Australasian Science: Australia's authority on science since 1938

A Diet that Calms the Schizophrenic Mind

Credit: tankist276/adobe

The ketogenic diet is preferred by bodybuilders who need a high energy intake that doesn’t promote the conversion of fat from excess carbohydrates. Credit: tankist276/adobe

By Zoltán Sarnyai

The ketogenic diet favoured by bodybuilders also normalises schizophrenia-like behaviours.

Schizophrenia has long been treated, with limited success, with drugs that block the brain neurotransmitter dopamine. Pharmaceutical companies have spent billions of dollars developing yet another drug with a slightly different mechanism of action to also block dopamine transmission.

Is it possible that a dietary intervention can help people suffering from this devastating mental illness, or is this idea no better than the fad diets promoted in the pages of celebrity gossip magazines?

Our research found that the ketogenic diet, which is very high in fat and extremely low in carbohydrates, effectively normalised a wide range of schizophrenia-like behaviours in a well-established mouse model of the disorder.

What does this diet do that makes the symptoms disappear? Is this approach safely translatable to humans?

I think it is translatable, and there is some fascinating new science behind it as well.

Beyond Dopamine

I was in the middle of my psychiatry rotation in the last year of medical school when my brother told me of the strange, highly disturbing behaviour of his classmate. “He locks himself in his room, closes the curtains and the shutters on the window. He does not talk to his mother and has not been taking much food for a week in the fear of being poisoned. When he does talk he is talking nonsense about some voices he is hearing.”

The very next morning, an 18-year-old man was taken to our psychiatry ward. He was very agitated and needed to be physically restrained by police. He attacked his mother because he thought she wanted to kill him.

He was my brother’s classmate. He was diagnosed with an acute psychotic episode, promptly received an antipsychotic injection and admitted to the ward.

I soon moved on to another rotation and a year later accidentally bumped into the psychiatrist who had admitted the young man. I asked him about his young patient.

“He died, unfortunately... committed suicide,” the psychiatrist answered.

“But he was looked after properly, received his antipsychotic medication and should have not ended up like this,” I protested in my youthful enthusiasm.

Perhaps now, one-quarter of a century later, we are in a much better position to help people who are suffering from this devastating mental illness.

For many years, schizophrenia research and drug development was driven by the idea that there seems to be an increased activity of dopamine in the brain. A lot of evidence pointed in that direction. For example, drugs used in the treatment of schizophrenia all blocked the receptor protein for dopamine, and drugs that stimulated brain dopamine tend to induce psychosis.

Animal models with hyperactive dopamine neurotransmission churned out drugs that all resembled the original antipsychotics used in the validation of these animal models in the first place. A bit of a Catch 22, isn’t it?

Then, in the dawn of the “genomic age”’ at the turn of the millennium, human genetics research started to uncover a number of other possible disease mechanisms. Researchers started to discover schizophrenia susceptibility genes that play a role in brain development and in the formation of the synapse that connects one nerve cell to the next. They also identified that environmental events associated with schizophrenia, such as maternal virus infection and malnutrition as well as exposure to stress and early adversity, influence the expression of these genes, resulting in abnormal brain development and pathological behaviour in experimental animals.

Schizophrenia has therefore been reconceptualised as a disease of abnormal neural development caused by the interaction of many genes and adverse early events.

Insulin Resistance

About 10 years ago researchers at The University of Cambridge discovered abnormal expression of genes responsible for the proper breakdown and utilisation of glucose in the prefrontal cortex of patients with schizophrenia. This brain region is involved in higher cognitive functions that are abnormal in schizophrenia, such as attention, planning and executive control of other brain areas. Other groups confirmed that these enzyme proteins are abnormally low in this brain region. They also identified changes in the structure and function of the mitochondria, the “power stations” that fuel all of the processes required for proper communication between nerve cells. These discoveries have given rise to the hypothesis that abnormal glucose and energy metabolism may contribute to the development of schizophrenia.

This wasn’t surprising. Maudsley, the famous 19th century British psychiatrist, had already observed that patients suffering from “insanity”, as well as their first-degree relatives, showed a disproportionately higher rate of diabetes mellitus. This was the first hint that glucose and energy metabolism might be abnormal in schizophrenia.

Before the discovery of modern antipsychotic agents, psychiatrists at the beginning of the 20th century employed insulin coma therapy to drive down circulating glucose in the hope that this would improve the symptoms of patients with a variety of mental illnesses. They found that patients with schizophrenia required much more insulin to achieve the same effect. This supports the notion that some sort of insulin resistance is involved in schizophrenia.

