A retrospective evaluation of the ketogenic diet (KD) was conducted comparing efficacy and tolerability of the diet with or without initial fasting and fluid restriction and involving university centers in Seoul, Korea. Of 41 children with intractable epilepsy treated with the nonfasting diet (NFKD), 14 (34.1%) were seizure-free for at least 3 months, and in 83 receiving the initial fasting diet (IFKD), 29 (34.9%) were seizure free. The patients’ mean age was 4.1 (+/-1.5) years in the NFKD group and 5.3 (+/-1.6) years in the IFKD group. The incidence of various seizure types was not significantly different in the 2 groups. Strong urinary ketosis developed within 2.4 +/- 1.2 days in the NFKD group, compared to 1.9 +/-1.3 days in the IFKD group (not significant, P>0.05). Among early complications, dehydration was significantly less frequent in the NFKD group (12.2%) than in the IFKD group (62.7%; P<0.05). The incidence of other complications was not significantly different in the NF and IF groups, including hypoglycemia (4.9 v 4.8%), nausea and vomiting (36 v 48%), elevated BUN (12 v 24%), elevated cholesterol (17% v 14%), elevated triglycerides (36 v 36%). Patients in the NFKD group were hospitalized for a mean period of 5.8 (+/-2.1) days and those in the IFKD group for 7.8 (+/-1.7) days (P<0.05). 
COMMENT. The nonfasting ketogenic diet, in the above retrospective study, was of equal efficacy, and its introduction was better tolerated than the initial fasting method. The KD was first introduced for the treatment of epilepsy at the Mayo Clinic . In the Mayo Clinic method, unlike the later Johns Hopkins protocol , the KD is introduced without initial fasting and usually, without admission to hospital. The ratio of ketogenic to antiketogenic (K:AK) items in the diet is reversed gradually, starting at a ratio of 1:1.1 on day 1 and increasing to 2.8:1 by the 4th day, or until ketones are found in the urine. In young children, a ratio of 3;1 is usually sufficient to maintain ketosis, and generally the higher 4:1 ratio is necessary only in older children. In my experience with the diet, beginning at the Mayo Clinic in the early 1960s and continuing at Children’s Memorial Hospital in Chicago, the NFKD method has proven satisfactory and free from serious side effects. [4, 5, 6, 7]
A flurry of papers on the ketogenic diet has appeared in recent months. A retrospective study has shown that the KD can be successfully initiated on an outpatient basis without starvation, fluid or caloric restriction . The worldwide use of the KD has been established by communicating with 73 academic centers in 41 countries using the Internet. The average number of patients enrolled was 72 per country, with 5 new patients annually. Common difficulties included failure to avoid rice intake, intolerance symptoms with the higher K:AK 4:1 ratio, using the Hopkins protocol. A website is now available. 
In a Korean multicenter study involving 199 patients (87 using the IFKD Hopkins protocol and 112 the NFKD), 66 (34%) successfully completed or maintained the diet. By modifying the protocol to omit the fasting period, especially in young children, acute dehydration was prevented, with no difference in the time to ketosis or in the efficacy of the diet. Five patients died related to lipoid aspiration pneumonia, serious infectious disease, and nutritional problems . The ketogenic diet was successful using gastrostomy tube feeding to ensure compliance, in the treatment of 12 children with static encephalopathy and intractable seizures.