The Protein Debate

About The Debate

What you're about to read is a protein needs debate between John Berardi, Ph.D. of the University of Texas at Austin and Stuart Phillips, Ph.D. of McMaster University in Ontario. Be prepared; this isn't your typical Internet debate between two partially brain damaged 17 year olds calling themselves biggunz252 and punkmass101. This is an intellectual argument between two established experts working in research labs on the very topic of the debate.

This debate came about via a "journal club." In Ontario, there's a small community of scientists, coaches, and students who get together periodically (via email) for this club. What happens is that each month one member of the journal club sends out a bunch of articles for the others to read and later discuss. Ideas, debates, future research and coaching strategies are derived from these sessions.

Recently, the journal club topic was protein needs and two articles were up for discussion. The first was a T-Nation article by John Berardi entitled The Protein Prejudice. The second was a scientific review article by Stuart Phillips entitled Protein Requirements and Supplementation in Strength Sports.

After discussing these articles, the members of the journal club invited Drs. Berardi and Phillips to debate the issue. This debate is what we've published below. To get the most out of it, you may want to follow the links above and check out what started the whole fracas.

The Debate

Dr. Phillips's Response to "The Protein Prejudice"

I've read Dr. Berardi's paper and have found it generally sound, but lacking in sufficient evidence to truly be considered "truth" in some areas (not many to be honest); however, I'm pleased to hear that John is in-line with the general thesis that lifting weights reduces protein "requirements."

As far as "optimization," this is a far more difficult question and even John knows that this hasn't been answered satisfactorily, or at least not to a degree that would make me offer the advice that lifters should "eat more protein–a lot more". While it's easy to say "eat more protein" and counsel athletes to do this, what you have to ask at this point is: What are the dividends of eating more protein over and above requirement? Likely few, but then since they'll never have tried doing anything else, do you get stronger, leaner etc. because of, or in spite of what you do?

In a recent review (Nutrition 20: 689-695, 2004) I examined the evidence for whether resistance trained athletes "need" more protein. The answer is likely the opposite. On this John and I can agree. I also pointed out in the same review, after taking relevant nitrogen balance data and regressing it through zero nitrogen balance, that resistance trained athletes require more protein than sedentary individuals. The problem is that, by all standards, most milk, egg and meat-eating North American males eat way more protein than they need.

Also, I take issue with Dr. Berardi's interpretation of the requirement (i.e., the RDA) being "eating just enough protein to prevent deficiency." In fact, the basis of the RDA is a much lower protein intake than 0.8g protein/kg/d, but addition of safety margins means that 97.5% of the population is covered. Thus, it may well be that your individual requirements are far less than the RDA, but you'd have to participate in a study to find out exactly where you are.

The problem with all of this discussion about requirements and using nitrogen balance is the method is badly flawed. Hence, trying to talk about a requirement or even optimizing protein intake based on findings from nitrogen balance studies is like trying to circumnavigate the globe using a map before we knew the earth wasn't flat–the instrument is badly flawed so it'll lead you, ultimately, in the wrong direction.

My feeling on the whole issue of protein consumption is that it appears to be beneficial to consume a higher than normal protein intake during a period of weight loss to prevent loss of lean body mass. Lifting weights will only augment this effect and will have the more potent effect in terms of allowing one to hang on to as much protein as possible during energy deficit. I know, the scientist finally agrees with the strategy tried and tested by bodybuilders for years.

Using the logic laid out by John in his argument for how much protein you should consume to "optimize" training, along with the flawed N balance approach, one could argue that athletes can simply consume more and more and more protein since it's impossible to define a true ceiling. I suppose to answer the question of optimization I'd define it thusly: a protein intake that allows maximal functioning of all synthetic processes that require amino acids and that prevents excessive oxidation of amino acids and production of urea.

The closest any study has come to measuring something like this is a study by Tarnopolsky et al in 1992 (J. Appl. Physiol. 93(5): 1986, 1995) in which he showed that despite an estimated protein intake to maintain nitrogen balance being somewhere around 1.7 g protein/kg/d, but that the synthetic of all body proteins (including muscle) was maximal at a protein intake of 1.4 g protein/kg/d, with no further increase up to 2.4 g protein/kg/d!

