Wednesday, 10 October 2018

Ketone Therapy in Autism (Summary of Parts 1-6)

Open the above file via Google Drive, so it is big enough to read. Click the link below. You can also take links from it to the relevant blog post.

In the mini series of posts on ketones and autism we have come across a long list of effects that will benefit certain groups of people.

1.     Change in gut Bacteria

2.     Ketones as a brain fuel    

3.     Niacin Receptor HCA2/ GPR109A

4.     NAD sparing

5.     CtBP Activation by reducing NADH/NAD+ ratio

6.     NLRP3 Inflammasome inhibition

7.     Class 1 HDAC inhibition

8.     Increase BDNF

9.     Ramification of Microglia

10.PKA activation

11.PPAR gamma activation
It was interesting that the beneficial effect of the Ketogenic Diet in epilepsy is driven by changes the high fat diet makes to the bacteria in your gut and seems to have nothing really to do with ketones. Well it took a hundred years to figure that one out.
In the case of Alzheimer’s, you can see that more than one effect is potentially beneficial. People with Alzheimer’s do have low glucose uptake to the brain, but they also have elevated inflammatory cytokine IL-1B.
In Huntington’s it is the HDAC inhibition effect that seems to be what helps.  This brings us back to HDAC inhibition as a potentially transformative therapy with long lasting effects. It appears that the small number of people who achieve long lasting benefit from short term use of sulforaphane or EGCG may have experienced HDAC inhibition changing the expression of up to 200 genes.  In the case of sulforaphane from broccoli, some people have gut bacteria that produces large amounts of the enzyme myrosinase, which means they convert very much more of the glucoraphanin in broccoli to sulforaphane (an HDAC inhibitor).
It does look like a low dose of a potent HDAC inhibiting cancer drug is what is needed by certain single gene autisms and perhaps some idiopathic autism. This was covered in a dedicated post where we saw the long-lasting benefit of short-term use of Romidepsin. Vorinostat, a very similar drug, but which is taken orally, should be trialled in Shank 3, Pitt Hopkins and Kabuki, to see if the same transformative long-lasting effect can be reproduced.
In Multiple Sclerosis (MS) the effect on Niacin receptor HCA2/GPR109A should help a lot, but so should PKA activation.
In mitochondrial disease it was suggested that increased ketosis will help conserve NAD, which may be deficient. Also, using ketones as an alternative brain fuel may bypass problems that occur when glucose is supposed to be the fuel and thereby boost brain function. The most important effect is likely to be activation of PPAR gamma by C10, which increases the number of mitochondria and boosts the enzyme complex 1.
Many of the people with autism and an overactive immune system stand to benefit from activating CtBP, inhibiting the NLRP3 inflammasome, or activating HCA2/GPR109A.
I think there should be clinical trials using a potent HCA2 activator in autism comorbid with immune over-activation. 
We can see that some people who respond to BHB, experience an immune rebound on cessation, so this helps narrow down the likely beneficial mode of action.  In this immune sub-group, the idea to using other activators of HCA2/GPR109A would seem worthwhile. 

PPAR gamma activation should help those with mitochondrial dysfunction, but this effect is produced only by C10, not BHB or C8. For C10 you eat a ketogenic diet or add it as a supplement (e.g. cheaper MCT oil, or coconut oil).

As recently highlighted by our reader Agnieszka, perhaps the fever effect in autism can be explained by short-term ketosis. Fever is known to sometimes raise the level of ketones, particularly in children (it is called non-diabetic ketosis).  So if your child's autism improves during, or just after fever, test the level of ketones in their urine.


We may have shown the benefits of a high fat ketogenic diet, but there are very many different fats and they do not all produce the same effects.

There are many saturated fatty acids, they are numbered based on how many Carbon atoms they have.

So, C8, known as Caprylic acid has the formula  C8H16O2

Eating C8 looks to be a great way to increase the level of ketones in your blood.

Eating C10 should be good for people with mitochondrial dysfunction and people with diabetes.

