KetoMental healthDepressionBipolar disorderBrain healthGuides

    Keto Diet and Mental Health: What the Evidence Shows, and What It Does Not Yet

    Shideh HeraviShideh Heravi·15 September 2026

    Someone in your family has bipolar disorder, or you have depression that two medications have only partly lifted, and a video appears: a psychiatrist explaining that the brain is an energy problem and that a ketogenic diet reached patients nothing else could. It is a compelling story. It is also the kind of story that makes people quietly skip a tablet to see what happens.

    Here is the short version. For the keto diet and mental health, the evidence is strong in one place, drug-resistant epilepsy, and early everywhere else. In bipolar disorder, schizophrenia and depression there are small pilot studies, mostly without a control group, that show promising changes in symptoms and in metabolic health. There is not yet a large randomised trial that tells us whether those changes come from ketosis, from weight loss, from the intense support in the studies, or from hope. If you want to try it for a diagnosed condition, it belongs inside a conversation with your treating psychiatrist, never in place of one. For the far more common question, "what should I eat to feel steadier?", the best-evidenced answer is still a Mediterranean-style pattern with stable blood sugar.

    What a ketogenic diet actually does

    A ketogenic diet keeps carbohydrate very low, usually under about 50 grams a day, with moderate protein and most energy from fat. After a few days the liver converts fat into ketones, small molecules the brain can burn instead of glucose. "Ketosis" means those ketones are measurable in blood or breath.

    Cutting bread at lunch does not put you in ketosis. The studies below used supervised protocols with ketone testing, which matters when comparing their results to what a person does alone. For the weight-loss side, see is keto good for weight loss. This post is about the brain.

    Where the evidence is strong: epilepsy

    The ketogenic diet was developed in the 1920s for seizures, and that is where its evidence base lies. A Cochrane review of randomised trials concluded that ketogenic diets can reduce seizure frequency in people, particularly children, whose epilepsy has not responded to medication (Martin-McGill et al., Cochrane Database of Systematic Reviews, 2020). It is offered in specialist epilepsy clinics, with a dietitian, blood monitoring and a plan for how long to continue.

    Even here, it is a medical diet delivered by a team, not a lifestyle. And the epilepsy evidence is why the psychiatric question exists: if changing the brain's fuel can calm seizures, what else might it calm? That justifies studies. It is not itself evidence for depression or bipolar disorder.

    Where the evidence is early: bipolar disorder, schizophrenia, depression

    What we know: a handful of small studies now exist and point in a hopeful direction. What is uncertain: whether the effect is real, how large it is, who it helps, and what it costs in side effects over years.

    The Stanford pilot. Researchers enrolled 23 adults with schizophrenia or bipolar disorder who also had metabolic problems, often a side effect of their medication, on a four-month ketogenic diet alongside usual treatment. By the end, none still met the criteria for metabolic syndrome. Among those who stuck to the diet, weight fell by about 12 percent and visceral fat by about 36 percent. Clinician-rated illness severity improved by an average of 31 percent, 79 percent of those who started with raised symptoms improved by at least one point, and participants with schizophrenia showed a 32 percent drop on a standard rating scale (Sethi et al., Psychiatry Research, 2024).

    A genuinely interesting result. It is also a single-arm study: everyone got the diet, nobody got a comparison, and everyone knew what they were eating. In psychiatry, where expectation moves symptom scores, that design can only generate a hypothesis.

    The Edinburgh pilot. A separate team recruited 27 people with bipolar disorder who were currently stable, 26 started a modified ketogenic diet and 20 completed six to eight weeks. Adherence was high, with 91 percent of blood ketone readings showing ketosis, and 11 minor adverse events were recorded, including fatigue, constipation, drowsiness and hunger (Needham et al., BJPsych Open, 2023). Its main contribution is showing the diet is feasible in this group, the necessary step before a trial that can test whether it works.

    A French inpatient series. A French clinic placed 31 inpatients whose symptoms were poorly controlled on medication onto a ketogenic diet, most with bipolar II or schizoaffective disorder, seven with major depression. All 28 who stayed on it for more than two weeks improved in depression and psychosis symptoms, typically within three weeks (Danan et al., Frontiers in Psychiatry, 2022). This is a retrospective chart review, the weakest design that still counts as evidence: a signal, not a proof.

