In brief
- A randomized, double-blind trial compared intravenous ketamine with midazolam in 68 people with bipolar depression unresponsive to standard medications.
- At fourteen days, the ketamine group scored 7.3 points lower on the MADRS depression scale than the comparison group.
- There were no cases of mania, hypomania, or psychosis—the very theoretical risks that had previously kept these individuals out of research.
For more than a decade, almost everything we have known about ketamine as an antidepressant has come from people with unipolar depression. Those with a bipolar disorder diagnosis were typically excluded from trials. A team from the University of Toronto has just published the missing research in JAMA Psychiatry.
Why this trial was missing
The reason for the exclusion wasn’t data, but caution. There was a fear that ketamine might push a person with bipolarity into a manic or psychotic episode, something that occurs with some conventional antidepressants. As Diana K. Orsini and her colleagues write, that patient complexity and those theoretical safety concerns were enough to leave the field unexplored. The result is an uncomfortable paradox: people with bipolar depression respond worse to available treatments and, precisely for that reason, had less access to those being investigated.
How it was done and what happened
The study involved 68 people between the ages of 21 and 65 (average age, 44.4; 55.8% women) with bipolar disorder type I or II and a moderate or severe depressive episode that had already resisted at least two evidence-backed pharmacological treatments. They were randomly assigned, half and half, into two arms: four 40-minute infusions over two weeks, with ketamine at doses of 0.5 to 0.75 mg/kg in one group and low-dose midazolam in the other. Everyone also maintained their usual mood stabilizer or antipsychotic.
Choosing midazolam as a comparator is not a minor detail. An inert placebo is easily identified against a substance that produces obvious physical sensations; midazolam, which also sedates, makes the blinding more credible. Even so, 47% of participants correctly guessed which group they belonged to after the first infusion, a fact the authors acknowledge straightforwardly.
At fourteen days, the mean score on the MADRS scale was significantly lower in the ketamine group, with a difference of 7.3 points between the two arms. The type of bipolarity, I or II, had no statistical relationship with the response. Five people dropped out before the primary measurement point, one of them from the ketamine arm.
The safety section is the most significant here: neither group recorded cases of mania, hypomania, psychosis, or suicide attempts. There was one case of mixed features—hypomanic symptoms below the clinical threshold—in each arm.
What it implies
This study involved 68 people and two weeks of follow-up. The authors themselves call for phase 3 trials with larger samples and longer treatment before drawing definitive conclusions, and they point out an honest limitation: the ketamine group had more people with prior experience with the substance, which could have influenced the results. None of this makes the trial a protocol change.
What it does do is move the starting point. The argument for excluding bipolarity from ketamine research was a supposed risk that this trial, specifically designed to look for it, did not find. It fits with what we already saw in the experience of three Spanish hospitals with esketamine: real-world use is ahead of published evidence, and closing that gap benefits, above all, those who make decisions with information in hand.
Source
- Healio. Ketamine shows promise for treatment-resistant bipolar depression (on Orsini DK et al., JAMA Psychiatry, doi:10.1001/jamapsychiatry.2026.2658). Healio, September 2, 2026.
Educational content written from a harm reduction perspective and with respect for individual freedom. It is not a substitute for advice from a healthcare professional and is not intended to encourage or condemn any drug use.
