Cholesterol or ketones, can we have both?

I wanted to know the relationship between and what drives production of cholesterol versus ketones. The simple reason being that 3-hydroxy-3-methylglutaryl Coa or HMG-Coa (HMG) is either reduced by HMG-Coa reductase (HMGr) into the pathway towards cholesterol or HMG-Coa lyase (HMGl) sets it on the pathway towards Acetoacetate. So 2 enzymes work on the same substance.

How are these enzymes regulated? Does one go up when the other goes down? Are both regulated linearly by the same or different triggers?

Ketones are crucial for longer term survival, for example under starvation to avoid muscle breakdown by serving as an alternative to glucose. Without ketones, your skeletal muscle would waste away much quicker being a substrate for gluconeogenesis.

So shouldn’t HMG be diverted towards ketones much more to support ketone production, at the expense of cholesterol production? And if this goes at the expense of cholesterol production, what would be the effect on our sex hormones for which cholesterol is a precursor? It seems that the HMGl must dominate. But is this domination at the expense of reductase, and does it therefor lower cholesterol production or is there simply ample HMG availability to accommodate both?

The feed into HMG can come from Acetyl-Coa which is build up in the mitochondria due to a slower running citric acid cycle thanks to low oxaloacetate which is due to the lower glucose availability due to lower glycogen levels in the liver as we understand it. The Acetyl-Coa comes from the fatty acids that reach the liver and enter the mitochondria where the acyls are cut up to form Acetyl-Coa.

Then there is the LMHR phenotype. Low carb, low body fat and high LDL cholesterol. Likely too low in body fat to have a sufficient continuous supply of fat leading to low ketone levels. It is only n=1 but I noticed for myself, also being LMHR, I had to drastically up my fat intake in order to generate sufficient ketones and see a suppression of glucose caused by the ketones.

Maybe there is a threshold of HMG availability required for ketone production. Perhaps until a certain level, most of HMG goes towards cholesterol and when this level is reached, there is a spill-over towards ketones? But it seems to conflict with my original idea of ketones being so important. They are, but thinking further, if ketones are not high enough the body will try and release more fat and also starts to break down skeletal muscle (low ketone -> low protection against catabolism) through catecholamines. So there must be a priority mechanism towards ketones. We need to hang on to our skeletal muscle as long as possible to increase our chances for survival.

But this is all guess work. What can I find in the studies?

HMG reductase (cholesterol production)

We do find that reductase is inhibited by the re-absorption of LDL. It seems to say stop producing cholesterol when more is returning from the circulation.

According to this study, insulin has an increasing effect on both mRNA and protein production. Glucagon and fasting causes a drop in HMGr activity affecting both protein levels and mRNA. This effectively means that going low carb slows down cholesterol production. This reduction in HMGr is new to me but a confirmation of the reduced ApoB production I wrote about:

“Insulin and glucagon modulate hepatic 3-hydroxy-3-methylglutaryl-coenzyme A reductase activity by affecting immunoreactive protein levels.” –

A further support for why cholesterol could be lower under high carb is that the LDL receptor in the liver seems to correlate with HMGr. Thus, the higher the cholesterol production (faster output), the higher the level of LDL receptors to (faster) reabsorb cholesterol. And keep in mind, cholesterol absorption via the LDL receptors will lower reductase activity thus creating an overall trend to lower circulating LDL-c levels. This mechanism could contribute to the reason why dietary cholesterol does not affect plasma cholesterol. There simply is a negative feedback loop when insulin is up.

“Low density lipoprotein receptor and 3-hydroxy-3-methylglutaryl coenzyme A reductase gene expression in human mononuclear leukocytes is regulated coordinately and parallels gene expression in human liver.”

Insulin is not the only stimulator of HMGr. Thyroid hormone and estrogen both have a similar stimulating effect.

“Feedback and hormonal regulation of hepatic 3-hydroxy-3-methylglutaryl coenzyme A reductase: the concept of cholesterol buffering capacity.”

Update 2021-03-02: Added the section below about AMPK

Further studies show us that when AMPK is activated, it causes a signaling cascade whereby HMGr is inhibited by ACC deactivation thus effectively blocking cholesterol synthesis.

“AMPK and cell proliferation – AMPK as a therapeutic target for atherosclerosis and cancer – Regulation of cholesterol synthesis pathway by AMPK”

“Liver AMP-activated protein kinase and acetyl-CoA carboxylase during and after exercise”

source: Activation of AMP-activated protein kinase in the liver: a new strategy for the management of metabolic hepatic disorders

AMPK is an important signal as it is also responsible for stimulating ketogenesis in the liver. And this article is about individuals on a ketogenic diet so we see here that AMPK itself stimulates ketogenesis and inhibits cholesterol production.

An external file that holds a picture, illustration, etc.
Object name is ijms-17-02093-g002.jpg
source: Regulation of Ketone Body Metabolism and the Role of PPARα

HMG lyase (ketone production)

There is surprisingly little to find on lyase except for a deficiency leading to hypoketotic issues. I could not find info on increase or decrease in expression of protein or mRNA. At best guess it is the reverse of HMGr, meaning HMGl would be down regulated under insulin and/or upregulated under glucagon. One thing I did find is that glucagon also increases HMG synthase thus increasing HMG supply and, as you could read above, in our low carb state this means that cholesterol production is down so we increase HMG availability for ketones not only by lowering reductase activity but also by increasing HMG availability itself through inhibition of the succinylation.

“Glucagon activates mitochondrial 3-hydroxy-3-methylglutaryl-CoA synthase in vivo by decreasing the extent of succinylation of the enzyme.”

“Treatment of rats with glucagon or mannoheptulose increases mitochondrial 3-hydroxy-3-methylglutaryl-CoA synthase activity and decreases succinyl-CoA content in liver”

What does this tell us about LMHR profiles?

Putting all the sourced material together, including from the wiki, in a situation where insulin is reduced to its minimum due to a low carb diet, it seems that the way to get very high LDL-cholesterol exists out of the following components. The more you tick, the higher you go and the increase is not due to an increase in production, rather due to a reduction in clearance. See the section on carnivore further down to understand the increase plasma levels:

  • Very low insulin
  • High to very high glucagon
  • Low free T3 (usually associated with high reverse T3)
  • Low fat mass
  • Very low estrogen

Insulin you can measure via blood test. Typical very low values would be 1 or 2 mIU/L.

Glucagon is not standard but you can ask for it. It is not straightforward to suggest a proxy. I’d go by being lean and have high ketone production but this can be confounded by having a high fat intake. Yet being very lean we usually end up having low ketones despite a possible naturally high glucagon level. Consider a blood test to be sure.

Thyroid hormone, free T3, can also be measured but doctors don’t always want to do this. Alternatively you can go by symptoms such as cold hand and feed, dry skin, irregular or missing menstruation. T3 is partially responsible for lipolysis. A reduction will lead to lower basal lipid availability, hence the lower ketones.

“Action of Thyroid Hormones, T3 and T2, on Hepatic Fatty Acids: Differences in Metabolic Effects and Molecular Mechanisms.”

Everybody has a fat mass, even lean people so when are you up for elevated cholesterol? There is no clear cut-off point but my guess is that for males it starts to be noticeable as of 15% body fat and for females probably around 20%.

Estrogen can also be measured in a blood test. I have no experience with this but if you are on the low side or lower than the reference values it will certainly be an additional checkbox.


If you fit the LMHR profile and don’t get high ketone levels then consider the following.

You have low body fat to begin with. Reaching high ketones requires sufficient fat availability for the liver. If you don’t get ketones in the range of >1, preferably >1.5mmol then you may not have sufficient protective effect against skeletal muscle breakdown. I’m basing this range on the glucose suppressive effect. Only when ketones are high enough to suppress glucose output from the liver, will it provide protection against muscle atrophy.

You should not be in a situation where you loose muscle mass or you are literally in starvation mode in my honest opinion.

If you decide to change something about this, consider a higher protein intake or higher fat intake. Higher fat will lead to higher ketones, helping to replace the need for glucose. Higher protein will lead to more glucose availability, also protecting against atrophy but potentially at the cost of ketones. You can also increase both.

There is a possible combination with temporarily giving up exercise but I don’t think that is sensible and desired. I prefer energy intake to match with what the body needs.

Why the increase in cholesterol when moving on to carnivore?

When leaving out plant food by replacing it with meat, more protein is absorbed. There are specific amino acids which stimulate glucagon production. HMGr, thus the pathway to cholesterol, is reduced by glucagon or at least associated with it. There are also insulin-stimulating amino acids.

I have to guess here because we need to look at things on a time scale. The slight increase in insulin, post-prandial, will increase cholesterol production a little bit more than a similar meal where part of the protein is replaced with plant foods.

At the same time, glucagon is stimulated more strongly due to the higher content of glucagon-stimulating amino acids in the meal. Insulin seems to dominate at all other levels so I suspect that it is also the case here so cholesterol production goes up.

