Neuropediatrics 2006; 37 - THP120
DOI: 10.1055/s-2006-945943

PREDICTORS OF A SUCCESSFUL OUTCOME IN CHILDREN ON THE KETOGENIC DIET

M Vila 1, A Macdonald 1, J Nedvidek 1, W Hopman 1
  • 1Queen's University, Kingston, ON, Canada

Objectives: The ketogenic diet (KD) has been shown repeatedly to be efficacious in treating intractable seizures in children. However, it causes many adverse effects and alters an entire family's lifestyle. Therefore, this study sought to identify potential predictors of a successful outcome in children treated with the KD.

Methods: A retrospective chart review of eleven children who had been treated with the KD. A baseline EEG and blood sample, consisting of CBC, electrolytes, creatinine, urea, glucose, triglycerides, cholesterol, HDL, LDL, and cholesterol/HDL ratio, were obtained prior to the start of the diet. Height, weight and the number of antiepileptic medications were also noted pre-KD. A successful outcome was defined as >50% reduction in seizure frequency.

Results: The effectiveness of the diet was similar to that seen in a previous meta-analysis: 55% of children had >50% reduction in seizure frequency. Children with higher cholesterol pre-KD and those who stayed on the diet longer were more likely to have a successful outcome. No significant difference in outcome was found related to age, gender, seizure frequency, seizure type, diagnosis, degree of developmental delay, or specific EEG characteristics. Of note, 3 parameters came very close to significance but failed to attain it: children with more antiepileptic medications, children with diffuse slowing on EEG, as well as those with higher LDL cholesterol and higher cholesterol/HDL ratio were more likely to be successful.

Conclusion: The results have the potential for directing KD therapy toward those who are most likely to succeed. The main limitation of this study is its small sample size. However, finding significant differences in such a small sample size is surprising, and if replicated in larger studies, can possibly transform patient selection for the KD.