Using a ketogenic diet for treatment of seizures in an adult--a case study. (Original Research).
This case study documents the use of a ketogenic diet for the treatment of intractable seizures in a previously healthy 32-year-old man. Although the use of the diet for children is well-documented, it is less common in adults. A carbohydrate-free feeding formula with a fat:protein ratio of 2:1 was administered via a nasogastric tube for 14 days, at which time EEG indicated minimal seizure activity. The patient resumed a normal oral diet. However an oral ketogenic diet was commenced four days later because seizures returned. After two weeks the diet was abandoned because of patient non-compliance. The ketogenic diet can be effective in controlling seizures in adults, however the diet is difficult to maintain and significant resources must be devoted to planning and preparing the diet. Adherence to the diet requires a high level of motivation.
Key words: ketogenic diet, ketodiet, epilepsy, seizure, adult
The ketogenic diet is an established option for the control of intractable seizures in children (1). There are few studies in the literature concerning its use and efficacy in adults.
A 32-year-old man, usually well and employed as a sales representative, was admitted to Royal Prince Alfred Hospital, Sydney, suffering from seizures thought to be related to recreational drug use. He had been using a wide variety of substances including cannabis and Salvia divinorum, both hallucinogens that can affect seizure threshold in susceptible people. For several days, he had experienced headache, fever and increasing confusion. Provisional diagnoses included tumour, viral encephalitis, and rabies. These were excluded after extensive investigations including brain CT, MRI and PET scans, brain and muscle biopsies, and a full-body PET scan. EEG indicated frequent focal seizures arising from both temporal lobes. For the next 45 days, the seizures were managed by keeping the patient in a thiopentone coma. He was intubated and ventilated, and fed by nasogastric tube using a standard feeding formula. EEG indicated that seizure activity was well controlled with anti-epileptic medications (phenytoin, sodium v alproate, topiramate, phenobarbitone and clonazepam), and the patient was extubated and discharged to a rehabilitation facility, before returning home. A few weeks later he was readmitted, now suffering from generalised seizures of unknown cause. It was unclear whether he had used any recreational drugs while at home. Again he was intubated and ventilated, and fed by nasogastric tube while a thiopentone coma was used to suppress seizure activity. Several attempts were made to wean the thiopentone while controlling the seizures with anticonvulsant medications in various combinations. Each time, EEG indicated a return of frequent seizure activity and thiopentone was recommenced. Thirty days later, after limited success in controlling the seizures with levetiracetam and phenobarbitone, it was suggested that a ketogenic diet might provide additional benefit.
A carbohydrate-free enteral formula (Ross Carbohydrate Free, Abbott Laboratories, Kurnell, NSW) was used as a tube feed. This formula provides 3.4 kJ per mL (0.8 cal per mL) with 80% of energy as fat (high-oleic safflower, soy and coconut oils). The amount of this feed that met the patient's energy requirements also met his recommended dietary intake for protein and all micronutrients excepting manganese, and magnesium and selenium were also marginally inadequate. The fat:protein ratio of the formula is 2:1 by weight, and it is recommended that oil is mixed with the formula to produce the desired ratio (usually 4:1 by weight). However, the patient had been experiencing large gastric aspirates, presumed to be a result of the heavy sedation, and it was thought that additional fat might further slow gastric emptying. Unmodified formula was given at first, with the intention of increasing the fat content gradually. However, the patient became ketotic within a few days of commencing the formula, so the unmodified formula was continued. This formula was administered via nasogastric tube for 14 days, during which time sedation was successfully withdrawn. EEG indicated minimal seizure activity, and the patient appeared alert. He then pulled out the nasogastric tube, and it was decided that he could try a normal diet. The patient received an ordinary oral diet for the next four days, during which time he appeared increasingly paranoid. He then experienced several significant seizures, and it was decided that he should resume the ketogenic diet, this time in oral form, in addition to continuing anticonvulsant medications. The possibility of percutaneous endoscopic gastrostomy insertion was discussed with the patient and family, but was not considered to be a realistic option.
A diet aiming for a 4:1 ratio (fat:protein+carbohydrate by weight) was planned and analysed using nutritional analysis software (Xyris Software, Brisbane, FoodWorks, version 2.10.146 2000). A carbohydrate-free supplement of water-soluble vitamins (Intravite, Roche, Frenchs Forest, NSW) and a sugar-free calcium carbonate supplement was prescribed daily because of the difficulty in meeting requirements on a ketogenic diet. It was difficult to achieve the required fat intake, given the limitations of the hospital food service system to supply a large variety of therapeutic foods for one individual patient. As special meals could not be prepared, the correct ratio was achieved by adding fat to existing hospital meals (with high-carbohydrate items omitted). However, the patient disliked this. Thus to increase the amount of fat in the diet, artificially sweetened and flavoured cream 'milkshakes' were provided six times per day. The patient enjoyed these, but the estimated ratio of his diet including these drinks wa s still only 3:1. His family expressed concerns about the effect on his blood lipid concentrations if he were to follow this diet in the long term. The patient's gradual decrease in weight was also a concern.
