Animal models of the ketogenic diet: what have we learned, what can we learn?

Despite its clinical use as a therapy for refractory epilepsy for more than 75 years, the ketogenic diet (KD) remains a therapy in search of an explanation. The mechanism of action of the KD is unclear and the optimal indications for its clinical use are incompletely defined. Animal models could help to elucidate these questions. Surprisingly, there have been very few animal studies of the KD, and those that have been performed are difficult to compare because of wide discrepancies in experimental methods. Earlier models concentrated on the effect of the KD on acute seizure threshold in normal (i.e. nonepileptic) animals. Recent studies are beginning to examine the longer term effects of the KD and its role in epileptogenesis. Some features of clinical experience have been replicated in animal models, including the role of ketosis, elevation of seizure threshold by both classic ketogenic and medium chain triglyceride diets, better effectiveness at younger ages, and rapid reversal of the seizure protective effect when the diet is discontinued. These parallels raise hope that pertinent clinical questions can be addressed in the more controlled setting of the research laboratory. As in the clinical arena, there has been a recent resurgence of interest in pursuing basic questions related to the ketogenic diet, using techniques of modern neuroscience. Experimental approaches such as brain slice neurophysiology, genetic models, dissection of metabolic pathways, and neurohistological techniques hold much promise in the effort to understand this intriguing alternative to standard anticonvulsants.