Tuesday, September 23, 2014

Ketones for ALS?

OK. I've been thinking a lot about ketogenic diets and motor neuron disease, which appears to me to be just one facet of Alzheimers, Parkinsons and a number of other neurodegenerative diseases.

The first thing to say is that they (ketones) don't seem to work terribly well. I picked up this paper via the Deanna Protocol website. I wrote an unpublished post at the time setting down what an abysmally written paper it is but I thought I would stick to the basics today. Feeding a ketogenic diet to the mice engineered to have an ALS-like disease delays their time to falling off a log but does not extend their lifespan:

"There was no statistically significant difference in the age at death between KD fed animals compared to SOD1-G93 transgenic mice fed a standard laboratory diet (133 ± 4 vs. 131 ± 4 days, p = 0.914)"

Some improved motor function, for a while, may be worth having if you suffer from ALS but I don't think it's exactly a cure or remission.

Although the methods section is very reticent about the diet, it is high in carbohydrate (20% of calories) and protein (20% of calories), so must be MCT based to achieve ketosis.

The findings are confirmed by a nicely written, very clear paper using caprylic acid as a supplement to standard CIAB mouse chow:

"SOD1-G93A animals on caprylic triglyceride diet had a median survival of 135 days. Although it was longer than the median survival of SOD1-G93A animals on control diet (129 days), it did not reach statistical significance (Mantel-Cox test, p=0.165)"

Things weren't much better here in the Deanna Protocol paper. Their ketogenic diet supplied 77% of calories from fat, type unspecified, and essentially zero from carbohydrate. Protein was high at 22% of calories. This is what you get as a result:

"Although the mean survival of SOD1-G93A animals was longer in all three treatment groups than the control group, this difference reached statistical significance only in the KD+DP (4.2%, p = 0.006) and SD+DP groups (7.5%, p = 0.001, Fig. 5, Table 5, Data S2)."

i.e. the ketogenic diet, without the alpha ketoglutarate of the Deanna Protocol, was no better than control mice on CIAB.

All of which is quite interesting and should be quite depressing for groups working with MCTs, ketone esters or ketone salts as managements for neurodegenerative diseases.

We can say certain things about the first two studies. Using MCTs on a moderate to high carbohydrate diet is unlikely to lead to the metabolic changes of a true ketogenic diet. Normoglycaemia is probably not on the menu. It will not lead to the sort of effects of minimal carbohydrate, just adequate protein, very high fat diet. The effect of such a diet has been described as unique.

Of course, a few grams of MCTs on a diet of standard lab chow will generate ketones. That is hardly equivalent to a true ketogenic diet with its reduced glycaemia and basement value insulin levels.

As the paper on the Deanna protocol reports:

"Blood glucose was not significantly different between the diet groups", not exactly what was reported for mice eating D12336.

Ultimately, no one yet appears to have looked at a true ketogenic diet in ALS.

The focus is on the ketones. Ketones are good, but they are not magic. There are people who believe that the ketones themselves are simply a surrogate for very low insulin levels, which is magic (You know who you are Wooo!) and that the benefits of ketogenic diets may stem from the low insulin levels rather than the ketones per se, certainly for obesity management. For neurodegenration I find this idea very appealing. I think that the low glucose/insulin might be particularly important within the brain. I can't see that the work has been done yet, too much of a focus on ketones.

As something of an aside, the Deanna Protocol is interesting in its own right. The core supplement is (arginine-linked) alpha ketoglutarate. From the Protons point of view, if the alpha ketoglutarate enters the TCA at alpha ketoglutarate dehydrogenase and leaves it at malate, it would appear to be a very FADH2 selective input at complex II, generating an NADH:FADH2 ratio of 1:1, i.e. it is functioning as a rather specific FADH2 input. We're all aware that complex I dysfunction is a hallmark of neurodegenerative diseases and, in the absence of beta oxidation (we're in neurons here), complex II is the primary route in to the CoQ couple for electrons via FADH2. Along with mtG3Pdh of course, if that happens to be active. I can see the logic to using this AKG to push complex II without the excess rise in non-usable NADH, which large amounts of acetyl-CoA provide. I'm not surprised AKG is the core component of the Deanna Protocol and hats off to her father for picking this up.

A further aside, Deanna tried coconut oil, caprylic acid and MCTs early on in her disease. Not a lot of help. Adding extra acetyl-CoA from ketones will be of limited help in a condition with complex I dysfunction. My interest still lies in ketones combined with low blood glucose, not as an add on to healthy starches.

It is quite clear from the last post featuring cardiac ischaemia and ketones that any old ketones will do when hypoxia is the problem: Bring on the MCTs. Logically ketones are fats, part pre-oxidised in the liver, so require less oxygen to complete their metabolism in the cardiac muscle. They do not uncouple protons from oxidative phosphorylation either, which we will probably come back to. And while normal fatty acids do uncouple ox phos, this effect is (under normal circumstances) completely lost when mitochondrial ATP levels fall. This probably happens rather quickly under hypoxia.

