The ketogenic diet in refractory childhood epilepsy

To report the efficacy and tolerability of the ketogenic diet (KD) in refractory paediatric epilepsy.

Twenty-six consecutive children were treated with the classical KD from 1996 to 2001. The epilepsy syndromes included symptomatic generalized epilepsy (15), idiopathic generalized epilepsy (4), symptomatic partial epilepsy (1) and unclassified epilepsy (6). One child was lost to follow up.

Median age at initiation of the KD was 6.1 years. Median duration of the treatment was 9 months. Twelve children (48%) were treated for >12 months; one still remains on the KD. Four children (16%) became seizure-free. Five children (20%) had 50-99% reduction in seizures, seven (28%) had <50% reduction in seizures and eight (36%) had no improvement. Age, seizure-type and aetiology did not predict response. The medications were decreased in 33% of the children. The KD was discontinued in 64% of the children because of poor efficacy and in 12% because of side-effects. Problems during initiation of the KD included asymptomatic hypoglycaemia (24%) and vomiting (12%). Later complications included poor growth (20%), hyperlipidaemia (16%), hypercalcuria (8%), hypocarnitaemia (8%), constipation (8%), pancreatitis (4%) and decreased bone density (4%). There were no deaths. A 3-month trial of the KD costs A3879 dollars. The first 12 months cost A7275 dollars with a cost of A4528 dollars each year, thereafter.

The KD is an effective treatment for some children with refractory epilepsy, being generally well tolerated and rarely associated with side-effects. Response is not necessarily predicted by age, syndrome or aetiology. A prospective study of the KD is presently underway.