It did appear that the RDA for protein wasn't sufficient to maximize the synthetic rate of body proteins. Hence, faced with these data I'm not sure that one can argue that a protein intake of more than 1.4 g protein/kg/d is necessary or even beneficial for protein requiring process.

Also, at the recent Experimental Biology conference, data from Mike Rennie's laboratory showed for the first time a dose response curve of muscle protein synthesis with oral dose of essential amino acids. They reported that oral dose of essential amino acids (EAA) at which muscle protein synthesis (i.e., the controlling and regulated variable determining muscle protein accretion) was only 10g – a similar dose of EAA is present in 500ml of skim milk!

Also, data from Mike Rennie's lab (Bohe et al. J. Physiol. 532: 575-579, 2001) shows that in the face of available amino acids that muscle protein synthesis is turned off after 1-2 hours. Hence, it doesn't appear that amino acid supply a) has to be large, and b) isn't limiting for the process of synthesizing new muscle proteins.

How about a simple calculation to hammer home my point? Let's say a hypothetical individual, Fred, weighs 90kg and in one year gains 10kg (22lb) of muscle. Not mass– muscle. That's a pretty impressive rate of gain and something most of us would be happy with.

Now, that means over a year Fred has gained 10kg of muscle, which is 75% water, so really he's gained 2.5kg of muscle or protein (i.e., 2500g). That means he's gained: 2500g protein / 90kg / 365d or ~ 0.076 g protein/kg/d. Now, if we assume that the RDA is sufficient to cover all of Fred's other protein needs, which is likely a large overestimate, then what Fred would be required to eat on a daily basis is:

0.8 g protein/kg/d + 0.076 = 0.876 g protein/kg/d to gain 22lb of muscle

Whoa, hold on you say, it takes six times as much protein consumption to lay down the equivalent amount of muscle (estimate from Mr. Colgan and the Colgan Institute–a hotbed of muscle research). In truth, I'm never sure where this number comes from since literature from the animal science area indicates that to lay down protein at rates sufficient to support weight gain in a growing steer–who in 180 days will lay down 226kg of mass, of which 60% is protein–you need only consume the mass gained x 2 in terms of protein content, but lets say it's six. Okay, so now you need to consume:

0.8 g protein/kg/d + 0.456 g protein/kg/d = 1.256 g protein/kg/d to gain 22lb of muscle

Now, what if this person were smart and consumed some extra energy? Well, we know that this spares protein (Garza et al. Am. J. Clin. Nutr. 29: 280-287, 1976). Also, we know that resistance exercise, consistent with its anabolic nature, is conservatory for muscle protein. Hence, resistance exercise, as John freely acknowledges, results in a reduction in protein requirements due to overall protein retention. This is particularly true in skeletal muscle when you haven't eaten (Phillips et al. Am. J. Physiol. 273: E99-E107, 1997).

Now, if all of this true, then my estimates above are even higher than they need to be. So protein requirements to "optimize" muscle mass gains aren't large, certainly no larger than what most lacto-ovo, meat-eating North American males consume on a habitual basis–this was my ultimate point in recent review.

However, as for the benefits John maintains accompany consumption of protein, my response is below (John's assertions in bold, my comments below):

John wrote: "Increased Thermic Effect of Feeding–While all macronutrients require metabolic processing for digestion, absorption, and storage or oxidation, the thermic effect of protein is roughly double that of carbohydrates and fat. Therefore, eating protein is actually thermogenic and can lead to a higher metabolic rate. This means greater fat loss when dieting and less fat gain during overfeeding."

True, and likely a benefit of eating more protein.

John wrote: "Increased Glucagon–Protein consumption increases plasma concentrations of the hormone glucagon. Glucagon is responsible for antagonizing the effects of insulin in adipose tissue, leading to greater fat mobilization. In addition, glucagon also decreases the amounts and activities of the enzymes responsible for making and storing fat in adipose and liver cells. Again, this leads to greater fat loss during dieting and less fat gain during overfeeding."

Effect is negligible. Associated with protein ingestion is actually a pronounced insulin response unless the protein load is enormous. John knows this.

John wrote: "Increased IGF-1 – Protein and amino-acid supplementation has been shown to increase the IGF-1 response to both exercise and feeding. Since IGF-1 is an anabolic hormone that's related to muscle growth, another advantage associated with consuming more protein is more muscle growth when overfeeding and/or muscle sparing when dieting."

