Nutrient Partitioning: …a *very* high protein diet.

Or: what happens when you eat a ton of protein?

RDA: 0.8 g/kg

Active individuals: 1.2-2.0 g/kg (via ISSN)
Comment (1): I think sedentary, physically inactive, and non-exercisers should be in this range to offset disuse atrophy.  And they should exercise.
Comment (2): Do athletes really need more protein than non-athletes?  They have exercise, a powerful anabolic stimulus.  More protein may improve performance or body composition, but they might not *need* it, in terms of nitrogen retention… there’s probably a study on this.


The effects of consuming a high protein diet (4.4 g/kg/d) on body composition in resistance-trained individuals (Antonio et al., 2014)

4.4 g/kg! for 8 weeks in resistance-trained individuals.


They recruited twice as many participants into the high protein group because they predicted a high dropout rate.

Aside: a group who studies bone metabolism tried to do this when I was in grad school.  There were some  problems: 1) the participants stunk, literally.  Gas.  A lot of it.  So much so that some dropped out of the study.  And 2) it’s difficult to consume over 300 grams of protein daily, consistently, for weeks at a time – some people didn’t drop out, but simply couldn’t force that much down. Even from protein shakes, which were used to supplement when necessary (also to selectively increase protein% sans other macro’s… protein-rich foods are usually very nutrient-dense, unlike protein shakes, but they wanted to look specifically at protein, only protein, and shakes served that purpose).

Changes in body composition after 8 weeks of consuming 4.4 g/kg protein were not statistically significant, but here goes it anyway:

body comp

Caveats: there was a lot of variation in the intervention group… as expected, because the control group was given relatively simple instructions: keep on doing what you’re doing; and the intervention group’s instructions were vast: double your protein intake.

Very few results reached statistical significance, likely due in part to small sample size.  So maybe I’m grasping at straws here.

Food intake data are probably fairly reliable – daily food diaries, every day.  That’s more comprehensive than most studies, which usually only take a few days.

And Bod Pod is a pretty good way to assess body composition (better than calipers; not as good as but less expensive than DEXA).

Nuance: the effect of the observer on the observed.


Everybody exercised more (they weren’t supposed to do this); however, the control group increased training volume twice as much as the high protein group.  The added anabolic stimulus may have closed the gap between the groups.

training volume

Counterpoint: the control group ‘went on a diet,’ which usually doesn’t bode well for muscle gains, although they did gain a bit of fat mass because there wasn’t enough dietary protein to match the increased anabolic stimulus with muscle hypertrophy…? 

The control group gained body fat despite eating less and exercising more. (NS)


This is speculative, but doesn’t seem too far-fetched when viewed comprehensively: the high protein group was eating WAY more calories, but since they were protein, and coupled with increased resistance exercise, the result was muscle hypertrophy (almost 50% more than the control group) and fat loss.  The added thermic effect of dietary protein also *could’ve* added a significant boost to energy expenditure, simply because it was a LOT of protein.  Given all the changes (food/macro’s, exercise volume, etc.), it’s difficult to predict the absolute difference in energy expenditure between the groups.  But that’s neither here nor there, because:


These findings are relatively similar to George Bray’s overfeeding study …they compared overfeeding low and high protein diets, and showed that high protein increased muscle mass whereas low protein had the opposite effect.  Both groups gained fat mass (more so on the low protein diet), although this may have been because neither group was exercising.


The high protein group lost fat despite EATING MORE and EXERCISING LESS. (NS)

Yes, it depends on how you look at things.


calories proper


Become a Patron!



Be Sociable, Share!
  • Pingback: New study on Protein intake and fat/muscle/weight gain - Page 2 | Mark's Daily Apple Health and Fitness Forum page 2()

  • alexmyers83

    I would love to see those food logs…

    • You’d see a lot of protein shakes! the authors mentioned protein shakes by MusclePharm and Adept Nutrition.

      • alexmyers83

        I might be enjoying a bit more whey tonight! One might postulate that whole food based proteins (e.g. chicken breasts) would surpass whey…if not in bio-availability but in thermic effect, which might be more relevant in overfeeding scenarios

  • CynicalEng

    Shame they did the mixed gender thing, adding a whole lot of variability to the numbers and hiding statistical significance. Why do people do that ?

