The effects of Kabuki Syndrome on my son Gus’ brain are varied and at times hard to nail down, but certainly include difficulties with math, spatial reasoning, drawing and handwriting. At 13, simply finding his way around the streets of our neighborhood can be a frustrating endeavor for him.
On the positive side, his memory for facts, music and stories is superb; these are some of the benefits that come with Kabuki. History fascinates him. He’s creative and fun, a passionate human being who engages deeply with his friends, teachers and family.
I initially felt overwhelmed by Gus’ Kabuki diagnosis, and made only token attempts to understand it on a deeper level.
However in 2022 I learned about a fascinating experiment that totally changed the way I viewed Gus’ challenges. A team of scientists at Johns Hopkins had bred a strain of mice with Kabuki Syndrome, who then displayed cognitive deficits generally comparable to Gus’, notably with processing spatial information.
Then, after a short course of a high-protein, low-sugar diet similar to the Atkins Diet, these mice showed a dramatic improvement in their formerly-deficient ability to navigate mazes.
This worked, the scientists posited, because of the “ketogenic” nature of the diet - that is, it was so low in sugar that it caused the body to switch over to using ketones instead, which are a kind of back-up power generator we all possess. When activated, ketones spread to every cell and cause a variety of changes in the body - perhaps priming it to change its pattern of activity for an environment where (sugar-containing) plant food is harder to find.
Ketones open DNA up
Ketones have the ability to enter the brain, then descend deeply into our neurons and form bonds with neural chromosomes, which respond by “opening up” and expressing proteins more freely. This makes them, it is thought, an effective foil for Kabuki, which tends to cause chromosomes to remain tightly bundled, blocking essential gene transcription.
Coming up with a plan
Putting Gus on this diet seemed daunting - but how about a ketone drink? Direct delivery of ketones - without the impossible diet. Could that work?
I’ve spent the past three years trying to get at an answer.
This proceeded on two fronts. First, finding the right ketone vendor: one with an effective, safe, product that was reasonable in price.
Keto Mojo in action
Making myself the guinea pig, I bought products from three vendors and drank them myself, measuring my own keto levels using a keto-mojo ketone meter. Ketone-IQ seemed to have the best bang for the buck, so I decided to put Gus on it for a week and measure his ketones several times a day. (This study and others gave me confidence that it would be safe to do this.)
The winner
Over a week in Sept 2023, I gave Gus a drink consisting of 2-3 servings of what is now marketed as the “Ketone-IQ Classic Multiserving” drink, mixed with diet coke. He didn’t LOVE the taste, but it was palatable enough with some (occasionally firm) encouragement on my part. I found it to raise his ketone levels immediately, from 0.1 or below, to between 0.9 and 3.4 mmol/L. I tested hourly after adminstration, and the levels fell relatively slowly. Here was a typical day (Sept. 1):
| time | action taken | measurement | ||
|---|---|---|---|---|
| 9:10AM | 2 servings of Ketone-IQ | |||
| 10:10AM | keto-mojo strip test | 0.9mmol/L | ||
| 10:50AM | keto-mojo strip test | 0.8mmol/L | ||
| 12:17PM | keto-mojo strip test | 0.7mmol/L | ||
| 12:25PM | 2 servings of Ketone-IQ | |||
| 1:28PM | keto-mojo strip test | 1.2mmol/L | ||
| 2:45PM | keto-mojo strip test | 1.8mmol/L | ||
| 3:54PM | keto-mojo strip test | 1.1mmol/L | ||
| 5:19PM | keto-mojo strip test | 0.7mmol/L |
The strip test, I learned, is always not the most accurate (blood draw being more accurate, more on that technique later), but I did similar tests over several days and this one was roughly representative of them.
However i wanted to do this under the supervision of a doctor who could help me validate the results and also affirm the safety of what I was doing. Finding this doctor was the “second front” of my plan.
Dr. Jackie Harris at the Johns Hopkins-affiliated Kennedy Krieger Institute has been at the forefront of many Kabuki-related activities and investigations and has a special interest in keto and Kabuki, so I naturally reached out to her. She was receptive to the bits of research I had done to that point and invited me to come down to her clinic in Baltimore, which Gus and i did at the beginning of last year, January of 2024.
There her team gave Gus a complete “neuropsych” test, which evaluates cognitive performance, and we agreed on a course of action: I would give Gus 8 servings a day of Ketone-IQ, divided into 4 in the morning and 4 after school. Later in the year, we would return to Kennedy Krieger for a follow up neuropsych.
We started with the regimen in January 2024 and finished in August, building slowly up to the target amount over a period of several weeks.
During this time we didn’t notice any changes in him that were clearly attributable to the ketone drink. Of course as a 12 year old turning 13 and entering puberty, during these several months he did change in many ways, including in his awareness of the world, interests, gradual improvement in certain skills, etc. But it was hard to point a finger and say that there was a dramatic improvement in particular cognitive areas. We informally reached out to his teachers and heard back similar things. In particular, his occupational therapist, who worked with him weekly, gave him standard tests that roughly mapped to the beginning and end of the trial. These didn’t show improvement.
Neuropsych #2 in August represented the culmination of our effort. We received the results over a video call with Dr. Harris, so the only record I currently have of them are my notes from the call. Here’s a summary of them.
