We are glad we have our son Marchant on the ketogenic diet. He is lively and alert and five months after starting the diet he became fit free. Marchant has been a ‘guinea pig’ in our learning about the diet, with our current knowledge he could have been fit free much sooner.

Unfortunately, we do not think the diet has very much chance of working in the UK as it is currently applied, although this could be simply rectified by providing training for paediatricians and dieticians, and better support for parents .

We sincerely hope our experiences can be used to help other children and we would be very happy to discuss our experiences further.

The ketogenic diet

Background to the diet

Over 70 years ago Geyelin found that many patients with epilepsy remained free of seizures while fasting and for several months after return to a normal diet. It was proposed that a ‘ketogenic diet’, high in fat and low in carbohydrate, should be used to treat epilepsy. Livingston used the classical diet at Johns Hopkins for a period of 41 years and reported that it controlled 54% of the 975 cases he studied.

The most common form of the diet is the classical diet as practised at Johns Hopkins. This has a restricted calorie intake, and its main variable is the ketogenic ratio, the ratio by weight of fat to carbohydrate plus protein. This ratio can vary from 2.5:1 up to 5:1, but is most usually in the range 3:1 to 4:1.

In the 1970’s Huttenlocher introduced a significantly different diet. This uses MCT oil to make it more palatable. Since MCT oil is more soluble than other fats it is more quickly metabolised and therefore ketosis is established even though the percentage of carbohydrate in the diet is higher. However some people were found to be intolerant of high amounts of MCT oil and in the 1980s Dr Schwartz introduced the MCT modified diet. It is the MCT and modified MCT diets which have mainly been used in the UK.

There are now three main versions of the ketogenic diet:

* Recommended Daily Allowance
** The percentages of fat, carbohydrate and protein vary with this diet as it is fine tuned for the individual - although the ratio by weight of fats to carbohydrate plus protein is usually about 4:1

Dr Schwartz (Consultant Paediatrician, Central Middlesex Hospital) has been one of the few researchers to compare all three diets. Although her original study⁵ reported all three diets to be equally effective, in a recent discussion with the authors (December 1996) she concluded that in her experience:

The classical diet gives better results than the diets containing MCT oil. Contrary to original thoughts the classical diet is better tolerated (MCT oil can cause stomach cramps and diarrhoea). In her experience patients often switched from the MCT to classical diet, because of taste, never the other way round. Contrary to previous belief, the cost of cream in the classical diet does not make it very expensive for patients.

Dr Schwartz feels that the ketogenic diet has not been popular in the UK because of the move towards low fat diets, even though in long term studies patients on the diet have not had raised cholesterol levels nor evidence of increased arteriosclerosis.

Dr Freeman⁷ is even more critical of the MCT oil diet:

Although the MCT diet has been reported to be equally as effective as the classical ketogenic diet, this has not been our experience at Johns Hopkins. . . Many parents tell us that their child has already been on the ketogenic diet without success. On further questioning, this prior diet usually turns out to have been the MCT diet. We have found some children who continued to have seizures despite tolerating the MCT diet, but who subsequently responded well to the classical ketogenic diet. We have also seen many children and families who could not tolerate the MCT diet, but who did well on the classical ketogenic diet.

There are at least 11 ketogenic diet centres in the US, and the number is growing rapidly. The main hospitals carrying out research are Johns Hopkins and Stanford. The Johns Hopkins Centre, under the supervision of Dr John Freeman has the most experience and to quote their results:

"The diet will completely control epilepsy in one-third of the children whose seizures are otherwise uncontrollable. In half of the remaining children the diet will either markedly decrease the frequency of seizures or enable medications to be reduced. Many children whose seizures are completely controlled can return to a normal diet in two to three years and remain seizure free without medication. They have either been cured of epilepsy or have recovered from their previously uncontrollable epilepsy. Permanent control often begins within days of diet initiation."

Dr Schwartz has found that some children stop having fits as soon as they are in ketosis, this group tend to be sensitive to the amount of ketones produced and can have fits if ketones are too high, another group can take over a month to show any results and these children’s fits decline gradually. Dr Freeman feels the diet can sometimes take 6 months to ‘fine tune’; in fact the diet has to be constantly fine tuned.

