Skip to content
MyFreud

DBT skills helped teenagers most, but that was a subgroup

A trial of DBT skills training for depression with self-injury found both therapies worked. The DBT advantage appeared only once the sample was split by age.

4 min read

Pop-art illustration of teenagers sitting and talking together outdoors.

Key takeaways

  • Both therapies worked. The overall comparison between them is not where the reported advantage lies.
  • DBT skills training beat the support group among adolescents specifically, and that came from splitting the sample by age.
  • Subgroup findings in an 80-person trial are hypotheses. They are the reason to run the next study, not a conclusion from this one.
  • The brain-imaging results are the least load-bearing part. The authors note they did not track medication changes, so the scan differences cannot be cleanly attributed to the therapy.
  • What is solid and useful is that a structured skills course and a support group both reduced depression and self-injury over 13 weeks.

If you are hurting yourself, or thinking about it, please talk to a doctor or contact a crisis helpline. This article is about a research result and is not a source of urgent help.

Self-injury frequently accompanies depression, and it is one of the places where the standard depression treatments are not obviously the right tool. A trial published in BMC psychiatry in 2026 tested a therapy built specifically for that combination, and produced a result that is easy to over-read. [pubmed-depression-jun-2026-source]

What the researchers did

Eighty people with major depressive disorder and a history of non-suicidal self-injury were recruited from four hospitals and randomly assigned to one of two weekly group programmes for 13 weeks: dialectical behaviour therapy skills training, or a social support group. Everyone continued their usual medication. Assessors did not know which group anyone was in.

Alongside the questionnaires, participants had resting-state brain scans before and after.

What they found

Both programmes worked. Self-injury and depression improved in both groups over the 13 weeks. That is the trial’s clearest result and it does not favour either therapy.

The difference appeared when the sample was split by age. Among the 13 to 17 year olds, DBT skills training produced substantially larger improvements in both self-injury and depression than the support group did.

The scans changed in the DBT group. Activity in the insula, a region involved in bodily awareness and emotional processing, dropped after the skills training, and the size of the depression improvement tracked one of those changes.

Why the age finding needs holding lightly

The headline here is a subgroup analysis, and subgroup analyses are the part of a trial where enthusiasm most often outruns the evidence.

Eighty participants split across two treatments gives 40 per arm. Split those again by age and the groups being compared are small, which makes each estimate unstable. A difference that looks dramatic in a group that size can be an ordinary consequence of chance, and the more ways a dataset is divided, the more likely it is that one of the divisions produces something striking.

None of that means the finding is wrong. Age-related differences in how people respond to skills-based therapy are plausible, and DBT has a considerable independent evidence base in adolescents. It means the appropriate description is that this is worth testing directly, rather than that DBT works better for teenagers.

Why the brain imaging is the weakest part

The insula results read impressively and carry less weight than they appear to.

The authors themselves flag the reason: medication was not tracked during the intervention. If some participants changed antidepressant dose over 13 weeks, that alone could produce the scan differences, and the trial cannot separate the two explanations.

There is a more general point too. Effective psychological therapy should be visible in the brain, because thoughts and brain activity are not two separate things. Finding a neural correlate of a therapy that worked confirms the therapy worked; it does not independently confirm anything else. Imaging tends to make a result feel more scientific than the same result stated in words, and it is worth noticing when that is happening.

Would a skills-based therapy suit how you are struggling?

This is not a diagnostic tool and it does not score risk. It is a way of seeing whether the kind of therapy this trial studied matches your difficulty.

0 of 6 ticked

What to take from it

That both a structured skills course and a support group reduced depression and self-injury over three months is the useful, well-supported finding, and it is quietly encouraging. Group treatment for this problem does something.

Whether DBT skills training is specifically better for adolescents is a good question that this trial raises and does not answer.

If this is your situation now

If you are hurting yourself, tell a doctor. It is far more common than people assume and it is treatable, and there are specific therapies for it rather than just general advice. If you are having thoughts of suicide, contact your local emergency services or a crisis helpline now.

Our depression overview covers the established treatments and how to get assessed.

The source

These findings are drawn from “Insular spontaneous activity changes after dialectical behavior therapy skills training for depressed patients with non-suicidal self-injury: a randomized controlled trial” (Lei M, Wu S, Hang Y, et al., 2026), published in BMC psychiatry. Read the full study on PubMed.

Frequently asked questions

What is DBT skills training?

Dialectical behaviour therapy was developed for people who feel emotions very intensely and act on them in ways that harm them. The skills training component is the taught part, run like a course rather than a conversation: distress tolerance, emotion regulation, mindfulness and interpersonal effectiveness, each with specific techniques to practise between sessions. It is often delivered without the individual therapy that full DBT includes, which is what happened here.

What is non-suicidal self-injury?

Deliberately hurting yourself without intending to end your life, most often as a way of managing feelings that have become unbearable. It commonly travels with depression, and it is a strong reason to get help even though it is not a suicide attempt. Anyone doing it deserves support rather than a lecture about it, and effective treatments exist.

Why be cautious about a subgroup result?

Because splitting a sample creates opportunities for chance to produce a difference. With 80 participants overall, an age split leaves small groups on each side, and small groups produce unstable estimates. The finding is worth following up. It is not the same as the trial having shown that DBT skills training is better for adolescents, because the trial was not designed or sized to test that.

Do the brain scans prove the therapy changed the brain?

No, and the researchers say so themselves. Activity in the insula changed in the DBT group, and depression improvement correlated with one of those changes. But medication use was not tracked during the trial, so a drug change could explain the same scan finding. Any effective psychological treatment should in principle be visible in the brain, since experience and brain activity are not separate things, so the imaging adds interest rather than proof.

Which of the two therapies should someone ask for?

Both helped, so the honest answer is whichever is actually available. DBT skills training has a substantial evidence base for self-injury built up well beyond this trial, which is a reasonable reason to ask about it first. A well-run support group is not a consolation prize, and this study is one more piece of evidence that it does something real.

References

  1. 1.Lei M, Wu S, Hang Y, et al. ( 2026). Insular spontaneous activity changes after dialectical behavior therapy skills training for depressed patients with non-suicidal self-injury: a randomized controlled trial. BMC psychiatry. pubmed.ncbi.nlm.nih.gov . doi:10.1186/s12888-026-08254-z