Ketogenic Diet for Treatment-Resistant Depression: What a New RCT Shows and How It Fits into Metabolic Psychiatry
by Kaida JIANG, Manfei XU, Wenxia ZHANG
Psychiatry rarely advances through clean, decisive turns. More often, the field moves by narrowing uncertainty, testing neglected hypotheses under stricter conditions, and learning which apparently promising ideas can withstand contact with controlled evidence. That is the frame in which a new randomized clinical trial of a ketogenic diet for treatment-resistant depression should be read. The study matters not because it settles the place of ketogenic therapy in depression care, but because it brings a line of reasoning that has often lived at the edge of psychiatry into a more rigorous clinical setting.
The trial, published in JAMA Psychiatry, enrolled 88 adults in the UK with treatment-resistant depression and compared a 6-week ketogenic diet with a carefully designed control diet. Both groups improved substantially. The ketogenic group improved somewhat more on the primary depression measure at 6 weeks, but the effect against control was modest, statistically borderline, and not clearly reinforced across secondary outcomes. The authors themselves therefore describe the clinical relevance as uncertain. That conclusion, however, should not obscure what the paper does contribute: controlled evidence that the metabolic hypothesis in depression is serious enough to test, and perhaps serious enough to keep testing.
Introduction
Treatment-resistant depression is one of the most difficult territories in clinical psychiatry. In ordinary practice, many patients with major depressive disorder do not respond adequately to first-line antidepressants. In this trial, participants were considered to have treatment-resistant depression if they had diagnosed depression, had received at least two antidepressants at adequate doses during the current episode, and still had at least moderately severe symptoms, defined as a PHQ-9 score of 15 or higher despite ongoing treatment. That is not a marginal population. It is a group for whom symptom burden remains high, functional impairment is often persistent, and treatment decisions become progressively more complex.
This matters for another reason as well. In treatment-resistant depression research, placebo effects can be substantial. The authors note meta-analytic evidence suggesting large placebo responses in this area, which is one reason uncontrolled improvement cannot be taken at face value. That is why randomized evidence carries disproportionate weight here.
Against that backdrop, the ketogenic diet paper is best understood as an evergreen question rather than a diet trend story. The enduring issue is not whether one nutritional approach has suddenly “solved” depression. It is whether at least some forms of depression may be meaningfully connected to disturbances in energy metabolism, inflammation, insulin signaling, or related biological systems, and whether shifting metabolism could have therapeutic consequences. The significance of the paper lies in how it brings that question into a better-controlled clinical trial.
What Treatment-Resistant Depression Means and Why Metabolic Psychiatry Entered The Discussion
Treatment-resistant depression is often discussed as though it were a single, stable entity, but clinically it is better understood as a difficult outcome state. A patient has persistent depressive illness despite multiple treatment attempts, yet the reasons for that persistence may differ substantially across individuals. This heterogeneity is one reason the field continues to search for interventions beyond the standard pharmacologic sequence. That search has helped create the space for metabolic psychiatry.
A ketogenic diet is a high-fat, very low-carbohydrate regimen that shifts cellular energy metabolism away from glucose and toward ketone utilization. In the paper’s rationale section, the authors describe several pathways through which ketone bodies might plausibly matter in depression: metabolic pathways, neurotransmitter systems, inflammatory signaling, and gut-brain mechanisms. None of this amounts to proof that depression is fundamentally a metabolic disease. But it does explain why the question has become more scientifically legible than it once was.
Importantly, the field did not begin with this trial. Interest in ketogenic diets for psychiatric disorders had already emerged through case reports and single-arm studies. The authors are explicit on this point: earlier studies were limited by small samples and lack of control groups.
There is also a more practical reason this trial was needed. Ketogenic diets commonly lead to weight loss, and weight loss itself can be associated with mood improvement. The paper is methodologically valuable partly because it was designed with these confounders in mind. The researchers aimed for weight stability rather than weight reduction, used a credible comparator diet, and matched the degree of support across groups. Those design choices do not eliminate uncertainty, but they make the central question sharper.