There is the problem of the LDL receptor which is upregulated by insulin which could counteract the increased production. I think producing and translocating it to the cell membrane takes more time. It therefor may be insufficient to prevent a rise in cholesterol.


This mechanism of temporarily increasing production is what leads to higher cholesterol levels in general. So it is the combined effect of getting a very low clearance with temporarily increased production.

The more stronger the cholesterol clearance via the LDL receptor is reduced, the stronger the increasing effect of plasma LDL-c will be when you take a meal which temporarily increases production.

And all this is to prioritize our ketone production !

How can you experiment with your LDL-c levels? Reduce glucagon and increase insulin (without insulin resistance). I haven’t tried this out myself but I would go for high starch, fiber and no meat. Starch to increase insulin, no meat to reduce its glucagon stimulating effect, fiber to help capture and clear the cholesterol and avoid re-absorption. Do this for 3 to 4 days and you should have the lowest LDL. Do the opposite and LDL goes up again.


Rodents on KD

There are quite a number of papers studying ketosis in rodents to measure outcomes. I just thought it would be a neat exercise to make an inventory of these studies and look at their properties such as feeding ad libitum, type of fat in the diet, level of ketosis reached etc.. to see if we can find a pattern in the outcomes and understand what it takes to get them into ketosis.

Study%protein%carb%fatfree/restrictedStarting age of dietBHB mmolOutcometypefat type
11018911.2kcal per day after weight gain12m0.7increased median lifespanC57BL/6 mice7% soybean oil, 56% lard
210090ad lib12m1.5reduced midlife mortality, preservation of memoryC57BL/6 mice31% crisco, 10.7% cocoa butter, 6% corn oil
35095ad lib8w0.8reduced amino acid catabolism, no adverse survivalC57BL/6J micelard, butter, corn oil
44.71.893.5ad libp21?1.77mTOR inhibitionSprague Dawley ratslard, butter, corn oil
54.71.893.5ad lib13w1.5enhanced neurovascular functionC57BL/6 micelard, butter, corn oil
6.a5.51.792.8isocal12w2.58measure ketosisWistar ratsbeef tallow
6.b11.81.986.3isocal12w1.18measure ketosisWistar ratsbeef tallow
6.c19.12.278.7isocal12w0.62measure ketosisWistar ratsbeef tallow

One study(7) I wanted to include but couldn’t fit it in the table above due to not being a longer term test, was the only one I could find that used a very different fat, hydrogenated coconut oil. Because of the oil used, we see ketone levels that are not observed in any of the above publications. Both the wild type and knock-out mice had levels above 5mmol on the high fat diet.

I have expressed my concern elsewhere for mouse models on a ketogenic diet not being representative for humans but study 7 shows the best approaching results that we see in humans. Higher levels of ketones are reached with sufficient protein (16% carbohydrates, 19% protein, and 65% fat) and normal similar weight gain for the wild type on the high fat diet as to the wild type on the control diet.

All other approaches are invalid as they feed too little protein, driving overfeeding and weight gain. Especially when different compositions are compared(6) we see that murines, and probably humans too, require sufficient protein. They’ll eat until they have obtained their required volume of protein or, when under restricted feeding, they’ll end up loosing lean mass(6) which is a sign of protein being broken down and converted to glucose.

Apart from the macro composition, the restriction or free feeding makes a difference. Interestingly, comparing 1 and 2 with virtually identical composition, we see that ad lib feeding on fat gives higher ketones. I have often mentioned this to people, if you measure low ketones, increase your fat intake. Otherwise your body does not have sufficient access to fat to generate ketones.

The type of fat certainly makes a difference as we’ve seen in study 7. For being quite identical in macro composition, 6.a does remarkably better in generating ketones versus 4 but to be honest, although both are rats, the breed is different so there is another influencing factor. It could be that Wistar rats have a higher lipolythic capacity. The research doesn’t cover that but they strains certainly have differences in lipolysis. 6.a was isocaloric thus restricted which we saw as a negative influence for ketones when comparing 1 to 2. We also know that Wistar rats are more susceptible to metabolic syndrome compared to Sprague-Dawley rats. How these factors influences their ability to produce ketones on a ketogenic diet is still a guess to me.

So for anyone looking into the results of a ketogenic diet on murines, do check out the fat they were given, the breed, mice or rats, protein amount, feeding regimen etc.. It influences the outcome and certainly cannot be extrapolated directly to humans.

(1) “A Ketogenic Diet Extends Longevity and Healthspan in Adult Mice”, Megan N. Roberts, Marita A. Wallace, Alexey A. Tomilov, Zeyu Zhou, George R. Marcotte, Dianna Tran, Gabriella Perez, Elena Gutierrez-Casado, Shinichiro Koike, Trina A. Knotts, Denise M. Imai, Stephen M. Griffey, Kyoungmi Kim, Kevork Hagopian, Marissa Z. McMackin, Fawaz G. Haj, Keith Baar, Gino A. Cortopassi, Jon J. Ramsey and Jose Alberto Lopez-Dominguez, 2017,

(2) “Ketogenic Diet Reduces Midlife Mortality and Improves Memory in Aging Mice”, John C. Newman, Anthony J. Covarrubias, Minghao Zhao, Yu Huang, Saptarsi Haldar, Eric Verdin, 2017,

(3) “Adaptive changes in amino acid metabolism permit normal longevity in mice consuming a low-carbohydrate ketogenic diet”, Nicholas Dourisa, Tamar Melman, Jordan M. Pecherer, Pavlos Pissios, Jeffrey S. Flier, Lewis C. Cantley, Jason W. Locasale, Eleftheria Maratos-Flier, 2015,

(4) “The ketogenic diet inhibits the mammalian target of rapamycin (mTOR) pathway.”, McDaniel SS, Rensing NR, Thio LL, Yamada KA, Wong M, 2011,

(5) “Ketogenic diet enhances neurovascular function with altered gut microbiome in young healthy mice”, David Ma, Amy C. Wang, Ishita Parikh, Stefan J. Green, Jared D. Hoffman, George Chlipala, M. Paul Murphy, Brent S. Sokola, Björn Bauer, Anika M. S. Hartz, and Ai-Ling Lin, 2018,

(6) “Induction of ketosis in rats fed low-carbohydrate, high-fat diets depends on the relative abundance of dietary fat and protein.” – Bielohuby M, Menhofer D, Kirchner H, Stoehr BJ, Müller TD, Stock P, Hempel M, Stemmer K, Pfluger PT, Kienzle E, Christ B, Tschöp MH, Bidlingmaier M – 2011 –

(7) “A high-fat diet reverses metabolic disorders and premature aging by modulating insulin and IGF1 signaling in SIRT6 knockout mice.” – Li Z, Xu K, Guo Y, Ping L, Gao Y, Qiu Y, Ni J, Liu Q, Wang Z – 2020 –

Protein and fructose

What could those 2 possibly have in common?  Both could be causing us to eat more.  The mechanism and the reason for each is very different though. Let’s have a look.


In a recent podcast, Peter Attia introduced professor Rick Johnson who is an expert in fructose metabolism.  He revealed that fructose is the only substance known so far that gets metabolized fully.  Meaning that normally a cell will have a feedback mechanism when cellular energy (ATP) goes low.  With fructose however, the cell doesn’t stop and ATP can get really low.  The details were not discussed but this would have the effect to signal to the rest of the body that there is an energy crisis.  As a consequence, hunger is stimulated and overall metabolism is reduced.  A perfect scenario to gain weight.  The effect is largely depending on 2 factors and that is dose and speed.  If either one increases then it triggers this effect.  Did you enjoy a tasty orange juice this morning? 

That is not all what was revealed in the podcast but I suggest you head on over and listen in as it is bursting with interesting info.  Kudos to Peter Attia for bringing this guy to the front.

To reveal a bit of the detail behind this, it is the fructose phosphorylation by fructose kinase in the liver for which there is no feedback (1). The phosphorylation happens through ATP donating one phosphor atom so you get fructose enriched with one phosphor.

With no feedback, you can see how taking in a bigger bolus of fructose can decrease ATP sufficiently to get the cell stressed out through too low ATP levels.

Continuous high fructose consumption looks detrimental and a key driver to insulin resistance, diabetes etc.

If you are developing cancer at the early stages or beyond, it will happily live on fructose (2) as much as it can on glucose. No wonder we find it associated with cancer (3). It is even less surprising if we know that fructose can cause vasoconstriction (4).

Vasoconstriction goes in hand with hypoperfusion which can lead to tissue hypoxia (5), a central regulator of the hallmarks of cancer. Of course I’m not talking about hypoperfusion at the levels you would find it in the emergency room. Just that lower than optimal level, a suppressing effect that has a very gradual effect over a long term. Speculation I know but given the physiological effects it is very plausible.


I’ve written about oxygen and CO2 before (6)(7) and how glucose may lead to a slightly lower oxygen saturation in the tissue. With fructose on top, I’m further convinced this lack of sufficient oxygen is taking place in our bodies with our standard diet today. Be it high fructose corn syrup or just plain sugar, they all have roughly an equal amount of glucose and fructose.


