By Eric Kossoff, MD
Editor, Keto News

Every December, the ketogenic diet research community gathers at the annual meeting of the American Epilepsy Society (AES) to discuss topics related to the ketogenic diet. At this “SIG” (special interest group), which I have coordinated since 2007, approximately 150 attendees will typically present new material, openly disagree, and debate more controversial topics. The AES SIGs are less regulated (and no continuing medical education credit is provided), hence often a more interactive and exciting setting than typical AES sessions. Attendees include neurologists, basic scientists, and dietitians.

This year’s ketogenic diet SIG in Seattle was no exception: a controversial topic was discussed and a lively debate ensued. This year, we had the issue of “Does ketosis really matter?” discussed from three perspectives: a basic science, clinical neurology, and dietitian perspective.

After I gave a brief introduction to the topic, Dr. Mac Burnham from University of Toronto presented his research regarding acetone as an anticonvulsant. He believed that the data does support that ketosis is important for seizure control in animal models, but interestingly not serum beta-hydroxybutyrate (BOH), which is often measured in clinical studies and occasionally by parents. He believed that acetone was the most important ketone in regards to seizure control. There was some disagreement voiced from basic scientists interested in caloric restriction and glucose modulation (as other mechanisms of action for the ketogenic diet) after his presentation.

Next, Dr. Elaine Wirrell from Mayo Clinic in Rochester, Minnesota gave a lecture on whether ketosis matters from a neurologist’s perspective. At this time, only one study by Dr. Gilbert in 2000, reported that higher levels of serum BOH correlated with improved seizure control (and were possibly more accurate than urine ketone measurement). This study, however, has never been replicated or confirmed. She highlighted that the low glycemic index treatment (See May 2007 Keto News) does not lead to significant ketosis, and yet still preliminarily leads to seizure control. Lastly, Dr. Wirrell discussed recent studies, including those on the modified Atkins diet, that ketosis does seem important during the first month…but not afterwards. This evidence seems to suggest that ketosis may be a marker of the metabolic state while on the ketogenic diet (and possibly correlates with early compliance and success), but does not clearly correlate with seizure control.

Lastly, Dr. Liz Neal from the Institute for Child Health in London, travelled across the Atlantic Ocean to present the dietitians’ perspective on this difficult topic. Dr. Neal discussed data from the recently completed randomized controlled trial of the ketogenic diet (See June 2008 Keto News). In this study, the classic (long-chain) triglyceride ketogenic diet led to higher levels of serum ketones at several time points when compared to the medium chain triglyceride (MCT) diet. However, this did not correlate with seizure control and both diets were equivalent. Interesting additional information regarding carnitine and the general lack of correlation with efficacy in her experience was also discussed.

In summary, I think the jury is still out on the role of ketosis in the ketogenic diet. Most basic scientists today believe that it is not the only (or even primary) mechanism of action of dietary therapy (See October 2007 Keto News). As was clear from both Drs. Wirrell and Neal, both child neurologists and dietitians are starting to agree. However, ketosis may be important during the initial several months based on some clinical research.

Topic Editor: Robert Fisher, MD, PhD
Last Reviewed: 1/30/09