PEDIATRICS Vol. 114 No. 6 December 2004, pp. 1627-1630 (doi:10.1542/10.1542/peds.2004-1001)
Benefits of the Nonfasting Ketogenic Diet Compared With the Initial Fasting Ketogenic Diet
* Department of Pediatrics, Ilsan Paik Hospital, Goyang, Korea
Department of Pediatrics and Epilepsy Center, Sanggye Paik Hospital, Seoul, Inje University College of Medicine, Seoul, Korea
Sowha Children's Hospital, Seoul, Korea
|| Department of Pediatrics, Severance Hospital, Handicapped Children's Research Institute, Brain Research Institute, Yonsei University College of Medicine, Seoul, Korea
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Objective. The ketogenic diet (KD) is traditionally introduced with an initial period of fasting and fluid restriction that is difficult and sometimes complicated by moderate dehydration. This investigation compares the efficacy and tolerability of the nonfasting ketogenic diet (NFKD) and the conventional initial-fasting ketogenic diet (IFKD).
Methods. Forty-one children with intractable epilepsy were treated with the NFKD, beginning with a gradual increase in calories with no initial fasting or fluid restriction. This NFKD population was compared retrospectively with 83 recent historical control subjects who were treated with the IFKD. Efficacy, tolerability, time until strong ketosis, and occurrence of complications were compared.
Results. Fourteen (34.1%) patients became seizure-free for at least 3 months after the NFKD, compared with 29 (34.9%) after the IFKD. There was no significant difference in days until strong urinary ketosis between the 2 groups. The incidence of hypoglycemia was also not significantly different between the groups as most other laboratory findings, although the blood urea nitrogen was elevated in 24.1% of the IFKD group and in only 12.2% of the NFKD patients without statistical significance. Conversely, moderate dehydration was significantly less frequent in the NFKD group (12.2%) than in the IFKD group (62.7%). Finally, these results were reflected to the shortening of the hospitalization period in the NFKD group.
Conclusions. These observations suggest that initial fasting and fluid restriction are not essential for the KD and that the tolerability of this treatment may be improved. These data support our intention to conduct a formal, prospective, randomized trial comparing 2 forms of the KD.
Key Words: nonfasting • ketogenic diet • childhood • epilepsy • fasting
Abbreviations: KD, ketogenic diet • AED, antiepileptic drug • IFKD, initial fasting ketogenic diet • NFKD, nonfasting ketogenic diet • BUN, blood urea nitrogen
The ketogenic diet (KD) was first described by Geyelin1 in 1921 as a treatment for epileptic seizures but subsequently fell from favor with the advent of the modern era of antiepileptic drugs (AEDs). However, the KD has recently regained a spotlight in the treatment of medically intractable childhood epilepsy.2 The KD traditionally starts with an initial period of fasting and fluid restriction until ketone bodies appear in the urine, followed by the introduction of high-fat, adequate-protein, low-carbohydrate foods.2–5 This conventional diet, the initial-fasting ketogenic diet (IFKD), has been the standard method for administering the KD.
Initial fasting has been considered essential to induce ketosis and to maximize the chance of therapeutic benefits. Furthermore, it has been suggested that initial fasting is advantageous in screening some underlying metabolic conditions associated with subsequent severe hypoglycemia and ketoacidosis.6 Unfortunately, initial fasting and fluid restriction have been regarded among the most difficult aspects of the diet and sometimes cause serious systemic complications, such as dehydration, which prolongs hospitalization. For these reasons, we conducted this retrospective study of the introduction of the KD with or without initial fasting and fluid restriction to compare their treatment efficacy and tolerability.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
A dedicated KD program has been present at the Epilepsy Center at Inje University Sanggye Paik Hospital from July 1995 to the present. All candidate patients for the KD experienced >4 seizures per month, were intractable to
3 AEDs, and were significantly disabled by their seizures. At this institution, the Johns Hopkins Hospital protocol2 for the traditional IFKD was used from July 1995 to July 1999 but was later replaced by a nonfasting protocol developed by Park and Kim3 and implemented from August 1999 to the present. In the traditional IFKD, patients undergo fasting and fluid restriction until strong urinary ketosis is achieved. The patients then are fed a small amount of a 4:1 ratio diet (amount derived from fat compared with amount derived from protein and carbohydrates) that gradually increases in size until the total daily required calories are provided. In contrast, the nonfasting ketogenic diet (NFKD) is initiated by gradually introducing high-fat foods with the regular diet and does not require fasting or fluid restriction. On the first day, the child is started with one third of the total daily required calories represented by high-fat foods, with two thirds of the required calories on the second day and full required calories on the third day of treatment. For both versions of the KD, all currently administered AEDs except acetazolamide were continued, and all patients remained hospitalized until acceptable tolerance was established and the education of caregivers regarding the diet was completed.
