Keto Schizophrenia Bipolar Trial and What Correlates Means

The keto schizophrenia bipolar trial randomized 58 adults with schizophrenia-spectrum or bipolar I disorder to a ketogenic diet or their usual diet for one month, then offered a keto extension that 25 people completed over four months. The randomized month showed metabolic changes such as weight, HbA1c and insulin resistance; the symptom and cognition gains showed up at four months, in a group that no longer had a randomized control. Higher blood ketones moved together with better blood sugar markers and lower depression scores, which is a correlation, not proof that ketones caused the change.

What did the keto schizophrenia bipolar trial actually test?

Picture a fit doctor sitting down to another burger, knowing the next one is already on the grill. That was Dr. Mike, and that was the start of 600 Burgers in 30 Days. I wasn’t watching to see whether he would suffer. I was watching to see what the numbers would say. I have brought that same habit to every study I’ve read since, and it’s the habit I’m bringing to this one.

The keto schizophrenia bipolar trial tested whether outpatients with schizophrenia-spectrum or bipolar I disorder could stick to a ketogenic diet, and what happened to their metabolism, symptoms and thinking when they did. The published abstract in Schizophrenia Bulletin calls it the first randomized controlled trial to ask those questions together.

Here is why that matters. Antipsychotic medication is hard on the body, and the same abstract opens by noting that psychiatric medications drive high rates of metabolic dysfunction in psychotic disorders. So the researchers were not asking whether keto replaces treatment. They were asking whether food could change the metabolic side of the picture, and whether the brain followed.

I start every experiment with three questions, and I read every study with the same three. One: who was tested? Two: for how long? Three: against what? If you aren’t testing, you’re guessing. This trial did test, and for families living with these illnesses, the design deserves a careful read. Hope built on a misread study costs real people real time.

How many people took part, and for how long?

Fifty-eight people were randomized: 28 to the ketogenic diet and 30 to their usual diet, for one month, according to the trial abstract. Partway through, both groups were offered a keto extension, and a subgroup of 25 finished four months on the diet.

Not everyone made it through the first comparison. Neuroscience News reports that 47 of the 58 completed the one-month randomized phase, and describes it as open-label, which means everyone knew which diet they were on.

Two time frames. Remember them. One month: randomized. Four months: not. The Brain & Behavior Research Foundation summary calls the diet feasible over four months, and that is a fair word for what was shown.

What changed on the ketogenic diet?

In the randomized month, the keto group lost weight and improved HbA1c and insulin resistance compared with the usual-diet group, per the published abstract. Their daily ketone levels also cleared the standard threshold for ketosis, so people really were in ketosis rather than just eating less bread.

The psychiatric changes came later. The same abstract reports that positive symptoms, negative symptoms, depression and cognitive performance improved after four months on the diet. The authors themselves, as Neuroscience News reports, say one month looks too short to capture mental health effects, and the four-month gains need replicating under controlled conditions.

Here’s the catch, and it comes in three parts. One: the strongest psychiatric numbers come from the part of the study with no control group. Two: they come from people who chose to keep going. Three: those people may already have felt it working. I saw the same thing on our set: the person who keeps showing up for the finger prick is usually the person who likes what the meter says.

What does correlates with mean in this study?

Correlates with means two numbers moved together; it does not mean one caused the other. In this trial, higher blood ketones in the keto group went along with better prediabetes markers and lower depression scores, according to the Center for Nutritional Psychology summary.

Let me explain with something I measured myself. The film shows me pushing my blood ketones down to 0.1 before a breath-hold test, then taking 12 grams of exogenous ketones and watching them rise by 0.7 two and a half hours later (in the film at 01:29:05 and 01:34:56). The book records my breath holds at 64, then 80, then 93 seconds across that day (600 Burgers in 30 Days). Ketones went up. Breath holds went up. Did the ketones do it? Practice, calm and a dozen finger pricks were all in the room too.

That’s the whole fight over one word. The headline says correlates because the team was careful. Some coverage was not: Neuroscience News wrote that the result proves ketosis drives the benefit. A correlation in 25 people proves nothing of the kind.