(1) “The negative and detrimental effects of high fructose on the liver, with special reference to metabolic disorders” – Brandon H Mai and Liang-Jun Yan – 2019 –

(2) “Increased utilization of fructose has a positive effect on the development of breast cancer” – Xiajing Fan, Hongru Liu, Miao Liu, Yuanyuan Wang, Li Qiu and Yanfen Cui – 2017 –

(3″) “The role of fructose in metabolism and cancer” – Charrez B, Qiao L, Hebbard L – 2015 –

(4) “The mechanisms underlying fructose-induced hypertension: a review.” – Klein AV, Kiat H – 2015 –

(5) “Choice of Pharmacological Agents in Hypoperfusion Syndrome” – J.-L. VincentE. Silva –

(6) “Ketones and oxygen” –

(7) “CO2 and nutrition” –


The effect is very different for protein.  It is hard to proof so consider this simply for what it is, a hypothesis. But I suspect that there is a protection mechanism to the amount of protein that is needed by the body.  Your body may sense how much is available for protein synthesis and will trigger hunger to satisfy that need if it is not met. I have a number of indications and some research pointing out the mechanism. 

First lets have a look at a recent mouse study where they wanted to investigate the carb-insulin hypothesis.  I’m not interested in the hypothesis itself but they tested a large variety of macro compositions (fat, protein, carbs).  Either fat or protein was fixed and when it was fixed they tested both a high amount and a low amount of the fixed component.

Before getting into the study, you have to know that mice on a ketogenic diet, when they want to keep the mice at equal weight, have to be fed about 12% extra.  They burn through that extra energy by generating more heat.  A pure waist of energy, why? 

Protein within the mice diet has to be kept very low because they would otherwise convert too much protein into glucose and that would hinder ketogenesis.  Already here I would ask the question: could it be that the mice try to burn so much energy so that they could eat more, thereby getting their needed daily amount of protein? 

Now, let’s look at the result of the study (1). There are a number of noticeable effects. 

Figure 4

Energy intake under fixed low protein is higher than under fixed high protein.  Is this to reach sufficient protein?

Under fixed high fat, shifting from low carb to high carb reduces the protein intake.  This leads to a higher energy intake.  Again, are we eating more to get to those protein?

Fixed high fat versus fixed low fat, there is a greater energy intake under high fat.  Fat does not produce as much glucose.  Could more glucose be sparing protein conversion to glucose?  And high fat intake automatically means lower protein intake versus the low fat.  Could it be a combination of higher glucose intake sparing protein conversion to glucose and the low fat allows more protein intake?

What carb intake does is stimulate insulin.  Insulin will oppose glucagon, preventing its production so that less amino acids are converted into glucose.

But insulin is not a fix-all.  Under fixed high fat we see a drop in lean mass as we get to the higher end of carb intake.  The protective effect of insulin is not sufficient to support a very low intake of protein, still leading to muscle catabolism.

We see the same is true for the fixed low fat, as carbs increase, protein reduces in the food.

What this all looks like is that there is a certain level of amino acid availability needed. There are several ways to obtain this level.

  1. Through dietary intake but if your food is low in protein then you need to eat more. In order to be able to eat more, you have to be able to burn off the excess energy.
  2. Through preventing conversion of amino acids to glucose. This can be achieved by taking in more glucose due to the insulin effect.

When we think of humans and look at the ketogenic diet, we do not have to be in the problematic zone of insufficient protein. If you produce high enough ketones, your protein demand may even go down. Ketones are both protecting the skeletal muscle from catabolism and also lower glucose output from the liver. Likely, the more full your liver is with glycogen, the more it can release through the action of glucagon. In such case insulin has to rise a little bit to counter that effect to keep plasma glucose nicely under control.

What applies to the mice, could that be appliceable to humans? Statistics don’t prove everything but let’s look at one anyway to get a hint of it.


Across time we see that the quantity of protein consumed remains more or less the same yet carbohydrate and certainly fat content increases. Purely speculating on the slight increase in protein… this may be because we get more obese which results in larger skin mass and cells which all require protein to maintain them.

On a standard diet, all these macros are combined in our meals. What this shows is that our meals are containing less protein in % of total caloric intake. So could it be that, because our meals are containing less protein, we naturally start to eat more in order to meet the protein need? Just like the mice seem to do?

If this is true then you could argue for prioritizing protein to meet your needs. This is essentially what the Protein Leverage hypothesis is about. Don’t think that this is just a vague idea though. There is extensive research that looked into this using animal studies. A review paper (a summary of research on a specific topic) looking at the effect in the brain concluded that feeding behavior is influenced by its amino acid sensing (2).

It doesn’t answer the question how much protein you should but you don’t have to know. Your body tell you!

I’m still curious though, how are the mice on a ketogenic diet able to upregulate their heat production, assuming this is in response to the low protein intake? Or is that a coincidence and purely ascribed to ketones which do cause the uncoupled metabolism in the mitochondria (through UCP1) (3).


(1) “The carbohydrate-insulin model does not explain the impact of varying dietary macronutrients on the body weight and adiposity of mice” – Sumei Hu, Lu Wang, Jacques Togo, Dengbao Yang, Yanchao Xu, Yingga Wu, Alex Douglas, John R.Speakman – 2020 –

(2) “Central Amino Acid Sensing in the Control of Feeding Behavior” – Nicholas Heeley and Clemence Blouet – 2016 –

(3) “Mitochondrial biogenesis and increased uncoupling protein 1 in brown adipose tissue of mice fed a ketone ester diet.” – Srivastava S, Kashiwaya Y, King MT, Baxa U, Tam J, Niu G, Chen X, Clarke K, Veech RL – 2012 –

Longevity (3)

Continuing on the information that David Sinclair provides in his book, we’ll have a look here at the 3 pathways he mentioned. Sirtuins, AMPK and mTOR and how these are influenced with a ketogenic diet.


According to his book, mTOR should be kept at low activity. Whenever mTOR is stimulated then the repair mechanisms can’t take be active. There are 2 main causes for activated mTOR and those are insulin and protein. So taking out carbs and protein from our food sets us up for longevity! Not so fast. While you can take out carbs completely, you do need protein. Don’t just think of protein as muscle meat. Every single cell in your body is made up and contains if not thousands of molecules made up from the individual amino acids that make up protein. These proteins have functions such as enzymes to make things happen in your body. There is a minimum that you need just to live.

A ketogenic diet is ketogenic simply because it restricts carbohydrates. It is a condition to produce ketones in your body. The protein side is a bit more grey zone in the amount that you can eat for ketogenesis but by increasing your fat intake you can limit your protein intake naturally as the fat will satiate you.

Eating zero carbohydrates, high in fat and not-more-than-sufficient protein will minimize mTOR and increase ketone production. Sorry to be so vague on the protein but in the science field numbers range wildely and there are no decent conclusions long term.

Personally I choose to aim for 1gr of protein per lean kg of body mass (total weight – fat weight).

Anyway, a reduced mTORC1 activity is needed in the liver to allow for ketone production so ketones can be taken as evidence of reduced mTOR.

The mechanism for the effect on cardiac hypertrophy appears to be inhibition of HDACs that suppress the activity of a mechanistic mTOR complex (88). This is one of several examples of intersections between BHB, its signaling effects, and mTOR/rapamycin, a canonical longevity-regulating pathway (55). As described above, mTOR is also a checkpoint in the activation of ketogenesis; inhibition of mTORC1 is required to activate the transcription factors and hormones that control ketogenesis (4117).

“β-Hydroxybutyrate – A Signaling Metabolite”, John C. Newman and Eric Verdin,

Wait a minute! I’ve read that ketones increase mTOR! That is correct. The ketone body beta-hydroxybutyrate (BHB) has different effects. Keep in mind that the study referenced below is about a ketone ester supplement (so not an endogenous increment) but nonetheless, and that it shows here a synergistic effect during protein intake AND the effect is noted in skeletal muscle where you want mTOR to be up temporarily in order to build muscle. It could be that this is happening in other organs as well.

Under normal circumstances insulin gets activated when eating. Only a very modest amount on a ketogenic diet but potentially enough to bring down blood BHB levels to lower than 1 or even 0.5. It would be a great experiment to see how fast BHB levels go up after feeding, what the plasma amino acids are across this time and how it affects muscle protein synthesis.

“Intake of a Ketone Ester Drink during Recovery from Exercise Promotes mTORC1 Signaling but Not Glycogen Resynthesis in Human Muscle”, Tijs Vandoorne, Stefan De Smet, Monique Ramaekers, Ruud Van Thienen, Katrien De Bock, Kieran Clarke, and Peter Hespel, 2017,

“Effect of beta-hydroxybutyrate on whole-body leucine kinetics and fractional mixed skeletal muscle protein synthesis in humans.”, K S Nair, S L Welle, D Halliday, and R G Campbell, 1988,

I’m of the opinion that longevity by restricting protein works but only if this is done while having elevated ketone levels in the blood. You need to protect those muscles from breaking down and preferably even build them up.

