a “dose” of any one intervention to result in change.

In adults who attempt suicide, dialectic behavioural therapy and cognitive behavioural therapy have been shown to reduce the rate of re-attempts.4 5 These two very different interventions both have focused models of suicidal behaviour, which aim to improve regulation of emotion and combat negative thoughts that lead to suicidal behaviour, respectively. Both treatments use chain analysis, which details the sequence of events, thoughts, feelings, behaviours, and context that led up to an episode of self harm, in order to develop a safety plan and to choose and prioritise strategies to help patients resist their urges to self harm.

Developmental group therapy recognises that many of these adolescents encounter social adversity, which it attempts to buffer through building their problem solving and emotion regulating skills. Sometimes these skills may not be enough to counteract a toxic social environment. For example, parental depression, family discord, and a history of abuse have been shown to wipe out the effectiveness of cognitive behavioural therapy for the treatment of adolescent depression.6 7 Therefore, sometimes direct intervention with the parent, family, or social system may be more effective than an exclusive focus on building skills. However, a randomised controlled trial of home based family treatment for adolescents who attempted suicide also showed no significant effects, so perhaps an exclusive emphasis on either individual or contextual factors will not meet the needs of many suicidal adolescents.8

Treatment studies of self harm and depression in adolescents have almost entirely focused on the remediation of emotional and cognitive weaknesses, rather than on the enhancement of personal and family resources that promote emotional health. Risky health behaviours, including self harm, are less likely to occur in the presence of a strong parent-child bond, consistent parental supervision and discipline, and a positive connection between the adolescent and the school.9 Interventions that augment family and individual resilience by improving the parent-child relationship can protect against mental health disorders and dysfunction up to six years after the intervention is delivered.10

We have not yet figured out how to protect adolescents from self harm. Treatments that have the best chance of success may be those that are based on a simple sharply focused model of suicidal behaviour that differentiates between non-suicidal self injury and suicide attempts, that personalise interventions on the basis of a chain analysis, and that focus on building protective factors within the person’s social context. Given the emerging biology of non-suicidal self injury, future intervention studies should target the opioid system with opioid antagonists.

If psychotherapy is “the art of wooing nature,” then self harm behaviour in adolescents is a condition that has spurned all suitors.11 12 Auden described this art as the ability to facilitate healing despite human variability; “all humans have prejudices of their own that can’t be foreseen.”12 If these unforeseen “prejudices,” that predispose to self harm, such as low tolerance of distress or hopelessness, could be identified they could be used to personalise treatment.


Cite this as: BMJ 2011;342:d592


  • Research, doi:10.1136/bmj.d682
  • Competing interests: The author has completed the Unified Competing Interest form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declares: no support from any organisation for the submitted work; DB has received grants from NIMH and royalties from Guilford Press; DB is an editor for UptoDate.

  • Provenance and peer review: Commissioned; not externally peer reviewed.