Review finds few reliable predictors of who responds to ketamine for depression

The question

Can doctors predict, from a person’s characteristics or history, who is most likely to feel better after ketamine or esketamine treatment for depression that has not responded to other treatments?

What they did

Researchers searched four research databases up to November 2025 for any study reporting clinical or demographic factors linked to how well ketamine or esketamine worked in treatment-resistant depression. They included 122 studies covering 12,674 people, published between August 2006 and September 2025, made up of 75 distinct samples totalling 6,902 people. Of these, 65 studies gave open-label ketamine (everyone knew what they were getting), 47 were secondary analyses of earlier trials, and 10 were randomized controlled trials. The team assessed each study for risk of bias and recorded whether each possible predictor showed a positive, negative or no link with response.

What they found

Most studies (62) included both unipolar and bipolar treatment-resistant depression, and 61 used intravenous ketamine at a fixed dose of 0.5 mg/kg. The reviewers looked at 77 different possible predictors, including body mass index, dissociation (a sense of detachment during the infusion) and past brain-stimulation treatments. For most of these variables, more studies reported no link with antidepressant outcomes than reported a positive or negative link. Two factors stood out as worth more study: responding early in the treatment course, and having a family history of substance use disorder.

The fine print

The review pooled a mix of study designs, and most of the evidence came from open-label treatment or re-analyses of existing trial data rather than randomized comparisons, so apparent predictors could reflect chance or expectation. Results were not combined into a single statistical estimate, and counting how many studies point one way is a crude measure. Non-English studies were excluded. This summary is based on the abstract only.

What it means

At present, there is no reliable way to tell in advance which people with treatment-resistant depression will respond to ketamine or esketamine, based on common clinical or demographic characteristics.

Evidence grade: Low. A careful systematic review, but most of the 122 included studies were open-label or secondary analyses rather than randomized trials, and results were not pooled statistically.

This summary is based on the study’s published abstract.