Ketogenic diet
From Wikipedia, the free encyclopedia
The ketogenic diet is a high fat, adequate protein, low carbohydrate diet, primarily used in the treatment of difficult-to-control (refractory) epilepsy in children. The diet mimics aspects of starvation by forcing the body to use fat rather than carbohydrate as an energy source. The body produces excess ketone bodies, a state known as ketosis. The "classic" ketogenic diet contains a 4:1 ratio by weight of fat to combined protein and carbohydrate. To achieve this, a number of foods are effectively eliminated (for example, starchy fruits and vegetables, bread, pasta, grains and sugar). A variant known as the MCT diet uses a form of coconut oil that is very high in medium-chain triglycerides (MCT; most dietry fat contains long chain triglycerides). This oil has a strong ketogenic effect, which allows some relaxation of the regime.
Developed in the 1920s, its popularity waned with the introduction of effective anticonvulsant drugs. In the mid 1990s the Hollywood producer Jim Abrahams, whose son's severe epilepsy was effectively controlled by the diet, created the Charlie Foundation to promote it. Publicity included an appearance on NBC's Dateline programme and ...First Do No Harm (1997), a TV movie starring Meryl Streep. The foundation funded a multicentre study that was published in 1996, which marked the beginning of renewed scientific interest in the diet.
The efficacy of the diet has not been tested in a large, double-blind, randomised controlled trial. Such a trial is regarded as unfeasible and possibly unethical as meta-analysis of the many uncontrolled prospective and retrospective trials provides sufficient evidence to recommend clinical use. In children with refractory epilepsy, the ketogenic diet is more likely to be effective than trying an alternative anticonvulsant drug. There is some evidence that adults may benefit too, and that a less strict diet (such as a modified Atkins) might be effective.
History
The ketogenic diet is a mainstream therapy that was scientifically developed to improve on the success and limitations of the non-mainstream use of fasting to treat epilepsy. Discarded as irrelevant in a world with numerous anticonvulsant drugs, the diet has once again found a role in the effective treatment of refractory epilepsy (epilepsy that cannot be brought under control after adequate trials of different drugs) in children.
The ancient Greek physicians employed dietetic regimens to treat disease, including epilepsy, but the treatment of seizures by fasting was not a popular therapy. In the Hippocratic collection, the author of On the Sacred Disease recommends diet and drugs over supernatural therapies. In the same collection, the author of Epidemics describes the case of a man whose epilepsy is cured as quickly as it had appeared, through complete abstinence of food and drink. Galen wrote, "one inclining to epilepsy should be made to fast without mercy and be put on short rations". He believed an "attenuating diet" might afford a cure in mild cases and be helpful in others.[8]
The first modern study into fasting as a treatment for seizures was in France in 1911.[9] Around this time, the exponent of physical culture, Bernarr Macfadden, popularised the use of fasting as a means of restoring health. His disciple, the osteopath physician Hugh Conklin, of Battle Creek, Michigan, began to treat his epilepsy patients by fasting. Conklin believed that epileptic seizures were caused when a toxin, secreted from the Peyer's patches in the intestines, was discharged into the bloodstream. A fast would allow this toxin to dissipate and typically lasted for 18 to 25 days. Conklin probably treated hundreds of epilepsy patients with his "water diet" and boasted of a 90% cure rate in children (a rate that steadily declined with patient age). Later analysis of Conklin's records show 20% achieved seizure freedom and 50% had improvement.
Conklin's ideas were adopted by neurologists in mainstream practice. In 1916, a Dr McMurray wrote to the New York Medical Journal claiming to have successfully treated epilepsy patients, since 1912, with a fast and then a starch- and sugar-free diet. In 1921, prominent endocrinologist H. Rawle Geyelin reported his experiences to the American Medical Association convention. He had seen Conklin's success first-hand and attempted to reproduce the results in 36 of his own patients. He had similar results, but there was no long-term follow-up. Further studies in the 1920s indicated that seizures generally returned after the fast. John Howland, professor of paediatrics at Johns Hopkins Hospital, received a gift of $5000 from his brother Charles, whose son had been successfully treated by Conklin. The money was to be used to scientifically study "the ketosis of starvation", a task undertaken by neurologist Stanley Cobb and his assistant William G. Lennox.
