The modified Atkins diet (MAD) is a relatively new, alternative dietary therapy for intractable childhood and adult epilepsy. Recent evidence suggests that a strict, highly ketotic, first month may be advantageous to both immediate and long-term efficacy. KetoCal® is a pre-mixed powder that can be used to create a 4:1 ketogenic diet shake as a meal substitute. The investigators hypothesize that substituting KetoCal® for a lunch during the initial month of the MAD will lead to improved seizure reduction over the MAD alone, as well as improved tolerability and lipid values.

The modified Atkins diet (MAD) is a treatment for intractable epilepsy that has been used in over 100 children and adults worldwide since 2002. It mimics the ketotic state of the ketogenic diet (KD), yet does so without fluid, calorie, or protein restriction, and without the need for an inpatient admission or fasting period. Children are started on a daily allotment of 10 grams of carbohydrates per day, with gradual increases to 20-30 grams per day after several months. To date, six studies have demonstrated the MAD as effective, with approximately 50% of patients having at least a 50% reduction in their seizure frequency. The MAD works relatively quickly, with beneficial results when present, within 2-4 weeks.

Over the past 5 years since its introduction, our center and others have researched methods to make the MAD both more tolerable and efficacious. Important clues came in 2007, in which an initial carbohydrate limit of 10 grams per day was compared to 20 grams per day in a randomized , prospective, crossover design. Although the hypothesis was that these two carbohydrate limits were equivalent, this was only true after the 3-month crossover period. At diet onset, there was a significantly higher likelihood of a >50% seizure reduction when a stricter, 10 grams/day MAD was used (60% vs. 10%, p=0.03). Ketosis was higher in that group at 1 month as well, similar to previous findings that the 1-month ketosis levels correlated with efficacy. These results suggested that a stricter, higher ketotic, start to the MAD was important - although loosening restrictions after 1-3 months was also reasonable and did not adversely affect seizure control.

That same year, a research study from South Korea evaluating the initial ratio (3:1 vs. 4:1) of the traditional ketogenic diet in a similar randomized, crossover design found identical results. A 4:1 initial ketogenic diet ratio was correlated with a higher likelihood of seizure freedom (55% vs. 31%, p<0.05), although ketosis at 3 months was not different. After 3 months, reducing to 3:1 did not lead to reduced seizure control in any of their children.

Recent results using the KD from our center and Chicago Memorial also confirm that the initial month of the KD may be the most crucial. In this study of 99 children on the traditional KD, nearly 90% of children who do respond to the KD do so within the first 28 days. This study suggests that early KD titration to improve efficacy is important.

The results from all of this data within the past year all confirm that the first month of either the MAD or KD may be the most critical. In addition, there is some suggestion that higher ketone levels at 1 month do correlate with better seizure reduction. Therefore, a strict, highly ketotic initial month of dietary therapy may result in better seizure reduction. Realizing that loosening of restrictions will automatically occur after one month, this "tough month" approach is likely to be followed in a compliant manner by most parents knowing it is time-limited.

What is the best way to then achieve higher ketosis levels initially? One option is to fast children before starting the MAD. However, this approach raises concerns regarding hypoglycemia and acidosis, especially in the setting of an outpatient approach, as is typically used for starting the MAD. Some families specifically ask for the MAD to avoid an admission or fasting period. Another option would be to give the KD first, then switch to the MAD after one month. However, the significant time and financial burden involved in a KD admission and training would decrease interest in the MAD approach by families contacting our center, and not truly test the MAD effectiveness as a primary therapy. Other potential ways to increase ketosis include carnitine supplementation, MCT oil inclusion, and calorie restriction. All of these options, however, have their own financial and safety implications.

Our proposal, therefore, is to use the 4:1 KetoCal®, and its proven ability to induce high levels of ketosis, as a partial supplement to the MAD. Our concept is to use this supplement for only the first month to boost ketosis and improve seizure control, as data suggests that this may be the key and only important time period. After this study is completed, it would be acceptable for a family to use KetoCal® as a periodic supplement to help with meal creation, should they wish to do so.

**The study involves 3 trips to Baltimore (travel not covered) over 2 months (baseline, 1 month and 2 months). Labs are not covered.

The clinic visits, printed information, carb-counting book, case of KetoCal powder/formula, and Ketostix are free.

Inclusion Criteria:

• Children ages 3-17 years
• At least daily seizures (or 7 per week). All seizure types allowed
• Tried at least two anticonvulsants

Exclusion Criteria:

• Unwilling to restrict carbohydrates
• Significantly underweight (BMI ≤ 5%)
• Prior use of the modified Atkins diet for ≥ 2 days
• Prior use of KetoCal® at any time for any duration
• Use of the KD within the past year
• Kidney disease
• History of hypercholesterolemia(>300 mg/dl) or hypertriglyceridemia (>200 mg/dl)
• Metabolic or mitochondrial disorder
• Aversion to shakes or inability to eat solid food