Your food contains many other saturated fatty acids and your gut bacteria produce even more.

Common Name Systematic Name Structural Formula Lipid Numbers
Propionic acid Propanoic acid CH3CH2COOH C3:0
Butyric acid Butanoic acid CH3(CH2)2COOH C4:0
Valeric acid Pentanoic acid CH3(CH2)3COOH C5:0
Caproic acid Hexanoic acid CH3(CH2)4COOH C6:0
Enanthic acid Heptanoic acid CH3(CH2)5COOH C7:0
Caprylic acid Octanoic acid CH3(CH2)6COOH C8:0
Pelargonic acid Nonanoic acid CH3(CH2)7COOH C9:0
Capric acid Decanoic acid CH3(CH2)8COOH C10:0
Undecylic acid Undecanoic acid CH3(CH2)9COOH C11:0
Lauric acid Dodecanoic acid CH3(CH2)10COOH C12:0
Tridecylic acid Tridecanoic acid CH3(CH2)11COOH C13:0
Myristic acid Tetradecanoic acid CH3(CH2)12COOH C14:0
Pentadecylic acid Pentadecanoic acid CH3(CH2)13COOH C15:0
Palmitic acid Hexadecanoic acid CH3(CH2)14COOH C16:0
Margaric acid Heptadecanoic acid CH3(CH2)15COOH C17:0
Stearic acid Octadecanoic acid CH3(CH2)16COOH C18:0
Nonadecylic acid Nonadecanoic acid CH3(CH2)17COOH C19:0
Arachidic acid Eicosanoic acid CH3(CH2)18COOH C20:0

C4, familiar as Butyric acid, helps maintain the integrity of the intestinal barrier and the blood brain barrier.  Butyric acid, or butyrate, is also an HDAC inhibitor and it seems that in animal models, and some humans, a small amount can be beneficial but large amounts can have a negative effect. A small amount in humans seems to be about 500 mg a day.  There are earlier posts is this blog on butyrate.

C3, familiar as Propionic acid, is bad for you and too much propionic acid will by itself cause autistic behaviours. NAC counters the effect of propionic acid in mouse models.

All those people eating coconut oil are consuming a 99% mixture of fatty acids with 1% phytosterols.

Phytosterols like β-SitosterolStigmasterolAvenasterol and Campesterol likely explain why coconut oil actually reduces "bad" cholesterol, rather than increasing it, as predicted by the American Heart Association and others. This counters the negative effect of the Palmitic acid (C16).

Lauric acid (C12) is thought to increase HDL ("good") cholesterol and may have a beneficial effect on acne.

Myristic acid (C14) is also thought to increase HDL ("good") cholesterol.

Palmitic acid (C16) raises LDL ("bad") cholesterol and large amounts have other negative effects.

Oleic acid is also found in olive oil and is seen as a fat with beneficial effects.

Fatty acid content of coconut oil
Type of fatty acid pct
Caprylic saturated C8
Decanoic saturated C10
Lauric saturated C12
Myristic saturated C14
Palmitic saturated C16
Oleic monounsaturated C18:1
black: Saturated; grey: Monounsaturated; blue: Polyunsaturated

So the only "bad" part of coconut oil is the Palmitic acid (C16).

As for MCT oil, what is in that?

In pharmaceutical MCT oil, like the one sold by Nestle, the contents are:-

Shorter than C8      1%
C8 (Octanoic)      54%
C10 (Decanoic)   41%
Longer than C10    4%

What is the effect of those fatty acids with more than 10 carbon atoms?  Nobody likely knows.

Cooking with MCT Oil? 

This is what Nestle has in mind for dinner.

Mct Spaghetti With Meat Sauce

4 Tbsp. MCT Oil® (Medium Chain Triglycerides)
1 lb. very lean ground veal or beef
1 tsp. salt
1/2 tsp. pepper
1/4 cup chopped onion
3 Tbsp. chopped green pepper
1 cup MCT Tomato Sauce (see recipe on site)
2 cups cooked spaghetti

Heat MCT Oil; add veal, salt and pepper.
Cook until meat is brown.
Add onion, green pepper, and tomato sauce. Cook for 30 minutes over low heat.
Add cooked spaghetti, stir and serve.


