Ketamine beat midazolam in treatment-resistant bipolar depression

Intravenous ketamine has a solid evidence base for major depressive disorder, but bipolar depression has been largely left out of that story — partly because the same drug that lifts mood can, in theory, tip a person with bipolar disorder into mania. A randomised trial in JAMA Psychiatry, published 2 September, provides some of the first controlled evidence in this population, and reports both an antidepressant benefit and, reassuringly, no manic switches [s1].

What Ket-BD tested

The Ket-BD study (Ketamine for Treatment-Resistant Bipolar Disorder) was an investigator-led, double-blind, midazolam-controlled randomised trial run at three sites in Ontario, Canada, from July 2022 to November 2025 [s1]. It enrolled adult outpatients aged 21 to 65 with a primary diagnosis of bipolar I or II disorder in a current moderate-to-severe depressive episode — a Montgomery-Åsberg Depression Rating Scale (MADRS) score of at least 21 — who had already failed at least two evidence-based drug treatments [s1].

Participants were randomised 1:1 to four flexibly dosed, 40-minute infusions over two weeks of either ketamine (0.5–0.75 mg/kg) or midazolam (0.02–0.03 mg/kg), given on top of a stable dose of at least one mood stabiliser or antipsychotic [s1]. The choice of midazolam as the comparator matters: it is a short-acting sedative that reproduces some of the immediate subjective effects of an infusion, making the placebo harder to distinguish from the drug and the trial more rigorous than a saline control [s1].

The result

Of 68 participants randomised (mean age 44.4 years; 38 female, 55.8%), 63 were included in the primary efficacy analysis, after five withdrew before the primary endpoint — one from the ketamine arm and four from the midazolam arm [s1]. On the primary outcome, change in MADRS from baseline to day 14, the ketamine group ended significantly lower: a between-group difference of −7.3 points (95% CI, −12.0 to −2.5; P = .003; Cohen d = 0.7), after controlling for sex, bipolar type and baseline severity [s1]. A Cohen d of 0.7 is a moderate-to-large effect for a psychiatric trial [s1].

On safety — the specific concern in bipolar disorder — the trial reported no cases of mania, hypomania, psychosis or suicide attempts in either group [s1]. One case of mixed features (subthreshold hypomanic symptoms) occurred in each arm [s1]. Within the bounds of a trial this size, that is an encouraging signal that adjunctive ketamine, given alongside a mood stabiliser, did not provoke the switches clinicians worry about.

The caveats the authors flag

Two limitations deserve emphasis. First, this is a small trial: 63 analysed participants is enough to detect a moderate effect but not to characterise rarer harms, and a manic switch that occurs in, say, 1 in 50 patients could easily be absent by chance here [s1]. Second, blinding was imperfect — after the first infusion, 31 of 66 participants (47%) correctly guessed their treatment allocation, which is the perennial problem with ketamine trials, since the drug's dissociative effects are hard to mask even against midazolam [s1]. When patients can tell what they received, expectation can inflate a subjective outcome like mood, so the true effect may be smaller than the point estimate.

For context, serial ketamine's antidepressant and anti-suicidal effects in unipolar depression have been reported in other recent controlled work, so a benefit in bipolar depression is biologically consistent rather than surprising [s2]. What Ket-BD supplies that was genuinely missing is controlled evidence in a bipolar population specifically — a group routinely excluded from ketamine trials precisely because of the switch concern, which has left clinicians extrapolating from unipolar data or relying on open-label experience [s1].

The trial design also answers a narrower methodological worry. Because every participant stayed on a mood stabiliser or antipsychotic throughout, the result speaks to ketamine as an add-on to standard bipolar treatment, not as a standalone — which is both how it would realistically be used and part of why no manic switches may have appeared, since the background medication is itself protective against mania [s1]. That makes the safety signal encouraging but conditional: it applies to adjunctive use under mood-stabiliser cover, and says little about ketamine given without it.

What to watch

Ket-BD is a meaningful addition because controlled data in bipolar depression are so scarce, but it is a beginning, not a verdict. The open questions are durability beyond two weeks, whether the safety picture holds in larger and longer samples, and how ketamine compares with the treatments already used for bipolar depression. Nothing here establishes a dosing regimen for routine care, and ketamine for bipolar depression remains an area where treatment decisions belong with a specialist.

This article is informational and does not constitute medical advice.

Sources

• [s1] Serial Ketamine Infusions for Treatment-Resistant Bipolar Depression: A Randomized Clinical Trial. JAMA Psychiatry, 2 September 2026. https://doi.org/10.1001/jamapsychiatry.2026.2658
• [s2] Ketamine Infusions and Rapid Reduction of Suicidal and Depressive Symptoms. JAMA Psychiatry, 6 May 2026. https://doi.org/10.1001/jamapsychiatry.2026.0612