“Although treatment strategies for children with autism spectrum disorders are under development, the ketogenic diet is available now and could offer multiple benefits. For example, children with autism and uncontrolled seizures have few options, and this research suggests a ketogenic diet could reduce seizures and improve behavior.” Susan Masino, PhD, Trinity College, Harford, Conn. 2013.

None of the commonly tried diets for autism, including gluten-free and casein-free, has proven effective in controlled clinical studies. The available medical treatments address the comorbid behavioral conditions related to autism but do not affect the core symptoms that result in the autism diagnosis. The occurrence of autism is currently one in 88 people in the U.S. and continues to increase.

The ketogenic diet, because of its very restricted carbohydrates and limited proteins, forces the body to use fat rather than glucose as an energy source and thus produces a metabolic state similar to fasting. Ketogenic diets have been used successfully to treat epilepsy in people since 1921 and epilepsy is common in people who have autism. A study published in 2003 demonstrated that some children with autism on an intermittent ketogenic diet experienced a significant reduction in autistic symptoms.

A recent case report published in The Journal of Child Neurology describes a child diagnosed at age four with autism who had limited response to other treatments but gained benefit on a gluten-free, casein-free diet.

She developed seizures after starting puberty and continued to have seizures despite the use of anti-seizure medication. With the assistance of our nutritionist, Beth Zupec-Kania, her gluten-casein-free diet was modified to make it ketogenically balanced using medium-chain triglycerides as the main source of fat. Significant improvements in multiple areas are described:

In addition to improvement in seizures, there was a 60-pound weight loss subsequent to initiation of the gluten-free casein-free ketogenic diet, as well as improved cognitive and language function, marked improvement in social skills, increased calmness, and complete resolution of stereotypes. Intravenous immunoglobulin treatments continued, since delays in delivery of immunoglobulin G were associated with illness accompanied by convulsion. Based on the clinical improvement, the electroencephalogram improvement, and development of a side effect involving extreme medication associated somnolence, anticonvulsant medication doses were reduced (first lamotrigine by 50%, and then ethosuximide by 25%) without worsening of seizures. Cholesterol was 152 mg/dl before starting diet, and was 160 mg/dl after more than a year on the diet.

Masino and her colleagues, who had been studying the ketogenic diet for a variety of neurological disorders, learned that certain biological compounds increase in the brain while on the ketogenic diet. She hypothesized that a compound called adenosine may be key to the diet’s effects and may also be helpful in alleviating symptoms of autism. Masino tested the behavioral effects of a ketogenic diet using a mouse model that has characteristics of autism, including low sociability, poor perception of social cues and highly repetitive behaviors. As expected, the mice fed a normal diet displayed behavioral symptoms of autism but after feeding a ketogenic diet to the autistic mice for three to four weeks, autistic behaviors reversed significantly and they behaved like normal mice. The mice fed a ketogenic diet were more social and spent more time with other mice; they were also able to perceive social communication cues regarding food choice. Ketogenic diet-fed mice also spent significantly less time grooming, indicating reduced repetitive self-directed behavior.  “Together, these behaviors represent the complement of core symptoms used to diagnose autism, and all were reversed by the ketogenic diet,” said Masino.

Because seizures are common in persons with autism spectrum disorders, it was important to determine if the effects of the diet were related to the ketogenic diet’s well-known ability to treat epilepsy. Based on tests, including EEG recordings and induced seizures, the researchers found that the mice used in the study did not have epilepsy. Therefore, the behavioral effects of the diet are independent of its anti-seizure effects.

In addition to Masino’s hypothesis regarding adenosine, inflammation is considered an important factor in causing autism: a general decrease in inflammation due to the ketogenic diet could underlie its effectiveness against autistic symptoms. Masino’s laboratory proved for the first time that a ketogenic diet can reduce inflammation in rats. Other research groups are now looking at the effects of the ketogenic diet on inflammation in autism.

Many clinicians have pointed out that people with autism have a narrow range of dietary preferences and therefore a restricted ketogenic diet may not be well tolerated. Our nutritionist, Beth Zupec-Kania, has found success by modifying the diet for children with autism to allow for preferences but still maintain ketosis. For example use of medium-chain triglycerides allows for generous amounts of carbohydrate. Although the classic ketogenic diet is gluten-free, many with autism follow casein-free restrictions as well. This added restriction is entirely feasible. Tasty beverages, ketogenic crackers and breads can be prepared from nut flours and oils. In addition, the volume of food on these high-fat, low-carbohydrate diets is much smaller than typical servings which assists with completion of meals.

Evidence from these studies and reports indicates that at least certain types of autism respond to metabolic diet treatments. At this time, with few treatment options, new treatment strategies for autism are needed. Additional research on the ketogenic diet may offer clues to reversing symptoms of autism.