    Depression in students. In a 2025 pilot, 24 university students with major depression received usual counselling or medication plus a 10 to 12 week ketogenic diet with food provided. Sixteen completed it, and their depression scores fell by 69 to 71 percent on two standard scales (Decker et al., Translational Psychiatry, 2025). A published critical appraisal made the point that matters: with no control group and a small, self-selected sample, placebo effects and the counselling alongside could explain much of that change (Critical appraisal, Translational Psychiatry, 2026). A third of participants did not finish, even with food supplied.

    What the reviews say. A systematic review of low-carbohydrate and ketogenic diets in mood and anxiety disorders found only eight studies to include. Ketogenic interventions were associated with lower depressive and anxiety symptoms and better metabolic markers, and the authors concluded that standardised protocols and longer studies are needed before safe clinical use (Dietch et al., BJPsych Open, 2023). To put that plainly: research suggests keto may help some people with serious mental illness, particularly those whose metabolism has been damaged by illness or medication. The evidence is still limited, and no guideline body currently recommends it for any psychiatric condition.

    How it might work

    Researchers have proposed several mechanisms, which explain why the idea is plausible rather than fringe (Norwitz, Sethi and Palmer, Current Opinion in Endocrinology, Diabetes and Obesity, 2020).

    A different fuel. Some imaging studies suggest that in schizophrenia, bipolar disorder and depression, parts of the brain use glucose less efficiently. Ketones enter brain cells by a different route and are burned by the same energy-producing structures, the mitochondria. If a struggling region is short of usable energy, an alternative fuel could help.

    Calmer signalling. Ketosis appears to shift the balance between the brain's main "go" signal, glutamate, and its main "slow down" signal, GABA, toward GABA. This is the leading explanation for the epilepsy effect and why it is plausible for mania or psychosis.

    Less inflammation, better insulin sensitivity. Low-grade inflammation and insulin resistance are common in serious mental illness and are associated with worse symptoms. Ketogenic diets tend to reduce both in the short term. Whether that is the mechanism or a passenger is unknown. The same pathways are why ordinary anti-inflammatory eating is studied for mood; see foods that lower inflammation.

    These mechanisms are reasonable and mostly shown in animals or small human samples. A plausible mechanism justifies a trial; it does not replace one.

    Why the hype runs ahead of the data

    Pilot studies produce large percentage changes because they are small, uncontrolled and full of motivated volunteers, and those percentages make easy headlines. The story, "your psychiatric illness is a metabolic illness you can eat your way out of", lands hardest on people other treatments have failed. None of that makes the researchers wrong. It makes the timeline wrong. The controlled trials that would show whether keto beats a well-supported Mediterranean diet, or beats losing the same weight another way, are only now being run. Until they report, "promising" is the strongest honest word.

    Safety, and the rule about medication

    If you have a diagnosed psychiatric condition and want to try keto, the one non-negotiable is this: do not stop or reduce any prescribed medication without your prescriber, and do not start the diet without telling them.

    The reasons are practical. Ketosis changes how your body handles fluid, salt and blood sugar, and several medicines used in psychiatry and diabetes are sensitive to exactly those things. Some mood stabilisers need blood-level monitoring that a large dietary change can disturb. Some diabetes medicines, which many people on antipsychotics also take, carry a risk of dangerous ketone build-up with very low carbohydrate intake. And if symptoms improve, dose changes should be made deliberately by the prescriber, not by missing tablets. In every study above, the diet ran alongside psychiatric care, not instead of it.

    The diet also has ordinary side effects: in the Edinburgh study, fatigue, constipation, drowsiness and hunger (Needham et al., 2023). Longer term it can raise LDL cholesterol in some people and it removes most fibre-rich foods unless planned carefully. Anyone with kidney or liver disease, a history of an eating disorder, pregnancy, or type 1 diabetes should treat keto as a medical decision.