What is the true significance of this? Even John would have a hard time finding data to support the thesis that greater circulating IGF-1 after exercise or feeding results in greater lean mass gains. (We have data showing that even with no change in IGF-1 you can get hypertrophy – Journal of Applied Physiology, in review.) This is hyperbole at best!

John wrote: "Reduction in Cardiovascular Risk–Several studies have shown that increasing the percentage of protein in the diet (from 11% to 23%) while decreasing the percentage of carbohydrate (from 63% to 48%) lowers LDL cholesterol and triglyceride concentrations with concomitant increases in HDL cholesterol concentrations."

True. Losing weight also reduces cardiovascular disease. Also, what about elevated homocysteine, an independent risk factor for cardiovascular disease, on high protein diets?

John wrote: "Improved Weight-Loss Profile–Brand spankin' new research by Layman and colleagues has demonstrated that reducing the carbohydrate ratio from 3.5:1 to 1.4:1 increases body fat loss, spares muscle mass, reduces triglyceride concentrations, improves satiety, and improves blood glucose management."

In the Layman study overweight women consumed either 68g protein/d (0.8 g/kg/d) and 239g CHO/d (2.8g/kg/d) or 125g protein/d (1.45 g protein/kg/d) and 171g CHO/d (2 g CHO/kg/d). The higher protein group saw the effects outlined. However, what I might point out is that most, if not all, lacto-, ovo-, meat-consuming North American males are getting that amount of protein anyway.

Hence, it would be foolhardy, based on these data, to conclude that eating more than this would have more benefit. Besides, if you're not losing weight but rather gaining weight, what's the effect of consuming extra protein? Do blood lipid levels go up? What about homocysteine? The Layman study does nothing to address these questions, which are very real possibilities.

John wrote: "Increased Protein Turnover–As I've discussed before in my article, 'Precision Nutrition For 2002 And Beyond,' all tissues of the body, including muscle, go through a regular program of turnover. Since the balance between protein breakdown and protein synthesis governs muscle protein turnover, you need to increase your protein turnover rates in order to best improve your muscle quality. A high protein diet does just this. By increasing both protein synthesis and protein breakdown, a high protein diet helps you get rid of the old muscle more quickly and build up new, more functional muscle to take its place."

This effect has never been shown in muscle. For whole body protein turnover, a study by

Tarnopolsky (Journal of Applied Physiology 73(5): 1986-1995, '92) shows that whole body protein turnover increases with increasing protein intake, but I suspect that this is due to more rapidly turning over tissues – gut and liver and not muscle.

In any event, it can't be stated as fact that higher protein intakes help "get rid of the old muscle more quickly and build up new, more functional muscle to take its place." This conclusion is simply unsupported by any real evidence.

John wrote: "Increased Nitrogen Status–Earlier I indicated that a positive nitrogen status means that more protein is entering the body than is leaving the body. High protein diets cause a strong positive protein status and when this increased protein availability is coupled with an exercise program that increases the body's anabolic efficiency, the growth process may be accelerated."

I'm not sure what increased N status means, but I think it means increased nitrogen balance–which is true, if you consume more protein you go into higher and higher nitrogen balance. Herein lies the greatest problem with the studies carried out by Lemon and Tarnopolsky: that is, reliance on nitrogen balance to determine protein needs.

Take for example the data from Tarnopolsky et al (J.Appl.Physiol. 68(1): 187-193, 1988), which is constantly cited and even Lemon et al (J.Appl.Physiol. 73(2): 767-775, 1992), another citation classic. Look at the nitrogen balances that these subjects are in when consuming intakes of protein of 2.4-2.5 g protein/kg/d – the mean in Tarnopolsky's '88 paper is 14g of nitrogen/d and in Lemon's paper at 1.6-1.7 g/kg/d, the Nbal is 9g N/d.

Now, if protein is 16% by mass nitrogen then 14g/d means that these subjects were gaining 87.5g protein/d–Tarnopolsky '88. Or according to Lemon, 56.25 g protein/d. Hence, by these calculations, these subjects, if they kept consuming these protein intakes, should gain 365d x 87.5g/d = 31.9kg of protein (i.e., muscle, unless your gut, liver, or skin are growing, which seems unlikely) per year. I think we can all agree this isn't possible even with good lifting, good nutrition and even steroids.