Diet
Rollin Woodyatt, reviewing the current research on diet and diabetes, reported in 1921 that "acetone, acetic acid and beta-hydroxybutyric acid appear … in a normal subject by starvation, or a diet containing too low a proportion of carbohydrate and too high a proportion of fat." Russel Wilder, at the Mayo Clinic, built on this research and coined the term ketogenic diet to describe a diet designed to produce ketonemia through an excess of fat and lack of carbohydrate. The idea was to maintain the benefits of fasting over a much longer period. His trial, in 1921, on a few epilepsy patients was the first use of the ketogenic diet as a treatment for epilepsy. Wilder's colleague, paediatrician Mynie Peterman, formulated the "classic" diet, with a ratio of one gram of protein per kg of body weight in children, 10–15 g of carbohydrate per day, and the remainder of calories from fat. Peterman's work, in the 1920s, established the techniques for induction and maintenance of the diet, as well as documenting both positive and negative side-effects. Meanwhile, the Massachusetts General Hospital, under Fritz Talbot, established their own ketogenic diet programme, which is very similar to the current one at Johns Hopkins Hospital. Talbot discovered that the ideal therapeutic ratio of fat to combined protein and carbohydrate was 4:1 and first used urine testing to monitor the level of ketosis.
Anticonvulsants
In the 1920s and 1930s, the only anticonvulsant drugs were the sedative bromides (1857) and phenobarbital (1912). The ketogenic diet was seen as an important and mainstream therapy, and widely used. This changed in 1938 when H. Houston Merritt and Tracy Putnam discovered phenytoin (Dilatin), and the focus of research shifted to discovering new drugs. With the introduction of sodium valproate in the 1970s, neurologists had drugs that were effective across a broad range of epileptic syndromes and seizure types. The use of the ketogenic diet, already restricted to difficult cases such as Lennox-Gastaut syndrome, declined further.
The ketogenic diet's severe carbohydrate restrictions made it difficult to produce palatable meals, leading to problems with compliance. During the 1960s, the properties of medium-chain triglycerides (MCT) were discovered: these oils are are considerably more ketogenic than normal dietary fats (which are mostly long-chain triglycerides); they are rapidly absorbed and have a high caloric value. In 1971, Peter Huttenlocher devised a diet with sufficient MCT oil to induce ketonuria and tested it on a dozen children and adolescents with intractable seizures. The oil was mixed with at least twice its volume of skimmed milk, chilled, and sipped during the meal or incorporated into food. About 60% of the diet's calories came from the MCT oil, and this allowed more protein and up to three times as much carbohydrate as a 3:1 classic diet. Most children improved both in seizure control and alertness; results were similar to the classic ketogenic diet. Gastrointestinal side-effects were a problem, which led one patient to abandon the diet, but meal preparation was easier and compliance generally good. The MCT diet replaced the classic ketogenic diet in many hospitals, though some devised diets that were a combination of the two.
Revival
The ketogenic diet achieved national media exposure in October 1994 when NBC's Dateline television programme reported the case of Charlie Abrahams, son of Hollywood producer Jim Abrahams. The 2-year-old had intractable epilepsy that remained undefeated by mainstream and alternative therapies. Abrahams discovered a reference to the ketogenic diet in an epilepsy guide for parents and brought Charlie to the Johns Hopkins Hospital, which was one of the few institutions to still offer the therapy. Under the diet, Charlie's epilepsy was rapidly controlled and his developmental progress resumed. This inspired Abrahams to create the Charlie Foundation to promote the diet and fund research. A multicentre prospective study began in 1994 and was presented to the American Epilepsy Society in 1996. There followed an explosion of scientific interest in the diet. In 1997, Abrahams produced a TV movie, ...First Do No Harm, starring Meryl Streep, in which a young boy's intractable epilepsy is successfully treated by the ketogenic diet.
