Keto Diet Could Be an Effective Treatment for Anorexia

By Irene Yeh 

Anorexia nervosa is a mental illness characterized by extreme food restriction, self-starving, significant weight loss, fear of gaining weight, and body dissatisfaction. Even after restoring weight, patients still experience psychological struggles, which increase the risk of relapsing. It is not enough to simply get to a target weight. Though diet interventions may raise concerns about promoting unhealthy eating patterns, the ketogenic diet might be an effective solution.  

Researchers from the University of California, San Diego School of Medicine conducted a pilot study to determine if the keto diet is a viable intervention for anorexia treatment (Communications Medicine, DOI: 10.1038/s43856-026-01644-0). They also measured the feasibility and safety of a weight-maintaining keto diet.  

Current anorexia treatment plans center around weight restoration via structured meal plans, but they do not address the fear of weight gain. With emerging evidence showing the connection between nutrition and mental health, as well as brain metabolic alterations, it is hypothesized that what patients eat may play an important part in psychiatric disorders.  

With the keto diet, the idea is to make the body break down fat through ketosis for energy. In addition to improving cell energy metabolism, ketosis therapy has been associated with decreased oxidative stress and inflammation, as well as regulation of key neurotransmitter systems. In anorexia, self-starvation can result in ketosis, but it leads to emaciation and death. The team proposes that the keto diet—which replicates the metabolic effects of starvation without caloric deprivation—could have therapeutic benefits. 

For the study, 22 participants aged 18 to 45 with a history of anorexia were recruited. They were either weight-restored or slightly underweight to allow flexibility for weight fluctuations. All participants were female and predominantly white and non-Hispanic. There was no placebo group. 

There was no target weight for the participants. Instead, the team aimed for weight maintenance and informed participants that they could not get below a 17.5 BMI score. The trial took place over 14 weeks, and four participants dropped out during the trial, resulting in 18 participants who completed the whole trial. The participants were also asked to complete self-assessment tools to measure various depression and eating disorder symptoms: the Eating Disorder Examination Questionnaire (EDE-Q), Eating Disorder Inventory-3 (EDI-3), Eating Disorder Clinical Impairment Assessment, Beck Depression Inventory 2, Spielberger State-Trait Anxiety Inventory, and Temperament and Character Inventory (TCI).  

The diet regimen consisted of three meals and two snacks per day and was composed of 70% fat, 20% protein, and 10% carbohydrates. The food was provided through a ketogenic meal delivery service, but participants were allowed to prepare their own meals with dietitian input. To monitor nutritional ketosis, participants submitted to daily blood draws over the first two weeks and then once a week afterward. The participants provided weekly self-assessments on anorexia symptoms, mood, anxiety, suicidal thoughts, ketone levels, and weight. They also received up to three hours of counseling support. Three months after the study, the participants answered questionnaires, provided their weight, and followed up with the researchers to provide further qualitative feedback and confirm whether they continued with the keto diet. 

The Potential of the Keto Diet 

The research team reported that there was a significant difference between the participants who completed the trial and the ones who did not. 72% of the completers reached the recovered range of eating disorder symptoms, according to their EDE-Q Global scores. Participants with lower novelty seeking scores (TCI) and higher low self-esteem scores (EDI-3) did not fall within normal range, indicating poor improvement in anorexia. The completers group had lower novelty-seeking and elevated harm avoidance, depression, and low self-esteem scores. According to the 3-month follow-up, 39% of participants reported continuing the keto diet while 61% did not.  

Some participants initially reported considerable anxiety and needed substantial coaching to maintain adequate intake, but their trust in the dietary intervention improved as the study progressed. By the end of the study, all participants showed improvement in depression scores. Post-hoc analysis also showed that low self-esteem was connected to a higher EDE-Q Global score.  

In the eating disorder field, there is always the concern of dietary interventions triggering symptoms or promoting unhealthy eating habits. With the keto diet often depicted as a weight loss diet, the researchers considered the possibility of participants relapsing. Fortunately, those concerns were not found in the results. The keto diet was generally well-tolerated and highly effective in reducing depression and eating disorder symptoms, except for the individuals who exhibited low self-esteem. Furthermore, anecdotal comments made by completers indicated an increase in confidence and surprise that the keto diet provided recovery levels they had not previously experienced.  

Another concern was whether the keto diet would increase the fear of carbohydrate intake.  However, none of the participants reported a fear of carbohydrates. Conversely, the participants felt “sluggish” when reintegrating carbohydrates.  

Other limitations to consider are the lack of direct supervision of the participants eating their meals. They were only required to send photos of their meals to the research team. Therefore, the possibility of participants restricting food intake must be considered. The study also did not consider expectation bias, placebo effects, changes in medication, ongoing psychotherapy, and eating behaviors. There were also no pre- or post-intervention labs to determine nutritional adequacy. 

Additionally, six participants fell out of ketosis, which could have affected the outcomes. It remains uncertain as to why the biomarkers of ketosis were associated with eating disorder and depression severity. The team encourages further research in order to fully confirm the keto diet’s effectiveness. 

The EDE-Q was also modified for the weekly assessment schedule, potentially introducing uncertainty in how the results compare to other studies. Finally, the small sample size needs to be considered. Though the team examined factors such as the use of counseling services, meal delivery, meal preparation methods, and supplements such as medium-chain triglyceride oil and carnitine, the small sample size could affect any conclusions made and need to be interpreted with caution. 

Despite the limitations, the keto diet shows promise as an effective treatment for anorexia patients. The team encourages further research, such as a controlled study to compare the keto diet with other interventions, including the Mediterranean diet, which was previously tested on a different group. Currently, there is an extension of the present study that is recruiting patients nationwide.