The medical records of all KD patients were reviewed retrospectively to measure treatment efficacy. Efficacy was calculated by comparing daily seizure counts from seizure calendars in the pre-KD baseline and after the initiation of the diet. Data were also collected regarding diet tolerability; time to onset of strong urine ketosis >3+ (50-150 mg/dL) in routine urinalysis; occurrence of moderate dehydration defined as a reduction of body weight >5% from the baseline, with poor skin turgor, dried mucous membranes, and an increased urine specific gravity >1.020; symptomatic hypoglycemia; and other acute or delayed complications. Finally, the results of standard laboratory values collected on all patients per protocol, including blood urea nitrogen (BUN), creatinine, electrolytes, liver enzymes, cholesterol, triglyceride, and uric acid levels, were recorded. Criteria for the laboratory abnormalities were referenced from the Nelson Textbook of Pediatrics.7
The SPSS program for statistical analyses was used for the 
2 test to compare categorical variables, and 2-tailed t test was used to evaluate the significance of differences of continuous variables between the NFKD and IFKD groups. P < .05 was regarded as statistically significant.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Patient Characteristics
A total of 124 patients with intractable epilepsy were treated for at least 3 months with 1 of the 2 forms of the KD at this institution from July 1995 to February 2001. All cases were intractable to >3 AEDs in combination with maximal tolerated doses, having >4 seizures a month. The traditional IFKD was used on 83 patients from July 1995 to July 1999, and 41 patients were treated with the NFKD from August 1999 to February 2001. There were no significant differences in the baseline (pre-KD) characteristics of these 2 groups (Table 1). The patients' mean (±SD) age was 4.1 (±1.5) years in the NFKD group and 5.3 (±1.6) years in the IFKD group (P > .05). Eighteen (44%) of the patients were male in the NFKD group, and 51 (61%) were male in the IFKD group.
|
Seizure Classification
Patients with a wide variety of intractable seizure types were included in the study (Table 1). In the NFKD group, infantile spasms were observed in 13 (31.7%) patients, Lennox-Gastaut syndrome in 5 (12.2%), severe myoclonic epilepsy of infancy in 1 (2.4%), other generalized seizures in 12 (29.3%), and partial seizures in 10 (24.4%). In the IFKD group, early infantile epileptic encephalopathy was observed in 1 (1.2%) patient, infantile spasms in 11 (13.3%), Lennox-Gastaut syndrome in 18 (21.7%), severe myoclonic epilepsy of infancy in 6 (7.2%), Landau-Kleffner syndrome in 1 (1.2%), other generalized seizures in 13 (15.7%), and partial seizures in 33 (39.8%).
Treatment Efficacy
Treatment efficacy was measured by comparing daily seizure counts (recorded in seizure calendars) in the baseline pre-KD period with seizure counts after the initiation of either form of the KD. During the 3 months of follow-up after the diet, 14 (34.1%) patients in the NFKD group became seizure-free, 8 (19.5%) showed a reduction in seizures of >90%, and 9 (22%) had a 50% to 90% decrease in seizures (Table 2). This did not significantly differ from the IFKD group in which 29 (34.9%) patients became seizure-free, 15 (18.1%) showed a reduction in seizures of >90%, and 19 (22.9%) had a 50% to 90% decrease in seizures. Among the 14 patients who were completely seizure-free after completing the NFKD, 11 (78.6%) had become seizure-free within 3 days of starting the KD, 2 (14.3%) within 4 to 7 days, and 1 (7.1%) within 8 to 14 days. Among the 29 patients who were completely seizure-free after completing the IFKD, 16 (55.2%) had become seizure-free within 3 days, 4 (13.8%) within 4 to 7 days, 3 (10.3%) within 8 to 14 days, 4 (13.8%) between 2 weeks and 1 month, and 2 (6.9%) after 1 month of treatment. The mean (±SD) duration until the cessation of seizures was 2.6 (±2.3) days in the NFKD group and 3.4 (±3.1) days in the IFKD group (P > .05; Table 3). Strong urinary ketosis developed within 2.4 ± 1.2 days in the NFKD group, compared with 1.9 ± 1.3 days in the IFKD group (P > .05; Table 4).