It's a little more reasonable in the Lemon study but still means the subjects would've gained 365d x 56g/d = 20.4kg/year. Quite simply, nitrogen balance can't be used as a good method to understand protein requirements. This has been recognized for a while now–Hegsted Am. J. Clin. Nutr. 21(5): 352-357, 1968.

John wrote: "Increased Provision of Auxiliary Nutrients – Although the benefits mentioned above have related specifically to protein and amino acids, it's important to recognize that we don't just eat protein and amino acids – we eat food. Therefore, high protein diets often provide auxiliary nutrients that could enhance performance and/or muscle growth. These nutrients include creatine, branched chain amino acids, conjugated linoleic acids, and/or additional nutrients that are important but remain to be discovered. This illustrates the need to get most of your protein from food, rather than supplements alone."

True enough. All are possibilities. In the final analysis, requirement or optimization, the argument is moot. Available evidence (Tarnopolsky Journal of Applied Physiology 73(5): 1986-1995, '92) as well as hypothetical calculations aren't consistent with the tremendously elevated protein needs John says must exist. Optimization can't, using nitrogen balance, be defined since nitrogen balance consistently (and falsely) shows a positive Nbal with increasing protein intakes.

Now don't get me wrong, I'm not saying to consume the RDA for protein – and that wasn't my message in my review–but supplementation with protein (above and beyond a good diet) appears unnecessary. Timing may be marginally important, but this is a comparatively minor part of the equation. In the end, protein supply is rarely limiting!

Accumulating evidence does suggest that during periods of weight loss, to prevent lean mass losses, higher protein intakes would help reduced protein loss – how high hasn't been determined, however. I'm sticking to my concluding statement in my review, since I see no true evidence to support any other recommendation otherwise; namely, that 12-15% of ones energy should come from protein.

If you're an 80kg male who's lifting 6d/week (1-2h/d) then you likely require something like 3800kcal/d, then that means that 12% = 114g protein or 1.4 g protein/kg/d, which incidentally has been shown in hard-training individuals to maximize the synthetic rates of all proteins in your body (Tarnopolsky Journal of Applied Physiology 73(5): 1986-1995, '92).

Dr. Berardi's Response to Dr. Phillips's Comments:

I appreciate you taking the time to construct an amazingly thorough and fair response to what can be, at times, a volatile debate. Your respectful demeanor is just what these types of exchanges need in order to stay on track and really make a meaningful contribution.

Sometimes these debates can degrade rapidly to slanderous mamma-slam fests. Hopefully, when faced with "opponents" who wish to take things to that level, those reading this exchange can follow your example, sticking to the facts and debating in a reasonable and rational way.

That said, let's roll up our sleeves and get down to the business at hand – discussing protein intake. And, since it's really difficult to discuss one macronutrient in isolation of the others, especially when our end is applied nutrition, we should probably touch on overall dietary strategy a bit also.

To begin with, though, I need to bring up a side issue that should be clarified before moving on. The "Protein Prejudice" article that was referenced in your response was actually published back in March of 2003, nearly two years ago. Why make this point? Well, in the two years that have passed, certainly new data have emerged that warrant discussion (some of which you referenced in your letter, yet I wasn't privy to at the time that the article was written).

This brings up a very important lesson for all of those reading this exchange. Many reading this are very interested in the applied side of this field and may want to publish articles in print or electronic formats. Here's a caution directed at them: publish something on the Web today and regardless of how you feel about what you wrote ten years from now, you'll have to live with it. Now, in this specific case, the "Protein Prejudice" article, I don't regret anything and I still stand behind my comments two years later.

Okay, that digression aside, one more note with respect to the original article. This was published on a site targeted to lay readers without the scientific training we have and are more interested in, if you will, applied nutrition. As a result, the article wasn't intended to be an exhaustive review of the literature. Therefore, comparing one of your more recent literature reviews to one of my short, two year old, magazine articles isn't exactly an apples to apples comparison. So I'm glad we're getting this chance to speak directly!

Now, let's get down to your comments. It should come as no surprise to anyone who's read many of my articles that I'm in full agreement with nearly everything Dr. Phillips has commented on. Therefore, for the blood thirsty, you won't get your blood here. However, I do believe that while some of the scientific interpretations of the muscle literature may be similar, there are a few outstanding issues that demand discussion. I'll address them below.