This is where metabolic psychiatry becomes more than a fashionable phrase. The ketogenic trial does not validate the entire framework. What it does is more modest and more important: it shows that a metabolic intervention can be tested under conditions serious enough to produce interpretable results.
What The Trial Actually Tested
The trial was a single-blind randomized clinical trial conducted in the UK between February 22 and June 15, 2024. It included 88 adults aged 18 to 65 years with diagnosed depression, a history of at least two adequate antidepressant trials during the current episode, and a PHQ-9 score of 15 or greater despite ongoing treatment. Participants were randomized 1:1 to the ketogenic diet group or the control group.
The ketogenic intervention was a modified ketogenic diet containing less than 30 grams of carbohydrates per day and 15–20% of energy from protein. Participants received three prepared ketogenic meals per day (or vouchers), snacks, urine ketone strips, and weekly 30-minute counseling sessions. Weight was actively managed to maintain stability.
The control condition was a phytochemical diet designed to increase intake of differently colored fruits and vegetables and replace saturated animal fats with unsaturated plant oils. Both groups received the same level of dietary support and were guided toward weight maintenance.
The intervention lasted 6 weeks, with follow-up to 12 weeks. The primary outcome was the between-group difference in change in PHQ-9 score at week 6.
What The Study Found And What Readers Should Not Overread
Depression severity decreased markedly in both groups over 6 weeks. The mean change in PHQ-9 score was -10.5 in the ketogenic group and -8.3 in the phytochemical control group. The between-group difference at 6 weeks was -2.18 points (95% CI -4.33 to -0.03, p = .05). At 12 weeks the difference was -1.85 points (p = .10) and no longer statistically significant.
At 6 weeks, 25% of participants in the ketogenic group and 9% in the control group achieved remission (PHQ-9 ≤ 4). At 12 weeks the figures were 18% and 9% respectively. The between-group difference in remission was not statistically significant.
There were no significant between-group differences on most secondary outcomes (anxiety, anhedonia, cognitive impairment, quality of life, functional impairment), except for an anxiety difference at week 12. In the per-protocol analysis, there was no statistically significant between-group difference at either time point.
In exploratory subgroup analysis, participants with more severe baseline depression (PHQ-9 20–27) showed a larger between-group difference (-4.73 points) than those with moderately severe symptoms.
No serious adverse events occurred during the trial.
What This Changes In Practice And Where The Limits Remain
The authors conclude that compared with a well-matched control diet, the ketogenic intervention showed antidepressant benefit at 6 weeks, but the clinical relevance remains uncertain because the effect size was modest and not evident in secondary analyses.
Under trial conditions the diet was feasible (high attendance and good early adherence), but durability was limited: by week 12, 48% had fully discontinued the ketogenic diet and only 20% were following it at least half the days.
The study has important limitations: modest sample size, short duration, use of urine ketones rather than blood measurement, and intensive support that included prepared meals and regular coaching. Results may not translate directly to ordinary clinical practice.
Nevertheless, this trial provides the strongest randomized evidence to date for a metabolic intervention in treatment-resistant depression. Metabolic approaches can no longer be treated as interesting only in theory or anecdote.
Conclusion
This trial is worth attention not because it offers a definitive solution to treatment-resistant depression, but because it tests a once-marginal psychiatric hypothesis under conditions serious enough to matter. A ketogenic diet outperformed a well-matched control diet on the primary outcome at 6 weeks, yet the advantage was modest, less convincing by 12 weeks, and not clearly supported by most secondary outcomes.
For now, the paper is best understood as a consequential contribution rather than a final answer. It makes the metabolic question harder to dismiss and gives the field a more disciplined basis for further research.
References
- Gao, M., Lam, T., Lahera, G., Vázquez, G. H., Price, M., Murphy, S. E., Wijeratne, A., Oke, S., Huhn, M., Comai, S., & Vieta, E. (2026). Ketogenic diet for treatment-resistant depression: A randomized clinical trial. JAMA Psychiatry. Advance online publication. Link to PDF