“Role of Dietary Protein and Muscular Fitness on Longevity and Aging”, Barbara Strasser, Konstantinos Volaklis, Dietmar Fuchs, and Martin Burtscher, 2018,

This makes the ketogenic diet an ideal fit. Low mTOR while fasted, extra driver of muscle protein synthesis when eating protein. Ideally after exercise when those circulating fatty acids still generate increased ketones.

Since one of the studies was about exercise… shortly after exercise your endogenous ketone production goes up and then gradually settles down again to baseline levels. So dietary protein intake is likely optimal shortly after exercise when your ketones are up highest. Do check out different types of exercise to see what it does to ketone levels though.

“Post-exercise ketosis”, J H Koeslag, T D Noakes, and A W Sloan, 1980,

“POST-EXERCISE KETOSIS”, R.H. Johnson, J.L. Walton, H.A. Krebs, D.H. Williamson, 1969,

Not only do we require low mTOR to produce ketones, ketones themselves lower mTOR. This has been found in the brain in animal experiments so caution to extrapolate this to human results. It will be hard to prove since we can’t take a sample from a living person 😦 But the pathway is described and for now I don’t see any reason this would not be the case for us as well.

“The ketogenic diet inhibits the mammalian target of rapamycin (mTOR) pathway”, McDaniel SS, Rensing NR, Thio LL, Yamada KA, Wong M, 2011,


We see that AMPK is elevated on a ketogenic diet. These studies are done on rodents which kind of makes it hard to accept as evidence. The reason for this is that these buggers have a high metabolism, around 7-fold higher than humans. For this reason their protein intake has to be severely restricted or they can’t get into ketosis. They’ll turn much of the protein into glucose otherwise.

“A high-fat, ketogenic diet induces a unique metabolic state in mice.”, Kennedy AR, Pissios P, Otu H, Roberson R, Xue B, Asakura K, Furukawa N, Marino FE, Liu FF, Kahn BB, Libermann TA, Maratos-Flier E, 2007,

However, the above publication also refers to other research showing that AMPK is upregulated when glucose metabolism goes down. That makes it more likely to happen in humans as well.

AMPK responds to ATP depletion and is activated by low glucose (27); AMPK may also be inhibited by insulin in some cases (50). Activation of AMPK leads to decreased fatty acid synthesis and increased fatty acid oxidation (20, 27)

Despite being another animal study, the researchers also decided to test via cell culture if ketones directly affect AMPK and found it to be the case.

“β-Hydroxybutyrate suppresses inflammasome formation by ameliorating endoplasmic reticulum stress via AMPK activation”, Ha Ram Bae, Dae Hyun Kim, Min Hi Park, Bonggi Lee, Min Jo Kim, Eun Kyeong Lee, Ki Wung Chung, Seong Min Kim, Dong Soon Im, and Hae Young Chung, 2016,

So it seems the we can safely assume in humans that AMPK goes up on a ketogenic diet. The diet keeps insulin low, there’s the ketones that induce AMPK and BHB also suppresses hepatic glucose output so there is a lower glucose level. BHB suppressing hepatic glucose output is not referenced here but this is visible in the human studies where they administer exogenous ketones. It is also a suppressive effect that must take place or you end up in ketoacidosis thanks to the CO2 contribution of metabolizing glucose.

Effects of ketone bodies on AMP-activated protein kinase (AMPK) actions in different tissues.

“Ketosis, ketogenic diet and food intake control: a complex relationship.”, Paoli A, Bosco G, Camporesi EM, Mangar D, 2015,


There are different sirtuins, 1 to 7, likely involved in longevity but SIRT1 is known to be directly involved in DNA repair so we’ll focus on that one. As explained in the book, these sirtuins require NAD+ to function. This gives ketones 2 possible ways to activate sirtuins. Either somewhat directly or via NAD+ production or saving.

Cockayne syndrome is a disease with accelerated aging caused through increased DNA repair. Similar to how the book describes they were able to accelerate aging in mice by deliberately breaking DNA. The next study also comments on how NAD+-dependent SIRT1 is required for this DNA repair. The disease causes higher activation of PARP1 which is also dependent on NAD+ resulting in a lower availability for SIRT1.

BHB caused an increase in SIRT1 activity. That in itself is encouraging but is it because it reduced PARP1 activity? Did it cause more NAD+ production? Did it directly stimulate SIRT1, bypassing the need for NAD+? Lots of questions we’d like to know an answer to. We want to make sure it is not an effect that is specific to Cockayne syndrome. They did experiment with PARP inhibition, leading to higher SIRT1 activity as it preserved NAD+ so this is another angle we can look at for BHB.

They did a lot of investigation and found BHB in cell culture would elevate acetyl-coa availability which regulate histone acetylase activity. So SIRT1 activity was increased indirectly by BHB but SIRT1 is a histone deacetylase.

“A high-fat diet and NAD(+) activate Sirt1 to rescue premature aging in cockayne syndrome.”, Scheibye-Knudsen M, Mitchell SJ, Fang EF, Iyama T, Ward T, Wang J, Dunn CA, Singh N, Veith S, Hasan-Olive MM, Mangerich A, Wilson MA, Mattson MP, Bergersen LH, Cogger VC, Warren A, Le Couteur DG, Moaddel R, Wilson DM 3rd, Croteau DL, de Cabo R, Bohr VA, 2014,

This needs to be explained a bit further. BHB is a histone deacetylase (HDAC) inhibitor. Sirtuins (1-7) belong to the class III HDAC’s while BHB inhibits class I & II HDAC’s. I was first thinking that, by inhibiting activity of other HDAC’s that BHB saves NAD+ consumption, leaving more available for the sirtuins. But those class I & II HDAC’s are zinc dependent rather than NAD+. So they do not oppose each other, neither does it help save NAD+.

“Suppression of Oxidative Stress by β-Hydroxybutyrate, an Endogenous Histone Deacetylase Inhibitor”, Shimazu T, Hirschey MD, Newman J, He W, Shirakawa K, Le Moan N, Grueter CA, Lim H, Saunders LR, Stevens RD, Newgard CB, Farese RV Jr, de Cabo R, Ulrich S, Akassoglou K, Verdin E, 2013,

So far it seems BHB increases histone acetylation through increased availability of acetyl-coa and causes increased expression of FOXO3A and MT2 (explained in the above paper) through the specific HDAC inhibition. Somehow this leads to increased SIRT1 activity.

As it turns out, it is FOXO3A that is required for SIRT1 expression (in combination with p53).

“Nutrient availability regulates SIRT1 through a forkhead-dependent pathway.”, Nemoto S, Fergusson MM, Finkel T, 2004,

This is really great because it shows the indirect effect BHB has on SIRT1 expression via FOXO3A thus a mechanism that is independent of our Cockayne syndrome.


As mentioned at the beginning, sirtuins are dependent on NAD+. As it turns out, Acetoacetate (AcAc) is the first metabolite in the pathway to BHB. In order to produce BHB, AcAc has to combine with NADH and H+. The result is BHB and NAD+. So at least in the liver there is extra support for sirtuin activity.

We also have evidence of ketone body production in astrocytes. Particularly in the ventromedial hypothalamus so at least that part of the brain could benefit from the addition in NAD+.

“Fatty acid-induced astrocyte ketone production and the control of food intake”, Christelle Le Foll and Barry E. Levin, 2016,

We could set our hope on a surplus in NAD+ availability, hopefully exported into the bloodstream and distributed across the body. There is a potential for it via the Connexin 43 transporter.

“A Pilot Study Investigating Changes in the Human Plasma and Urine NAD+ Metabolome During a 6 Hour Intravenous Infusion of NAD+”, Ross Grant, Jade Berg, Richard Mestayer, Nady Braidy, James Bennett, Susan Broom and James Watson, 2019,

“Exogenous nicotinamide adenine dinucleotide regulates energy metabolism via hypothalamic connexin 43”, Eun Roh, Jae Woo Park, Gil Myoung Kang, Chan Hee Lee, Hong Dugu, So Young Gil, Do Kyeong Song, Hyo Jin Kim, Gi Hoon Son, Rina Yue, Min-Seon Kim, 2018,

Update: 2020-01-01

As mentioned above, BHB production leads to the generation of NAD+ but BHB conversion to acetyl-Coa consumes NAD+ because the conversion back to AcAc requires this before it can be converted to acetyl-Coa.

But this is still far better than glucose metabolism which utilizes 2 NAD+ molecules per acetyl-Coa generated while dietary glucose doesn’t create any NAD+.

In addition AcAc also circulates in the body. If it gets used directly for energy so skipping the conversion to BHB then it doesn’t consume NAD+ when converting to acetyl-coa. This means ketone metabolism is a neutral effect on NAD+. Although, by replacing glucose metabolism, it preserves NAD+ resulting in a higher NAD+/NADH ratio.

