When is a ketogenic diet not a ketogenic diet?

I've listened to a few youtube/podcasts videos recently. Seyfried, Veech and D'Agostino appear to besettled in to what looks like a "Ketone Ester" corner. There are many, many things which make me splutter a little in some of the things they say but I think it's hard to decry ketone bodies too badly. They clearly do things. It's quite possible that the main effect of flooding the TCA with acetyl-CoA is the inhibition of glycolysis. If it does nothing else, that seems worth doing.

The flip side is Ron Rosedale's view, shared by a few others. Towards the end of his presentation, largely about mTOR, he takes a position on ketones. I don't think he has anything against them per se, but what he really wants is a metabolic state based around beta oxidation of fatty acids. Just enough protein, minimal carbs, oxidise fats. If that throws off some ketones, so be it.

I clearly recall Wooo posting some time ago about the ketogenic diet for epilepsy and pointing out that what matters is compliance with the diet, not levels of a given ketone in blood or urine. If fatty acid oxidation dominates I'd be willing to bet the glial cells generate a ton of ketones which enter neurons without a dipstick in sight.

I think I might be in that camp. So I'm a little uncomfortable with medium chain triglycerides, octanoate alone, ketone salts and ketone esters. They clearly have benefits but they are not a route I would take currently.

I like ketones as a surrogate for fatty acid oxidation. Or should I say that I like beta oxidation, and ketones are a reasonable surrogate.

For the beta oxidation "camp" we have this paper:

Induction of ketosis in rats fed low-carbohydrate, high-fat diets depends on the relative abundance of dietary fat and protein.

It's a no nonsense sort of a paper. For the LC rats the diet was beef dripping. With added casein at 5.5%, 11.8% or 19.1% of calories and a few vitamins and minerals. The only carbs came from the vitamin/mineral mix. Anyone could get any rodent diet manufacturer to formulate it:

Rule one of a scientific paper: The methods must supply enough information to replicate the study.

Replication = validation. Without it your paper is worthless.

So rats on a LC diet are only in ketosis when over 90% of calories are supplied as beef dripping:

A BHB blood level of 28mg/dl is reasonable ketosis. That would be around 3.0mmol/l in new money. That's for the rats on 5.5% casein in their beef fat. All else was ns compared to the chow fed rats. There we have it. Decent ketosis in a rat is reliably achievable by feeding beef tallow. No MCTs, no ketone esters, no octanoate. Anyone could do this with their rodents. One niggle:

For the CICOtards: The LC animals were pair fed to the calories eaten by the chow fed rats, a feature of the paper I dislike a little. It's equal calories all round but the low carb rats probably ate less than their appetite would have dictated. This might have accentuated ketosis (but we'll never know...) and ad lib feeding might have blunted ketosis.

The weight gains themselves are fun:

Obviously the red squares are the ketogenic animals. There is a table of blood insulin, glucose, FGF-21 and FFAs but, well, we all know what the number have to be so there's not much need to go in to it in any detail. KDs work.

So if you want to know what a ketogenic diet does in a given medical condition, this is the one diet you have to use on a rat model. Maybe use butter instead of beef dripping (I'd prefer this for myself) but it looks like a gold standard to me... Ketones as a spin off of a whole body FFA based metabolism. The metabolic state is what interests me. Replicate at will.


Let's compare this with a ketone ester camp paper. This is the diet:

"...mice received KD-USF, a custom ketogenic diet designed by the authors and produced by Harlan Laboratories, fed ad libitum".

What's it made of? Dunno. Here are the macros:

What sort of fat? Dunno. Just fat. Maybe: Crisco? Fish oil? Canola? Coconut? Mmmm, butter? Replication anyone? If you can't replicate the study how can you tell whether the results were made up or real? Personally, I think the results are absolutely true. I have no doubt. Why? Because this is the level of ketones generated by the diet, where it says KD:

The ketogenic diet generated, estimated by zooming in on the above chart, something around 0.1-0.2mmol/l of BHB. No one would make up a figure that low, these are honest results. With added ketone esters we get up to almost exactly the level of ketones found in Bielohuby's truly ketogenic tallow fed rats, without the crippling expense of the ketone esters.

People shouldn't get me wrong. I have nothing against trying to use a ketogenic diet for management of cancer. I even think using ketone esters might be reasonable for folks who can't cope with cream, butter, eggs and a wide variety of meats and non starch veggies.

What I would prefer is for a diet described as ketogenic to actually generate some ketones. You cannot describe a diet generating 0.1mmol/l of BHB as ketogenic! Especially when close on 3.0mmol/l is easily achieved on a beef tallow based true ketogenic diet. It's not exactly surprising that adding 3.0mmol/l BHB derived from ketone esters should out performed a non-ketogenic "ketogenic" diet for cancer management! And the non ketogenic high fat diet did help a little, presumably by eliminating starch triggered insulin signalling...

I was a bit shocked that non of this was discussed in the discussion. Being driven by a love of ketone esters is no excuse for sloppy science. When you are on the winning side there is no need for this.

Peter