Dietary protein does not negatively impact blood glucose control.

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

At a reasonable level of dietary intake, protein is used for the maintenance of organs & tissues.  Lean body mass.  It’s functional.  Protein isn’t stored in any appreciable capacity, and most excess is either oxidized or stored as glycogen.  Theoretically, about 50-60% of protein-derived amino acids can be converted into glucose, mathematically, but it’s not what you think…

“At a reasonable level of dietary intake.”  A recent publication took a look at this (Fromentin et al., 2013).   They set out to determine how much protein is converted to glucose under “optimal gluconeogenic conditions.”  That is, the subjects were 12 hours fasted, which is a physiologically relevant, optimal gluconeogenic condition.  They were then given 4 eggs (~23 g protein) that were labeled with two stable isotopes (15N & 13C, derived from hens fed isotope-enriched diets!).  Throughout the entire study duration, the subjects were infused with a third isotope, 2H-glucose.  By collecting and analyzing the enrichment of isotopically-labeled metabolites like expired CO2, urea, and glucose, the researchers were able to determine the fate of those 23 grams of protein.

Some of the dietary protein-derived amino acids were used for protein synthesis, others were oxidized.  But blood glucose levels did not change.  Nor did the rate of gluconeogenesis.


Over the course of 8 hours, their livers cranked out ~50 grams of glucose, and about 4 of it came from the protein.  8% of the blood glucose produced under “optimal gluconeogenic conditions” came from dietary protein.  8%.  And it wasn’t enough to elevate blood glucose levels.

It could be argued that since only about a third of liver glucose production after an overnight fast comes from gluconeogenesis (the rest from glycogenolysis), 25% of the gluconeogenesis-derived glucose came from dietary protein.  My response: it still doesn’t matter because blood glucose levels didn’t change!

23 grams = ~0.3 g/kg for a 80 kg person.

Do you even lift, bro?

How about 2 g/kg?

Thank you, Conn and Newburgh, for doing this study prior to the existence of Institutional Review Boards (1936).  Best. Study. Ever.  They were going off of the hypothesis that half of dietary protein can be turned into glucose and this should cause profound hyperglycemia if they were to give something ridiculous like 160 grams of protein… right?  …because that would be the same as 80 grams as glucose.  In the figure below, line #1 is blood glucose after 80 grams of glucose.  Line #2 is 160 grams of protein (theoretically equivalent to 80 grams of glucose).beef vs glucose

They repeated this ad nauseum, with a variety of different sources of carbohydrate and protein (up to 1.3 pounds of lean beef in one experiment!), but the picture didn’t change.

What if your metabolism is “broken?”  Type 2 diabetic patients from Gannon 2001 given a half a pound of steak (~50 grams of protein)  (closed circles) or water (open circles) first thing in the morning:glucose response to protein diabetics

Blood glucose levels decline in the water-only group because they’re essentially fasting all day until 4 pm.  Similarly, blood glucose levels decline in the 50 grams of protein group because dietary protein DOES NOT CAUSE HYPERGLYCEMIA (might get a small non-significant bump when compared to water ingestion, but no red flags).  Glucagon was secreted but the rate of gluconeogenesis did not change. glucagon and gng

If this surprises you, then you’re barking up the wrong tree.  About 2 grams of the blood glucose produced over 8 hours came from those initial 50 grams of protein.  This is considerably less than in healthy people (see above), and still far less than the theorized 50%.

See an excellent piece by Amber and Zooko for more on this; they’ve gone into it in considerably more detail.  Also here.

Dietary protein is safe and even beneficial for diabetics… for ketogenic dieters trying to maintain a certain level of ketosis, it’s a gray area, ymmv…  Please do not confuse the two and recommend a type 2 diabetic to reduce their protein intake (especially without knowing their baseline intake!).  There are no studies to support such a recommendation and it will likely do more harm than good.  Maintenance of lean body mass is more important then deep ketosis for type 2 diabetics.  They are carbohydrate-intolerant, not protein-intolerant.  Brain fog is another issue entirely and is not in the same ballpark as diabetic hyperglycemia… and as evidenced above, dietary protein does not cause hyperglycemia.  In anyone.  Even if they eat a pound of steak.