  1. Great summary, thanks Peter!
    Started yesterday with the C8 oil (didn't receive the BHB yet). Will keep the board posted on the results.
    Interesting comment about the fever effect. Reason we are trying the C8/BHB is that we did the ketogenic diet some years ago (actually it was SCD), and the improvements were huge (especially the hyperactivity was way down). But after a year and a half we couldn't do it anymore, and as soon as we went back to a normal diet the hyperactivity/stinking/etc... went back up. So we are looking forward to seeing if the C8/BHB will bring us similar results. Thanks a lot again.

  2. Obviously meant "stimming" not "stinking" :/

  3. Here is another idea for an intervention that is cheap, almost certainly safe, and based upon some new research which shows that the amino acid L-Alanine changes glucose metabolism in a manner similar to metformin by activating AMPK:

    Press Release:


    A quick search on Google of this site for AMPK turns up many, many search results so I won't go into all of them, but it is thought in many autisms AMPK activation can improve autism symptoms. Here are a few I dug up pretty fast:

    AMPK from exercise improves symptoms in autistic children

    AMPK inhibits mTOR

    The Naviaux study which shows suramin therapy increased blood levels of AICAR which activates AMPK.

    The original study with L-Alanine was done on rats, however, it did show that oral L-Alanine supplementation raised blood levels of L-Alanine which is obviously necessary for it to be therapeutically useful today as injecting L-Alanine is both impractical and would take time to study in humans as to its safety just as megadoses of B-12 are safe to take orally because B-12 is poorly metabolized, but if injected you can do serious nerve damage with the wrong dose.

    Also, in autism and with respect to the recent topics, glucose issues in autism would obviously benefit from exogenous ketones or the ketogenic diet, but perhaps L-Alanine could be of significant benefit to those not on a ketogenic diet, especially with respect to its apparent mTOR inhibition.

    1. To follow up on this research as well as the recent blog posts on the ketogenic diet, here is some new research showing benefits of the ketogenic diet in helping to prevent cognitive decline in aging mice:

      Press Release:


      The takeaway here is that it is shown that the primary cognitive benefits of a ketogenic diet, at least with respect to two mouse models of brain aging and Alzheimer's disease, are due to mTOR inhibition brought about by the ketogenic diet and the effects mTOR inhibition have on increased cerebral blood flow.

      Even though this study has nothing to do with autism directly, mTOR hyperactivation in autism is kind of becoming a theme now in a lot of research so anything to reduce it may improve symptoms which may be useful as a ketogenic diet is hard to maintain compliance on, and there is some concern that from recent evidence that if the ketogenic diet is high in palmitic acid (saturated fat), that the risk of cardiovascular disease increases greatly. This means that the best ketogenic diet is going to need to be very high in unsaturated fats, such as foods high polyunsaturated fats which are generally not as palatable as eating bacon and eggs.

    2. Thanks for highlighting that Tyler, Ive always been skeptical about long term effects of keto diets, especially with regards to benefits:side effect ratio.
      I like the idea of certain recipies such as the spagetti posted above.

      As an adult with ASD/aspergers myself its easy to get over the fact that healthy food doesnt always taste the best, for example I take daily on average 1-2 full tablespoons of oliveoil and wash it down with a bit of water, as cooking the oil some of the healthy ingredients is lost.

      I can see this being problematic for kids with autism who are often very picky eaters.

  4. Here is a kind of far out there theory on how valproate might help induce autism based on some new research which accidentally discovered a method of converting GABAergic striatal neurons into dopaminergic neurons:

    Press Release:


    The original goal of the research was to induce glial cells into being epigenetically reprogrammed into striatal dopamine neurons. However, what the researchers discovered was that nothing happened to the glial cells, rather some of the GABAergic striatal neurons changed into dopaminergic via the use of three transcriptional regulators plus valproic acid.