|
|
|
Transient Early Complications of the 2 Diets
Initial gastrointestinal disorders manifested as nausea, vomiting, constipation, or diarrhea occurred in 15 (36.0%) patients of the NFKD group and in 40 (48.2%) of the IFKD group (Table 4). Moderate dehydration was significantly more frequent in IFKD patients (52 [62.7%]) than in NFKD patients (5 [12.2%]) despite the constant intravenous administration of normal saline (P < .05). During the introductory period of the KD, symptomatic hypoglycemia developed transiently in 4 (4.8%) patients of the IFKD group and in 2 (4.9%) of the NFKD group (Table 4).
Laboratory Measurements
Hypernatremia was encountered in only 5 (6.0%) patients of the IFKD group. Elevated BUN values were more commonly seen in 20 (24.1%) patients in the IFKD group than 5 (12.2%) of the NFKD group, but this difference was not significant. Uric acid was elevated in 10 (24.4%) patients in the NFKD group and 23 (27.7%) in the IFKD group. Hypercholesterolemia occurred in 7 (17.1%) patients in the NFKD group and 12 (14.5%) in the IFKD group. Hypertriglyceridemia occurred in 15 (36.6%) patients in the NFKD group and 30 (36.2%) in the IFKD group. No decrease in serum high-density lipoprotein cholesterol was seen in either group (Table 4.).
Delayed Complications of the 2 Diets
Delayed complications seemed unrelated to the method of KD initiation. Renal stones occurred in 1 patient, acute hepatitis in 1, and cardiomyopathy in 1 of the NFKD group. Renal stones appeared in 1 patient and lipoid pneumonia in 6 of the IFKD group.
Hospitalization Period
Patients in the NFKD group were admitted for a mean (±SD) period of 5.8 (±2.74) days and in the IFKD group for a mean (±SD) period of 7.8 (±1.76) days for close observation to evaluate their initial tolerance of the diet, to monitor any acute complications, and to educate their families about preparation of the diet at home (P < .05).
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The antiepileptic efficacy of the KD has been acknowledged by clinicians for many years, although no large-scale, placebo-controlled, randomized study has unequivocally demonstrated its efficacy. Nevertheless, it is remarkable to witness in clinical practice the total cessation of disabling seizures in some children who had been previously medically intractable. The apparent efficacy of the KD has been reported in numerous uncontrolled series.4–6,8–11 Despite its widely accepted efficacy, many families and physicians remain reluctant to initiate the KD because of its inherent difficulties and potentially serious complications. The obstacles that face a successful initiation of the KD include the initial fasting and fluid restriction, the frequent occurrence of gastrointestinal symptoms and hypoglycemia, and its poor palatability. Any alteration in the traditional KD, as exemplified by the Johns Hopkins Hospital protocol,2 that would enhance its ease of use and tolerability without sacrificing efficacy is worth exploring.
In the traditional initiation of the KD, initial fasting and fluid restrictions were empirically believed to be necessary for seizure control by inducing more rapid ketosis and for metabolic adaptation to the state of ketosis.8 However, dehydration reduces the transportation of ketone bodies into the brain and restricts the supply of this essential fuel source.12–14
Although the use of the KD has been considered as an effective therapy for glucose transporter defect and for pyruvate dehydrogenase deficiency,15,16 initial fasting was also considered important because it provided an opportunity to screen for some metabolic diseases that predispose to severe hypoglycemic and for some mitochondrial cytopathies that can be exacerbated by the KD.8,17 Fatty-acid-oxidation disorders have been considered contraindicated for KD treatment because of increased stress on the respiratory chain and tricarboxylic acid cycle functions.8,17 Examples include carnitine deficiency and pyruvate carboxylase deficiency.8,17
In this study, the occurrence of early hypoglycemia was also similar between the 2 diets. There were no instances of severe hypoglycemia in either group. However, it is unlikely that patients with rare metabolic disorders would be included in this relatively small sample of patients. Nevertheless, most patients with underlying metabolic diseases can be identified by screening laboratory analyses and a characteristic clinical course.18 It is noteworthy that the KD has been used successfully in some patients with oxidative phosphorylation disorders and mitochondrial disorders such as Leigh's encephalopathy. Furthermore, some mitochondrial cytopathies can be relieved by the KD.19
In an effort to improve the traditional KD, the NFKD protocol was developed at our institution. However, it was important to establish early on that efficacy was preserved as methods to improve tolerability were sought. In the present study, seizure freedom and seizure reduction were remarkably similar between the IFKD and the NFKD (Table 2). Furthermore, the efficacy for both groups compares favorably to previously reported case series.4–6,8 Similarly, the timing of onset of seizure reduction (Table 3) and the time to achieve strong urinary ketosis were similar between the 2 diets.