    One red flag: if a change in eating coincides with new thoughts of self-harm, a manic episode, confusion, fainting, or vomiting that will not stop, that is a same-day call to your clinician or emergency services.

    The everyday version: what the evidence supports for most people

    Most people reading this have low mood, anxiety, brain fog or the 3 pm slump, not treatment-resistant bipolar disorder. For them the evidence points somewhere much less dramatic.

    Blood sugar stability. In a study of more than 70,000 postmenopausal women, those eating the highest glycaemic index diets had 22 percent higher odds of developing depression than those eating the lowest; added sugar showed a similar pattern, while fibre, fruit and vegetables went the other way (Gangwisch et al., American Journal of Clinical Nutrition, 2015). This is observational and cannot prove cause, but it fits what people feel. You do not need ketosis to flatten blood sugar swings; you need protein and fibre at each meal and fewer refined carbohydrates eaten alone.

    A Mediterranean-style pattern. A review of 41 observational studies found closer adherence to a healthy diet, including the Mediterranean pattern, was associated with lower risk of depressive outcomes (Lassale et al., Molecular Psychiatry, 2019). Trial evidence began with SMILES, a 12-week randomised controlled trial in 67 adults with moderate to severe depression: those given dietitian-led support to eat a modified Mediterranean diet alongside usual treatment improved significantly more on a standard depression scale than those given matched social support (Jacka et al., BMC Medicine, 2017). It was small and could not be blinded, but it had a fair comparison group, which no ketogenic study in psychiatry can yet say.

    The better-evidenced baseline is not glamorous: vegetables, legumes, whole grains, fish, olive oil, nuts, less ultra-processed food and regular meals. Depression and food covers that pattern in detail and brain fog and food the cognitive side. A lower-carbohydrate version of it is a reasonable personal choice; it is not a therapeutic ketogenic diet and does not need to be.

    A plan you can start this week

    1. Decide which question you are asking. "Could keto treat my condition?" is for your psychiatrist. "Could food help me feel steadier?" you can start on today.
    2. Anchor breakfast with protein and fibre. Eggs, yoghurt, beans, oats with nuts. This does more for mid-morning mood than any single food.
    3. Choose a Mediterranean-style base. Vegetables and legumes at most meals, fish twice a week, olive oil as the main fat, fruit rather than sweets when you want something sweet.
    4. Track mood next to meals for two weeks. Not calories: timing, what you ate, and how you felt two hours later. Patterns show up quickly.
    5. Skip the DIY therapeutic diet if you take psychiatric medication. If you want to explore keto, ask your prescriber to refer you to a dietitian who works with it, and agree monitoring before you start.
    6. Check the basics that mimic low mood. Iron, vitamin D, thyroid and sleep. A blood test is cheaper than a diet overhaul; see vitamin D before winter.
    7. Reassess after eight weeks. If nothing has changed, the answer is not more restriction. Discuss it with someone who can look at the whole picture.

    Ordering out

    Stable blood sugar needs no special menu, only a brief the kitchen can act on.

    Say: "Could I have the grilled fish or chicken with the vegetables and a side of the beans or salad instead of the fries? Bread on the side, not before, please."

    Not: "I am on a diet for my mental health."

    On a supervised ketogenic protocol, use the brief your dietitian gave you; sauces and dressings are where a well-meant order usually goes off track.

    Where our AI nutrition coach fits

    A coach's useful job here is to hold the pattern. Tell it what you ate and how you felt, and it can notice that your worst afternoons follow your most refined-carbohydrate lunches, suggest a Mediterranean-style swap you like, and remind you what to ask the kitchen. A personalised dietary AI agent suits that kind of steady support and can plan meals around any restriction your clinician has set.

    What it cannot do is diagnose a mood disorder, judge whether keto suits your condition, adjust a medication, or see inside a kitchen. Those belong with your psychiatrist and a dietitian. For the split between what a coach and a professional each do well, see AI nutrition coach vs a dietitian.


    Frequently asked questions

    Does the keto diet help with depression? There is growing interest, but the evidence is still limited to small, uncontrolled pilot studies. In a 2025 pilot, students who completed a supervised ketogenic diet alongside usual care saw depression scores fall by around 70 percent, but with no control group that cannot be separated from placebo effects and counselling (Decker et al., 2025). The better-evidenced approach for depression is a Mediterranean-style pattern.