As of 2007, the ketogenic diet is available from around 75 centres in 45 countries. The form of classic or MCT ketogenic diet offered varies with the hospital and also culturally. Less restrictive variants, such as the modified Atkins diet, have emerged as alternatives, particularly among older children and adults. The ketogenic diet is also under investigation for the treatment of a wide variety of disorders other than epilepsy.
Efficacy
Early studies showed high success rates: in one study in 1925, 60% of patients became seizure free, 35% had a greater than 50% decrease in seizure frequency. These were retrospective studies, that only report on patients who stuck with the diet. Recent studies show that patients and their carers give up if the diet is not effective. Therefore, even modern retrospective studies show a high success of around 29% seizure freedom. The patient group in older studies is also different; modern patient groups tend to study children with refractory epilepsy. Older protocols generally had a much longer initial fast (with the aim of losing 5–10% body weight) and had restricted calorie intake.
The biggest modern study with intent-to-treat prospective design was published in 1998. The Johns Hopkins Hospital studied 150 children for at least 12 months. By three months, 25 patients had dropped out, 26% had a good reduction in seizures (50–90% reduction), 31% had an excellent reduction (90–99% reduction) and 3% became seizure free. By twelve months, 67 patients had dropped out, 23% had a good reduction, 20% had an excellent reduction and 7% were seizure free. In the same year, a multicentre study of 51 children showed similar efficacy, proving that the results could be repeated by other institutions.
A meta-analysis of studies for the Blue Cross and Blue Shield Association in 2000confirmed that about half the children starting the diet will achieve at least a 50% reduction in seizure frequency. About half drop out by twelve months, and these are overwhelmingly patients who had less than 50% reduction.
The success of the diet is measured clinically by reported seizure reduction. Although urinary ketone levels are checked daily, the levels do not correlate with seizure effect. They are useful in detecting that ketosis has been achieved, and for spotting issues with compliance. Electroencephalogram (EEG) changes are also not a reliable indicator of seizure protection. There is no relationship between outcome and age, sex, principle seizure type or initial EEG. Adults can benefit too, though adherence to the regime becomes more difficult with adolescence. Despite this, if the patient achieves a good reduction in seizure frequency, they will stick with it. For patients who benefit, half will achieve a seizure reduction within five days (if the diet starts with an initial fast of one to two days), three quarters achieve a reduction within a fortnight and 90% achieve a reduction within 23 days. If the diet does not begin with a fast, the time for half of the patients to achieve an improvement is longer (a fortnight) but the long-term seizure reduction rates are unaffected. The initial fast has been likened to an intravenous loading dose of anticonvulsant, and may be particularly beneficial where there is some medical urgency that outweighs the increased risk of acidosis and hypoglycemia. If no improvement is seen within two months, it is likely that the diet has failed.
The lack of randomised controlled trials meant that a Cochrane review in 2003 concluded that the diet was merely "a possible option" in the treatment of intractable epilepsy. Long-term blinding is made difficult by to the nature of the diet. However, a short-term blinded study is possible, by spoiling the ketogenic diet with the use of a glucose-sweetened drink. Children may be randomised to receive a drink containing glucose or one containing an artificial sweetener. Other study options being trialled include a controlled parallel-group where patients are randomised to receive the diet after a short interval (4 weeks) or a long interval (16 weeks). A long-term randomised placebo-controlled trial is not feasible and may be unethical, as meta-analysis of the many uncontrolled prospective and retrospective trials indicates sufficient evidence to recommend clinical use. Children with refractory epilepsy are more likely to find the ketogenic diet to be effective than trying an alternative anticonvulsant drug.
