Background: Although a larger number of antiepileptic drugs became available in the last decades,
epilepsy remains drug-resistant in approximately a third of patients. Ketogenic diet (KD), first proposed at the
beginning of the last century, is complex and has anticonvulsant effects, yet not completely understood. Over the
last decades, different types of ketogenic diets (KDs) have been developed, namely classical KD and modified
Atkins diet (MAD). They offer an effective alternative for children and adults with drug-resistant epilepsies.
Methods: We review several papers on KDs as an adjunctive treatment of refractory epilepsy of children and
adults, discussing its efficacy and adverse events. Because of the heterogenous, uncontrolled nature of the studies,
we analyzed all studies individually, without a meta-analysis.
Results: KDs may be considered first choice treatment in some specific metabolic conditions, such as glucosetransporter
type 1 and pyruvate dehydrogenase deficiencies, and mitochondrial complex I defects. Preliminary
findings indicate that KDs may be specifically effective in some epileptic syndromes, such as West syndrome,
severe myoclonic epilepsy of infancy, myoclonic-astatic epilepsy, febrile infection related epileptic syndrome,
and drug-resistant idiopathic generalized epilepsies or refractory status epilepticus. Short term adverse events are
usually mild in both children and adults, including gastrointestinal symptoms, hyperlipidemia, and hypercalciuria;
potential long term adverse effects include nephrolitiasis, decreased bone density, and liver steatosis. Possible
atherosclerotic effects remain a concern. Patients on KDs should be carefully monitored in specialized centers
during initiation, maintenance and withdrawal periods, in order to minimize such adverse events, and to improve
compliance. Although the majority of KD trials on children and adults with drug-resistant epilepsies are openlabel,
uncontrolled studies based on small samples, an increasing number of randomized controlled trials have
provided better quality evidence on its efficacy in recent years.
Conclusion: There is a need for future randomized clinical trials aimed to confirm the efficacy of KDs in specific
epileptic syndromes, and to provide further information about some practical unsolved problems, i.e. for how
long KD treatment should be continued.