    • another source of variability that they clearly didn’t need!

  • Wenchypoo

    Wait `til Jimmy Moore sees this–he’s always going on about how meat is the “other” sugar bomb…maybe if you eat store-bought meat!

    • he’s pretty busy right now with keto clarity and his convention 🙂

  • johnnyv

    It is not pretty when a large bolus of protein reaches the large intestine!
    I wonder how much and the proportions of nutrients that made it through the small intestine.

    • Also, this would’ve been a great study to test the impact of a *very* high protein diet on the gut microbiome.

      • Jack Kruse

        Bill you’d die at some of my dinners…….3 lbs of meat/seafood per meal.

        • that’s a lot of protein! I’d be sweating. or in a food-coma.

          • Jose Antonio PhD

            Many of the subjects felt that they were ‘hot’ (body temp elevated) much of time. One female subject mentioned laying in bed and just feeling hot.

          • “The results of the current investigation do not support the notion that consuming protein in excess of purported needs results in a gain in fat mass. Certainly, this dispels the notion that ‘a calorie is just a calorie.’ That is, protein calories in ‘excess’ of requirements are not metabolized by the body in a manner similar to carbohydrate.”


          • Wenchypoo

            What makes a person “feel” hot like that–could it be her thyroid? Her hypothalamus? Her adrenals? And what in protein would cause that–maybe tyrosine? Too many aminos in general? I’m just spitballin’ here.

          • In this context (50% more calories, mostly from protein), it could be the thermic effect of feeding and some uncoupling.

            Not 1 specific protein, but amino acid metabolism in general (protein turnover, urea cycle, gluconeogenesis, etc.)

          • Galina L.

            Then a swim in a cool swimming pool or a lake would be nice after such meal. Icy bath is not for everyone.

          • Wenchypoo

            Poor kidneys!

  • there’s so much to say on this…I’m not sure how they handled eating so much protein…on my n=1 experiment, when I try to overfeed on protein (more than 2g/kg) I feel stomachly disturbed…any good way to overcome this guys?

    • you could try spacing it out in more [smaller] meals throughout the day…

      • I wouldnt like to get into the 6-8 meals/day dogma Bill 🙂

        • Thomas Hemming Larsen

          For protein I would rather spread it out to stimulate MPS more times during the day than having >50g in one meal.
          The subjects just drank whey shakes.

  • valerie

    The numbers in the last table don’t add up, especially the delta for the control group. Do they explain why in the paper? (Maybe fiber is included in carbs, or alcohol had been omitted, or they didn’t use the traditional 4-4-9 approximation, etc.)

    I have a hard time taking conclusions seriously when the authors can’t do simple arithmetics.

    • Jose Antonio PhD

      The data is what it is. Sometimes the coolest studies are the ones where the data is somewhat unexpected.

      • valerie

        The math doesn’t work. That’s not “somewhat unexpected.” It’s a mistake. Do you not see the problem with the numbers in the last table? The carbs go down 9 grams, the proteins go down 11 grams, the fats go down 2 grams. How can the total energy go down 243 kcal?

        • Valerie, I’d take it up with MyFitnessPal or Nutribase!

          • valerie

            Sorry, I don’t understand what you mean. Why would you need MyFitnessPal or Nutribase? They already give you the macronutrients in the table. Confused.

          • That’s where the numbers in the table came from. See “Methods” section:

          • valerie

            Thanks for the link to the full paper, but it still doesn’t explain the bad math. Do you think their numbers are accurate? Or did someone mess up the data entry? The second option seems more plausible to me. Any statistical analysis after that would be garbage.

          • I don’t know! …which is why I suggested contacting MyFitnessPal and Nutribase 🙂

            I ran the numbers as you suggested and they only differed by about +/- 100 kcals… it doesn’t seem like fibre was excluded because some of the calorie counts were actually higher than the mathematical 4-4-9 sum.

          • Jose Antonio PhD

            We input whatever is provided. When you are dealing with 56 daily food logs for 30 subjects, there is always a bit of background ‘noise.’ Having done these types of studies several times, you rarely if ever get numbers to add up exactly. What’s important is that the baseline data is based on a week of food logs. The ‘post’ data is based on 8 weeks (daily log). So if an error exists, it should be consistent from subject to subject. What’s ironic is that there are published exercise and diet studies in which ‘diet’ is determined by a single 24-hour food recall. Sometimes they measure three days. Of course it is easier to get numbers to add up if you are only looking at one-three days.

