The effectiveness, tolerability, and adverse effects of the ketogenic diet were reviewed retrospectively in 31 infants (18 male; 13 female) with refractory epilepsy treated at Columbia-Presbyterian Medical Center, New York, NY. A 3:1 or 4:1 (fat:nonfat) ketogenic regimen was introduced slowly after a 12-38 hour inpatient fasting period to initiate ketosis. The ratio of the diet was adjusted to produce moderate to strong ketosis. Adequate calories and protein to sustain growth were provided. Mean age at start of diet was 13.8+/-5.7 months; 14 were <12 months and 17 >12 months. The duration on the diet was <3 months in 3, 3-5 mo in 7, and =/>6 mo in 21 (67%). Epilepsy etiology was idiopathic in 12(39%) and symptomatic in 19(61%); 4 patients had progressive metabolic and degenerative disorders. Seizure types were generalized in 10 (32%), partial in 6, and mixed in 15 (48%). Seventeen (55%) had infantile spasms. Twenty four (77%) were developmentally delayed.
Six patients (19%) became seizure-free, 11 (35%) were improved (>50% decrease in frequency), and 14 (45%) were not benefited. Antiepileptic medications were reduced concomitantly with improved dietary control. Patients with consistent moderate to severe ketonuria (71%) showed a trend toward better seizure control than those with variable ketonuria (29%). Level of ketonuria and level of seizure control were not significantly correlated. Favorable seizure control was increased sixfold in infants with infantile spasms; response was not related significantly to gender, developmental delay, or etiology. Parent questionnaires revealed that 75% noted a decreased frequency of seizures and 35% a lessening in seizure intensity. The majority reported improvements in the child’s attention, alertness, activity level, and social interactions. An overall beneficial effect was reported by 50% of parents. Five infants developed adverse events that were reversible, including severe vomiting, renal stones, gastrointestinal bleeding, hyperlipidemia, and ulcerative colitis (the time on the diet before complication had been 7 to 26 months). One additional patient developed coma with hypoglycemia and acidosis after 8 days on the diet. Three patients died, 1 with a multisystem disease and pulmonary infection, 1 with aspiration, respiratory failure and congenital heart disease, and the third from complications related to an underlying pyruvate dehydrogenase deficiency. The authors recommend the ketogenic diet as safe and effective in infants with intractable epilepsy. 
COMMENT. The ketogenic diet, if closely monitored by a major medical center team, can be an effective treatment in infants with refractory epilepsies, and especially for those with infantile spasms and myoclonic epilepsies. In addition, it is the recommended treatment for infants with pyruvate dehydrogenase deficiency , and glucose transporter protein deficiency . Adverse events attributed to the diet are mainly reversible. These include nausea, vomiting and hypoglycemia, particularly prevalent during the initiation of the 4:1 ratio diet and with fasting, as recommended in the Johns Hopkins method. The omission of the period of fasting, and the introduction of ketosis using lesser fat/nonfat ratios, as practiced in the Mayo Clinic method, is efficacious and associated with fewer reports of adverse events . Serious complications associated with high fat/nonfat ratios include renal stones, cardiomyopathy (Best et al, 2000; see Ped Neur Briefs July 2000; 14:50), and platelet dysfunction with bleeding (see Ped Neur Briefs (Feb 2001;15:12-13) . Serious adverse events were also reported in 10% of 52 children treated with a high 4:1 ratio/fat:nonfat/diet; complications included hypoproteinemia, lipemia, hemolytic anemia, renal tubular acidosis, and liver dysfunction with concomitant thrombocytopenia . Despite the rigors and adverse events associated with the diet, in selected patients it can be a worthwhile method of treatment. Its effectiveness has been documented in numerous studies and reviews, from the initial report by Wilder RM (Mayo Clinic Bull 1921;2:307) to a recent confirmation in 150 children treated at Johns Hopkins University . The present report is the first emphasizing the effectiveness of the ketogenic diet in infants. The attending adverse events and those reported in older children demonstrate the importance of close monitoring by a specialized medical team and support staff.