An external file that holds a picture, illustration, etc.
Object name is fnmol-10-00377-g0001.jpg

“Ketone-Based Metabolic Therapy: Is Increased NAD+ a Primary Mechanism?” – “Mechanisms of Ketogenic Therapy: Evidence for Increased NAD+”, Marwa Elamin, David N. Ruskin, Susan A. Masino, and Paola, Sacchetti, 2017,


We found evidence that the ketogenic diet works on all 3 levels expressed in the book of David Sinclair. This is very encouraging.

What still remains open is how well we can maintain ketosis. And if we can generate sufficient ketones, do they, and the NAD+, end up in every cell of our body or are there only specific organs affected? It is easier for rats in a lab to be in lifelong ketosis but humans don’t live there and are not rats.

At least the available data makes me hopeful and so do all animal experiments where they show the life extending or at least health-span extending abilities of the ketogenic diet.

— The End —

Breathing Exercise

Improve your oxygen uptake capacity

I have written about tissue oxygen throughout various articles (1, 2) so be sure to check them out for a background of why this type of exercise would be beneficial. In short, apart from a ketogenic diet, regular acute exposure to oxygen deprivation will make your body adapt to be able to take in more oxygen. It is one of those hormesis things such as exercise, cold exposure and heat exposure.

If you have any trouble with breathing or are somehow at increased danger of abnormalities, always first discuss doing this with your doctor. As mentioned, this is a hormesis thing, a stressor so you need to be sure you are fit enough to handle it.

So what do we have to do as an exercise and why is it composed that way?


The beauty of it is that it takes less than 10 minutes and it creates the same adaptation as staying at high altitude.

  1. Lay down stretched and fully relaxed
  2. Breath in and out normally for a few times just to get you relaxed
  3. Exhale and don’t breath in anymore and hold for as long as possible.
  4. Breath in deeply in and out and repeat this 3 times
  5. Repeat step 3 and 4 about 5 times
  6. After the 5th time, breath in and out deeply for 1 minute

For point 3, breath out to the point where there is no more air naturally being pushed out. There’s no need to force the last bit out.

For point 3, you’ll reach a point where you start to get the reflex to breath in again. Get used to this feeling and try to overcome it. The feeling that we are looking for before we start breathing again is a tingling sensation in your lungs, arms and possibly your legs. This is the dropping of the pH. People who engage in intensive exercises can recognize this as when they go for a maximum effort feeling their legs burn. This is the point where you breath in again.


I’ll address these by point:

  1. We want an evenly distributed blood flow so that the increase in pH drop is distributed throughout the whole body. Every part has to experience the oxygen shortage so that adaptation takes place everywhere.
  2. This is just part of step 1, making sure there is no tension in your body that may create a less than optimal blood flow.
  3. This is the key point. Exhaling and not breathing anymore will create oxygen shortage and trigger both an immediate and longer term adaptation to oxygen deprivation.
  4. We can’t continue like this forever so 3 deep breaths are sufficient to restore oxygen levels
  5. Repeating it 5 times is enough to trigger a response. You’ll also notice that if you time the exhaled holding, you’ll increase the time towards the end.
  6. Doing those 5 repeats your body has already adapted to enhance oxygen uptake! The 1 minute kind of hyperventilation will saturate your body with the increased oxygen.

Why exhale and not inhale and hold your breath? The research I’ve come across showed that for some reason it doesn’t work as well.

What are the adaptations to look for as a sign that you are doing it right?

  • You could experience a reduction in heart rate
  • You will get an increase in red blood cells
  • Your hemoglobin will increase
  • Your hematocrit will increase

These changes will already take place after about 1 week and can be measured through a blood panel. Your body will adapt to carry more oxygen. At the same time, the heart rate will go down because your blood pH will not be pushed down so much. The blood flow now has a bigger capacity to buffer against a low pH.

There are likely other adaptations or things that you could notice but they are too much under influence of other lifestyle factors that I leave them up to you to discover and comment about at the bottom of this page 😉

Personally, I’ve seen my hematocrit go from +/-42% to >47% (a 12% increase) and hemoglobin from around 14 to 15.5 (a 10.7% increase) in about a week time. Anyone in aerobic fitness wouldn’t mind seeing this happening!


I’ve been able to compose this type of exercise based on the inspiration from Wim Hof (the iceman) and the published scientific literature on breathing and exercise.

The reason I like it is because the essence seems to be in the exhaled breath holding. This way I have a routine that takes a minimal amount of time with clear measurable improvements.

— The End —

Demand or supply

Conversion of protein to glucose

Among the keto crowd we have people who go full carnivore.  Especially in the carnivore community they have posed that the conversion of protein to glucose (gluconeogenesis or GNG) is demand driven so excess amino acids (from the protein eaten) do not convert to glucose if there is no demand for it.

For example Amber O’Hearn has written about this in 2012:

Studies such as the following seem to further support the idea.

“Dietary Proteins Contribute Little to Glucose Production, Even Under Optimal Gluconeogenic Conditions in Healthy Humans”, Claire Fromentin, Daniel Tomé, Françoise Nau, Laurent Flet, Catherine Luengo, Dalila Azzout-Marniche, Pascal Sanders, Gilles Fromentin, and Claire Gaudichon, 2013,

However they are not complete by only looking at blood glucose that appears from the tracer amino acids.  It does not account for any possible GNG in the liver which could be converted to storage as glycogen. And when it comes to the question of supply vs demand, it needs to take into account the hormonal response noted by glucagon and insulin.

Glucagon does cause the breakdown of glycogen (glycogenolysis) from the liver so it is reasonable to think there is no glycogen refill during stimulation of breakdown.

The clarity on this demand/supply question is important for the keto-carnivore and anyone who eats a high protein diet.  By that I’m not saying eating carnivore means eating a high protein diet but there are people who do just that.

The reason it could be of importance, for those who care, is that if there is a supply driven conversion of excess dietary protein to glucose then you risk going out of ketosis.  Whether this is good or bad is not the point of this article. What I want to do here is look at available knowledge and help those who are in this situation of wanting ketone production, to evaluate whether it is demand or supply driven.

You’ll want to adjust your protein intake to avoid stepping too much out of a ketogenic state if that is what you are looking for.  That is, if GNG is supply driven.

The level of glycogen in the liver is a determining factor for getting into and out of ketosis.  So we need to look into this and see if there is any effect on the glycogen storage.

For the majority, the mindset in the community is already determined that it is demand driven so I’ll take the stance of trying to disprove it.  Full disclosure, I have previously written about this, taking that same stance. I have in the meantime found a few more relevant things which I’ll discuss in this article.

Previous post:

I think I’m able to further convince due to a number of points that are missed in the demand driven arguments.  These points below come on top of what I have already written down.

  • intravenous (IV) studies versus actual dietary protein intake & cortisol
  • kidney GNG
  • longevity
  • Type 1 diabetes

Apart from those points I also don’t see any clarity on where the excess protein end up if it is demand driven.  I have not seen any material that shows there is a pool of amino acids somewhere in the body that is refilled with the excess protein intake.  If there is no such pool then what happens to the excess? All used for extra protein synthesis? There is also no data showing that protein synthesis is linear to amino acid availability indefinitely. 

These are important questions that need an answer when GNG is demand driven.

But OK, let’s start with the way tests are done.

Real life studies vs IV

There’s a couple of things wrong with IV studies.  Usually they test the addition of a substrate (for example a single amino acid) directly in the blood, through IV.  This is not how we eat protein. We eat protein as a whole food and thereby trigger a set of hormones that could be of influence which are not triggered through IV.  Testing whole food protein makes it more difficult to isolate protein specific effects but that is less relevant. We want to look at the effects of eating meat or other protein sources in real life.

If you checked my previous writing on the subject then you already know that incretins, which are triggered during meal ingestion, make a difference in the insulin (and maybe also glucagon?) response.

In a study, further down this article, we noted a slight increase in insulin.  This increase in insulin cannot be simulated with just an IV injection of a single amino acid resulting in a failure of IV to be representative for the real world. Likewise it is with the rate of absorption of amino acids and the effect of additional fat in the meal.

IV studies also fail to account for the full labeled amino acids, only noting there is little appearing in the plasma glucose. Plasma glucose is only output from the glucose producing/releasing organs. This will be further clarified when we look at the kidneys.

Update 2020.04.12:

Further studies found clearly demonstrate the effects of incretin and how they are important for the level of secretion of glucagon and insulin. Especially the second link clearly explains the interplay between incretin and the alpha- and beta-cells in the pancreas to regulate the hormonal secretion.

Post image

“Incretin and islet hormonal responses to fat and protein ingestion in healthy men”

“Expanding the definition of the incretin axis”


One other such dietary effect is cortisol.  Cortisol is elevated when eating a meal. This is independent of the amount of protein and is significant.  For women the effect is much less pronounced.

“Influence of Consumption of a High-Protein vs. High-Carbohydrate Meal on the Physiological Cortisol and Psychological Mood Response in Men and Women”, Sofie G. Lemmens, Jurriaan M. Born, Eveline A. Martens, Mieke J. Martens, Margriet S. Westerterp-Plantenga, 2011,

Glucocorticoids to which cortisol belongs appear to have a synergistic effect with glucagon on GNG meaning that together they are able to increase GNG more than each individually can.  This synergy takes place in the liver and in the kidney.  Both glucagon and cortisol result in an increased expression of PEPCK-C which is a rate limiting enzyme for GNG.