Dr. Phillips wrote: "I have read Dr. Berardi's paper and have found it generally sound, but lacking in sufficient evidence to truly be considered 'truth' in some areas..."

I agree. In fact, much of my "lay" writing is more speculative than most scientists would feel comfortable with. If I were writing a literature review, certainly, there would be leaps that I simply couldn't and wouldn't take.

But in this type of forum, where I'm writing to an "applied" crowd, I feel that I do have free rein to take the literature that's available and make a few speculative leaps at some things beyond what the literature tells us explicitly. In this sense, without the ability of science to demonstrate verisimilitude, my speculations are just that – my best guesses based on what we do know in the literature and what I see in practice.

Any scientist worth his or her own salt would agree that we're barely scratching the surface of what's yet to be known, explored and discovered. Yet we still have to make decisions every day, based on incomplete evidence. And that's where much of my writing leads me.

I look to the scientific literature but since, in parallel to my scientific training, I run a successful coaching/consulting practice – one in which I attempt to rigorously control what's possible to be controlled in free living clients and document what I see – I also look to the "clinical" results that I see every day. Finding this balance is difficult but that's my mission and the express mission of my company, Science Link.

Our tag line is this: "However Beautiful the Strategy, You Should Occasionally Look at the Results." By striving to find the balance between research and results, applied and basic, clinical and laboratory, I know that I can find support for ideas that work and drive new ideas in the process. This is what I get excited about!

Of course, it's critical to be true to the rigorous data we've got:

"Science is merely an extremely powerful method of windowing what's true from what feels good; without the error-correcting machinery of science we are lost to our subjectivity..."

But science alone doesn't always get it right–nor is it always there with an easy answer to "what's true" – nor is it capable of supporting or refuting many hypotheses that we continually test in our own daily lives. So it's my impression that the dialogue between "clinical" and "laboratory" observations is critical in making recommendations/decision for daily life (including nutrition and exercise research).

Heck, this is why I wanted to open up this conversation. Dr. Phillips might be one of the smartest guys around in our field and his input on this question is invaluable. I think I'm a pretty smart cookie too. What a cool chance to really break new ground, get a bunch of new ideas stimulated, and open up a running dialogue that blends clinical and laboratory.

So even if my article isn't true in the scientifically verifiable sense of the word true, I'm just happy that it might stimulate enough discussion so that we can better get after the "truth."

Dr. Phillips wrote: "Hence, trying to talk about a requirement or even optimizing protein intake based on finding from nitrogen balance studies is like trying to circumnavigate the globe using a map before we knew the earth wasn't flat – the instrument is badly flawed so it'll lead you, ultimately, in the wrong direction."

This is a fantastic analogy. I agree that the N balance studies are flawed, but unfortunately until you start publishing your more recent data, Dr. Phillips, they represent the bulk of what's out there and available for review (especially two years ago when I wrote the article).

This area is poorly researched. That's due to a lack of governmental funding for areas considered more recreational than of major public health concern. So think of us as miners with small head lamps in a dark cavern that's not been well explored. We can make guesses as to what's ahead of us in the cavern. These guesses are based on the small amount of light we can see with and the small amount that we're seeing.

So I remain open to the suggestion that one day someone will find a way to rip the roof off that cavern and I might feel very, very silly about my guesses when it's exposed in its entirety. But I don't believe that I can be faulted for reporting on what I see, what I guess, and how I interpret both.

Dr. Phillips wrote: "Also, data from Mike Rennie's lab shows that, in the face of available amino acids, muscle protein synthesis is turned off after 1-2 hours."

Now, although very interesting, these data aren't a good justification for a lower protein requirement/optimization point. In fact, I distinctly remember bringing these data up at ACSM at a session Dr. Phillips was moderating. I believe the consensus was that it would be nearly impossible to duplicate (with the diet) the "steady rate" amino acid infusion program of this study. In essence, when you eat protein you get peaks and troughs and it's probably the change in amino acid concentrations that stimulate the protein synthetic mechanisms, not a threshold.