    Now, one very interesting thing is that in the striatum there seems to be a deficit of GABAergic transmission and an excess of extracellular dopamine floating around. Why is this the case? Well there are many theories, but this research suggests that the ratio of GABAergic expressing cells (95% of the striatum in a healthy human), could be shifted in a human exposed to valproate during pregnancy via excessive dopamine expressing neurons.

    The other interesting thing about this research with respect to autism is that it suggests that it may be true that no neuron is in a permanent finalized state, rather they could be induced to change into a completely different neuron if the right transcription factors, plus a series of HDAC inhibitors such as valproate.

    All of this I stress is just an idea I thought about and has no real research to directly back it up via showing how valproate exposure could cause various brain cells to be reprogrammed into different types of brain cells in numbers undesirable for healthy neural development. Nevertheless, it also suggests therapeutic hope for literally changing to brain cell composition of the brain when certain developmental windows have already closed naturally.

  5. The amount of spam and ads that get posted is getting a bit out of hands and its starting to annoy me.
    Peter is there any way you can go back to the old posting system, where a post would be checked before published?

    1. Most spam is on the older posts, so I have made comments on posts older than 14 days subject to moderation. If you have too much moderation on posts, you get less comments, because people expect to see their comment published immediately.

  6. Hi Peter and all,
    My boy (18) is autistic and struggles against hypoglycemia. At first, I thought that MCT would be good for him, but I think that it has worsened the hypoglycemia. Do you have any thoughts, please ?

    1. Luis, non diabetic hypoglycaemia has various possible causes, some of which are relevant to autism. Have you established with your doctor/endocrinologist the cause of the hypoglycaemia?

      Via its effect on PPAR gamma there will be an effect from MCT oil. It may be that MCT oil is not a good choice for your son.

      You could try BHB and see if that gives a benefit, without any negative effects.

    2. Hi Luis,
      Some people get little or no symptoms from low glucose levels, it seems to be related to how well adapted the body is to using fatty acids and/or ketones. See e.g.
      (The comments are interesting.)
      I guess if alternative fuels can be used, then the low glucose levels may be less of a problem.

  7. Hello Peter and all,

    This is un tio. I have a question which has nothing to do with this post but I will use this section to do it:

    does anybody give paracetamol to his autist son or daughter, and check if autism goes worse or better?

    I'm not saying anybody should try, I'm just asking if anybody did try for any reason (for example fever) and could see any change.



    Un tio

    1. Un tio, I cannot say about paracetamol, but NSAIDS like ibuprofen do seem to improve some people's autism.

      Paracetamol depletes GSH, so in theory should worsen autism.

  8. Here is some very interesting research on brain development consequences from Caesarian section, and in particular a reduction of vasopressin expressing cells in the hypothalamus:

    Press Release:


    Many factors associated with autism such as age of the mother, obesity, and gestational diabetes are also associated with Caesarian delivery, but vasopressin issues seem to be associated as well. The brain has two separate nuclei in each hemisphere which secrete vasopressin. One is the paraventricular nucleus in which in this study there was a reduction in vasopressin expressing cells in mice delivered by caesarean section, and the other is the supraoptic nucleus which was not look at in this study.

    Also, even though the PVN had the greatest increase in cell death among brain regions investigated, many other primary subcortical regions involved in stress regulation had increased rates of cell death as well, many of which express CRH which drives the HPA axis.

    It is also worth noting that if there is a reduction in vasopressin levels in the brain shortly after birth, vasopressin receptors may not be regulated to the appropriate number. In addition, caesarean section born mice also have reduced distress ultrasonic call amplitude as well which in effect means they are less interested in their mother's attention and this feature was more biased in males than females.

    Last but not least, the oxytocin mediated exitatory to inhibitory GABAergic switch (the reason for Bumetanide therapy) is speculated upon with respect to the decline in vasopressin as the oxytocin and vasopressin peptides are very similar and more or less are co-released in the same areas of the hypothalamus.

  9. This comment has been removed by the author.


Post a comment