“Factors that control the tissue-specific transcription of the gene for phosphoenolpyruvate carboxykinase-C”, Chakravarty K, Cassuto H, Reshef L, Hanson RW, 2005, 

Kidney GNG

The kidneys are heavily involved in the regulation of plasma glucose.  It seems they have a compensating role. I’ll first explain and then we pull up some papers.

Upon feeding, protein stimulating glucagon would normally increase glucose release from the liver, coming from both glycogenolysis and gluconeogenesis.  Yet studies show no such increase in plasma glucose. There is however also an increase in insulin in normal meals. Very mildly when the protein is eaten with fat and more severely when ingested with carbohydrates.  Despite the elevated insulin, we don’t see a drop in glucose but insulin is very potent at blocking glycogenolysis. The reason we don’t see this drop is because the kidneys are also stimulated in gluconeogenesis by glucagon. The kidneys increase output while the liver reduces output and builds up its glucose reserve.  This is an ideal situation when feeding as it gives the liver the chance to increase its reserve!

Let’s first have a look at this interesting fact from 1964.  They found that the kidneys increased their GNG in the presence of ketones and short-chain fatty acids.


That is really interesting as it could again indicate that the kidneys help the liver to maintain/save its glycogen reserve as much as possible by assisting in the maintenance of plasma glucose through GNG.

The following review on kidneys is very extensive and interesting in total but I want to know about the postprandial part. They came to the same conclusions when investigating postprandial GNG from the kidneys.

Seemingly, renal glucose release paradoxically increases postprandially and it accounts for >50% of the endogenous glucose release for several hours. These observations suggest that increased renal glucose release may play a role in facilitating efficient liver glycogen repletion by permitting the substantial suppression of hepatic glucose release.

“Renal Gluconeogenesis”, John E. Gerich, MD, Christian Meyer, MD, Hans J. Woerle, MD and Michael Stumvoll, MD, 2001,

When running the test on rats we can dissect and see what effect it had on the liver.  They increase GNG, contributing for about 15%~20% of the post-absorptive endogenous glucose production (EGP, what is found in circulation).  This raises blood glucose levels and causes a slight rise in insulin. Enough insulin to reduce the output of glucose by the liver so that the liver can fill up its glycogen store.

Protein feeding ameliorated the suppression of EGP by insulin and the sparing of glycogen storage in the liver but had no effect on glucose uptake… Protein feeding increases kidney gluconeogenesis without increasing global endogenous glucose production, and improves insulin suppression of the latter, likely at the hepatic site.

“Protein Feeding Promotes Redistribution of Endogenous Glucose Production to the Kidney and Potentiates Its Suppression by Insulin”, Bruno Pillot, Maud Soty, Amandine Gautier-Stein, Carine Zitoun, Gilles Mithieux, 2009,

They provided a low dose of insulin and found that with the high protein diet, the liver glycogen was significantly higher (almost 4-fold) than the normal diet.

Important here is that we are talking about a rat study.  They have roughly a 7x higher metabolism. In order for them to be ketogenic, protein has to be reduced to about 10%~15% depending on the addition of carbs in the diet or not, so they may be more optimized to GNG and therefore able to fill up their liver faster.  The authors of the study also referred to their study on humans where they found a similar contribution of the kidney’s to glucose production.

In another rat study, we note similar results.  An increase in gluconeogenesis and a reduction in glycolysis when shifting to a zero carbohydrate diet.

“Gluconeogenesis from dihydroxyacetone in rat hepatocytes during the shift from a low protein, high carbohydrate to a high protein, carbohydrate-free diet.”, Azzout B, Chanez M, Bois-Joyeux B, Peret J, 1984,

Taking everything together we see that protein is converted to glucose by both the liver and kidneys.  The liver reduces output to increase its glycogen reserve while the kidney provides a backup to maintain blood glucose.

The supply is used to fill up the liver glycogen reserve as much as possible.  Storing it as glycogen requires first to pass through GNG.


This is not at all backed up with research, just ideation.  After reading the book on longevity from David Sinclair where he explained how cells wait for the ideal circumstances to proliferate, I thought about how that would translate to our bodies and protein ingestion (or actually food in general).  We live our lives fasted most of the time. Eating breaks that situation for a brief moment and signals the body to make the best use of it. We live in plentiful times now but having food was not always obvious in our past. As such, cell proliferation is a key event and we know this by increased mTOR activity for growth, activated by both insulin and certain amino acids.  However, such growth doesn’t come without an energy cost. Part of the food itself must serve as fuel and part of that food must be stored as backup for later energy requirements. In such light it makes sense that any excess is converted and put into our energy reserves which is our adipose and glycogen. The glycogen in our liver is a constantly fluctuating buffer and must be filled at times when supply is plenty.  This does not happen easily under a ketogenic diet when carbohydrate sources are heavily reduced or even eliminated as on a carnivore diet.

The body can’t afford to have completely depleted glycogen in the liver.  It has to keep it at a minimum level to respond to exercise activities and also in response to pathogens to regulate the immune system.  As such, again, it would seem logic it would take any opportunity to refill the glycogen store as much as possible whenever there is excess.

The glycogen buffer in our liver is key for survival.  As with any buffer, you refill whenever supply is available.

Type 1 Diabetes (T1D)

Update 24 April 2020: fully added the T1D section

To demonstrate the higher GNG rate we need to disable insulin because, as you could read above, insulin causes a reduction in output of glucose from the liver to build up the glycogen.

If GNG would be demand driven then insulin would not make a difference. You could say it makes a difference for glycogenolysis so to rule this out we can look at T1D who do not produce (or very little) insulin. Their fasting insulin and post-meal insulin is pretty much the same so there would be no difference in glucose output from the liver when comparing high versus low protein content in the meal.

But it does, the next reference tested exactly this point. When the protein content in a meal is higher, T1D need to adjust for it by increasing their insulin. There is not a single study that shows that protein cause a greater glycogenolysis so where is the extra glucose coming from? Indeed, from the glucogenic amino acids through the GNG process.

The 60 gram of protein versus the 5 gram of protein resulted in a 50% greater need for insulin. Carbs and fat were kept the same.

“Dietary protein affects both the dose and pattern of insulin delivery required to achieve postprandial euglycaemia in Type 1 diabetes: a randomized trial”

Amino Acids directly stimulate glycogen synthesis (added 2021.01.21)

Another study found that amino acid availability induces glycogen synthesis. This happens through inhibition of glycogen synthase kinase 3 (GSK-3). It was shown in muscle cells which does not guarantee it also happens in the liver…

“Regulation of glycogen synthesis by amino acids in cultured human muscle cells”

But it does.. A study was done by inhibiting GSK-3 in diabetic rats. They found that the inhibition caused a 3-fold increase in liver glycogen synthesis and lowered plasma glucose.

“Effects of a novel glycogen synthase kinase-3 inhibitor on insulin-stimulated glucose metabolism in Zucker diabetic fatty (fa/fa) rats”

Not only amino acids inhibit GSK-3, insulin does this as well and we have seen that dietary protein also induce an insulin response.

“Glycogen Synthase Kinase-3 in Insulin and Wnt Signalling : a Double-edged Sword?”


It looks like both glucagon and insulin are active on the liver at the same time.  Glucagon triggers glycogen breakdown and GNG, 2 sources to release glucose into the bloodstream.  Insulin however opposes the release, the glycogenolysis and causes the generated glucose to be stored in the liver.

During this phase, the kidneys cover for some of the effect by increasing its glucose production.

“Plasma Glucagon and Insulin Responses Depend on the Rate of Appearance of Amino Acids after Ingestion of Different Protein Solutions in Humans”, Jose A. L. Calbet, Dave A. MacLean, 2002,

The degree to which insulin opposes the situation will dependent on the level of amino acids that is reached in the plasma but the rise in insulin noted in the different experiments show it is sufficient to increase the liver glycogen levels.

As such, if you are concerned about ketone production you have to properly balance out protein with your activity level.  

Being very active in a way that demands extra glucose will pull glucose from the liver glycogen store during activity and afterwards to replenish muscle glycogen store.

Not being active you need to temper your protein intake.  Your demand would be lower anyway as you don’t have as much muscle repair and growth going on.

How much protein?  As always, measure your ketones to know how much you are producing. It’s individual, nobody can tell for you personally where the tolerance level is and then it will still depend on different situations, varying day by day.

— The End —

Ketogenesis and blood glucose

It is explained everywhere that the creation of ketones (ketogenesis) is done by the liver, converting fat in ketone bodies such as beta-hydroxybutyrate (BHB). Yet when measuring ketones, after being on the ketogenic diet for a longer period, people struggle with this concept when they measure low ketones.

