Modified Ketogenic Diet for Childhood Epilepsy

When I started this blog several years ago, we were in the midst of successfully treating our daughter’s epilepsy with a modified ketogenic diet. At the time of this writing, she is 9 years old, 5 years seizure-free, and no longer on a diet therapy!

For the full story, recipes, etc., please see

My goal here to to describe our “modified” version of the ketogenic diet that gave Nora great seizure reduction. Although still very rigorous, I think of it as keto-lite in practice. To describe what I mean, I think I need to tell the story of how we got to this modified version of the keto diet.

The ketogenic diet for epilepsy hit the mainstream in 2010 with an article in the NYT Magazine, “Epilepsy’s Big Fat Miracle.” I recommend it for a history and description of the typical keto experience;  I will not repeat those details here. Read the article here and now:

When Nora started having myoclonic seizures and was not responding to medication, several people pointed me to the article. I finally, reluctantly, read it. Reading stories of other kids’ struggles with epilepsy was difficult for me. One the one hand, I felt that we had it so much better than other families because Nora was not having hundreds of seizures a day like other kids. On the other hand, I was afraid that I was looking into our future of intractable epilepsy. Many kids started out like Nora.

A few things in the article were daunting to me. First, the image of the weekly intake of bacon, butter and cream was a little shocking:

Stephen Lewis for The New York Times; Food Stylist: Brett Kurzweil

Second, this line resounded in my head: “To jump through these arithmetic hoops, Evelyn [Sam’s mom], [ ] gave up her career to take on the now full-time job of feeding Sam.” Could we do this? I knew I had the skills. I love to cook, have a working knowledge of nutrition, better than average math skills, and an anal attention to detail. But I am the first to admit that I am a happier person and better mother when I have a nominal amount of intellectual work and time away from my children. If one mother had to give up a CAREER to feed her kid, did I really want to do this? Because as we knew, food preparation is my department in our house.

The article gives you a good description of the diet: 3 grams of fat for every 1 gram of carb + protein. No more than 10 grams of carbs per day. Weighing food to the gram to calculate the carb/protein/fat breakdown of each portion. Scraping out the weighing bowl and getting every bit into the kid. Being sure that each meal meets the diet’s requirements to keep the ratio constant. And not to be overlooked, a hospital stay to initiate the diet with a fast and monitoring.

You should also know that there 3 diet options for epilepsy treatment. The least strict is a low glycemic index diet, which allows 40-60 carbs per day, 1:1 fat to carb+protein ratio, and estimates of these values. Some people benefit greatly from this less-strict approach. The medium approach is the Modified Atkins Diet for Seizures (MAD). Like the Atkins weigh loss diet, you start with an “induction” phase of 10 carbs per day for 1 month, then liberalize to 20 grams of carbs per day. Fat ratio is 2:1, generally. Values are also estimated not weighed.

When we started the diet approach, we agreed to start with the medium-level MAD diet to see if we got any results. I will describe that first month in another post, but in time we found that Nora was doing better on a higher ratio of fat to carbs + protein. When we kept it closer to 3:1, her seizures decreased. When I discussed this with the dietician, I always wanted to know what other families experienced. The dietician’s somewhat amused response was, “other MAD families don’t keep records like you do. It sounds like you could do the ketogenic diet.” I couldn’t imagine NOT keeping records. How would you know that your kid is getting only 10 carbs per day? And enough fat? We were hardly given guidance to start the diet, just a website:, a page run by parents, not medical professionals.

In addition to record keeping, we moved from pure estimation to weighing with a gram scale for our own peace of mind. And honestly, it is easier. Some foods are difficult to estimate. An online nutrition database will tell you the macronutrients in 1/4 avocado, but avocados range in size. I was not always confident in my estimates, so the gram scale became my friend, not a burden.

But I was reluctant to move to the full ketogenic diet. There were a few things that still concerned me. First and foremost, using the keto diet database sounded like a burden. Although I could see how it would ease record keeping and meal planning, it also prescribed particular brands of foods to reduce any variation in the estimates. I could not see myself buying Hormel bacon for my kid, and feeding her 12 slices per week. I strongly believe in organic and locally sourced foods for personal health and systemic food system reasons, and I expected to be able to continue to use the foods that I found most wholesome for her growing body and brain. I had been successful in estimating the carb-protein-fat breakdown of our regular foods, judging by the seizure-reduction Nora experienced, and felt that I could continue without the guidance of the official database.