Why must protein remain in the lower end of the spectrum for ketosis?  Dietary protein and/or amino acids are necessary and essential for survival <– fact.  When taken in excess, some are stored as glycogen, some are oxidized.  Ketosis relies on a need for “fat-derived fuels.”  If liver glycogen is full, then some of that glycogen reduces the need for said fuels.  Further, ketones are derived from hepatic fatty acid oxidation; if liver is getting plenty of energy from oxidizing amino acids, it certainly won’t need any from fat.  When it comes down to prioritization of hepatic substrate oxidation: big influx of diet-derived amino acids > fatty acids.  And insulin generally helps to prevent a head-on collision.

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

calories proper

Be Sociable, Share!
  • charles grashow

    So – if you do not want to be in ketosis and you workout regularly how much protein do you need??

    • William Lagakos

      The key part of that question is “need.” YMMV! Also, it’s difficult to put an absolute number on this.

      For healthy people who work out recreationally, 1-1.2 g/kg is probably a reasonable starting point. But how much more or less likely depends on a lot of factors. For example, higher total calorie intake reduces the amount of protein required to maintain nitrogen balance… but that’s “need,” not “optimization.” If you’re doing high intensity resistance exercise with the goal of physical performance and body composition, I’d say that amount may be considerably higher.

      • Jack Kruse

        Bill I think the true amount of protein is directly correlated to how bad reversal of flow is on the inner mitochondrial membrane. This is where the resevoir of cytochrome 2 reduction couple meets quantum tunneling and why so few biology folks understand how ketones and fats can do what they do when the membrane potential appears too low.

        • William Lagakos

          seems likely related to brain fog as well. Better energy metabolism via ketone use –> increased “tolerance” for dietary protein.

  • George

    In the Steffanson Bellvue experiment, protein wasn’t restricted, and urine ketones were being measured to check on compliance, and were always present.

    • William Lagakos

      agreed, protein intake wasn’t experimentally restricted… but they didn’t eat a “high protein diet.” Fat intake hovered around 80%. That should be enough to produce detectable urinary ketones in a wide variety of contexts (especially if carb intake is nil).

    • Jane Plain (Woo)

      matches my experience, protein does not terminate keto. I restrict it only because it seems to promote agitation.

      • George

        What if it’s the same with epilepsy? Protein needs to be restricted not to increase ketones, which are already sufficient, but because protein metabolites also trigger seizures?
        this is so plausible..

        • Ash Simmonds

          Hence the reasoning behind the ketogenic ratio:

          Formula: (0.9*fat + 0.46*protein) / (1*carbs + 0.1*fat + 0.54*protein) = ketogenic potential, >1.5 = more ketones

          • Ash Simmonds

            I did a mini write-up recently on the ketogenic potential of steak, here’s a copy/pasta:

            Let’s take a 1lb/450g ribeye/scotch fillet steak – pretty much a perfect Intermittent Feasting meal size I reckon – and see how it stacks up on the keto regime.

            Firstly, here in Oz we call this cut “ribeye” when it’s on the bone, and “scotch fillet” when the bone is absent (insert viagra joke). This is one of the most versatile and naturally fatty cuts of meat, and difficult to stuff up.

            What you have to understand is the “drier” the meat, the higher the protein and lower the fat – well, it’s a good rule of thumb anyhoo.

            I like to work with round numbers for easy head figgerin’ on the fly, so I round off meat in general as being 50% food and 50% water. Then to break that down, the 50% that’s food is half protein and half fat – so in any given steak 1/4 is protein and 1/4 is fat. Let’s see how that works for our delicious little ribeye:

            * Steak: 450g/16oz

            * Water and stuff: 225g/8oz

            * Protein: 115g/4oz

            * Fat: 115g/4oz

            Did I mention I like to work with round numbers? I find no joy in unnecessarily complicating life.