Unfortunately, despite a lot of speculation, nobody knows how the diet works. Dr Freeman feels the diet's effectiveness lies in other influences on metabolism. For example, while the brains of infants, children, and adults burn glucose almost exclusively, the foetus and new-born are able to exist on the metabolism of fats. Does the ketogenic diet enable the brain to revert to a more primitive form of metabolism? Is the ketogenic diet more effective in children than in adults because of the younger brain’s capability to metabolise fats? Dr Schwartz et al studied the biochemical effects of the ketogenic diet in 59 patients. Their comprehensive assessment failed to identify any obvious clue to the mechanism of the beneficial effect of the diet. They suspect it may be mediated by changes either in the nerve cell lipid membranes which decrease neuronal excretion or in the local production or clearance of neurotransmitters. Clearly, more research is needed to explore the mechanisms by which the ketogenic diet achieves its results.

When we looked at the three diets we wondered:

Why the percentage carbohydrate was so high in the MCT diets (19% cf 5% for the classical diet)?

Why MCT oil was not used in the classical diet since it raises ketones and thereby may allow 100% RDA of calories to be used.

Latest thinking on the diet

Our information has been received from Dr Freeman, Dr Schwartz, a conference attended in the USA and from the 20+ e-mails we received each day from parents all over the world fine tuning their children’s diets:

• For children whose fits are difficult to control Dr Freeman is advocating total elimination of carbohydrates from the diet and replacement with protein.

• John Hopkins have also started using some MCT oil in the classical diet to increase ketones.

• Carnitine has been used by some centres as a supplement and parents have reported that it has increased ketones. Carnitine is important because it is necessary to transport fat through the cell walls to the mitochondria. Carnitine also helps alleviate constipation (a side effect of the diet). Valproate and depakote can cause a deficiency in carnitine. The ketogenic diet causes deficiency of carnitine because reduced protein means less carnitine direct from the diet. Carnitine is not degraded in the metabolic process but it is lost by excretion (~ 25% daily) and so must be replaced on a continuous basis.

Best candidates for the diet

The following appear to be criteria for the use of the diet:

• 1 - 12 years old (although Johns Hopkins are increasingly working with children under one)

• Organised, capable parents

• Fit type less crucial (however from our experience with Marchant it is easier to assess and fine tune the diet when children are having frequent fits)

The normal practice in the US is not to consider the diet unless at least two drugs have been tried up to toxic levels and do not work. We would question this approach.

Side effects of the diet

A number of side effects of the diet have been reported:

• Acute acidosis, dehydration, vomiting and hypoglycaemia
• Constipation
• Renal stones, gall stones (use hem sticks to measure blood in urine to get early indication of kidney stones). Stones can be caused if calcium supplements are too high)
• Optic neuropathy
• Hyperlipidemia
• Decreased appetite
• Slow linear growth

However, in comparison with anti-epileptic drugs, the diet does not have narcotic effects, nor interfere with mental development.

Fine tuning Marchant’s diet

Background

Marchant was a term baby weighing 2.7kg. His birth had no complications but he had a low apgar score (7 at 15 mins) and was taken to the NICU. He had fits for the first three days, these were controlled by phenobarbitone and paraldehyde. His MRI scan was normal but the EEG was grossly abnormal. He developed well for the first 4 - 5 months.

He was diagnosed with infantile spasms at 6½ months, although he was probably having ‘twitches’ from 4 months. The fits involved all four limbs and were clonic with up to 10 cycles each. He was treated with vigabatrin for two weeks, this increased his fits. He was then given daily injections of ACTH (60 units) for 5 days, his fits stopped but the side effects were horrendous, Marchant screamed for 5 days and nights and needed sedation. After a week without ACTH Marchant’s fits returned. The week before he started the diet Marchant was having 50-90 fits a day, the fits affected all his limbs and his eyes flickered; they were less violent than before, but the number of cycles in a fit increased. Some of these fits were salaam attacks and we began to see further retardation.

Whilst Marchant was on vigabatrin and his fits were one every 5 - 10 minutes, he was tube fed for one night. Marchant was sick all night and so effectively starved for 12+ hours. We noticed the next morning that his fits were greatly reduced and they did not increase again until after he was breast fed. It was that experience that finally convinced Sue that we should try the ketogenic diet. Interestingly, Dr Freeman in his book ³ states that it has been noticed that when a child’s fits reduce after vomiting or starving, this is a good indication that the diet may work.

Prior to the diet Marchant had been a poor, erratic feeder. As a baby he would often scream and come off the breast. He had reflux and was given gavascon. At 5 months old he started solids. From 4 months to 8 months, ie prior to the diet, Marchant had only put on 1lb.

Results and experience

When Marchant’s fits returned subsequent to the ACTH treatment, Dr Jardine, our neurologist, confirmed that he could start the ketogenic diet. However, a bed was not available, so we had to wait a further two weeks, during which time Marchant’s fits reached 50 - 90 a day. In retrospect, we feel we should have pressed to have Marchant admitted sooner.

