If you have a brain, you need to know about ketogenic diets. The fact that these specially-formulated low-carbohydrate diets have the power to stop seizures in their tracks is concrete evidence that food has a tremendous impact on brain chemistry and should inspire curiosity about how they work. I first became interested in ketogenic diets as a potential treatment for bipolar mood disorders, given the many similarities between epilepsy and bipolar disorder.
Ketogenic diets have been around for about 100 years, and have proved to be invaluable tools in the treatment of stubborn neurological conditions, most notably epilepsy. They have also shown promise in the management of other brain-based disorders such as Parkinson’s Disease, ALS, Traumatic Brain Injury, Multiple Sclerosis, and chronic headaches, as well as in metabolic disorders like obesity, cancer, and type 2 diabetes.
But where does the science currently stand on the ketogenic diet and psychiatric disorders like bipolar disorder, schizophrenia, and Alzheimer’s Disease? How many human studies do we have, and what do they tell us? If you are struggling with mood, attention, or memory problems, should you try a ketogenic diet? If you are a clinician, should you recommend a ketogenic diet to your patients?
A recent review article “The Current Status of the Ketogenic Diet in Psychiatry” by researchers at the University of Tasmania in Australia [Bostock et al 2017 Front Psychiatry 20(8)] brings us nicely up to date on all things ketogenic and mental health. I summarize the paper below and offer some thoughts and suggestions of my own. [Full disclosure: I am a psychiatrist who studies nutrition and eats a ketogenic diet.]
First, some basics for those of you who are unfamiliar with these special diets.
Definitions vary, but what all ketogenic diets have in common is that they are very low in carbohydrate (typically 20 grams per day or less) and relatively high in fat. The goal is to lower blood sugar and insulin levels; when these are nice and low, the body naturally turns to fat (instead of sugar) as its primary source of energy. Most ketogenic diets also limit protein (to no more than the body requires), because excess protein can raise blood sugar and insulin levels to some extent. Body fat and fat from the diet then break down into ketones, which travel through the bloodstream and can be burned by various cells throughout the body, including most brain cells. Ketone levels rise in the blood, urine and breath within days, and can be measured using various home test methods, but it can take weeks for the body to become efficient at burning fat for energy, and for full benefits to be realized.
When a person is “in ketosis”, fasting morning blood glucose levels tend to average between 60 and 85 (mg/dl), and blood ketone levels rise to at least 0.5 mM (with much higher levels recommended for certain conditions). These parameters distinguish ketogenic diets from other low-carbohydrate diets, which may contain too much protein and/or carbohydrate to produce these metabolic effects.
It remains unclear how ketogenic diets work to control seizures, let alone how they may improve psychiatric symptoms. On a fundamental level, we are not even sure whether it is the presence of ketones, the reduction in blood sugar, the reduction in insulin and other growth-promoting hormones, or the combination of all of these which are responsible for the brain-stabilizing effects of these diets. Theories abound, and include altered neurotransmitter levels, changes in electrolyte gradients (lower intracellular sodium and calcium), reduction in markers of inflammation, and improved mitochondrial function. The general consensus is that the brain functions more cleanly and efficiently when a significant portion of its energy comes from ketones, calming overactive and overly-reactive brain cells.
You can raise your blood ketone levels without changing your diet, by either taking expensive ketone supplements or by ingesting fats high in medium chain triglycerides (MCTs), which the liver rapidly transforms into ketones. Purified MCTs are available for purchase, or you can simply take coconut oil, which is naturally rich in MCTs. You’ll see below that these approaches can be effective short-term, but my opinion is that they simply mask the underlying disease, which continues to worsen due to ongoing high insulin and/or blood glucose levels.
SUMMARY OF THE SCIENCE
I’ve listed all relevant studies covered in the 2017 Bostock review below, paying special attention to the human studies, and supplementing with original source material where helpful, as there were some minor errors in the text of the review.
2002: A one-month case study of a woman with unspecified, treatment-resistant bipolar disorder noted no improvement after two weeks on a ketogenic diet followed by two weeks of MCT oil supplementation. Urine testing found ketosis was never achieved.
2012: A case study of two women with bipolar II disorder who ate a ketogenic diet (one for two years, the other for three years) found that the diet was superior to the anticonvulsant/mood stabilizer lamotrigine (Lamictal) in management of symptoms. Ketosis was documented using urine test strips.
A 3-week mouse study showed that a ketogenic diet normalized pathological behaviors.
1965: A 2-week study of 10 women with treatment-refractory schizophrenia found a significant decrease in symptoms when a ketogenic diet was added to their ongoing standard treatments (medications + ECT). Ketone monitoring was not reported.
2009: A 12-month case study details the experience of a 70-year old overweight woman with chronic schizophrenia who was prescribed a diet limited to 20 grams of carbohydrate per day. She noted significant improvement in severe symptoms beginning only eight days after starting the diet, which consisted of “beef, poultry, ham, fish, green beans, tomatoes, diet drinks, and water.” [ Kraft and Westman 2009 Nutrition & Metabolism 6:10.] She reported complete resolution of auditory and visual hallucinations--with which she’d suffered since age seven. Ketone levels were not monitored.
Comment: This diet is best characterized as a low-carbohydrate, primarily whole foods diet. As protein wasn’t limited and fat intake wasn’t manipulated, this may or may not have been a truly ketogenic diet.
A rat study found that adding ketone supplements to a standard high-carbohydrate diet reduced anxious behavior.
A rat study found that a ketogenic diet reduced depressive behaviors.