People who have been on a ketogenic diet for a while tend to measure lower ketones. The high values they measured at the beginning ranged somewhere around or above 2 mmol easily and the longer they are on this diet, the measured values hardly reach 1 mmol with only exceptionally a higher value but never as high as before where values above 3 or 4 were less exceptional. There are exceptions to this which I’ll briefly mention at the end.

The lower measurement results are often explained as the body being more adept at utilizing ketones.

I have previously written about this before with a focus on fat mass in conjunction with metabolism and how that relates to ketogenesis but want to explore it a bit further with some more material related to glucose.

Previous post:

Why glucose? Because BHB, the ketone we are measuring in our blood, has a suppressive effect on plasma glucose. So by looking at how much it is able to suppress glucose we can also obtain an idea of how much BHB we are producing and see if that is in line with an effect on glucose. Meaning that if we measure low BHB but have high BHB production, explained by better utilization, then we should still see the glucose lowering effect.

“Suppression of glucose production and stimulation of insulin secretion by physiological concentrations of ketone bodies in man”, Miles JM, Haymond MW, Gerich JE, 1981,

The reason why BHB would be pressing down on glucose can be found in the pH of the blood. The ketone bodies Acetoacetate and BHB are acidic. If we would not press down on glucose, then glucose is utilized as a fuel resulting in increased CO2 production. The addition of CO2 would further lower the pH while, with BHB we are already on the borderline. See my previous posts on the topic of CO2 for more info.


BHB trend over time

First lets review again how BHB averages over time. The Virta health study, to help people controle their Type 2 Diabetes (T2D), included regular measurement of BHB. The result of it are shown in the picture below. One thing to keep in mind here is that people are T2D so they already started with elevated insulin and also gradually went off their insulin medication. Insulin prevents BHB production. This means we won’t see high BHB values at the beginning but towards the end we have less people on insulin meds and having better control over their own insulin production.

Fig. 4

A lot can be said about the reasons why the trend goes down such as non adherence by carb creap etc but my argument is about total fat availability. This is the combination of both your own body fat and dietary fat that is able to reach your liver (while being low on glucose). The study also shows weight trend over time.


Note that the weight change above is expressed in percentage, not in kg.

Another example is of an individual who has been tracking his BHB over a long time. This person is not a diabetic and considered a normal healthy person. He is of course an individual so not a scientific study but I’m just using his data to explain the general concept.

Don’t mind the trend line. As you can see here we have the same high values at the beginning with a downward trend. Let’s have a look at how that lines up with the blood glucose. Both solid lines are trend lines, 4 measure points average centered. For reference, the horizontal line at the level of the 70 (on the left) corresponds to 1.2 mmol BHB.

You can already see it instinctively but I’ve mapped that to a Pearson correlation, for factual data, where zero means there is no correlation and 1 means they are fully correlated. The first, big, circle on the left is what matches below with Part 1. Part 2 is everything else behind it.

Part 1 – Pearson correlation coefficient: -0.72

Part 2 – Pearson correlation coefficient: -0.509

As you can see, when we have higher levels of BHB, the suppressive effect on glucose is much stronger giving us a better correlation. The 2 other circles show this as well. Just by looking at the graph itself you can’t state that correlation but the point of those 2 markers is that we don’t see blood glucose going up. We should find a case where high BHB is matched with increasing blood glucose.

BHB is of course not the only thing that controls blood glucose but a high level does give it more weight in the final result of your blood glucose.


At the beginning I mentioned about exceptions when it comes to the correlation with fat mass. Some people may find that they easily get into the higher numbers despite being lean. As far as I can tell, this has to do with your personal rate at which you can free up fat from your adipose.

For me personally, one of those influencing components is the level of gene expression of IL-6. During exercise and in general during inflammation, this cytokine will signal to the adipose to release fatty acids. My genes will result in less IL-6 production.

IL-6 is not the only thing that frees up fatty acids but that goes beyond the point of this post.


The point I was trying to make, in addition to the data in my previous post, is that by looking at your glucose levels you can understand if you are producing high or low levels of ketones. When you are out of ketosis yet eat a low carb high fat style diet, you’ll notice a fairly stable blood glucose, always hovering around the same value. Whenever you notice a significantly lower blood glucose value you can assume a high enough ketone production in the range of >1 mmol.

You can of course have deviating numbers such as shortly after exercise where you still have an elevated glucose but already produce higher ketones thanks to the circulating fat so keep in mind when you measure and under what conditions.

Make sure it is a situation that resembles what you will experience most of the time. This is your fasted state so for example if you eat breakfast and skip lunch then measure right before dinner. If you skip breakfast then measure right before lunch etc.. Then you can see how your ketones relate to the time since your last meal. A normal meal will always create a drop in ketones.

If you want to be in ketosis or not is up to you but consider measuring <1 mmol (as a general reference!) that you are indeed not producing as much ketones.

The End

Ketones and oxygen

If you have read my previous work regarding oxygen and carbon dioxide (CO2) you probably would expect, as I did, that on keto your oxygen saturation in the blood would be huncky dory. But it was not as expected to be 98% or 99%. Instead it is hovering around 96%. Yes there is a certain margin of error on these devices (pulse oximeter). So what is that all about? My wife measures 98%, not in ketosis so I decided to dig into it to understand what is going on.

Previous writings:

Actually after investigating and writing about all this stuff it made it easier to find out.

First we get back to basics and that is the oxygen dissociation curve.

The graph plots percent oxygen saturation of hemoglobin as a function of oxygen partial pressure. Oxygen saturation increases in an S-shaped curve, from 0 to 100 percent. The curve shifts to the left under conditions of low carbon dioxide, high pH, and low temperature, and to the right in conditions of high carbon dioxide, low pH, or high temperature.

For the same partial pressure of oxygen (x-axis) in the blood, a lower pH will cause a drop in haemoglobin oxygen saturation (y-axis). There are other factors such as 2,3 DPG content and temperature but our focus will be on pH for now.

Why zoom in on pH? Because the ketone bodies Acetoacetate (AcAc) and beta-hydroxybutyrate (BHB) are acidic bodies with AcAc having a 3.59 pKa and BHB with a pKa of 4.41. In comparison, lactic acid falls in between with a pKa of 3.86 and stomach acid (HCl) has a pKa of around – (sources vary) so the lower the pKa the more acidic. The ketone bodies are considered weak acids. We’re also interested in carbon dioxide or CO2 (? pKa) which react with water or H2O (15.7 pKa) to form carbonic acid or H2CO3 (6.1 pKa) which then further reacts to bicarbonate or HCO3 (10.3 pKa) by shedding hydrogen (-1.74 pKA) which further decreases the pH. They are all taking part in the system. (some more pKa’s for reference)

To give you a bit of an idea of the quantities of H2CO3 and HCO3, this graph shows you their balance across the pH range. Just keep in mind our ideal range is 7.35~7.45.


OK, so they are acidic and that could cause a lowering of the pH and could explain the lowered oxygen saturation. Not so fast… we need evidence. Let’s look at some trials that recorded values for ketones and pH together.

The following first paper shows us that ketones do have an effect on the pH, although very mild.

Acetone is a neutral compound, and, unlike AcAc and 3-OHB, it does not affect blood bicarbonate concentration, arterial blood gases or pH (Sulway and Malins, 1970)

However, after a few days of starvation a new steady state develops, and the arterial pH stabilizes at about 7.35 (Reidenberg et al, 1966)

“Ketosis of starvation: a revisit and new perspectives.”, Owen OE, Caprio S, Reichard GA Jr, Mozzoli MA, Boden G, Owen RS, 1983, ;

After a few days of fasting, we see in this publication that BHB has reached a level of about 1.5 mmol to only go up further after that. Interesting is that the pH is re-positioning itself at 7.35. There was no comment on what the original pH was. In any case 7.35 is at the bottom of the range which is considered OK. The body tries to keep it tightly within 7.35 to 7.45. Below 7.35 you are considered to be in acidosis but this is of course not a binary thing.

The next paper states a definition of metabolic acidosis. Useful to see if we can notice similar changes under ketosis (but in a more subtle form).

Clinically, metabolic acidosis is characterized by a decrease in serum HCO3– levels, accompanied by an increased arterial partial pressure of carbon dioxide (PaCO2) and pH of blood.

There isn’t that much data available but here’s a nice one that looked at exogenous ketones. The study compared ketone esters and ketone salts but we only want to focus on the esters because salts influence the pH as you can see in the graphs below.

When we administer ketone esters, we see a drop in pH and bicarbonate. Surely the time frame is short so we can’t say much about it long term but it is clear what effect there is with ketones. The pH even drops below the ideal value to stabilize at 7.35. The same value as observed under starvation. By the 2 hour mark we still have a value around 1.3 mmol

“On the Metabolism of Exogenous Ketones in Humans”, Brianna J. Stubbs, Pete J. Cox, Rhys D. Evans, Peter Santer, Jack J. Miller, Olivia K. Faull, Snapper Magor-Elliott, Satoshi Hiyama, Matthew Stirling, and Kieran Clarke, 2017,

So we saw the natural production of ketones under starvation and here supplemental ketones. I also found an infusion of ketones. Unfortunately it was in a salt solution which increased the pH so effects cannot be derived to understand endogenous production. I’ll leave the reference because it is an interesting read anyway.