The dietician did not think it was necessary to see us before the diet started. This was a pity because we were anxious and had lots of questions about the details of the diet, how to prepare food, scales to buy and, most importantly, when to stop giving him food before the hospital admission.

When Marchant was admitted, the hospital did not have food for the diet, not even cream, so Sue had to go to the shops from the hospital to buy the food specified by the dietician. Subsequently, Marchant’s feeds were delivered from central catering; this was not satisfactory, because the food could not be relied upon to be compliant with the diet, and the timing was not helpful for feeding a difficult baby. After three days, the dietician agreed that it would be easier for us to manage the diet from home and Marchant left without having achieved sustained ketosis.

In the hospital Marchant was initially put on the MCT modified diet. However we found that it was extremely difficult to keep Marchant in ketosis because the diet was not balanced at each meal (ie percentage fat: carbohydrate: protein). The main problem was that he was given formula milk which is high is carbohydrate and he was bottle feeding in the night.

Marchant was given the full quantity of MCT oil after day two of the diet. This made him very sick and he rejected all solid food. We subsequently found out that many children are intolerant of MCT oil and it must be introduced into the diet very gradually.

After only a week on the diet we decided to reduce the carbohydrate to 10%. We also took formula milk out of Marchant’s diet. As he refused solids we decided to make a liquid blend of cows milk, cream, egg (best quality protein) banana (a complex carbohydrate, so more slowly released) and MCT oil.

We told the dietician that we could not work with formula milk. She was not happy giving Marchant cows milk as he was under one year but realised the problem created by the carbohydrate in formula milk. We suggested contacting Dr Schwartz, Dr Jardine was happy with this.

Dr Schwartz confirmed that it was not possible to use formula milk in the diet. She said Marchant should go on to the classical diet. We were reluctant to do this because of his poor weight; the classical diet only provides 75% of the daily recommended calorie allowance. Dr Schwartz suggested we could follow the diet for his ‘expected’ weight. The results we had on our own diet had been good (down from 50-90 fits to 20 very much less severe fits) so we decided to continue with our own ‘Pringle diet’. We reduced the percentage of MCT oil to improve Marchant’s tolerance and reduced the carbohydrate in the diet. Marchant’s fit control was good when his ketones were high but unfortunately it was difficult to get high ketones in the morning. This is a common problem as the natural pattern for ketone production is high in the afternoon and evening and low during the night and morning. The problem of low ketones in the night can be overcome by giving the last feed as late as possible. However it is interesting to note that it is particularly hard to get high ketones for children under one and we suspect this is the reason the diet has generally not been used for under ones.

We eventually had Marchant on the classical diet for his expected weight but still used MCT oil to increase his ketones. Unfortunately in the middle of January Marchant had a viral infection and would not eat. We worried that this sickness was due to the MCT oil and removed this from his diet, so he was on the simple classical diet. When he first refused food his fits stopped but he became increasingly ill and vomited. His fits went back up to 10-15 a day, but his fits were no longer in all limbs but just minor twitches in one arm. He was conscious through these jerks and even watched his arm twitch!

Marchant went into hospital for a reflux test which showed he had mild reflux. Unfortunately during his stay in hospital he was nearly given an electrolyte drip containing glucose and was given cisapride containing sucrose which did increase his fits. The feeding situation worsened and Sue had no other option but to force feed Marchant using a syringe. During this period his fits were erratic and did increase up to ~ 20 a day but the fits were still confined to one arm and Marchant was conscious throughout.

Eventually, in March, when Marchant was one year old he was given an ng (nasal gastric) tube. His fits went up, possibly because of the increased intake, so we eventually reduced his carbohydrate to zero, replacing the carbohydrate by protein (on the advice of Dr Freeman at Johns Hopkins). Dr Freeman also advised us that children under one should have a minimum protein intake of 1.5 g/kg, not the 1g/kg advocated earlier in the literature; he also told us that babies should start on the diet at a 3:1 ratio, not a 4:1 ratio. During this time Marchant had a 24 hour EEG. On the day of his EEG his fits were the most frequent we had seen since we started the diet. However the next day Marchant had a temperature of 103º and developed an awful rash so probably he was going down with a virus when he had his EEG. We were absolutely astonished and delighted that the EEG report stated that, despite thorough examination no seizure activity could be detected.

In April Marchant was finally given a carnitine supplement . Carnitine makes the diet more effective and it also helps with constipation problems. In retrospect, it is a pity that the carnitine supplement was not given much earlier, and that it was delayed to test for deficiency. By the beginning of May Marchant was fit free.