Most measures of tolerability seemed similar between the 2 groups except for moderate dehydration, which appeared significantly more often in those who were initially fasted (62.7%) despite the use of intravenous normal saline than those who were nonfasted (12.2%). As expected, elevated values of BUN also occurred more frequently in the fasted group (24.1%) than in the nonfasted group (12.2%), but this difference was not statistically significant. In addition, there were no important differences in the delayed complications of the 2 forms of the KD. Therefore, the NFKD seemed more tolerable than the IFKD in 1 important outcome measure with no loss of therapeutic efficacy.
Finally, fasted and nonfasted patients do not differ in the onset of ketosis, in the incidence of complications such as hypoglycemia, or in the efficacy for seizure reduction. The fasting period is emotionally and physically difficult, and it is important to know that it can be avoided. Naturally, these results can be reflected for shortening of the hospitalization period, if required, in the nonfasted group, although hospital cost for admission is not so high in South Korea comparing the benefits obtainable from the hospitalization.
It is recognized that this study has important strengths and weaknesses. This is a large study population at a single institution with a long-standing interest in and prescribed protocol for the initiation and maintenance of the KD. That the KD can be used successfully in an Asian population, whose customary diet contains substantially less fat than traditional Western ones, is evidence of the broad applicability of the KD. A major limitation of the study is its retrospective nature, using recent historical control subjects at the same institution to compare the efficacy and tolerability of the traditional KD with the newly introduced NFKD. There may be unrecognized differences in the characteristics of these 2 populations that influence efficacy and tolerability because patients were not randomized into one form of the diet or the other. Nevertheless, these early data support our intention to conduct a prospective comparison trial in which eligible patients with intractable seizures are randomly assigned into either the NFKD or the IFKD group.
|
ACKNOWLEDGMENTS |
|---|
This study was supported by a grant from the Korea Health 21 R&D Project, Ministry of Health and Welfare, Republic of Korea (02-PJ1-PG10-21301-0001).
We appreciate Dr Robert R. Clancy (Division of Neurology, Children's Hospital of Philadelphia, Philadelphia, PA) for review of this article.
|
FOOTNOTES |
|---|
Accepted Jun 4, 2004.
Reprint requests to (H.D.K.) Department of Pediatrics, Severance Hospital, Yonsei University College of Medicine, 134, Shinchon-dong, Seodaemun-gu, Seoul 120-752, Korea. E-mail: hdkimmd{at}yumc.yonsei.ac.kr
No conflict of interest declared.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
- Geyelin HR. Fasting as a method for treating epilepsy. Med Rec. 1921;99 :1037 –1039
- Freeman JM, Kelly MT, Freeman JB. The Epilepsy Diet Treatment: An Introduction to the Ketogenic Diet. New York, NY: Demos; 1994:65–100
- Park JC, Kim HD. Benefits of nonfasting ketogenic diet in intractable childhood epilepsy as compared with initial fasting ketogenic diet. Epilepsia. 2001;42(suppl 7) :214 –215
- Vining EPG, Freeman JM, Ballaban-Gil K, et al. A multicenter study of the efficacy of the ketogenic diet.
Arch Neurol. 1998;55
:1433
–1437
[Abstract/Free Full Text] - Hemingway C, Freeman JM, Pillas DJ, Pyzik PL. The ketogenic diet: a 3- to 6-year follow-up of 150 children enrolled prospectively.
Pediatrics. 2001;108
:898
–905
[Abstract/Free Full Text] - Katyal NG, Koehler AN, McGhee B, Foley CM, Crumrine PK. The ketogenic diet in refractory epilepsy: the experience of the Children's Hospital of Pittsburgh.