            BTW, these numbers I tend to round UP just so if anything you are OVER-estimating stuff. In reality by the time a 450g steak hits the plate I reckon it’s closer to 100g protein/fat in reality – but who’s counting…

            So in the end how does this work out ketogenically speaking?

            Well, let’s take our trusty’ish ketogenic formula:

            * (0.9 * fat + 0.46 * protein) / (1 * carbs + 0.1 * fat + 0.54 * protein) = ketogenic potential

            * >1.5 = MOAR KETONAGE

            Basically, a number of 1.5 in a given food says it’s keto, anything above 1.5 is ace for ketards. Now, run it through the ketobobulator…

            * (0.9 * 115g + 0.46 * 115g) / (1 * 0 + 0.1 * 115g + 0.54 * 115g)

            * = (156.4) / (73.6)

            * = 2.125

          • William Lagakos

            Ha! Ash the mathematician. Awesome.

          • Michael

            I like this post, and the math is easy to understand. The Viagra joke about the bone, on the other hand, is hard.

          • William Lagakos

            i see what you did there

        • William Lagakos

          it’s very possible… but epileptics also seem to benefit from fake-keto diets: moderate protein supplemented with MCTs.

          • George

            True, but there are some who respond to low-GI diets – 75g carb daily with meal limits, moderate protein – possibly insulin-sensitive. So there may be a few classes of epilepsy based on diet effects.

          • Jane Plain (Woo)

            good point

          • William Lagakos

            lol shortest Woo comment ever!

        • Jane Plain (Woo)

          I agree very much George,I suspect insulin metabolising amino acid is excitatory and affects both mood abd seizure risk.

  • rand18m

    Excellent piece William. I look forward to reading the 1936 paper and both studies will go in my files.

    • William Lagakos

      Thanks, Rand. That 1936 paper is great! All individual data are reported.

  • Ash Simmonds

    Awesome write-up. Ever since Jimmy’s brainfart last year where he said something like “eating too much protein is basically just like eating a chocolate cake” us long-term ketards/carnivores have had a hard slog.

    Every newbie and their bacon fuelled dog were freaking out about “being kicked out of ketosis” and spiking their blood sugar by a damn steak, ignoring the overriding determinant – liver glycogen levels and ketone *needs*.

    I like the elegant way it’s put here, but I think we’re still gunna have a hard time for a while with people confused as hell about protein and obsessing about getting the exact right amount to maintain ketosis.

    • William Lagakos

      Thanks, Ash. That’s pretty much the exact mindset I had writing this article. Don’t want to see “low protein” taken out of context… especially because reducing protein AND calories is a recipe for muscle wasting.

      • Thomas Hemming Larsen

        Spot on Bill! I have tried restricted protein on top of restricting calories and that was the straight path to losing a lot of muscle and generally not being able to do anything. I had edema in my legs and stomach and had to wear compression all day just to get around.

        • William Lagakos

          “Protein has a purpose, and that purpose is not carbs” !!!

          important for maintaining lean body mass; even more so when restricting calories.

  • PhilT

    So excess protein doesn’t adversely affect glucose homeostasis, but may impair ketosis and will reduce fat loss in dieters by providing a preferential energy source ? Did I get it right.

    What about insulin levels ?

    • Chris King

      nope, you didn’t get it right.

    • William Lagakos

      Excess protein compared to excess carbs? excess fat? In either case, I don’t think you’ll see a big impact on the rate of fat loss.

      “And insulin generally helps to prevent a head-on collision.”
      A collision of adipose-derived fatty acids and dietary protein-derived amino acids in the liver would likely just result in fatty liver. Insulin prevents that.

      • PhilT

        So if my aim is to maximise lipolysis (for fat loss) then excess protein above that required for maintenance isn’t helpful, which is I think where the “lower protein” camp are coming from.

        If low blood ketones due to higher protein point to less use of fats then lower protein (let’s say 100g instead of 200 g/day at 100 kg BW) to promote ketosis seems helpful, even if as you explain it’s nothing to do with the glucose levels.