          • annique

            There’s no error, the discrepancy is typical of any mutlivariant analysis with standard deviations. It’s because of ‘error propagation’ when doing calculations that have +/-std. Think of it like ’rounding errors’ but magnified every time you do a calculation. This means it matters what order you do them in.
            So, counting macros for each individual, then averaging over the time period, then adding for all individuals, then averaging that and Then multiplying by 4-4-9 to get calories… will give one answer.
            This, btw, is what you are effectively doing by trying to xply the in that table with 4-4-9.

            However, adding macros for each individual, then xplying by 4-4-9 for calories, then averaging the individual calories over a week, then averaging for all individuals…will give another answer.
            This is more likely the calories number in the line of the table to do with calories.
            …and yes, there’s variations in each of the two sequences you can do, those were just two examples – the point is that there’s no one ‘right’ way to do it.

            I don’t know if I made this any clearer or added to confusion! Just trying to help, because I know it’s a very frustrating topic for bio-stats students who often don’t encounter it until that course in year 3-4, while chemists, physicists and engineers get it as freshmen….seeing as it’s rather fundamental 🙂

          • valerie

            Those discrepancies are way too big to be rouding errors (98kcal vs 243kcal). Measurement errors (errors as in imprecisions, not as in mistakes), maybe, but again, the discrepancies are way too big for that.

            Stdv’s have nothing to do with it. The average of sums or the sum of averages give the same results, no matter what the stdv’s. Stdv’s will affect the p-values, but not the averages.

            To me, it looks much more like a mistake in data entry/manipulation. Which of course ruins the statistical analysis completely. Such an easily spotted mistake leads me to question the care taken in the whole thing, plus the review process.

  • Jose Antonio PhD

    The subjects in this study (I’m the PI btw) literally had to ‘force’ themselves to eat that much protein (most of it in whey protein shakes). You have to be a professional eater to accomplish it. I tried it myself…after 2 days, forget it. Can’t do it. I like rice too much. I’d have to drown myself in protein shakes. We’re following up this study with one looking at 1) blood markers (e.g. liver/kidney function, lipids, etc) 2) using a slightly lower dose [3-4 g/kg/d] 3) changing their training so that volume load goes up.

    • Hi Jose, thanks for the comment. I’m not surprised; that’s a lot of protein!

      Have you considered collecting some fecal samples & sending them off to uBiome or American Gut? I haven’t seen any good data on high protein diets and your study design is perfect for this.

      • Jose Antonio PhD

        I haven’t considered that. I’m not so much interested in mechanisms, but pure outcomes. I think body comp can/will change with a high protein diet only if the training stimulus changes.

        • I agree, but this seems like it could get complicated! Will the increase in training volume be similar between the groups? Also, are you expecting the high protein group to experience greater strength gains?

          • Jose Antonio PhD

            I honestly don’t know what to expect. I’d speculate that if you train folks hard enough and really manipulate training volume and intensity…AND feed them a lot of protein, they should experience significant skeletal muscle hypertrophy!

    • Thomas Hemming Larsen

      Hi Jose,

      Did any of the subjects mention anything about stool frequency, consistency and volume?

      • Jose Antonio PhD

        We didn’t ask subjects about this. Nor did any provide that information.

        • Thomas Hemming Larsen

          Thanks Jose. I was just thinking if all that protein caused constipation, diarrhea or something else.

          • Jose Antonio PhD

            Nobody mentioned it when reporting side effects. The only ones mentioned were “feeling hot” (elevated body temp) and bloated from eating a lot of protein.

          • Thomas Hemming Larsen

            That’s interesting. I believe I’ve heard bodybuilders complain of diarrhea because of the protein.

            I’m not surprised they felt hot. A lot of the calories must have gone to thermogenesis.

          • MichaelGWS

            Thomas I think it’s because so many people can’t handle lactose and the whey shakes typically have lactose in them.