Did weight loss explain the results?

The researchers report that weight loss did not account for the links they found. Medical Dialogues reports that weight loss showed no statistical association with the metabolic or symptom measures, while ketone levels did.

Good clue. Not a verdict. Taking weight out of the picture knocks down one obvious explanation. Plenty of others are still standing: sleep, routine, a study team checking in on you, and cooking your own food every single day.

I’ve watched it happen: the structure of an experiment changes people before the food does. You wake up to a test. You eat to a schedule. You write it all down at night. The film suggests the same thing about Dr. Mike: once the burgers, the 30 days and the cameras arrived, his days became a protocol, and a protocol is its own intervention.

What would a stronger trial look like?

A stronger trial would randomize people for the full four months, use a comparison diet, and be large enough to see who does not respond. One is already designed: a published protocol describes a randomized placebo-controlled trial of 100 adults with bipolar disorder, schizoaffective disorder or schizophrenia on a modified ketogenic diet over 14 weeks.

That’s how evidence gets built, one rung at a time. Rung one: a pilot shows people can do it. Rung two: a short randomized trial shows the body changes. Rung three: a longer controlled trial asks whether the mind changes because of it. You don’t get to skip rungs.

Individual results vary, and in serious mental illness that sentence carries weight. Nobody in these studies quit their care team to try a diet. If this topic touches your family, read the source with your doctor. For how thin the evidence runs on all-meat eating, see our carnivore scoping review.

Frequently asked questions

No. The trial tested a ketogenic diet, which limits carbohydrate enough to raise blood ketones, and it did not require eating only meat. Carnivore is one extreme way to get there. The Keto Project compared those approaches directly, which we unpack in Keto vs Carnivore: What the 600-Burger Experiment Tested on this site.

Diet as usual means the control group keeps eating the way it already eats, with no new instructions. It shows what happens with time and attention alone. It does not test keto against another structured diet, so it cannot tell you whether keto beats a healthy low-fat plan with the same support.

Not that anyone has shown. And this trial used food, not supplements. In my own test, the film shows I said the exogenous ketones did not raise my levels any more than just not eating all day. The book adds that the ketone salts gave me about a day and a half of gastric distress.

In an open-label trial everyone knows which diet they are on, and that can colour how people report mood and symptoms. Blood markers like HbA1c are harder to sway with expectation. That is one reason the metabolic results in the randomized month carry more weight than the later symptom scores.

The Bottom Line

The evidence says a ketogenic diet is doable for outpatients with schizophrenia and bipolar I disorder, changes metabolic markers within a month, and moves alongside symptom gains over four months. It does not yet say keto causes those gains. The careful word in the headline is the most honest thing in it.

I care about getting this right because the people reading about this trial aren’t reading for fun. They’re reading for a son, a sister, a spouse. That’s exactly why the careful word matters. I have spent years asking one question about food: what actually happens when you measure it? A fit doctor ate 600 burgers in 30 days so we could find out, and the book walks through every test we ran and every day of his journal. How it turns out is in there and in the film. Get the book, then ask yourself: what are you guessing about that you could be testing?

Sources

• Ketogenic Diet, in a First Randomized Trial, Correlates With Symptom Improvement in Schizophrenia & Bipolar Disorder With Psychosis (BBRF)
• Metabolic Improvements with a Ketogenic Diet Correlate with Symptom Improvement in Psychosis: A Randomized Controlled Trial (Schizophrenia Bulletin)
• Ketogenic Diet Eases Schizophrenia and Bipolar Symptoms (Neuroscience News)
• Impact of Ketogenic Diet on Metabolic and Psychiatric Health in Schizophrenia and Bipolar Disorder (Center for Nutritional Psychology)
• Ketogenic Diet May Improve Symptoms and Cognition in Schizophrenia and Bipolar Disorder (Medical Dialogues)
• The effects of ketogenic metabolic therapy on mental health and metabolic outcomes in schizophrenia and bipolar disorder: a randomized controlled clinical trial protocol (PubMed)