Emma Lawson
If you follow the keto world at all, you’ve probably seen the headlines this week. A peer-reviewed trial just found that a ketogenic diet significantly reduced core symptoms of anorexia nervosa in adult women, with 72% of completers scoring in a “recovered or normal” range by the end of the intervention. The reaction in clinical nutrition circles has been, to put it mildly, a lot. Anorexia nervosa is one of the most treatment-resistant psychiatric conditions we know of, and it carries the highest mortality rate of any eating disorder. The idea that a high-fat diet, of all things, could move the needle is genuinely surprising, and it deserves a careful look rather than either breathless enthusiasm or reflexive dismissal.
What the Trial Actually Did (and Didn’t Do)
| Aspect | Trial Details |
|---|---|
| Lead Researcher | Dr. Guido Frank, UC San Diego School of Medicine |
| Publication | Nature’s Communications Medicine, June 2026 |
| Sample Size | 22 women aged 18-45 |
| Inclusion Criteria | History of anorexia nervosa, BMI >17.5 |
| Intervention Duration | 14 weeks |
| Macronutrient Target | 70% fat, 20% protein, 10% carbohydrates |
| Weight Goal | Maintenance (no caloric restriction) |
| Completion Rate | 18 of 22 (82%) |
| Primary Outcome | 72% of completers in “recovered or normal” range |
| Statistical Significance | p < 0.001 (intent-to-treat analysis) |
| Follow-Up Trial | NCT06000774 (60 participants, partial hospital and intensive outpatient) |
The study, led by Dr. Guido Frank at UC San Diego School of Medicine and published in Nature’s Communications Medicine in June 2026, enrolled 22 women aged 18 to 45 who had a history of anorexia nervosa and a BMI above 17.5. That last detail matters. These weren’t acutely medically compromised patients at dangerously low body weights. The intervention was a 14-week ketogenic plan targeting 70% fat, 20% protein, and 10% carbohydrates.
Here’s the part that cuts against the obvious worry: the explicit goal was to maintain body weight, not lose it. Participants weren’t being asked to restrict calories. They were being asked to change fuel source. Eighteen of 22 participants (82%) completed the trial, which is a solid completion rate for any dietary intervention, let alone one in a population that has historically struggled to stay in treatment. The intent-to-treat analysis showed significant overall treatment effects, with a p-value below 0.001, and symptom scores declined across both eating disorder-specific and general psychopathology measures.
What the trial didn’t do: it had no control group, it was small, and the participants were mostly higher-functioning adults who could manage intensive outpatient participation. The authors are clear about these limitations, explicitly calling for larger, more diverse follow-up trials including metabolic PET imaging of brain glucose metabolism. This was a feasibility study. It tells us the approach is worth pursuing seriously, not that keto is a treatment for anorexia nervosa.
Why Keto? The Brain Metabolism Hypothesis
The mechanism being proposed here is more interesting than “keto suppresses appetite,” which would be entirely the wrong effect for this population anyway. The hypothesis, based on Frank’s broader research into the neurobiology of anorexia, centers on brain glucose metabolism. People with anorexia nervosa show altered glucose uptake in areas of the brain involved in reward processing, anxiety regulation, and body image perception. The thinking is that ketone bodies may provide an alternative energy substrate that bypasses some of that disrupted glucose signaling.
I’ve worked with clients long enough to know that the metabolic effects of ketosis aren’t just about weight. People consistently report changes in mood stability, anxiety, and cognitive clarity when they’re in sustained nutritional ketosis, and those effects are real even if we don’t fully understand the mechanisms. Whether those same mechanisms are therapeutic in a clinical eating disorder context is a different and much harder question. But it’s a scientifically coherent one. The fact that UCSD has already launched a larger follow-up trial (NCT06000774) targeting 60 anorexia nervosa participants across partial hospital and intensive outpatient settings, with a separate arm for bulimia nervosa, suggests the research team sees enough signal here to invest serious resources.
The Counterintuitive Part Deserves Honest Attention
What most people don’t realize, especially those outside clinical eating disorder work, is how few effective treatments actually exist for anorexia nervosa. Cognitive behavioral therapy helps some patients. Medication has limited evidence. Family-based treatment works reasonably well in adolescents but is harder to implement in adults. The treatment landscape is genuinely thin. So when a small feasibility trial shows a p-value below 0.001 and a 72% recovery-range response rate, clinicians pay attention even with all the caveats about sample size and study design.
The counterintuitive tension here is real, though. Prescribing a diet to someone with a restrictive eating disorder, even a high-fat, weight-maintaining one, raises legitimate clinical concerns. Food rules are often central to the disorder. Adding a new set of macronutrient targets could, in some patients, reinforce rigid food thinking rather than challenge it. The trial authors acknowledge this, and it’s one reason the follow-up study involves more structured clinical settings where patients are monitored closely. This is not something to try at home with a macro calculator and a YouTube video.
What This Means for Clinicians and Patients Right Now
If you or someone you know is currently in treatment for anorexia nervosa, this trial is not actionable yet. What it is: a genuinely promising research direction that may eventually expand the treatment toolkit for a condition that desperately needs more options. The appropriate response is curiosity, not a dietary change made outside of supervised clinical care.
For dietitians and clinicians following this research, a few things stand out. The weight-maintenance framing of the ketogenic intervention is critical and should be central to any future clinical implementation. The population selected, adults with BMI above 17.5 and the capacity to engage in outpatient treatment, was relatively stable. Applying this outside those parameters would be a significant clinical leap that the current evidence doesn’t support. And as with all ketogenic dietary therapy in clinical populations, medical supervision isn’t optional. It’s the entire point.
I’ve seen keto get oversold plenty of times, for conditions where the evidence is thin and the hype is thick. This isn’t one of those cases. The research team here is rigorous, the mechanistic hypothesis is grounded, and the trial results are striking enough to justify serious follow-up. But striking early results in small trials have disappointed before. The honest answer is that we’re in the “this is genuinely interesting and we need more data” phase, not the “here’s your treatment plan” phase.
The larger trial underway at UCSD will tell us much more, particularly if the PET imaging arm reveals what’s actually happening in brain glucose metabolism during ketosis in this population. That’s the data worth waiting for.
Sources
- Frank et al., “Ketogenic dietary therapy for anorexia nervosa,” Communications Medicine (Nature Portfolio) (June 2026)
- Fox News Health: “Keto diet may help reduce anorexia nervosa symptoms” (June 4, 2026)
- GeneOnline News: “UC San Diego Pilot Trial Evaluates Ketogenic Diet for Anorexia Nervosa” (June 3, 2026)
- UCSD Clinical Trials: Ketogenic Diet Studies 2026 (April 2026)
Photo: Huzaifa Bukhari 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.