Indications and contra-indications
The ketogenic diet is indicated as an adjunctive treatment in children with drug-resistant epilepsy.[16][17][18] The ketogenic diet is endorsed by national guidelines in Scotland,[18] England and Wales[16] and by US insurance companies.[14][19][20] The ketogenic diet is a first-line therapy for patients with seizures due to pyruvate dehydrogenase (E1) deficiency and glucose transporter 1 deficiency syndrome. Both these disorders prevent the body using carbohydrate as fuel, resulting in a dependency on ketone bodies.
In the UK, the National Institute for Health and Clinical Excellence state that the diet should not be recommended for adults with epilepsy due to insufficient evidence of efficacy. The conditions pyruvate carboxylase deficiency and porphyria are absolute contraindications to the ketogenic diet. Other conditions that generally contraindicate are defects in fatty acid oxidation, certain mitochondrial cytopathies, and known carnitine deficiencies.
Interactions
Most children who start the ketogenic diet have already tried six or seven anticonvulsants and are typically taking two. There are no harmful or beneficial interactions between anticonvulsant drugs and the ketogenic diet. A trial in 2007 of 30 children studied the combination of ketogenic diet and the vagus nerve stimulator (VNS). About half the children were already on the ketogenic diet and had the VNS added to their therapy; the other half had the opposite sequence. About two thirds of the children had a greater than 50% reduction in their seizures as a result of combining these therapies. Those who responded well, generally did so within a month. No significant side effects were noted and as with other studies, the children who did not respond well tended to be the ones who subsequently discontinued the diet.
Adverse effects
The ketogenic is not a benign holistic or natural treatment for epilepsy; as with any serious medical therapy, there may complications. These are generally less severe and less frequent than with anticonvulsant medication or surgery. Common but easily treatable side effects include constipation, lack of appropriate weight gain for age, low-grade acidosis, and hypoglycemia if there is an initial fast. Cholesterol may increase by around 30%.
About 1 in 20 children on the ketogenic diet will develop kidney stones (compared with 1 in several thousand for the general population). A class of anticonvulsants known as carbonic anhydrase inhibitors (topiramate, zonisamide) are known to increase the risk of kidney stones, but the combination of these anticonvulsants and the ketogenic diet does not appear to elevate that risk. The stones are treatable and do not lead to discontinuation of the diet. Oral potassium citrate is preventative and had no clear disadvantages; its routine use is under investigation. Kidney stone formation (nephrolithiasis) occurs on the diet for four reasons. Excess calcium in the urine (hypercalciuria) occurs due to increased bone demineralisation with acidosis (bone phosphate acts as an acid buffer) as well as increased calcium excretion by the kidney. There is an abnormally low concentration of citrate in the urine (hypocitraturia), which normally helps to dissolve free calcium. The urine has a low pH, which stops uric acid from dissolving, leading to crystals that act as a nidus for calcium stone formation. Many institutions restrict fluids on the diet to 80% of normal daily needs.
Initiation
The best documented protocol for initiating the diet is the one practised at the Johns Hopkins Hospital (JHH). At initial consultation, patients are screened for conditions that may contraindicate the diet. Dietary history is obtained and the parameters of the diet selected: the ketogenic ratio, the calorie requirements, and the fluid intake. The day before admission to hospital, carbohydrates are decreased and the patient begins fasting after their evening meal. On admission, they may drink but not eat until dinner, which consists of an "eggnog" restricted to one-third of the usual calories for a meal. The following breakfast and lunch are similar, and on the second day, the dinner is increased to an "eggnog" with two-thirds of the usual calories. By the third day, dinner contains the full calorie quota and is a standard ketogenic meal (not "eggnog"). After a ketogenic breakfast on the fourth day, the patient is discharged.