The Results
(Note: The neuropsych test is divided into several sections. These are structured notes i made after the fact based on my in-call notes, laying out the results of the subtests and Dr. Harris’ comments on the meaning of those results).
| test name | “before” score | “after” score | Dr. Harris’ comments |
|---|---|---|---|
| SimRaw lang | 25 | 32 | |
| block design | 22 | 18 | |
| matrix reasoning | 14 | 20 | “substantially better” |
| digit span | 29 | 33 | “29 already a good number” |
| coding | 19 | 32 | “executive function… marked improvement … this+matrix: wow!” |
| vm1 | 17 | 17 | “copy shapes” |
| verbal memory | 35 | 45 | “quite good already … ‘very good’ to ‘great’” |
| memory for designs | 64 | 96 | “select picture seen before but modified … substantial jump” |
| flanker inhibitory | ? | ? | “executive function” (didn’t get the numbers, but Dr. stated they were essentially unchanged) |
| route finding | 2 | 9 | “find route between 2 points on map. really really big jump” (but still well below normal range, said Dr.) |
| narrative memory | ? | ? | Didn’t get the numbers here again but Dr. described them as substantiall increased, quote: “off the charts - 1.5+ standard deviations … from borderline to average” |
To summarize this table – out of 11 subtests, there were:
- 8 where the score increased (highlighted in bold)
- 2 that stayed the same
- 1 that declined.
Dr. Harris seemed pleased, but did say there was no “smoking gun” per se. She noted that the improvements were in the areas of executive function and memory.
I try to be measured in my interpretation of all of this, after all we were lacking many of the elements that good experiments tend to have, like a way of measuring placebo effect, or an accounting for natural random changes that can occur in a child, or improvements that may have happened anyway as part of his development.
Despite that, I allowed myself to be encouraged by the numbers that increased, and by what I saw as Dr. Harris’ generally positive assessments. She seemed excited to me, and wants to invest more of her own time in this investigation, which is closely linked to her funded research anyway. Next steps she discussed on the call included evaluation of the before and after methylation tests that were a part of our visit, writing up a case study, and, were we to continue with the keto drink, doing another neuropsych in a year.
And we are continuing with the keto drink. It’s quite expensive, so we will be doing only 5 servings a day. I’ve found it helpful to pour it out in the morning and let Gus sip on it throughout the day, with the understanding that it needs to be done by bedtime.
Dosing
The dosing question in general is a tough one. As you can see from the first table, 4 servings kept his levels around 1mmol/L for several hours on that day. Dr. Harris stated to me the number of 2.0 mmol/L as a target, though the search for the right number there is ongoing in her lab so that was not meant to be authoritative I don’t think.
About midway through the 6 month trial, Dr. Harris prescribed a “venous” keto test to Gus’ family doctor. This involved going into his office and getting a blood draw, and sending the results to a special clinic.
On the day we got the test done, we were particularly careful to note timing and dosage. That day Gus had 4 servings of Ketone-IQ, equivalent to 40g of ketones, at 7:30am, and an identical dose at noon. The blood draw was noted to be at 2pm. Here were the results:
I had a hard time parsing the details of this form, but Dr. Harris told me that it indicated a reading of 2.8mmol/L. Given that the venous test is the gold standard test, and it was taken 2 hours after the drink, i found it quite encouraging as an indication that we were likely meeting or exceeding the 2 mmol/L target for several hours per day throughout the 6 month period.
(I’ll just quickly note here that our compliance with our regimen was not 100%. On the weekends we often slacked and only gave him one dose a day, or sometimes none on Sunday, and there were probably 1-2 other missed doses throughout a typical week.)
Bringing me back to our current regimen of 5 doses of Ketone-IQ per day, 1/2 bottle. The bottles themselves currently cost $36 apiece in bulk, or $3.60 per dose:
Ideally I would have liked to increase Gus’ dose from 8 daily doses to an even much higher number to see if we could juice his numbers.
However this would have been prohibitively expensive for us, and there would of course be questions about his willingness to take that much, and the health of doing so.
How much bang for the buck will we get with 5 servings? Let’s take the venous measurement as a basis of comparison. On the day of the venous test, Gus took 4 servings at noon, and his ketones were at 2.8mmol/L at 2pm. If we are giving him 5 servings, we can hazard that after 2 hours his ketone level will be 2.8* 5/4 = 3.5 mmol/L. Applying the slow reduction in ketone level which I observed in him while taking hourly keto strip tests during our initial testing, it seems reasonable to suppose that his ketone level would still be above 2mmol/L several hours after administration.
How to compare that to the scenario we are in, where he takes 5 servings in small bites throughout the day, i can’t say. But it feels like we’re in the ballpark to be above 2 mmol/L, Dr. Harris’ rough target, for a good chunk of each day.
So we are settled for now on the half a bottle a day, which is definitely not nothing to us cost-wise but can be worked into our monthly budget with a little engineering. I plan on continuing this indefinitely, and continuing to measure progress with regular neuropsych evals.
Note: if you create an account here at Kabuki Brain, you should receive an email notification when I post any updates here!
Other note: I’m publishing this just on my own lights - have no agreement to promote this product or anything like that!
Other note: Gus has what is known as a “clinical” Kabuki diagnosis, meaning doctors have as yet been unable to track down a genetic cause.
P.S.
For a slightly deeper dive into the science, this video is a good friendly overview of the basic contours of the epigenetic landscape Kabuki disrupts.