Now, Rennie's more current data may suggest otherwise; I can't wait to read it in print. But as of our last conversation, it was Dr. Phillips's opinion that amino acid changes in the blood are what govern synthesis, not having an amino acid concentration above a certain threshold. This means that each time we eat more protein we'd get a bump in PS. So I don't really see how this comment is relevant – interesting, yes – relevant, I'm not sure.

Now, concerning what you wrote about "Fred," of course, these are all guesses – but the point is well taken. However, just how meaningful are they? Probably about the same as suggesting that burning an extra 20kcal/day by gardening would lead to an increase in monthly caloric expenditure by 600kcal and in yearly expenditure by 6000kcal. This surely means the person would lose just short of two pounds per year from gardening for three minutes a day!

Forgive the absurd example, but the point is that the body makes constant adjustments, adjustments designed at the maintenance of homeostasis in the face of such subtle perturbations. As a result, with all the "slop" room in the calculation above, it's really hard to say what exact amount of "extra" protein is needed to pack on more muscle mass. I do get your point that according to even very liberal assumptions, it doesn't seem like much, and perhaps athletes don't need much more protein for enhanced protein accretion. But let's not forget the big picture here (below).

Dr. Phillips wrote: "As far as optimization, this is a far more difficult question and even John knows that this hasn't been answered satisfactorily, or at least not to a degree that would make me offer the advice that lifters should 'eat more protein – a lot more.'"

Again, I agree completely that "optimal" is a loaded word and nearly impossible to define. One man might define "optimal protein intake" as a protein intake that allows maximal functioning of all synthetic processes that require amino acids and that prevents excessive oxidation of amino acids and production of urea (Dr. Phillips's choice). And another might define "optimal protein intake" as the protein intake that, when combined with sport-specific physical training, produces ideal body size and composition for that particular sport (which I would choose when working with an athlete or writing an article with athletes as a target audience).

But regardless of the choice of the "optimal" definition, both are pretty damn hard to test and with the current literature there's very little evidence as to what is the optimal intake for either definition. Of course, there's some evidence, but as Dr. Phillips rightfully pointed out, it's based on some old N balance data that's probably not valid anyway. So we're back to square one. Not much evidence.

So what do we do in these circumstances? Dr. Phillips gets back to the lab and begins measuring this stuff with some cool measurement techniques he's got at his disposal. Me, since I don't have my own lab yet, I wait for Dr. Phillips to publish his work. But, in the meantime, I experiment with my athletes.

1) Improvements in body composition

Now, I concede that protein intakes above what the typical North American lacto-ovo, meat eating male gets are probably not going to pack on additional muscle mass directly from the additional amino acid load. However, as stated in the article, perhaps the other components of the protein foods are making an impact with respect to muscle mass.

But even more importantly, muscle mass isn't the only endpoint I choose when measuring body composition. For most performance (aerobic or anaerobic) and physique athletes, the relationship between fat mass and lean mass is critical – not just the absolute amount of muscle mass. Consider this: some of the athletes with the highest absolute amounts of lean mass on the earth are sumo wrestlers. Need I say more?

If I have an athlete that needs to continue to train hard and perform at a high level yet needs to lose body mass/body fat to achieve the right body comp for his sport, would I alter his macronutrient recommendations or would I keep them the same as I would with an athlete who has no fat to lose? It's a toughie...

Of course, we don't want to put this athlete on a calorie restricted diet (or, at least, in a large calorie deficit) as that might compromise performance ability and/or recovery. So how do we get the fat off?

Here's what I'd do. My solution (and it's worked time and time again) is to replace some carbohydrates from the diet with protein. The thermic effect of the protein may allow for more energy expenditure while still taking in a large amount of total energy and, importantly, micronutrients. Perhaps the balance of glucagon to insulin may alter nutrient partitioning. Or there might be something else at work (perhaps something on the neurochemical/neurohormonal level) that we don't even have enough evidence to speculate on.

Now, let me give a specific example of how this manipulation might be carried out and what results one might see. I have a cross country skier (national level) who went from 165 lbs to 135 lbs (with no net loss in lean body mass; measured via Bod Pod) in three months without appreciably altering workout volume or intensity by increasing calorie intake by 1000kcal/day (via weighed food records with the staff nutritionist), but going from a 70% carbohydrate, 15% protein, 15% fat diet to a diet containing 40% carbohydrate, 30% protein, and 30% fat (based on food selections I recommended).