          • EdW

            hey just thought I’d chime in here.
            since reading Chaos and Pain blog I started upping my protein, this study is around about my normal protein intake consistently.
            I weigh around 109KG and consume 440g protein a day readily.
            it was difficult to start with but I think I’ve adapted now and any less would feel a bit odd.
            pre wo: scoop whey
            pwo 2 scoops whey
            meal 1 250g bacon
            meal 2 same
            home from work meat and nuts
            evening meal meat and eggs or similar
            sometimes pre bed greek yogurt and 2 scoops whey.
            I went low carb some years back but now I back load and cycle my carbs along with a big refeed every now and then when I feel useful, normally on a Friday.
            I have had time off full body training this year due to injury but now I am back on full body workouts 5 times a week and strength is returning.
            as for protein/meat sweats I totally get this and have elevated body temperature all the time.
            when I started C&P predator diet it was like 5-6 shakes a day plus heavy evening meal which averaged protein intake of 400g a day and it was then I started playing around with it, because despite probably spending most of each day on a caloric deficit I gained a whole lot of muscle mass.
            but maybe I am a freak? I gained muscle mass during a 3 month keto period this year too, training upper body only and mostly back, due to issues with muscular imbalances which are largely resolved now.
            love your blog

          • “I weigh around 109KG and consume 440g protein a day readily.”

            that level of dietary protein is nothing short of epic. Kudos

    • Thomas Hemming Larsen

      I read you’re working on a follow up study with blood markers etc. Do you have any hypothesis/speculations so far regarding the long term effects of eating this amount of protein? Some claim that it reduces aging by activing mTOR and so on.

  • Galina L.

    I tried different diets during my life-time, but I missed experimenting with the high protein one somehow. May be it is not too late for me to give it a try some day (not now). My culinary instinct points me into the adding sauerkraut and vegetables with vinegar to the meat . I am absolutely not into shakes. My guess the easiest way to consume an extra protein as a meat would be in a ground form.

  • Thomas Hemming Larsen

    I wonder how much the resistance training has affected their body composition. It would have been interesting to have a high protein group performing no exercise/training (I’m also thinking if this could partly explain how Sam Feltham was able to eat so much fat and protein while gaining a negligible amount of weight which was maybe even muscle).
    After I’ve increased my protein intake (thanks to you Bill convincing me that it will not kill me) and started resistance training I’ve increased my muscle mass and lowered my body fat (the increased satiety is also a nice things :)).
    I’m really looking forward to the next study looking at the impact on the liver and kidneys, I can’t imagine that that much protein is a good thing long term. It must also cause some stress on the stomach…

    • resistance training: my guess is a LOT. in this context (combined with high protein), I think, it was uniquely set up to have a pretty significant impact on nutrient partitioning. And yeah, the perfect study would’ve had more control groups, but understandably, that can get pretty expensive.

      As to the long-term effects of high protein… my “opinion” is that ‘excess’ dietary protein rarely occurs because protein intake is self-limiting. That is, I think you’d be way too satiated to continually overeat protein… like, food would begin to get disgusting after a point.

      P.S. thanks for pushing me to study CBL! I read his book and will be writing a blog on it soon.

      • Thomas Hemming Larsen

        I know its a pain setting up these studies both financially and practically. I remember when I participated in one – it was quite clear that it was done by a private company 🙂

        Yeah, I agree. Paul Jaminet also had a study which shows that both humans and animals eat ‘adequate’ protein and then carbs/fat to fulfill their energy needs.

        Haha, I don’t think I was pushing you 🙂 I’m really looking forward to hearing your opinion on it and seeing if you have some tweaks. The book is outdated on some points like caffeine, preferred type of fat etc. which Kiefer has mentioned in recent podcasts (he’s working on CBL 2.0). I really hope that he will see your podcast and invite you to his podcast. You would have a lot to discuss and you could challenge him on his views.

  • Eddie

    Hi Bill, what would be your opinion on the following scenario.

    I have been trying o lose fat for an exact month, my approach was the following:

    1- Get my BMR (

    2- Multiply BMR it by a activity factor of 1.2 since I work on a desk job, 8 hours a day and only workout in the gym 4 times a week 2 hour sessions, it didn’t seem to me to be substantially enough to increase it.

