does better mentalizing in mbt make people stop self-harming and stop having suicidal thoughts?
“They stop self-harming, they don't have suicidal thoughts, they don't have big emotional thoughts with people or themselves and symptoms get better.”
a guest, a clinical psychologist said on The Carlat Psychiatry Podcast, 2026-10-05, at 37:44. hear it in context →
the episode: Treating BPD Series, Episode 4: Mentalization-Based Treatment—From Reacting to Reflecting with Carla Sharp, PhD (at 37:44). we link the show's own page; we do not re-host audio.
first, the part that holds
the direction is right. a 2020 cochrane review found mentalization-based treatment did better than usual care on self-harm and suicidality at the end of treatment, and people in mbt programmes do report much less self-harm over time.
what we found when we went and looked
we read the 2020 cochrane review of psychotherapies for borderline personality disorder. compared with usual care, mbt cut the risk of self-harm by about a third (rr 0.62, 3 trials, 252 people) and lowered suicidality (rr 0.10, 3 trials, 218 people), both rated low-quality evidence. so self-harm went down but did not stop. a 2024 meta-analysis of 14 mbt studies found large before-and-after drops in self-harm, but in randomized comparisons mbt was not better than control treatments for self-harm. the authors concluded that putting mbt first for self-harm "is not supported". the mechanism in the claim, that better mentalizing is what produces these gains, has not been tested properly: a 2021 systematic review found no study had run a proper mediation analysis.
assertion by assertion
| assertion | verdict | what the record says |
|---|---|---|
| patients in mbt stop self-harming | overstated | cochrane 2020: mbt vs usual care, self-harm rr 0.62 (95% ci 0.49 to 0.80; 3 trials, 252 participants), low-quality evidence. that is about a 38% relative reduction at the end of treatment, not an end to self-harm. the 2024 meta-analysis found a large pre-post drop (g = -0.82) but no advantage over control treatments in randomized trials. [1][2] |
| patients in mbt stop having suicidal thoughts | overstated | cochrane 2020: mbt vs usual care, suicidality rr 0.10 (95% ci 0.04 to 0.30; 3 trials, 218 participants), low-quality evidence, measured at the end of treatment. this is a large relative effect from only three small trials. it shows fewer patients with the outcome, not that suicidal thinking disappears. [1] |
| these gains come from improved mentalizing capacity | unverifiable | a 2021 systematic review of 10 studies found that none ran a state-of-the-art mediator analysis, so the mediating role of mentalization "remain[s] without empirical support". none of those studies examined mbt. a 2024 naturalistic mbt study (n = 46) found mentalizing and symptoms both improved, but the authors say they "cannot make causal claims" because there was no control group. [3][4] |
| popular framing: mbt is the treatment of choice for self-harm because it beats other therapies | overstated | the 2024 meta-analysis of 14 studies (612 participants) found that mbt and its adolescent version were not more effective than control interventions for self-harm. the only advantage it found was on borderline symptoms in adults. the authors concluded that prioritizing mbt for self-harm "is not supported". [2] |
both directions
for the speaker: the cochrane review found effects in favour of mbt over usual care on self-harm (rr 0.62) and suicidality (rr 0.10) at the end of treatment, and it named mbt as one of the two most-studied therapies for borderline personality disorder. before-and-after data are strong: the 2024 meta-analysis pooled a large drop in self-harm (g = -0.82), and a 2024 naturalistic mbt study found mentalizing, psychiatric symptoms and personality functioning all improved together over 18 months. a clinician who watches patients through mbt will see many of them stop self-harming, and that observation fits these data.
against: "stop" goes further than the numbers. a 38% relative drop in self-harm means many patients still self-harm. cochrane rates both the self-harm and the suicidality findings low-quality, based on three small trials each. compared with active control treatments rather than usual care, mbt has not been shown to beat them on self-harm. the link the claim depends on, that improved mentalizing causes these changes, has not been tested with proper mediation analysis. in the studies that exist, mentalizing and symptoms improve at the same time, which is correlation and does not show cause. the claim also covers the speaker's experience in the clinic, and no trial data can confirm or rule that out.
what this cannot say
the trials are small, mostly in adults with borderline personality disorder, and measure outcomes at the end of treatment, so they say little about other diagnoses or long-term follow-up. "suicidality" in the cochrane analysis is a dichotomous trial outcome and may not match "suicidal thoughts" in everyday speech. we did not read the full text of every included trial. we relied on abstracts and review summaries for the effect sizes. we found no mbt trial that directly tested whether gains in mentalizing explain the drop in self-harm. that study may exist and we missed it. the speaker named no source, so we cannot tell what evidence they had in mind.
sources, each one fetched and read
- Cochrane Database of Systematic Reviews, Psychological therapies for people with borderline personality disorder, 2020, CD012955.pub2, PMID 32368793 What we read there: MBT appears to be more effective than TAU at reducing self-harm (RR 0.62, 95% CI 0.49 to 0.80; 3 trials, 252 participants) and suicidality (RR 0.10, 95% CI 0.04 to 0.30; 3 trials, 218 participants); effects for MBT on self-harm and suicidality at end of treatment 'were all based on low-quality evidence.' Fetched 2026-10-30 (HTTP 203). https://pubmed.ncbi.nlm.nih.gov/32368793/
- Efficacy of mentalization-based therapy in treating self-harm: a systematic review and meta-analysis, 2024, PMID 38279664 What we read there: 14 studies, 612 participants; pre-post self-harm g = -0.82 (95% CI -1.15 to -0.50); compared with control interventions MBT(-A) 'did not prove to be more efficacious, with the exception of MBT showing superior effects on BPD symptoms in adults'; 'prioritizing the application of MBT (-A) for the treatment of self-harm is not supported.' Fetched 2026-10-30 (HTTP 203). https://pubmed.ncbi.nlm.nih.gov/38279664/
- Systematic review on mentalization as key factor in psychotherapy, International Journal of Environmental Research and Public Health, 2021;18(17):9161 What we read there: 10 studies; 'none of the included studies conducted a state-of-the-art mediator analysis'; 'all assumptions and postulates on the mediating role of mentalization remain without empirical support'; MBT was not examined in any included study. Fetched 2026-10-30 (HTTP 200). https://pmc.ncbi.nlm.nih.gov/articles/PMC8431324/
- Mentalization based treatment for a broad range of personality disorders: a naturalistic study, BMC Psychiatry, 2024 What we read there: n = 46, uncontrolled, assessed at baseline, 6, 12 and 18 months; mentalizing improved over time alongside symptom improvement (OQ-45.2 d = 1.00); 'As the study is not experimental in design, we cannot make causal claims.' Fetched 2026-10-30 (HTTP 200). https://pmc.ncbi.nlm.nih.gov/articles/PMC11157867/
speakers are described by role, not named. one sentence is quoted under fair use, with attribution and a link to the original at the timestamp; we do not re-host audio or video. nothing on this page is a reason to start, stop or change a medication. if you have the file behind anything marked unverifiable, the research board is where people can answer. adam