Protein therefore Glucose

There has for some time been a disturbance in the keto-force around the production of glucose from protein excess to maintenance requirements.

A steak is the same as a cake

One one side you have early adopters of a ketogenic diet like Jimmy Moore (Adam and Jimmy’s Keto Talk podcast) who has observed his own ketones go down and glucose go up when he goes over a specific limit of protein.  And you have the clinical experience of doctors who treat thousands of bariatric patients who have seen the same thing over and over.

The hypothesis given is that;

If you increase protein, you increase the substrate for making glucose, and you end up with more glucose and less ketones in your blood

A steak is not a cake


On the other side you have  Dr. Bill Lagakos who started the ball rolling in 2013 with his blog post “Dietary protein does not negatively impact blood glucose control” who prosecuted the case that gluconeogenesis is not substrate driven.

Dietary protein-derived amino acids have a purpose, and that purpose is not carbs.


This has blown up more recently as some members of one of the larger Keto communities on facebook have piled on with the post

Gluconeogenesis – The worst name for a rock band ever

I tend to think both sides are correct, and I can explain the biochemistry if you really want to get into the weeds.

We spill ketones when we have 2 things going on;

  1. A build up of Acetyl-CoA in our mitochondria because we have plenty of fuel (in this case beta oxidation of fatty acids)
  2. a depletion of a key metabolite in the Krebs cycle – Oxaloacetate (OAA).


When we make glucose in our livers the process consumes OAA and so the mitochondria in our livers spill ketones. The substrates for GNG can be glucogenic amino acids, but they can also be glycerol, pyruvate and lactate.

The rate limiting factor of GNG is not availability of substrates it’s an enzyme (fructose 1,6-bisphosphatase or FBPase-1) that is dis-inhibited by the hormones glucagon (and epinephrine but that’s another longer story).fbpase

Glucagon is inhibited by Insulin – Sooooo to cut a long story short when Blood glucose is low, Insulin goes low, glucagon goes high, FBPase-1 goes high, and GNG substrates get converted into glucose, oxaloacetate get’s “permanently borrowed” from liver mitochondria which are burning fat and we spill ketones.

Glucose↓→Insulin↓→Glucagon↑→New Glucose↑→Ketones↑

Note: When we eat protein we actually secrete glucagon so there is some new sugar made (gluconeogenesis), we also secrete insulin to use the new sugar – so although that does increase our exposure to insulin it normally has no net effect on our serum glucose.

So from a 1000 yd view that all should make sense – blood glucose goes low, we make more glucose. When we make glucose we make ketones, and if we are keto adapted our brains run almost 80% on ketones and 20% on glucose. This is kind of important because if you are in ketosis this is what is keeping you from slipping into a coma.

Let’s play a thought experiment – what happens if for ANY reason your ketones drop. Let’s say your available production of ketones dropped by 25%. So instead of your brain running 80% on ketones it now runs 60%, so your brain has to increase the amount of glucose it draws from 20% of its energy needs to 40%.

See the leverage point? If you lower your production of ketones by 25%, you increase the glucose required by 100%.


Anything that lowers your ketone production a small amount will increase your glucose demand by a significantly larger amount – and as we have seen GNG is a demand driven process.

Remember that depletion of OAA is critical to ketone production. What if OAA is replaced? Ketones production would plummet, brain consumption of glucose would skyrocket due to the leverage, and more production of new glucose will occur and if OAA is still getting replaced you get no new ketones.

So if anything can make more OAA for us then we’re screwed right?

screwedGlucose can make OAA via condensation of pyruvate with carbonic acid. But as I learned from ( The Daily Lipid Podcast Episode 13 ) that process is inhibited in liver mitochondria by methylglyoxal which we metabolize from the ketone acetone. So making ketones prevents any glucose laying about from stopping the production of ketones. So that is a good thing.

There is another way that we can make OAA
The amino acid aspartate can also be transaminated enzymatically to make OAA.

Note: aspartate can come from the metabolism of the sweetener aspartame – so that’s another reason not to use aspartame if you like ketones

Anyway this explains how Adam and Jimmy can observe that people who eat protein see a drop in ketones and a significant increase in glucose – and yet GNG is still not a substrate limited process.

If this is true then I would change the trite Cake-Steak statement to

“A Steak therefore A Cake”


ps: … and yes I know that is a drumstick which is also not a steak.  Surely there should be a steak emoji.

4 Responses to “Protein therefore Glucose

  • If I understand this theory correctly, wouldn’t it imply that a steak plus exogenous ketones or MCT oil does not equal a chocolate cake? Has anyone checked to see what happens to their blood glucose if they go over the protein limit but also make sure their keytone levels do not drop by supplementing with exogenous keystones or MCT oil?

  • this is a great explanation. super simple and very clear. thanks

  • Fred Tittaferrante
    6 years ago

    “Gluconeogenesis – The worst name for a rock band ever”. “Gluconeogenesis” could translate, loosely, to “Sweet New Genesis”. Sounds like a pop band to me!

    Love the show. I know it is a late comment but I am starting at the beginning and just got to this episode.

Leave a Reply

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

This site uses Akismet to reduce spam. Learn how your comment data is processed.