It was also unclear to me if I could use recipes to deliver her food. I can get a pretty good estimate of the breakdowns in the recipes I make by using some online recipe analyzers, or doing the math myself. But if every portion is not mixed uniformly, the proportions will be off and it could violate the keto diet.

I was also skeptical about the need to calibrate every meal perfectly, weigh, scrape out every atom of food weighed, and get each atom into her body. Again, we had good results with our less exacting methods at home and I didn’t see the need to add more work for marginal difference in outcome.

After about 6 weeks of progressing from MAD to a more rigorous (but not full-blown) keto diet, we had an appointment with Nora’s doctor that specializes in the ketogenic diet. He was happy to let us continue with our current approach, considering the great results and our ability to handle the diet. Nora was almost seizure-free after a few weeks of a consistent 3:1 ratio. Because of the 3:1 ratio, he was calling it the ketogenic diet, even though Nora did not start on a hospitalized fast, we were not using their keto database, and we were not measuring and planning as rigorously as the method described by the NYT article cited above.

Therefore, I am calling our approach the modified ketogenic diet. Maybe we can get away with it because Nora’s seizure count was not so very high, but on the other hand myoclonics are a notoriously difficult epilepsy to control. We know that there are different “levels” of this diet approach that work for different people, so maybe we’ve just found the right mix for Nora. I don’t suggest that these results are generalizable. But I do think that the diet can be started with a less rigorous approach and increased if “stronger medicine” is needed. I hope that the practice eventually allows for this kind of gradual approach, because it may be much more manageable for a broader range of patients.

Finally, I would call our approach “keto-lite” because Nora is not eating anywhere near the amount of bacon suggested by the NYT article. I will describe Nora’s food choices in another post, because that is the most frequent question that I field about her “special diet,” as she calls it.

Print Friendly, PDF & Email

7 Responses to Modified Ketogenic Diet for Childhood Epilepsy

  1. brekkent says:

    You are an amazing parent.

  2. Ellen says:

    I loved this post. I talk about these very issues with the “clinical” version of the ketogenic diet on my website. I think you are smart to take a leap of faith with whole foods and stick to the spirit of the diet rather than jump through the medical hoops. Good for you. I hope it’s okay to link to this post from my epilepsy page here:

    I think it would be helpful for other parents facing the same issues. Bravo to you.

    • Christy Anderson Brekken says:

      Thanks Ellen–I checked out your site, and will spend more time with it later (I’m a tired parent tonight). Bravo to you as well.

      I want to refer you to our even newer site that will focus on Nora’s diet. We decided to make a dedicated site as a resource, rather than fit it in with a personal blog. We have added more about Nora’s typical diet there. I would suggest that you share the link to that site instead, and thanks!

      By the way, how did you find this so fast?

      • Ellen says:

        I have a Google Alert set to send me a email when a new site is posted with the keyword “ketogenic diet”. I’ll change my link to the site you suggested.

  3. HopefulMom says:

    Wondering if have any additional information (I’m sure from your attention to detail that you have scratched every surface) regarding Myclonus? My son has developed these (previously tc) and they seem to be somewhat calculated…not that something triggers them but the events and side effects are all the same. Any information you can provide would be greatly appreciated.

    • Christy Anderson Brekken says:

      Hi, I want to refer you to our other blog that focuses on Nora’s diet. We decided to make a dedicated site as a resource, rather than fit it in with this personal blog. We have added more about Nora’s typical diet there, her history, etc.

      Myoclonics have a typical presentation, which is how you identify them. Nora’s were centered around sleep–many in the morning, fewer later in the day, and clustered around times when she was getting tired. But they could also show up at any time. Maybe check out the other site and we can answer more questions there if you have them. We are very happy to help!

  4. Pingback: Epilepsy/Alzheimer’s Disease Treatment: Ketogenic Diet vs Drugs | What's Up Doc?

Leave a Reply

Your email address will not be published. Required fields are marked *