Food selection and timing improved, of course, with my recommendations so that's another variable. But, in the end, the major change was an increase in kcal and an increase in protein at the expense of carbs. (Again, I concede there could be other factors at work, error in dietary reporting, etc., but these errors would also be present in any chronic training/nutrition study and therefore shouldn't create a knee-jerk dismissal mentality. These data are interesting and need to be considered!)

To a similar end, I also have examples of athletes maintaining body mass (while losing fat mass and gaining lean mass) after an increase in protein intake at the expense of some carbohydrate. Importantly, this isn't a low carb diet. These are alterations that simply reduce carb intake percentages from the 60s and 70s to the 40s and 50s while often increasing energy intake.

So my point is that for most athletes, body composition improvement is often the goal – not absolute muscle mass. And from what I've seen, increased protein intakes (with the appropriate dietary shifts) are much more effective at achieving the alterations most athletes need – less fat plus preserved muscle, or less fat plus more muscle while still training hard and recovering appropriately.

Perhaps the best paper demonstrating the potential for an effect here is the Forslund paper (Effect of protein intake and physical activity on 24-h pattern and rate of macronutrient utilization. Am J Physiol. 1999 May;276(5 Pt 1):E964-76.) They showed increased protein intakes (at the expense of carbohydrate) lead to a negative fat balance and a positive protein and carbohydrate balance vs. "normal" protein intakes.

Surely this gives some indication that an increased protein intake at the expense of carbohydrate can lead, at least, to acute shifts in macronutrient balance, the cumulative effect of such may be altered body composition via nutrient partitioning.

2) Improvements in recovery and the ability to handle higher training densities

Now, like "optimal," the word "recovery" can be a nebulous term. So let me clarify. Using both subjective and objective markers of recovery (POMS, resting heart rate, heart rate variability, quality of sleep indices, and daily performance), I've found that my athletes can train with more density when the typical sports nutrition recommendations of 60-70% carbohydrate, 15% protein, and 25-35% fat are altered in favor of an increased percent protein intake and reduced percent carb intake as mentioned above (as long as rapid post-exercise carbohydrate replenishment strategies are utilized in order to ensure quick resynthesis of muscle glycogen during the most "glucose tolerant" part of the day – post-exercise).

Of course, I often see large changes in body mass/composition with these shifts so perhaps it's not necessarily the protein increase or dietary shift that's creating the effect but the simple loss of "nonfunctional baggage" as I call it (i.e. body fat). Again, interesting reports none the less.

Now, a clarification has to be made in response to Dr. Phillips's comment "...but then since they will never have tried doing anything else, do you get stronger, leaner etc. because of, or in spite of what you do."

While some coaches may fall victim to this, with each athlete I attempt rigorous reporting and outcome-based experimentation. I don't put all athletes on the similar diet; instead I use the first month or two as a trial period to see which strategies/changes produce the best results and then I do more of what has seemed to work.

With that said, these two benefits are the basis of my recommendation for "more protein – a lot more." But let's remove the "a lot" ambiguity and put some numbers on it. After all, an 80kg athlete eating 4000 kcal/day and taking in 30% of those calories from protein would be getting about 300g protein/day (3.75g/kg), and this is a lot more than 0.8 or even 1.4g/kg.

Interestingly, even at your recommendation of 15%, a 65kg Tour de France cyclist, requiring in excess of 7,000 to 8,000 kcal would be getting 300g protein/day or a whopping (4 - 4.6 g/kg). That, to me, is a lot more than the 0.8g/kg recommendation or even the 1.4g/kg suggestion by many sports nutritionists.

But mathematics aside, my point remains firm – myself and many of those I teach through my seminars and writing report very similar effects with this type of dietary shift. I have some guesses as to why these things occur but I have no firm scientific support on all of these benefits.

However, it's important to note that there aren't any firm scientific objections either! And this is critical. If I'm doing things that fly in the face of scientific evidence, I'm willing to reappraise and throw out the chaff. However, in the absence of both scientific approval and rejection, there's no cause to suggest that my methods are wrong – simply that they might not be completely scientifically verified or even verifiable as of yet.

Would I love the support, yes. It would silence some of the critics (although they'll still yap away; some people just want to criticize). However, I don't really need it. The best test of my consulting efforts is this – success with clients and repeat business.