A mouse study found that feeding pregnant animals a ketogenic diet reduced offspring susceptibility to depressed (and anxious) behaviors.
A 70-day mouse study found that a ketogenic diet improved behavior.
A 10-14 day rat study found that a ketogenic diet improved complex social behaviors and mitochondrial function.
A 3-4 week mouse study found that a ketogenic diet improved behaviors in ways that were different for males than females.
2003: A 6-month inpatient study evaluated the effects of a cyclical ketogenic diet (4 wks on, 2 wks off) on 30 children with ASD. Of the 18 children who completed the study, eight showed moderate improvement, and two showed “significant” improvement. Benefits appeared to persist even during the 2-week “diet-free” periods. Urine and blood ketone monitoring confirmed that all children were in ketosis.
Comment: blood ketone levels ranged from 1.8 to 2.2 mMol during ketogenic phases and from 0.8 to 1.5 mMol during “diet-free” periods, meaning that the children actually spent the entire 6-month study period in ketosis.
2013: A 14-month detailed case study of one child with ASD, epilepsy, and obesity who was placed on a ketogenic diet in combination with anti-epileptic medications, noted numerous improvements. “In addition to improvement in seizures, there was a 60-pound weight loss…as well as improved cognitive and language function, marked improvement in social skills, increased calmness, and complete resolution of stereotypies.” [Herbert and Buckley 2013 J Child Neurol 28(8)]. Ketosis was confirmed (presumably by urine testing).
A 6-month study of dogs with ADHD and epilepsy found significant improvement in ADHD behaviors on a ketogenic diet.
2009: A 90-day randomized, double-blind, placebo-controlled, parallel study of 152 people with mild to moderate Alzheimer’s Disease tested the effects of a daily MCT supplement (previously marketed under the name Axona) on cognitive test performance. People continued their usual diets and took either the MCT supplement or a safflower oil placebo. Regular medications were continued throughout the study. At 45 and 90 days, patients taking MCTs showed significant improvement on a cognitive test known as the ADAS-Cog scale, unless they carried a gene called ApoE4, which is associated with higher risk for Alzheimer’s disease. Cognitive benefits did not persist after MCTs were discontinued.
Not mentioned in the Bostock review are the following two studies:
1) A 6-week study of a simple low-carbohydrate diet (protein and fat unrestricted) in people with mild cognitive impairment (MCI, aka “pre-Alzheimer’s” disease) demonstrating improvement in verbal memory, with greater benefits seen in those who achieved higher ketone levels [Krikorian R et al 2012 Neurobiol Aging 33(2):425].
Please also see my Psychology Today article Preventing Alzheimer's Is Easier Than You Think.
If you are a ketogenic diet skeptic, you will find the above summary unimpressive. I can’t blame you—there is precious little human data about ketogenic diets and psychiatric disorders, and what is there is flawed: small sample sizes, no controls, ketosis unconfirmed in some cases, diet composition and length of treatment variable between studies, etc. However, as a low-carbohydrate diet enthusiast, I find much to be excited about in this ragtag fugitive fleet of papers.
Nutrition studies are hard. Blinding is nearly impossible, funding is hard to come by, compliance is challenging, and controls are difficult to design. Fear, bias, and nutrition miseducation limit the number of scientists interested in and willing to conduct studies of low-carbohydrate, high-fat diets. We clearly need high-quality human studies exploring the effects of ketogenic diets on mental health disorders, as many clinicians and patients will be afraid to utilize this diet without stronger evidence. If only more people held the USDA Dietary Guidelines to the same scientific standard and eyed it with the same healthy skepticism...
Nevertheless, I believe in the therapeutic potential of low-carbohydrate diets to stabilize brain chemistry, and feel strongly that people should be made aware of dietary strategies as an option. For people who don’t want to take medication, haven’t responded to medication, can’t tolerate medication or can’t afford medication, nutritional intervention can offer real hope and empowerment. I personally noticed an overall improvement in concentration, mood, energy, and productivity when I changed my own diet years ago, and have witnessed people in my own practice whose mood stabilized by switching to a low-carbohydrate, high-fat, whole foods diet. My philosophy regarding the dietary treatment of mental health problems is as follows:
It could be many years before we see high-quality studies of ketogenic diets in the treatment of psychiatric disorders. For those of you who don’t want to wait that long—what do you need to know?
The ketogenic diet is safe for most people, but there are clear exceptions to consider that go beyond the scope of this post, so please do not embark on a ketogenic diet yourself nor recommend it to someone else without first reading Is the Ketogenic Diet Safe for Everyone? and seeking additional guidance and resources to educate yourself about the diet. Two of the best books on the subject are The Art and Science of Low Carbohydrate Living by Drs. Phinney and Volek, and The New Atkins for a New You by Dr. Eric Westman.
Some studies report side effects such as constipation, leg cramps, and increased risk of kidney stones. There are very real shifts in salt and fluid balance that occur in the early stages of the diet that may explain some of these side effects, but most people tolerate the diet well, including myself. My opinion is that the majority of side effects are not due to the low-carbohydrate, high-fat nature of the diet, but rather due to food choices. Unfortunately, most ketogenic diets, particularly those prescribed to children with epilepsy, are high in processed/artificial ingredients, and foods that many people are sensitive to, such as nuts, dairy, eggs, and biogenic amines.
If you are a psychiatric clinician interested in recommending the ketogenic diet to a patient, I highly recommend collaborating with a dietitian, primary care provider, or specialist with expertise in ketogenic diets. If there isn’t such an individual in your area, there are other options, including specialty dietitians and medical centers offering on-line consultation.
I’m happy to address any questions below in the comments section.