During their stay in hospital, the patient has their glucose levels checked and is monitored for signs of symptomatic ketosis (which can be treated with a small quantity of orange juice). Lack of energy and lethargy are common but disappear by two weeks. The parents attend classes over the first three full days, covering nutrition, managing the diet, preparing meals, avoiding sugar and handling illness. Medicines are exchanged for carbohydrate-free formulations, where possible.
Deviations from the Johns Hopkins protocol are common. If there is no initial fast, the time to reach ketosis is longer (but still achieved within five days), and there were fewer initial complications. The initiation can be performed using outpatient clinics rather than requiring a stay in hospital. Fluid restriction may be relaxed, leading to fewer cases of dehydration. Rather than increasing meal sizes over the three day initiation, some institutions maintain meal size but alter the ketogenic ratio from 2:1 to 4:1.
Maintenance
At Johns Hopkins Hospital, outpatient clinics are held at 3, 6, 12, 18 and 24 months. Throughout the diet, telephone contact with the nutritionist helps with fine tuning. A short-lived increase in seizure frequency may occur during illness or if ketone levels fluctuate. The diet may be modified if seizure frequency remains high, or the child is losing weight.
Discontinuation
About 10% of children on the ketogenic diet achieve seizure freedom and many of them also manage to reduce or discontinue anticonvulsant drugs. At around two years on the diet, or after six months of seizure freedom, the diet may be gradually discontinued over a two to three month period. This is done by lowering the ketogenic ratio until urinary ketosis is no longer detected, and then lifting all calorie restrictions. Children who discontinue after achieving seizure freedom have about 20% recurrence risk of seizures. The length of time until recurrence is highly variable but averages two years. This recurrence risk compares with 10% for resective surgery (where part of the brain is removed) and 30–50% for anticonvulsant therapy. Of those that have a recurrence, just over half regain their seizure freedom either with anticonvulsants or by returning to the ketogenic diet. Recurrence is more likely if, despite seizure freedom, an EEG shows epileptiform spikes, or if an MRI shows focal abnormalities (for example, children with tuberous sclerosis). Such children may remain on the diet longer than normal, and it has been suggested that children with tuberous sclerosis who achieve seizure freedom could remain on the ketogenic diet indefinitely.
Variants
Classic
The ketogenic diet is calculated by a dietician for each child; age, weight, activity levels, culture and food preferences all affect the meal plan. A computer program may be used to help generate meals. A typical day of food for a child on a 4:1 ratio, 1500 calorie ketogenic diet:
- Breakfast: egg with bacon
- 28 g egg, 11 g bacon, 37 g of 36% heavy whipping cream, 23 g butter, 9 g apple.
- Snack: peanut butter ball
- 6 g peanut butter, 9 g butter.
- 28 g tuna fish, 30 g mayonnaise, 10 g celery, 36 g of 36% heavy whipping cream and 15 g lettuce.
- 18 g of 36% heavy whipping cream, 17 g sour cream, 4 g strawberries and artificial sweetener (e.g., Splenda).
- 22 g minced (ground) beef, 10 g American cheese, 26 g butter, 38 g cream, 10 g lettuce and 11 g green beans.
- 25 g of 36% heavy whipping cream, 9 g egg and pure vanilla flavouring.
A ketogenic "eggnog" is used during induction and is a drink with the required ketogenic ratio. For example, a 4:1 ratio eggnog would contain 60 g of 36% heavy whipping cream, 25 g egg, vanilla and saccharin flavour. This contains 245 calories, 4 g protein, 2 g carbohydrate and 24 g fat (24:6 = 4:1).
MCT oil
Normal dietary fat contains long-chain triglycerides (LCT). Medium-chain triglycerides are more ketogenic than LCTs. Their use allows the fat content to be lowered and consequently greater protein and carbohydrate intake. The MCT ketogenic diet is identical in efficacy to the classic diet; abdominal bloating and diarrhoea are more common, but constipation is less of a problem. A combination of the classical and MCT diet may be used, which aims to avoid the disadvantages of either.