        • William Lagakos

          It also depends on your goals. Protein promotes nutrient partitioning; good for lean body mass… and is especially important if you’re in an energy deficit.

          Deeper ketosis doesn’t necessarily mean more rapid fat loss; that is, I don’t think the difference between 1 mM and 2 mM is going to have an appreciable impact. Jane Plain wrote about this recently:

    • DrewStearns

      Yes. From the article you JUST READ, lol (just giving you a hard time PhilT):

      “When it comes down to prioritization of hepatic substrate oxidation: big influx of diet-derived amino acids > fatty acids.”

      • PhilT

        reading and understanding aren’t always the same thing ;-)

  • Jane Plain (Woo)

    Fantastic dr bill. In crazy world where keto nuts are telling wt loss dieters to restrict protein to 40 g per day, a factual common sense article like this is needed.

    This doesn’t even address the superiority of protein for energy and fat losses; even if drinking the keto koolaid, restricting protein still does not make sense.

    In my experience, keto may even darken with protein because of increase total sns thyroid and lipolysis from energy consumption. Most protein sources are predominantly fat, as well as many aminos being directly ketogenic.

    • William Lagakos

      40 g/d = negative nitrogen balance!

      Even if it a little more protein does reduce ketone levels, I don’t think this will negatively impact body comp or BG control… probably the opposite.

      Wrt your experience (?protein ? ?ketostix)… post that to FB! A lot of FB commenters arguing the opposite and there aren’t many human studies.

    • Jack Kruse

      Dont expect that message to come out of the low carb community that favors opinions of lay folks over people who understand science.

  • Pingback: Podcast Interview with Dr Bill Lagakos – Nutritional Biochemistry Ph.D – | The Ketogenic Lifestyle

  • Kindke

    the main reason to eat plenty of protein is to control hunger. nearly all research illustrates that low protein increases overall food intake. i.e. protein leverage. I can tell you from first hand experience that low protein is quite a horrible diet and leads to a constant nagging feeling in the stomach for more food.

    • William Lagakos

      also good for lean body mass, body composition, and energy levels… especially on hypocaloric diets.

    • Wenchypoo

      It also triggers meat cravings–I tried controlling Hubby’s gout by lowering his protein intake, and all it did was trigger intense meat cravings. I thought maybe he was suffering from iron or B-12 deficiencies, and supplemented with both, but to no avail. I even tried using protein substitutes, like beans, soy, and the like (pre-paleo days).

      Nothing works like good old-fashioned MEAT. Lowering his vitamin C levels eventually worked for the gout (the “eventual” part is my speed of learning this trick).

  • Wenchypoo

    Has anybody to date clarified if the protein they used was from
    grain-fed animals or grass-fed pastured ones? I think it goes without saying that grain-fed ones would naturally have a higher percentage of glucose in their meat. So what’s the percentage of GRASS-FED, PASTURED meats as compared to grain-fed meats? I bet the difference is dramatic!

    • johnnyv

      A higher multiple of a very low proportion is still a rather small amount, ruminant skeletal muscle is a poor source of carbohydrates. Animals tend to lose glycogen during transport and the wait before being slaughtered anyway.

    • William Lagakos

      In fresh biopsies, glycogen is less than 1 g / 100 grams of muscle… so by the time the meat is frozen, it should be considerably less. This is just a guess, but I’m pretty confident it’s close to nil in both cases.

  • Pingback: You Can Better Care For Your Diabetes With These Great Tips | True Wellbeing

  • Pingback: Diabetes Control Diet – Blood Sugar Patients Must Be Careful In Food Selection | Home Loan Interest Rates.Info: The Leading Home Loan Interest Rate Site on the Net

  • Raymund Edwards

    I think the people who find too much protein is a factor are not concerned with the immediate post postprandial effects on blood sugar.