“Ketone Body Infusion With 3‐Hydroxybutyrate Reduces Myocardial Glucose Uptake and Increases Blood Flow in Humans: A Positron Emission Tomography Study”, Lars C. Gormsen, Mads Svart, Henrik Holm Thomsen, Esben Søndergaard, Mikkel H. Vendelbo, Nana Christensen, Lars Poulsen Tolbod, Hendrik Johannes Harms, Roni Nielsen, Henrik Wiggers, Niels Jessen, Jakob Hansen, Hans Erik Bøtker, and Niels Møller, 2017,

The following paper makes reference to lower pH

Most research projects indicate a tendency towards lower blood pH and reduced blood base excess and bicarbonate levels after a ketogenic diet at rest; and especially after exercise with maximal intensity [34]

“The effects of a ketogenic diet on exercise metabolism and physical performance in off-road cyclists.”, Zajac A, Poprzecki S, Maszczyk A, Czuba M, Michalczyk M, Zydek G, 2014, ;

I followed the reference but was unable to obtain full access. The abstract talks about post-exercise lower pH values and bicarbonate. They did report about 2 mmol post-exercise and stated 0.8mmol 1 hour after exercise so it looks like the post-exercise is a couple of hours after exercise.

OK now what does it all mean?

So the point is that being in ketosis presses down on the pH, give or take around 7.35. If you went through my previous write-ups (see links at the top) then you would understand that I derive a bad effect from the lowered pH due to CO2 production by glucose metabolism. My main interest is the tissue oxygen which, under insufficient oxygen, leads to diseases.

Under ketosis, I speculate, that due to the pressing effect on pH, there is a slight shift in the oxygen dissociation curve. The slightly lower pH also causes a bit less affinity for oxygen making it easier to release. There is however a difference with how CO2 affects this, namely that there is a lower production of tissue CO2 (!). So my theory is that we get better oxygenation in the tissue due to being in ketosis.

It’s a long shot given the little data but it would make sense as well because the oxygen is needed for fat metabolism. You certainly need that extra oxygen when exercising as a fat-adapted athlete but you also benefit from from that effect at rest. If I’m right, this is a very important factor in preventing chronic diseases. This could have direct implications for Alzheimer’s, cancer, CVD, COPD etc. These diseases are referenced in those write-ups at the top.

Next up I will write about how I further try to increase tissue oxygenation besides being in ketosis so stay tuned. You can also subscribe to get notified about new posts.

I would also want to invite you to leave a comment if you like it, if you don’t agree with some of the things I wrote etc.. There is so much information I’m sure I always miss out on something important.

Further reading:

“Carbon dioxide transport”, GJ Arthurs, M Sudhakar,

Longevity (2)


Oh dear, do we dare to write about being pro- or anti-vaccination? I think I MUST do it. There is no such thing as censorship in science. And if there is one topic that is very heated at the moment that deserves a lot of attention from science then it is vaccination because the outcome of your stance has big consequences. When it is heated, the answer probably lays somewhere in the middle.

To tell you the truth, I’ve seen pretty good arguments against as well as for vaccination.

The book on longevity talks about vaccination but I didn’t think much of writing about it until I saw this latest news report on measles. Sorry it is in dutch but the main point they say is that the measles outbreak is on the rise again due to lack of vaccination, increasing child death.

That is what I mean with big consequences. We are talking about child death here. If anything, we have been able to raise our average life expectancy because we were able to save our children from dying of all sorts of causes like infections, accidents etc.

So what can we look at when talking about vaccination? Don’t expect to have a clear decision on yes they are safe or no they are not. Instead I want to have the following questions answered or at least looked at. They would help me to evaluate how to deal with the whole situation.

  • Do vaccines get tested for safety in the same (rigorous) way as other drugs?
  • Do vaccines have known side effects? What are they?
  • If they have side effects, are they worth it versus the protection that they bring?
  • Are all vaccines effective, tested and safe or do we find a mixed landscape?
  • Even if vaccines would not be very safe but are effective, what consequence would there be not having them?
  • If vaccines have side effects, who experiences them and can the side effects be mitigated somehow?
  • If I decide to vaccinate, can unvaccinated people still have an effect me? Or if I don’t, am I a threat to others who are vaccinated?

Those are not questions that can be answered with a quick google. I will have to take my time digging into them and will spread across different posts as I’m sure each question will provide enough material.

When vaccinations are effective, it would be senseless not to consider them if they prevent you from dying early. Hopefully we can get some more clarity because the article on measles is something to be concerned about.

Longevity (1)

I am almost finished with the book “Lifespan: Why we age – and why we don’t have to” from David Sinclair. In this book he explains the result of his research in a very readable way, avoiding much of the scientific language so that it is accessible to everybody. While reading the book, I recognized a lot of effects that can be obtained in a different way. Namely through a ketogenic diet.

So I’ll be going through the points raised in the book where I saw the ketogenic diet having a similar effect. The first point is the uncoupled metabolism.


One of the reasons why uncoupling would be health promoting and life extending is because it would be protective against oxidative damage by reducing ROS production. This is also what the ketone body beta-hydroxybutyrate (BHB) does. It stimulates the expression of the uncoupling protein.

“Mitochondrial uncoupling and longevity – A role for mitokines?”, Klaus S, Ost M, 2019,

“Ketones drive mitochondrial uncoupling in adipose tissue”, Chase Walton, and Benjamin T. Bikman, 2018,

“Mitochondrial biogenesis and increased uncoupling protein 1 in brown adipose tissue of mice fed a ketone ester diet”, Shireesh Srivastava, Yoshihiro Kashiwaya, M. Todd King, Ulrich Baxa, Joseph Tam, Gang Niu, Xiaoyuan Chen, Kieran Clarke, and Richard L. Veech, 2012,

So ketones may promote longevity by the same mechanisms. This could make sense as it is a molecule associated with long-term fasting and starvation. Plenty of discussion on what exactly is fasting versus starvation because starvation has a negative connotation but what I mean here is going without food long enough so that your level of BHB rises significantly to support energy requirements while not eating.

But that definition doesn’t really matter.

What I want to get to is, to understand the effect, how much BHB do you have circulating and where is it used? Because, even if you have sufficient quantities of it circulating around, only where it is used as fuel it will be able to exert its effect.

For example the liver is unable to use BHB as fuel. Curiously, the liver is the only organ that can recover fully despite tremendous damage.

In order for BHB to reach the cells in the organs, it has to be carried across the endothelial cells. BHB is transferred using MCT1.

Now also keep in mind that there is an inverse correlation between BHB and blood glucose in part due to the suppressing effect BHB has on the release of glucose from the liver. The brain is a huge consumer of glucose so it must be able to obtain BHB more easily when glucose goes down. And for that we see an 8-fold (!) increase of MCT1 in the endothelial cells in the blood-brain-barrier. The same effect in humans can be expected as shown by starvation experiments.

“Diet-induced ketosis increases monocarboxylate transporter (MCT1) levels in rat brain.”, Leino RL, Gerhart DZ, Duelli R, Enerson BE, Drewes LR, 2001,

So can we expect all of this uncoupling to take place in the brain? Probably, but I want to cast some doubt here. Uncoupling results in heat production. In the past I have raised a question to professor Thomas Seyfried asking why fatty acids are not used as fuel in the brain. He suspect it is because of the heat production. A skull is a closed area making it difficult to cool off. It is not proof uncoupling doesn’t take place but I would be careful to take that assumption for true without any research.

What is already available is mRNA that indicates uncoupling does take place to some degree. We don’t know by how much though but the following research tells us it is even vital for health, and I would say maximum healthspan of the central nervous system (cns). The uncoupling proteins in the CNS are not the UCP1 version which is important for thermogenesis. Instead, it are different versions that have health improving aspects. Sounds good but… UCP1 was stimulated by ketones, is the same true for these other versions (UCP2, UCP4, BMCP1/UCP5)?

“Mitochondrial uncoupling proteins in the cns: in support of function and survival”, Zane B. Andrews, Sabrina Diano & Tamas L. Horvath, 2005,

The little evidence that I could find seems to suggest it is the case for at least UCP4 and UCP5.

“A Ketone Ester Diet Increases Brain Malonyl-CoA and Uncoupling Proteins 4 and 5 while Decreasing Food Intake in the Normal Wistar Rat”, Yoshihiro Kashiwaya, Robert Pawlosky, William Markis, M. Todd King, Christian Bergman, Shireesh Srivastava, Andrew Murray, Kieran Clarke and Richard L. Veech, 2010,

So we’ve seen an effect in brown fat and CNS. It is reasonable to suspect that the brain will be the main consumer of BHB due to its energy requirements and evidenced by the upregulation of MCT1 so I don’t think other organs will make increased use of BHB and instead rely more on fatty acids for fuel. Perhaps there are some other mechanisms that the rest of the body can benefit from. We’ll explore them in the next posts.

Create your website with
Get started