Unfortunately Marchant has suffered from coughs and chest infections almost continuously since January. The mucus from the infections caused him to be sick; this in turn increases his ketones. High ketones is good for his fit control but if the ketones are too high this also causes him to vomit.

Current resources

The diet could be more widely used and would successfully treat many children whose fits are currently not controlled. Parents all over the world have worked hard to instigate the diet for their children. These people are ‘ordinary parents’, not biochemists, professors or wealthy. Their commitment has come from the enthusiasm of ketogenic diet teams at centres like Johns Hopkins, from seeing other children become fit free, and from the support and encouragement from parent groups. We have had contact with parents who have tried the diet when it has not worked, these parents do not regret trying the diet.

To offer the diet as an option to many more parents will require resources. However, we would like to make some simple suggestions which might improve implementation of the ketogenic diet using the resources currently available.

• Parents need to be properly briefed about the diet and its operation before they commit to the diet.

• The video produced by the Charlie Foundation should be shown to parents (if classical diet followed).

• The dietician should see the family before a child goes into hospital to start the diet to help plan for the child’s homecoming - eg, what types of food to buy/prepare, favourite foods considered, scales to be bought etc.

• The hospital needs to have food available for the diet. The child’s urine should be checked for ketones before starvation is ended.

• Once on the diet, a child’s notes should be carefully marked and the pharmacist briefed to avoid drips and drugs being prescribed which contain carbohydrates.

• There should be clear procedures for when children are ill. We had a brief check list but a more comprehensive procedure should be available for the parents and for other doctors and should be in the child’s notes for hospital admission.

• Carnitine should be prescribed as a supplement at the outset of the diet.

• Better results may be obtained if the classical rather than the MCT modified diet is followed.

Making the diet a success

Neurologists need to fully understand the diet for it to be used successfully. The ketogenic diet should be regarded as a ‘treatment’ rather than a ‘diet’, it is not sufficient, or fair, to leave it in the hands of dieticians. Ketone production has a major effect on the body: it affects how other medications work (eg phenobarbitone) and can cause problems when patients suffer other illnesses. Neurologists also need to know that results with the diet are at least as good as with many of the drugs used.

Parents are crucial to the success of the diet and commitment of parents should not be underestimated. Parents need to be enthused and supported by experienced dieticians and other parents with children on the diet.

Dieticians need training and support to fully understand the diet, their input is imperative. They need to enthuse and educate parents.

The UK would benefit from having a Ketogenic Diet Centre. Such a centre should have as its objectives:

• To promote the use of the ketogenic diet, particularly within the medical community.

• To provide training in the ketogenic diet for doctors and dieticians.

• To maintain a source of knowledge and expertise on the ketogenic diet.

• To co-ordinate support activities for parents across the UK.

• To carry out research on the diet and to promote research on the diet elsewhere in the UK.

• To maintain international links.

Summary of Marchant's diet treatment

References

1. Wilder RM, The Effect of Ketonuria on the Course of Epilepsy. Mayo Clinic Bull, 2:307, 1921
2. Livingston S, The Diagnosis and Treatment of Convulsive Disorders in Children, Springfield, Thomas, 1954
3. Livingston S, Convulsive Disorders in Infants and Children. In Levine, S.Z. (Ed.): Advances in Pediatrics. Chicago, Year Book, 1958
4. Livingston S, Comprehensive Management of Epilepsy in Infancy, Childhood and Adolescence. Charles C Thomas 1972
5. Schwartz RM, Boyes S, Aynsley-Green A: Metabolic Effects of Three Ketogenic Diets in the Treatment of Severe Epilepsy. Dev Med Child Neurol 1989; 31:152-60
6. Kinsman SL, Vining EP, Quaskey SA, Mellits D, Freeman JM, Efficacy of the Ketogenic Diet for Intractable Seizure Disorders: Review of 58 Cases. Epilepsia. 33(6):1132-6, 1992 Nov-Dec
7. Freeman JM, Kelly MT, Freeman JB, The Epilepsy Diet Treatment: An Introduction to the Ketogenic Diet. Demos, 1994
8. An Introduction to the Ketogenic Diet-A Treatment for Pediatric Epilepsy. Video Tape. Charlie Foundation to Help Cure Pediatric Epilepsy, 501 10th Street, Santa Monica, CA 90402. 1994
9. Wheless JW, The Ketogenic Diet: FA(c)t or Fiction. J Child Neurol. 10:6 419. Nov 1995

return to   Marchant's story