Education on the diet was provided to the patient and to his partner and mother, who visited frequently, however, compliance was difficult. Psychiatric assessment revealed that the patient was often confused, with occasional psychotic features, and this was attributed to seizure activity revealed on ambulant EEG. He was also on a high dose of levetiracetam, which can have side effects such as depression, loss of appetite, emotional instability, hostility and psychotic symptoms. He found it difficult to recall the foods allowed on the diet, and quickly became bored with the limited selection. He was also angry about his continuing weight loss, but unwilling to eat more. His intake was estimated to be between 7500 and 8000 kJ, meeting approximately 80% of his energy requirement. His partner was encouraged to bring in foods from home to increase the variety of the diet, and was provided with recipe ideas (bacon and low-carbohydrate vegetables in a cream sauce; oily stir-fry of fatty meat with low-carbohydrate ve getables; egg scrambled with butter and cream).
The patient's weight began to increase. EEG indicated reduced seizure activity. However, the foods brought in were not always appropriate, and at times the patient managed to obtain unsuitable food items himself. His erratic cognitive state, presumed due to occasional seizure activity, made treatment difficult, and he repeatedly absconded from the ward. After two weeks, the diet was abandoned. The patient remained in hospital for a further two weeks, during which time he appeared to improve somewhat, and his weight steadily increased. He was then discharged home, with medications partially controlling the seizures (levetiracetam, phenytoin, sodium valproate, topiramate, phenobarbitone and clonazepam, as well as an anti-psychotic, haloperidol). The consultant neurologist predicted that the patient was unlikely ever to be seizure-free. The long-term use of an oral ketogenic diet, while possibly beneficial, did not appear to be a realistic strategy for this patient.
Data and laboratory investigations
On the second admission to hospital, weight was estimated at between 75 and 85 kg and serum albumin was 37 g/L (normal range 40-50 g/L). On day five of the admission prealbumin was 0.15 g/L (normal range 0.17-0.35 g/L) and albumin was 22 g/L. By day seven, prealbumin had reached 0.17 g/L and remained within normal range for the remainder of the admission; albumin was 24 g/L. The carbohydrate-free formula was commenced on day 48 of admission, providing 8030 kJ per day. On day 50, serum concentrations of the ketone bodies were as follows: serum [beta]-hydroxybutyrate 974 [micro]mol/L; acetoacetate 408 [micro]mol/L. total 1384 [micro]mol/L (usually total would be 100 [micro]mol/L in the fed state). By day 54, serum [beta]-hydroxybutyrate was 1549 [micro]mol/L and acetoacetate was 934 [micro]mol/L (total 2483 [micro]mol/L) On day 60, weight was 55 kg. Albumin was 27 g/L and increased steadily for the remainder of the admission. By day 80, albumin had reached 42 g/L. Weight was 60 kg and remained between 60 and 63 kg for the remainder of the admission. No abnormally low blood glucose concentrations were recorded and serum calcium concentrations remained within the normal range during the admission. Low serum magnesium concentrations, requiring replacement, occurred persistently during the admission, both before and after the ketogenic diet period.
The ketogenic diet has long been known to help control seizures. By 400BC Hippocrates had reported a case of epilepsy that was cured by fasting. In the early 1920s, several researchers reported that this effect could be achieved without starving the patient, by inducing ketosis with a high fat, low carbohydrate diet in a specific ratio (2-4). The ketogenic diet increases the seizure threshold, and may also improve alertness, cognition and mood.
Despite its long history, the ketogenic diet is still poorly understood. There have been many suggestions as to its mechanism of action. The ketone bodies themselves (acetoacetate, beta-hydroxybutyrate and acetone) have been shown to have anticonvulsive properties. However, in many cases seizure threshold does not appear to change until more than a week after ketosis develops. This suggests that the effect might not be due to the ketones alone, but rather the metabolic adaptation of the brain to using them as a fuel.
Younger children (aged under ten years) respond better to the diet, because they become ketotic much more readily, and their brain adapts earlier to using ketone bodies. Early researchers found that a ratio of at least 3:1 (fat:carbohydrate+protein by weight) was needed for seizure control (3). Later a modified diet using medium chain triglycerides (MCT) was suggested. This allowed a more liberal ratio, as MCT produces more ketones, but appeared to be less effective in older children (5). Reviews of the literature estimate that the ketogenic diet appears to produce significant improvements in about 30 to 70% of cases (6,7). However, there are no blinded studies of the diet's effectiveness.
Growth retardation has been demonstrated in children maintained on the diet for two years (8) despite provision of the recommended intake for protein. Loss of lean body mass is inhibited during ketosis, although the mechanism is poorly understood. It appears that the breakdown and turnover rate of muscle protein is reduced, possibly in response to the increase in blood ketone concentrations. Weight maintenance may be difficult because the total energy intake decreases on an unpalatable diet.