    Can keto treat bipolar disorder or schizophrenia? Not yet as an established treatment. Pilot studies at Stanford and Edinburgh found the diet feasible, with better metabolic health and clinician-rated symptoms (Sethi et al., 2024; Needham et al., 2023). Controlled trials are needed before anyone can say it works, and no guideline recommends it.

    Can I try keto while taking antidepressants or mood stabilisers? Only with your prescriber involved. Ketosis changes fluid, salt and blood sugar handling, which affects how some psychiatric and diabetes medicines behave and are monitored. Never stop or reduce a prescribed medication on your own, even if you feel better.

    How is keto supposed to help mental health? The proposed mechanisms are an alternative brain fuel (ketones instead of glucose), a shift toward calmer GABA signalling and away from excitatory glutamate, and lower inflammation and insulin resistance (Norwitz et al., 2020). These are plausible but have not been shown to be the reason any patient improved.

    What should most people eat for better mood? A Mediterranean-style pattern with steady blood sugar: vegetables, legumes, whole grains, fish, olive oil and nuts, with protein and fibre at each meal and fewer refined carbohydrates. This is supported by the SMILES randomised trial and by large observational studies (Jacka et al., 2017; Lassale et al., 2019).


    Sources

    1. Martin-McGill KJ, et al. Ketogenic diets for drug-resistant epilepsy. Cochrane Database of Systematic Reviews, 2020.
    2. Sethi S, Wakeham D, Ketter T, et al. Ketogenic Diet Intervention on Metabolic and Psychiatric Health in Bipolar and Schizophrenia: A Pilot Trial. Psychiatry Research, 2024.
    3. Needham N, Campbell I, Grossi H, et al. Pilot study of a ketogenic diet in bipolar disorder. BJPsych Open, 2023.
    4. Danan A, Westman EC, Saslow LR, Ede G. The Ketogenic Diet for Refractory Mental Illness: A Retrospective Analysis of 31 Inpatients. Frontiers in Psychiatry, 2022.
    5. Decker DJ, et al. A pilot study examining a ketogenic diet as an adjunct therapy in college students with major depressive disorder. Translational Psychiatry, 2025.
    6. Critical appraisal of a pilot study examining a ketogenic diet as an adjunct therapy in college students with major depressive disorder. Translational Psychiatry, 2026.
    7. Dietch D, Kerr-Gaffney J, Hockey M, et al. Efficacy of low carbohydrate and ketogenic diets in treating mood and anxiety disorders: systematic review and implications for clinical practice. BJPsych Open, 2023.
    8. Norwitz NG, Sethi S, Palmer CM. Ketogenic diet as a metabolic treatment for mental illness. Current Opinion in Endocrinology, Diabetes and Obesity, 2020.
    9. Gangwisch JE, et al. High glycemic index diet as a risk factor for depression: analyses from the Women's Health Initiative. American Journal of Clinical Nutrition, 2015.
    10. Lassale C, Batty GD, Baghdadli A, et al. Healthy dietary indices and risk of depressive outcomes: a systematic review and meta-analysis of observational studies. Molecular Psychiatry, 2019.
    11. Jacka FN, et al. A randomised controlled trial of dietary improvement for adults with major depression, the SMILES trial. BMC Medicine, 2017.

    Educational content, not medical advice. A ketogenic diet for a psychiatric condition is a medical decision that belongs with your treating psychiatrist and a dietitian, with monitoring agreed in advance. Do not stop, reduce or skip any prescribed medication without talking to the person who prescribed it. If you have thoughts of harming yourself, contact your clinician or emergency services today.

    Shideh Heravi

    Written by

    Shideh Heravi

    Newsletter

    Gut health & clean eating, in your inbox

    New posts on food intolerances, gut health and safer eating no spam, unsubscribe anytime.

    By subscribing you agree to receive emails from HealthBlokk about gut health, food intolerances, and product updates. You can unsubscribe anytime.