One hundred years later, researchers using modern analytical tools and diagnostic criteria have reported that unmedicated patients experiencing their first episode of schizophrenia showed resistance to the effects of insulin. This also suggests that there might be a problem with glucose metabolism in and beyond the brain.

This all makes a lot of sense. The brain is an extremely energy-hungry organ that uses a disproportionately high amount of glucose compared with other parts of the body, such as the heart and the muscles.

Glucose in the brain is converted into chemical energy in the form of adenosine-triphosphate (ATP) molecules in order to:

  • maintain communication between nerve cells;
  • synthesise the main excitatory neurotransmitter, glutamate;
  • synthesise the main inhibitory neurotransmitter, gamma-amino-butyric acid (GABA); and
  • synthesise molecules that deal with toxic free radicals produced by normal neuronal activity.

If the utilisation and processing of glucose is abnormal, neurons cannot function properly, will not communicate efficiently and their connections may get damaged due to high levels of toxic free radicals. If these changes occur in brain areas that underlie key cognitive functions such as the prefrontal cortex, the symptoms of schizophrenia may emerge. These include hallucinations, delusions, impaired attentional and other cognitive functions.

Here Comes the Diet

If schizophrenia is driven, at least partly, by abnormal glucose metabolism, we hypothesised that we might be able to normalise the disease process if we provide energy sources in a way that circumvents glucose metabolism. If this energy supply is adequate, nerve cells would be able to communicate properly through the synapse and remove harmful free radicals more efficiently.

One such roundabout way of feeding nerve cells with molecules that can be used for the production of energy and the key neurotransmitters glutamate and GABA is the use of fatty acids instead of glucose. In the absence of glucose and other carbohydrates, fatty acids are broken down to beta-hydroxybutyrate (BHB) and acetone in the liver.

BHB travels easily to the brain, where it’s taken up by nerve cells. Here BHB is further broken down to molecules that can enter into the energy production pathways downstream of the point where energy metabolism breaks down in schizophrenia.

This metabolic pathway is a normal physiological process that takes places during starvation, when the body accesses its fat deposits to produce energy. In fact, it is likely that this could have been a dominant metabolic process during most of human evolution, when food was always scarce and carbohydrates played a very small role as energy substrates.

My student, Ann Katrin Kraeuter, tested our hypothesis by putting mice on a ketogenic diet for 3 weeks. She then administered a drug that has been known to induce psychosis in humans and schizophrenia-like behaviours in mice.

Mice on the normal diet promptly exhibited a variety of schizophrenia-like behaviours, such as hyperactivity, excessive stereotyped behaviours, abnormal social interactions as well as impaired working memory. The latter indicates that prefrontal cortical functions are compromised.

Mice on the ketogenic diet, however, showed none of these symptoms. They were lean, weighed less than mice on a normal diet and had lower circulating blood glucose levels.

These results suggested that the ketogenic diet might be a useful approach to manage schizophrenia, not only because it normalised schizophrenia-like behaviours but also because of its beneficial metabolic effects.

Antipsychotic drugs used in the treatment of schizophrenia produce weight gain, elevated blood glucose, insulin resistance and ultimately metabolic syndrome. These can contribute to the patient’s early death due to cardiovascular disorders.

The ketogenic diet seems to counteract these deleterious metabolic effects.

Are These Findings Translatable to People?

We know that the ketogenic diet can be safely given to humans. In fact, it has been used for the management of drug-resistant epilepsy in children for a long time. The ketogenic diet has also been used as a weight loss diet, and seems to be preferred by bodybuilders who need a high energy intake that doesn’t promote the conversion of fat from excess carbohydrates.

Long-term feeding trials of the ketogenic diet in mice showed no deleterious consequences when used for up to a year, which is about half of the life span of a mouse. This is encouraging, but we need to demonstrate the efficacy of the ketogenic diet in other animal models of schizophrenia before we can move to clinical trials.

One might argue that the ketogenic diet is not easy to follow and can be especially challenging for patients who are suffering from schizophrenia. This prompts us to seek alternative ways to deliver the diet, perhaps in the form of a supplement that mimics its effects. This requires a better understanding of what is happening in the body and brain of someone on a ketogenic diet.

Using dietary means to manage schizophrenia is not unheard of. Recent pioneering work at The University of Melbourne has shown that omega-3 fatty acids in fish oils decrease the development of psychosis in individuals who have a high genetic risk of developing schizophrenia.

Perhaps we are entering into the era of nutritional psychiatry.


Zoltán Sarnyai is Associate Professor and Head of the Laboratory of Psychiatric Neuroscience at the Australian Institute of Tropical Health and Medicine, James Cook University.