Mark Chen
A new clinical trial is putting a genuinely provocative question on the table: could changing what you eat actually help when antidepressants have already failed you? For anyone who’s cycled through multiple medications and still feels stuck, that question isn’t academic. It’s personal. And the research that dropped in early 2026 is the most credible evidence yet that the answer might be yes, at least for some people.
In February 2026, JAMA Psychiatry published a randomized clinical trial out of Oxford University that tested a ketogenic diet against a plant-based control diet in 88 adults with treatment-resistant depression. The keto group showed a 10.5-point improvement on a 27-point depression scale over six weeks. The plant-based group improved too, by 8.3 points. That’s a real but modest difference, and lead researcher Min Gao concluded the results “suggest ketogenic diets may be effective as an adjunctive treatment for treatment-resistant depression.” A concurrent systematic review and meta-analysis published in JAMA Psychiatry in January 2026, from researchers at St. Michael’s Hospital and the University of Toronto, pulled data across 10 randomized controlled trials and found a significant association between ketogenic eating and reduced depressive symptoms, with a standardized mean difference of -0.48. That’s a small-to-medium effect. Not nothing. Not a cure.
What surprised me was how quickly a JAMA Psychiatry correspondence in July 2026 pushed back, scrutinizing both the Oxford trial’s effect size and its methodology. The scientific community is not letting this one slide into the mainstream quietly.
- The Oxford RCT showed a 10.5-point vs. 8.3-point depression scale improvement: keto beat plant-based, but modestly.
- A 10-RCT meta-analysis found keto linked to reduced depressive symptoms (SMD -0.48), strongest when ketosis was verified.
- The meta-analysis found no consistent anxiety benefit; larger trials are still needed.
- Benefits appear strongest as an add-on to medication, not a replacement for it.
- JAMA correspondence published July 2026 is still actively debating the trial's clinical meaning.
Why Treatment-Resistant Depression Is the Right Place to Look
About 30% of people with major depression don’t respond adequately to antidepressants. This is the population researchers call “treatment-resistant,” and it’s enormous. These aren’t people who gave up after one bad experience. Many have tried three, four, five different medications. The standard next steps, things like augmentation strategies, switching drug classes, or adding psychotherapy, help some people but leave a lot of others still struggling. That’s the gap researchers are trying to fill, and it’s why a dietary intervention is worth taking seriously even if the effect sizes feel small. A 2.2-point additional benefit on a 27-point scale might not look dramatic in a table, but to someone who has spent two years on medications that barely moved the needle, it can matter.
The Oxford trial was notable for keeping participants on their existing psychiatric medications throughout. This wasn’t keto versus antidepressants. It was keto added to antidepressants. That distinction shapes everything about how to interpret the results.
What the Numbers Actually Show
I’ll be honest: the effect sizes here are real but not enormous, and the head-to-head comparison is worth seeing clearly.
| Measure | Keto Group | Plant-Based Control |
|---|---|---|
| Depression scale improvement (6 weeks) | 10.5 points | 8.3 points |
| Scale used (max points) | 27-point scale | 27-point scale |
| Trial size | 44 participants | 44 participants |
| Meta-analysis SMD for depression | -0.48 (significant) | N/A |
| Meta-analysis finding for anxiety | No consistent benefit | No consistent benefit |
The meta-analysis finding on anxiety is worth pausing on. The research is mixed there, and the Toronto/St. Michael’s team was clear: keto didn’t reliably move the needle on anxiety symptoms across the 10 trials they reviewed. If anxiety is your primary struggle, the evidence base here isn’t supporting much optimism yet. Larger, longer trials are needed.
The meta-analysis also found that effects were strongest when ketosis was biochemically verified, meaning participants who were actually in ketosis showed more improvement than those who reported following the diet but weren’t confirmed to be in ketosis. That’s a practically important detail. It suggests compliance and actual metabolic state matter, not just the intention to eat low-carb.
How Keto Might Be Affecting the Brain
The honest answer is that researchers are still working this out. Several mechanisms are being investigated. Ketone bodies, especially beta-hydroxybutyrate, appear to have anti-inflammatory effects and may influence neurotransmitter balance, particularly GABA and glutamate. There’s also evidence that ketosis reduces neuroinflammation, which is increasingly being studied as a factor in depression. Mitochondrial function is another area of interest. None of this is settled science, but it’s not speculation either. It’s active investigation.
Stanford Medicine published guidance in April 2025 on ketogenic therapy and serious mental illness, signaling that at least some major academic medical centers are taking this seriously enough to advise clinicians on it. UCSD launched a clinical trial in 2026 specifically testing therapeutic ketogenic diets for eating disorders including anorexia nervosa and bulimia nervosa. The research interest is widening.
What This Means If You’re Considering It
This is where I want to be direct with you. The trial was six weeks long, in 88 people, and a JAMA correspondence published in July 2026 is still picking apart its methodology. That is not a foundation for overhauling your psychiatric care on your own. Anyone managing treatment-resistant depression is, by definition, dealing with a medically complex situation, and dietary changes that induce ketosis can affect medication levels, electrolytes, and energy metabolism in ways that need monitoring.
What the research does support is a genuine conversation with your psychiatrist or a dietitian who understands therapeutic ketogenic protocols. This is emerging enough that many clinicians haven’t engaged with the 2026 literature yet. Bringing it to the appointment is reasonable. Treating it as a proven therapy is premature. The distinction matters.
If you do pursue a ketogenic diet in this context, the meta-analysis data suggest that actually achieving verified ketosis, not just reducing carbs somewhat, is where the potential benefit seems to live. That means stricter carbohydrate restriction, consistent monitoring, and support from someone who knows both the diet and the psychiatric complexity.
The 2026 data is the most credible signal yet that diet-brain connections in depression are worth taking seriously. It’s also early. Staying curious without getting ahead of the evidence is the right posture here. That balance is genuinely hard to hold, but it’s what the research actually earns right now.
Sources
- Keto Diet a Potential Treatment for Depression, Trial Shows (February 6, 2026)
- A Ketogenic Diet for Treatment-Resistant Depression: A Randomized Clinical Trial (February 4, 2026)
- Ketogenic Diets and Depression and Anxiety: A Systematic Review and Meta-Analysis (January 1, 2026)
- A Closer Look at Ketogenic Diet Effects in Depression (July 2026)
- Ketogenic Diets May Improve Depression Symptoms (June 17, 2026)
Photo: MART PRODUCTION via Pexels
This article is for general informational purposes only and does not constitute medical or dietary advice. Always consult a licensed healthcare provider or registered dietitian before making significant changes to your diet, especially if you have a medical condition.