There are very few studies of adult response to a ketogenic diet. Those that exist suggest that the benefits may be similar to those in children, although it is more difficult to induce ketosis and to maintain compliance to the diet (6,9,10). Reductions in seizure frequency may be more modest than those seen in children, however, many adults also report improvements in subjective indicators such as mood and cognition. More adults complain of difficulty in following the diet, and side effects also appear to be more common, including gastrointestinal complaints such as constipation or bloating, rise in serum cholesterol, LDL and triglycerides, and, in women following the diet, menstrual irregularities.
Side effects of the ketogenic diet can inhibit compliance. Constipation, which is the most common complication (10,12), may be relieved by the use of some MCT oil that initially has a laxative effect. In the longer term, lipid control may be a problem because cream and butter are more versatile cooking ingredients than oils, and thus the saturated fat content tends to increase with the palatability of the diet. Oil emulsion products, which can be used in drinks, can help to improve the fatty acid profile of the diet. Kidney stones, usually of calcium oxalate, are relatively common, occurring in 5 to 8% of children on the diet (11) and this may be exacerbated by the requirement for calcium supplements. Micronutrient status may be of concern in the long term. Standard watersoluble vitamin supplements do not include antioxidants like carotenoids.
Other factors in compliance include the patient's ability to manage the diet. This may be difficult if the patient is cognitively impaired, which can occur as a result of seizure activity or as a side effect of some anticonvulsant medications. Anticonvulsants can also cause nausea or loss of appetite, which may also affect motivation. Cooking skills are important for producing palatable meals, and the patient needs to be able to read the labels on food packaging. Medications should be carbohydrate-free. The patient in this study found compliance difficult despite a supportive family and close supervision. Even with good patient compliance, the diet can be difficult to maintain.
The ketogenic diet can be used to control intractable seizures in adults. Tube-feeding is an easy way to provide an effective diet. Providing an oral ketogenic diet within a hospital food service system may be difficult. Many people find high fat diets unappealing, and it can he a challenge to render such diets sufficiently palatable for long-term compliance. Side effects of the diet may include gastrointestinal discomfort, raised blood lipids or kidney stones. A high level of motivation is important, and this may be lacking if the patient's neurological issues include cognitive impairment. Patients using this diet at home require ongoing support and information in order to maintain ketosis and enjoy the benefits of living without seizures.
The contribution of Dr Armin Mohamed, in commenting on neurological aspects of the case, is gratefully acknowledged.
(1.) Wheless JW, Baumgartner J, Ghanbari C. Vagus nerve stimulation and the ketogenic diet. Neurol Clin 2001;19:371-407.
(2.) Geyelin HR. Fasting as a method of treating epilepsy. Med Rec 1921;99:1037-9.
(3.) Wilder RM. The effects of ketonemia on the course of epilepsy. Mayo Clin Proc 1921;2:307-8.
(4.) Peterman MQ The ketogenic diet in epilepsy. JAMA 1925;84:1979-83.
(5.) Huttenlocher PR, Wilbourn AJ, Signore JM. Medium-chain triglycerides as a therapy for intractable childhood epilepsy. Neurology 1971;21: 1097-103.
(6.) Prasad AN, Stafstrom CF, Holmes GL. Alternative epilepsy therapies: the ketogenic diet, immunoglobulins and steroids. Epilepsia 1996;37(Suppl 1):81S-95S.
(7.) Vining EPG Clinical efficacy of the ketogenic diet. Epilepsy Research 1999;37:181-90.
(8.) Williams S, Basualdo-Hammond C, Curtis R, Schuller R. Growth retardation in children with epilepsy on the ketogenic diet: a retrospective chart review. J Am Diet Assoc 2002;t02:405-7.
(9.) Barborka CJ. Epilepsy in adults: results of treatment by ketogenic diet in one hundred eases. Arch Neurol 1930;6:904.-14.
(10.) Sirven J, Whedon B, Caplan D, Liporace J, Glosser D, O'Dwyer J, et al. The ketogenic diet for intractable epilepsy in adults: preliminary results. Epilepsia 1999;40:1721-6.
(11.) Herzberg GZ, Fivush BA, Kinsman SL, Gearhaxt, JP. Urolithiasis associated with the ketogenic diet. J Pediatr 1990;117:743-5.
(12.) Swink TD, Vining EPG, Freeman JM. The ketogenic diet: 1997. Advances in Paediatrics 1997;44:297-329.
Royal Prince Alfred Hospital, Sydney
S. Ferrie BSc, GradDipFoodSci, MND, APD, Critical Care Dietitian
Correspondence: S. Ferrie, Department of Nutrition and Dietetics, Royal Prince Alfred Hospital, Missenden Road, Camperdown NSW 2050. Email: email@example.com