Modified Atkins
A modified Atkins diet is effective in children and adults. The diet consists of 60% fat, 30% protein and 10% carbohydrate by weight; calories are not restricted. Carbohydrate is limited to 10 g per day for at least one month, and gradually increased to 10% if this limitation is not tolerated. Consistently strong ketosis is more difficult to achieve than on the ketogenic diet; patients with wildly fluctuating urinary ketones have unfavourable seizure outcomes. Achieving the balance of fat, protein and carbohydrate can be difficult; patients may consume the appetising protein (meat) and leave or vomit the fat. Older children and adolescents who refuse the ketogenic diet's restrictions may tolerate the modified Atkins diet.
Prescribed formulations
Infants, or patients fed via a gastrostomy tube can also be fed a ketogenic diet. A prescribed powdered formula, such as KetoCal, can be made up into a feed that has none of the palatability issues. KetoCal is a nutritionally complete feed containing milk protein and supplemented with amino acids, fat, carbohydrate, vitamins, minerals and trace elements. It is used to administer the 4:1 ratio classical ketogenic diet in children over 1 year. Each 100 g of powder contains 73 g fat, 15 g protein and 3 g carbohydrate, and is typically diluted 1:5 with water. The formula is available unflavoured or in an artificially sweetened vanilla flavour and is suitable for tube or sip feeding.
Mechanism of action
Many hypotheses have been put forward to explain how the ketogenic diet works; it remains a mystery. Disproven hypotheses include systemic acidosis, electrolyte changes and hypoglycemia. Changes in neurotransmitter levels occur and cerebral energy state is improved. Although many biochemical changes are known to occur in the brain of a patient on the ketogenic diet, it is not known which of these has an anticonvulsant effect. The lack of understanding in this area is not dissimilar to the situation with anticonvulsant drugs.
On the ketogenic diet, carbohydrates are severely restricted so cannot provide for all the metabolic needs of the body. Most energy is instead produced by a high rate of fatty-acid oxidation in the cell mitochondria. This produces large amounts of acetyl-CoA, which the liver uses to synthesize the three ketone bodies ß-hydroxybutyrate, acetoacetate and acetone. The brain is normally fueled solely by glucose; fatty acids do not cross the blood-brain barrier. Ketone bodies can enter the brain but they are not used preferentially to glucose. The ketone bodies are converted to acetyl-CoA and subsequently to adenosine triphosphate (ATP) as part of the Krebs cycle within brain mitochondria.
The ketone bodies are possibly anticonvlusant in themselves; acetoacetate and acetone protect against seizures in animal models. The ketogenic diet results in adaptive changes to brain energy metabolism that increases the energy reserves; ketone bodies are a more efficient fuel than glucose, and the number of mitrochondria is increased. This may help the neurons to remain stable in the face of increased energy demand, and may also confer a neuroprotective effect.
The ketogenic diet has been studied in at least 14 rodent animal models of seizures. It is protective in many of these models and has a different protection profile to any known anticonvulsant. This, together with studies showing its efficacy in patients who have failed to achieve seizure control on half a dozen drugs, suggests a unique mechanism of action.
Anticonvulsants suppress epileptic seizures but they neither cure nor prevent the development of the inherent seizure susceptibility. The developement of epilepsy (epileptogenesis) is a process that is poorly understood. A few anticonvulsants (valproate, levetiracetam and benzodiazepines) have shown antiepileptogenic abilities in animal models of epileptogenesis. However, no anticonvulsant has ever achieved this in clinical trial in humans. The ketogenic diet has been found to have antiepileptogenic properties in rats.
Other applications
A number of rare metabolic disease may benefit directly from the ketogenic diet. Case reports on indicate a possible use in treating brain tumours (astrocytomas). Migraine headaches, autism and depression have been shown to benefit in small case studies. Animal models of Alzheimer's disease and amyotrophic lateral sclerosis (ALS) show benefit.
