A diet high in fat and low in carbohydrate and protein has been used to treat seizures in children with epilepsy for nearly 90 years. A number of studies have shown the diet is effective, but none of these have been of the type considered to provide the best quality of evidence: randomised controlled trials. These are recommended for all epilepsy treatments, including anti-epileptic drugs (AEDs).

A controlled trial is one in which participants are divided into two groups. Participants in one group try the new treatment; those in the other group (the control group) do not, though they might take a placebo. The control group is a comparison group. If the trial is randomised, then participants are allocated at random to either the treatment group or the control group. This avoids any bias in allocating particular types of patient to either group.

Ideally, the best trials are also blinded, so that neither the doctors nor the participants but only the trial administrators know who's getting the treatment or the control. However this is not possible for a trial of a diet, for obvious reasons! Read more about how epilepsy trials
are designed

In May 2008, the results of the first randomised controlled trial of the ketogenic diet were published in the Lancet Neurology.

The 145 children in the trial were divided into two groups, one of which started the diet three months later than the other. For that period, the delayed group were the control group.

The children in the trial were aged between 2 and 16 years. They all had active epilepsy, with seizures at least once a day, and had not responded to at least two AEDs. All children in both groups kept on taking any AEDs they had been on before the study.

At the end of the three month comparison period, children on the diet had on average a 38% drop in seizure frequency, whereas seizure frequency in the control group had increased. Just over a third of children in the diet group had seen their seizure frequency halve, compared to only 6% of the control group. Seven percent of children in the diet group had a greater than 90% drop in seizures; no children in the control group had this.

Side effects of the diet included vomiting, diarrhoea, constipation and feeling hungry, most of which could be treated by adjusting the diet. The diet didn't suit everybody: six children dropped out because of side effects and two because they didn't like the foods allowed in the diet. Parents' attitudes were important too: three children dropped out because their parents were unhappy about the restrictions of the diet.

This is the first trial which clearly shows that the ketogenic diet is a better treatment option for children with drug-resistant epilepsy than no change. The rates of response to the diet are similar to those seen in trials of new AEDs. The diet worked equally well in all types of epilepsy. The reductions in seizure frequency seen in this study are smaller than those seen in other studies, but this may be due to the fact that the children in this study had epilepsy that was difficult to treat.

The researchers, led by Dr Elizabeth Neale and Professor Helen Cross of the Institute of Child Health, University College London, are now following up the children in both arms of the study for a longer period, to see what the long-term effects of the diet are. They are also investigating whether there is any difference in efficacy between the classical ketogenic diet and the diet supplemented with MCT oil.

Professor Helen Cross is the Chair of Trustees of Epilepsy Research UK. She spoke about this study at the 2008 Epilepsy Research UK Evening Lecture in Dundee. A video of this talk will be available later this year.

More information about the ketogenic diet can be found here and
here