Purpose: The ketogenic diet (KD) has been considered a highly potent antiepileptic treatment for intractable childhood epilepsy. In this study, we compared the antiepileptic efficacy and diet tolerability of two different diets with lipid:nonlipid ratios of 3:1 and 4:1.

Methods: Seventy-six patients with refractory childhood epilepsy were randomly placed into two groups and were started on KD diets with nonlipid:lipid ratios of either 3:1 or 4:1. Antiepileptic efficacy and diet tolerability were evaluated 3 months after initiating the diet. Patients showing seizure-free outcome with the 4:1 diet were changed to the 3:1 diet, and those without a seizure-free outcome on the 3:1 diet were changed to the 4:1 diet, for three more months, after which time their progress was monitored.

Results: (1) Antiepileptic efficacy was higher for the 4:1 than the 3:1 diet (p < 0.05). Twenty-two (55.0%) of 40 patients on the 4:1 diet and 11 (30.5%) of 36 patients on the 3:1 diet became seizure free. Seizure reduction of over 90% was observed in 2 (5.0%) patients on the 4:1 diet, and 2 (5.6%) on the 3:1 diet. (2) Dietary tolerability was better for the 3:1 than the 4:1 diet. Gastrointestinal symptoms were observed in 5 (13.9%) patients with the 3:1 diet and 14 (35.0%) patients with the 4:1 diet (p < 0.05). (3) For seizure-free patients who started on the 4:1 diet, antiepileptic efficacy was maintained after changing to the 3:1 diet, while 10 (83.3%) of 12 patients who were not seizure free with the 3:1 diet showed increased seizure reduction after changing to the 4:1 diet. (4) Complications from the KD and laboratory data were not significantly different between the two groups.

Conclusions: The 4:1 KD showed greater antiepileptic efficacy than the 3:1 diet with higher seizure-free outcome. In most cases, seizure free outcome was maintained even after changing the ratio to 3:1. Dietary tolerability was better in the 3:1 diet than the 4:1 with less frequent gastrointestinal symptoms.