Clin Pediatr (Phila). 2000;39
:153
–159
[Abstract/Free Full Text] - Nicholson JF, Pesce MA. Reference ranges for laboratory tests and procedures. In: Behrman RE, Kliegman RM, Jenson HB, eds. Nelson Textbook of Pediatrics. 16th ed. Philadelphia, PA: WB Saunders Co; 2000:2181–2224
- Nordli DR Jr, DeVivo DC. The ketogenic diet. In: Wyllie E, ed. The Treatment of Epilepsy, Principles & Practice. 3rd ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2001:1001–1006
- Freeman JM, Vining EPG, Pillas DJ, Pyzik PL, Casey JC, Kelly MT. The efficacy of the ketogenic diet—1998: a prospective evaluation of intervention in 150 children.
Pediatrics. 1998;102
:1358
–1363
[Abstract/Free Full Text] - Wheless JW, Baumgartner J, Ghanbari C. Vagus nerve stimulation and the ketogenic diet. Neurol Clin. 2001;19 :371 –407[CrossRef][Web of Science][Medline]
- Vining EPG. Ketogenic diet. In: Engel J Jr, Pedley TA, eds. Epilepsy, a Comprehensive Textbook. Philadelphia, PA: Lippincott Raven; 1998:1339–1344
- Haymond MW, Howard C, Ben-Galim E, De Vivo DC. Effects of ketosis on glucose flux in children and adults. Am J Physiol. 1983;245 :E373 –E378
- Haymond MW, Karl IE, Clarke WL, Pagliara AS, Santiago JV. Differences in circulating gluconeogenic substrates during short-term fasting in men, women and children. Metabolism. 1982;31 :33 –42[Web of Science][Medline]
- De Vivo DC, Pagliara AS, Prensky AL. Ketotic hypoglycemia and the ketogenic diet.
Neurology. 1973;23
:640
–649
[Free Full Text] - Klepper J, Diefenbach S, Kohlschutter A, Voit T. Effects of the ketogenic diet in the glucose transporter 1 deficiency syndrome. Prostaglandins Leukot Essent Fatty Acids. 2004;70 :321 –327[CrossRef][Web of Science][Medline]
- Wexler ID, Hemalatha SG, McConnell J, et al. Outcome of pyruvate dehydrogenase deficiency treated with ketogenic diets. Studies in patients with identical mutations.
Neurology. 1997;49
:1655
–1661
[Abstract/Free Full Text] - De Vivo DC, Haymond MW, Leckie MP, Bussman YL, McDougal DB Jr, Pagliara AS. The clinical and biochemical implications of pyruvate carboxylase deficiency.
J Clin Endocrinol Metab. 1977;45
:1281
–1296
[Abstract/Free Full Text] - DeVivo DC, DiMauro S. Mitochondrial diseases. In: Swaiman KF, Ashwal S. Pediatric Neurology, Principles & Practice. 3rd ed. St. Louis, MO: Mosby; 1999:494–509
- Kim YR, Kim HD. Clinical and electrographic characteristics of intractable childhood epilepsy with complete seizure remission over 6 months after ketogenic diet. J Korean Child Neurol Soc. 2000;8 :242 –249
PEDIATRICS (ISSN 1098-4275). ©2004 by the American Academy of Pediatrics
CiteULike 
Connotea 
Del.icio.us 
Digg 
Facebook 
Reddit 
Technorati 
Twitter What's this?
This article has been cited by other articles:
|
|
 |
|
  B. A. Zupec-Kania and E. Spellman An Overview of the Ketogenic Diet for Pediatric Epilepsy Nutr Clin Pract, December 1, 2008; 23(6): 589 - 596. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. S. Mastriani, V. C. Williams, T. C. Hulsey, J. W. Wheless, and B. L. Maria Evidence-Based Versus Reported Epilepsy Management Practices J Child Neurol, May 1, 2008; 23(5): 507 - 514. [Abstract] [PDF]
|
|
|
|
 |
|
 J. M. Freeman, E. H. Kossoff, and A. L. Hartman The Ketogenic Diet: One Decade Later Pediatrics, March 1, 2007; 119(3): 535 - 543. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Rizzutti, A. M. Figueiredo Ramos, M. Muszkat, and A. A. Gabbai Is Hospitalization Really Necessary During the Introduction of the Ketogenic Diet? J Child Neurol, January 1, 2007; 22(1): 33 - 37. [Abstract] [PDF]
|
|
|
|
 |
|
 J. G. Millichap Ketogenic Diet in Epilepsy: To Fast or Not To Fast AAP Grand Rounds, May 1, 2005; 13(5): 54 - 55. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