    3- Get the 80% of that number in order to create a calorie deficit

    4- Get my protein needs based on a 1.5 gr per kilo, on lean mass (total body weight – (BW * %BF))

    5- Multiply Protein gr by 4 to get # calories from protein

    6- Add the calories from 100 gr of carbs (100 * 4)

    7- Subtract protein calories and carb calories from the 80% deficit and that would be my fat calories

    I have been steadily losing 1 kilo per week, which it is due to daily readings 80% fat and 20% muscle.

    After reading your article, my question is, do you think that the calorie deficit is needed for fat loss in conjunction with the max of 100 gr of carbs? or just sticking to the 100 gr rule and increase the gr of protein per kilo would spare more muscle?



    • Hi Eddie,

      That’s a little too much math for my liking 🙂 …but your progress (-1kg/wk, 80% fat) speaks for itself! 80% of the weight coming from body fat is pretty good, I’d keep doing what you’re doing.

      Some resistance exercise or HIIT might help to spare more muscle. Also, there’s a lot of evidence that sleep quality plays a critical role in determining the ratio of fat:muscle loss… eg,

      • Eddie

        Thanks a lot for the feedback Bill, you are right, too much math, but what would you say the macros woud be for someone trying to get muscle, lose fat in a Paleo/natural bodybuilding approach?
        How much protein grams per kilo?
        Would you impose a Carb grams per kilo limit?
        As an example I am 36yrs, 1.73 cms and 15% BF

        thanks in advance for your time.

        P.S. on your book you mention Leucine as an alternative to induce Insulin peaks in the Post-workout meal with high glycemic carbs, at least, that what I got 😉 . If that is the case, how much Leucine would you recomend and what timing would you say its better?

        • “How much protein g/kg?

          Would you impose a Carb g/kg limit?”

          Eddie, as mentioned above, the answer to these questions is: however much protein/fat/carb you’re eating now. You’re making good progress, keep it up!

          Leucine: not *with* high glycemic carbs, *instead of* them. Results with leucine have been mixed, but 5-10 grams w/ a small meal should suffice. Timing = with a meal, around when you exercise.

          • Eddie

            Thanks again Bill, I was just trying to get further improvements, trying to not ingest more protein than needed, same for carbs.
            Last days I have reach a plateau in my progress and wanted to check if it was a consequence of too much calories, even though they were “good”.
            Its strange but I seem to feel that if I want to see a fat storage decrease, I have to drop carbs dramatically, like no more than 2 cups of green beans and carrots even on workout days, does this means I have a poor insulin sensitivity?
            On the other hand I have been tracking sleep and using the Blue blockers, but everything seem to point out that even a banana on a post-work out shake, seems to stop progress (this means less fat, muscle has being the same or increasing). Sorry to keep bugging you with this, but your feedback its being the most useful in a long time of trying new things, no other author replies as you do, thanks for that too!

          • Again, I think the answer was in the question! If fat loss is stalled by postworkout 2 cups green beans & carrots, or a banana, then skip ’em!

          • Eddie

            Thanks for the cyber-slap Bill, you are totally right, sometimes we have the answer in front but we hesitate too much, since your feedback I have cut the carbs on my PWO shake, which now consist in a protein blend (Phase8), glutamine, creatine HCL and 3.5 grs of BCAA’s. The rest of the day I am mostly eating omelettes with veggies.
            Fat loss Plateau is gone.
            Regarding Leucine, I found these article from which led me to add the the BCAA not only in the pre and during workout, but also in the post.

            Thanks again for you time 😉

  • Thomas Hemming Larsen

    I wanted to revisit this epic study. I realised that your table with bodyfat percentages doesn’t match the one in the study. Why the difference?

    • I calculated body fat percentage by dividing fat mass by body weight. Not sure how they calculated it, but the final numbers are relatively close.

      • Thomas Hemming Larsen

        Everything is relative… In the study it has 15.1 -> 14.2 for the control and 16.9 -> 16.3 for the HP. It does change the picture a little.

        • Right.

          Sorry, what I meant to say was that I think the discrepancy is their 16.9, 16.3, 15.1, and 14.2.
          How were those numbers calculated?

          • Thomas Hemming Larsen

            I agree, your numbers look correct. It’s just that your numbers make the HP diet look even better 🙂 You should ask Jose how he has calculated the numbers 🙂