The ketogenic diet is a medical therapy for epilepsy that is used nearly predominantly for refractory epilepsy (after 2-3 drugs have been tried and failed). However, there is both published evidence for first-line use (infantile spasms, Glut1 deficiency syndrome) and also anecdotal experience (families choosing to change the child's (or the family' own) diet rather than use anticonvulsant medications). Childhood absence epilepsy (refractory) has been published as being responsive to ketogenic diet therapy by the investigators' group previously. This is a small, prospective, 3 month trial to assess if using a modified Atkins diet is a feasible and effective option for new-onset childhood absence epilepsy. The investigators will compare to a group of children in which the parents have declined and chose to start anticonvulsant medications.

The ketogenic diet has been in continuous use since 1921 for children and adult with medically-refractory epilepsy. One of the major unanswered questions is whether it would be as effective for children with new-onset epilepsy. Although logically, this would be the case, it remains to be shown in clinical trials. Additionally, it is much easier to take a medication than to change dietary habits and there is doubt whether families would truly wish to try dietary therapy first (or stay on dietary therapy if not effective for a 6 month trial period).

There is limited published evidence supporting the use of the ketogenic diet as a first-line therapy for infantile spasms, myoclonic astatic epilepsy, and in some situations where a family member had success and the family wishes to start it first. However, these are relatively rare conditions. The emergence of the modified Atkins diet as an outpatient, quickly-initiated, non-fasting approach since 2003 has changed the concept of dietary therapy towards a much less restrictive, potentially emergent therapy. In this way, using dietary therapy could potentially be started before medications for a willing family.

The use of dietary therapy (including the modified Atkins diet) for childhood absence epilepsy goes back decades, but was recently profiled in a review article from the investigators' group. In this publication, 17 studies were identified, and 69% of 133 children with refractory childhood absence epilepsy had a >50% seizure reduction and 34% were seizure-free. At the investigators' center, 21 children as of 2011 had been treated with dietary therapy with 19% seizure-freedom. The question of whether results would be similar (or better) for children with new-onset absence epilepsy was unanswered.

The standard treatments for childhood absence epilepsy (ethosuximide, valproate, lamotrigine) are effective in ~50% of children by 16-20 weeks. However, side effects exist and include stomach upset, inattention, mood disturbance, rash, liver function test abnormalities, and fatigue. Families at times do ask about avoiding treatment completely, especially as this epilepsy usually resolves in puberty and convulsions only occur in 20% (most children have brief staring spells only). In addition, families do also ask about "nonpharmacologic" treatment, but to date the investigators have not recommended it due to lack of data.

This study will have 20 children in each arm (diet and drug) with ability to crossover. Parents with a child with new-onset absence epilepsy will choose between the two therapies. Visits will be at baseline, 1 month and 3 months. EEG, labs and clinic visits will be paid by the parent's insurance (not free).

Inclusion Criteria:

• Children ages 3-12 years at seizure onset with classic childhood absence epilepsy clinically.
• Normal intellect or mild disability
• EEG with confirmed 3/second spike-wave discharges, usually with hyperventilation
• Daily reported absence seizures.
• Generalized convulsions allowed

Exclusion Criteria:

• Previous treatment with any anticonvulsant drug
• Previous use of a ketogenic dietary therapy for epilepsy or any other condition
• Glut1 deficiency syndrome
• Metabolic disorder known that would preclude dietary therapy
• Dietary restrictions for which a high fat, low carbohydrate diet would be precluded.
• Prior history of epilepsy (febrile seizures allowed)
• Unwilling to consent to study procedures or return for visits