The energetic failure of neurodegenerative diseases is only partially amenable to ketones. We are looking at a rather different phenomenon to ischaemia and it might be worth looking at the problems of burning glucose in neurons next. And the problems from failing to generate adequate superoxide for maximal health. There's a lot to think about.



August said...

This Deanna protocol is very interesting. I have some AAKG just sitting around because I forget to take it, and I've ordered some picamilon, which is a compound of niacin and GABA.

I didn't intend to attempt to use either as a therapy for the unexplained pain on the left side of my face, but now I shall. Been in pain since 2010 and doctors are of no help.

Lacie said...

Thank you so much, Peter - I've been waiting for you to explain AAKG as no one else inclined to do so.

As I've posted previously in various neighborhoods we hang around in, my partner with Parkinson's decided to refuse L-dopa and instead take AAKG, 1:1 ratio Primaforce, 18g per day. We heard about it on Deanna Tedone's website, winningthefight.net in a document by Dominic D'Agostino, Ph.D., University of South Florida, who helped Deanna's father develop the protocol.

My partner is six years past onset of PD and has a slight tremor in his right thumb that's worse in the morning (dawn glucose/cortisol dump, probably), when he's hungry (IF isn't a good idea for patients with neurodegeneration), or in times of stress. Insulinogenic foods also wreak havoc with his movement, but when he stays low carb <40g per day and pumps the AAKG, the tremor is barely noticeable. No problems with walking or balance. He should be in far worse shape.

The AKG is the active ingredient but the arginine is needed as a buffer for the stomach. The nitric oxide produced by arginine is a concern as NO seems to trigger apoptosis but is also neuroprotective. See http://www.ncbi.nlm.nih.gov/pubmed/16005074 http://www.ncbi.nlm.nih.gov/pubmed/15691583 and http://www.ncbi.nlm.nih.gov/pubmed/17408564 (PM if you want full text.)

NY said...

Hi Peter,

Would love to know your take on the study which came out apparently showing that some protein sources are more insulinogenic than even carbs eg., lean beef and fish raised quite a bit of insulin as compared to pasta. If keeping insulin low is the whole point of cutting carbs then do you think we should also be careful about lean sources of protein? Also, what do you think optimal fasting insulin levels are? Mine came at 3.5 iu/ml.

Peter said...

NY, lean protein is for paleotards. I eat fat! Adequate protein hidden in fat and some carbs as flavourings.


Peter said...


Back in my days of anaesthesia examination endurance the rumour was that if you were stumped for an answer you just said nitric oxide. If wrong it might still get a laugh in the viva. It does everything. It reminds me of superoxide!

Physiologist's T shirts had "Just Say NO" written on them. Nothing to do with recreational drugs.


Guillermo Fernandez said...

Great post Peter, as usual!

Ash Simmonds said...


If it's the study I'm thinking of from like decades ago, you should actually read it.

It compares the insulinogenic AND glucogenic profiles of protein, fat, and carbs.

Long story short - fat does fk all, protein causes a *mild* increase of BOTH insulin and glucagon, and carbs spike insulin and suppress glucagon.

It's really dumb how it's painted in sound bites like "beef spikes insulin just like pasta".

twitchyfirefly said...

Peter, I'm so glad you're writing about ketones and ALS. I have a particular interest, as I have something called multifocal motor neuropathy (MMN) which is often misdiagnosed as ALS. Most of the stuff I read online treats neurodegenerative disease as primarily in the brain, which makes it hard for someone like me to extrapolate, as my problem is strictly peripheral. Any little crumbs you can drop about whether effects are similar in the brain and in peripheral nerves would be greatly appreciated.

I looked at that Deanna Protocol a couple of years ago. I actually wrote to them asking why they did not recommend reducing carbs in conjunction with the AAKG. Didn't get a useful answer back that I recall. Perhaps they figure that it would just be too difficult for people to change their diet, despite being under what is essentially a death sentence.

That said, a lot of the supplements in the Deanna Protocol align with what I have been able to discover independently.

You might also find this of interest: Reversal of cognitive decline: A novel therapeutic program

Amy Berger said...

Spot-on, Peter: "Using MCTs on a moderate to high carbohydrate diet is unlikely to lead to the metabolic changes of a true ketogenic diet. Normoglycaemia is probably not on the menu."

I wrote a paper about the therapeutic use of low-carb/ketogenic diets for Alzheimer's, and the way I phrased it was: "MCTs could certainly be a powerful *adjunct* to a reduced carbohydrate diet. The latter *would* reset the metabolic machinery, ease oxidative stress, and reduce glycation. In the absence of dramatic dietary overhaul, however, the administration of KBs is akin to bailing water out of a leaky boat without stopping to patch the hole: you merely manage the effects while the root cause continues wreaking havoc."

Russell said...

Can someone please elaborate on Lacie's comment that intermittent fasting isn't a good idea for patients with neurodegeneration