    Metabolic response of people with type 2 diabetes to a high protein diet

    We also determined the serum insulin response to the ingested protein and in confirmation of the studies of Berger [7], Fajans [8] and others, we observed a modest increase in the insulin concentration in the non-diabetic subjects [6]. However, there was a relatively large increase in insulin concentration in the subjects with type 2 diabetes [5]. Indeed, it was about four-fold greater than in the non-diabetic subjects (Figure (Figure2).2).

    ••• We also determined that the rise in insulin following the ingestion of 50 g of beef protein was just as potent in raising the insulin concentration as was the ingestion of 50 g of glucose [5]. That is, meat protein and glucose were equipotent in stimulating insulin secretion. In addition, we also demonstrated a linear dose-response relationship between the amount of beef ingested and the insulin response [5].

    • RS711

      Raymund Edwards

      maybe it would be worth mentioning those results were obtained on a diet where “dietary protein was increased from 15% to 30% of total food energy [&] The carbohydrate content was decreased from 55% to 40%”….How protein is metabolised in this experiment is not in the context of nutritional ketosis, thus our conclusions should keep this in mind.

      The conclusion of your study is worth repeating: “From these data we conclude that increasing the protein content of the diet at the expense of carbohydrate can reduce the 24-hour integrated plasma glucose concentration, at least over a 5-week period of time. The reduction was similar to that of oral agents. Renal function was not affected significantly. Thus, increasing the protein content of the diet with a corresponding decrease in the carbohydrate content potentially is a patient empowering way of reducing the hyperglycemia present with type 2 diabetes mellitus, independent of the use of pharmaceutical agents.”

      • William Lagakos

        I don’t expect protein metabolism to be markedly different during nutritional ketosis.

        “Dietary protein has a purpose, and that purpose is not carbs.”

        Amino acids are used for the synthesis of enzymes and proteins. If during nutritional ketosis they were instead diverted into primarily gluconeogenesis, then the diet would cause muscle wasting.

        • Raymund Edwards

          Bill , I find If I eat too much protein esp after fasting for say 18 hours it puts me to sleep and then I feel hungry.

          It is very obvious if I have been fairly active as well. I am as sharp as can be before eating the protein , but 20 mins after if I am sitting relaxing I will nod off for maybe 15 to 20 mins..

          I will then feel sharp again BUT will feel hungry.

          The insulin release ? and a slightly delayed adjustment to the lower Blood Glucose ?

          If I eat less protein then this effect does not happen.

          This sort of response ( insulin )
          is where some people I think find less protein = a more even appetite and energy.

          In a diet with more carbs this effect maybe does not show up, hence more protein is only beneficial.

          • William Lagakos

            I find this very interesting. Do you think it is detrimental? I don’t see anything wrong or unhealthy with a 20-minute post-meal nap… does the hunger push you to over-eat? interesting, nonetheless.

          • Raymund Edwards

            The hunger can make you over snack on things if you react to it.If I wait the hunger passes. Yes The nap is not unpleasant.

            But if I eat a little less protein then this effect does not happen.

    • William Lagakos

      Hi Raymund, thanks for the link.

      No doubt lean beef stimulates insulin secretion, but it has a negligible glycemic impact.

  • Pingback: Cope With Diabetes By Following These Strategies | True Wellbeing

  • Pingback: Grounds Well.Biz | Diabetic Diet – Get Your Blood Sugar Level Lower Rapidly – Part One

  • Pingback: Grounds Well.Biz | Diabetic Diet – Get Your Blood Sugar Level Lower Rapidly – Part One

  • persistentone

    So if I get a large glucose rise at breakfast after eating a protein-only meal, this is being caused by something other than the protein? It’s just a Dawn Effect rise in glucose for example? I have had days where I get fasting glucose of 90 on rising, and an hour after an all protein breakfast I see 140. I am prediabetic, A1C around 5.7%

    • William Lagakos

      I don’t know.

      Conn & Newburgh showed huge amounts of protein don’t impact blood glucose; and Ganon showed something similar in type 2 diabetics.

      Have you tried measuring during any other conditions? eg, high fat breakfast, no breakfast, etc.