Dr. Phillips wrote: "Hence, it doesn't appear that amino acid supply a) has to be large and b) isn't limiting for the process of synthesizing new muscle proteins."

Agreed, but again, I'm not merely interested in the synthesis of new muscle proteins alone. It's the balance between fat mass and lean mass I'm most concerned with. And here's a major problem with the lower protein recommendations. When you have an athlete overfeeding to gain mass or simply eating a lot of total dietary energy to support energy expenditure, the current recommendation is to shove carbs down their throats! At 4500kcal/day and 65% carbohydrate, we're talking over 730g carbohydrate per day. Now, while carbohydrate energy does impact athletic performance and does spare protein, could there be too much of a good thing?

Getting 730g carbohydrate in the typical 4-5 meals an athlete might eat/drink means huge carbohydrate loads that, with subsequent huge releases of insulin, could lead to reductions in fat mobilization and oxidation throughout much of the day – especially if this chronic high insulin dump leads to neutralization of the improved insulin sensitivity most athletes enjoy. Now, you said the following: "Associated with protein ingestion is actually a pronounced insulin response unless the protein load is enormous. John knows this."

Certainly not more than the associated insulin response accompanying a huge carbohydrate load. At least with the protein load, some counter-regulatory glucagon is released. So my question to you is this: If you had to overfeed an athlete (or simply have an athlete ingest a very large daily energy intake) in an attempt to maximize the lean mass to fat mass ratio and you had to choose a macronutrient to "overeat," what would it be and why?

(Keeping in mind "overeat" might not be well defined in this context – another way of saying it might be "choose a macronutrient to eat more of.") This question assumes that they're getting adequate daily energy intake. That's a big assumption for most recreational athletes and some elite athletes, but not much of one for the athletes I work with directly.

As for your remarks about higher protein consumption being beneficial during weight loss, I agree 100%. During periods of weight loss, there's no question or really any reason for equivocation – the research is clear.

Dr. Phillips wrote: "Also, what about elevated homocysteine, an independent risk factor for cardiovascular disease, on high protein diets?"

Is this a legitimate concern or simply a secondary or tertiary debate point? Admittedly, I'm not an expert on homocysteine and heart disease so forgive any ignorant comments. I'm assuming that the speculation here is that high dietary methionine would lead to higher serum homocysteine concentrations as a result of methionine's demethylation via the methyltetrahydrofolate reductase enzyme pathway?

If so, recent literature shows that a) dietary methionine increases homocysteine to a lesser extent than free methionine and b) the dietary cysteine and serine attenuate any increases in homocysteine as a result of methionine intake. (Verhoef – Dietary serine and cystine attenuate the homocysteine-raising effect of dietary methionine: a randomized crossover trial in humans. Am J Clin Nutr. 2004 Sep;80(3):674-9.)

Furthermore, if this is the only potentially serious health consequence of a higher protein diet (and increased homocysteine isn't universally regarded as a "real risk"), and the ingestion of cysteine and serine (present in high protein diets also) doesn't take care of the homocysteine, daily folic acid supplementation should take good care of that (and would offer other benefits) as discussed in: Treatment of coronary heart disease with folic acid: Is there a future? Am J Physiol Heart Circ Physiol. 2004 Jul;287(1):H1-7. Review.

Here's the conclusion from that paper:

The view that a raised plasma tHcy level is causal in the development of vascular disease is an attractive hypothesis if only because folic acid offers an easy, inexpensive, and generally safe means of lowering it. This review challenges the hypothesis that tHcy is causal and raises the possibility that an increased tHcy is an epiphenomenon.

Moreover, there's evidence that the beneficial vascular effects with folic acid are only achieved in pharmacological doses. Low-dose folic acid will reduce plasma tHcy, but a high dose may be required to produce the beneficial effects on vascular function, which occur before and apparently independently of homocysteine lowering.

The current clinical trials are on the whole, designed to test the homocysteine hypothesis of vascular disease using relatively low doses of folic acid. While these trials will undoubtedly show that folic acid lowers tHcy effectively, it's unlikely that the expected reduction in cardiovascular events will be seen. However, it's important therefore not to discount treatment with folic acid if these trials are negative, because it's possible that high-dose folic acid may have a beneficial effect on outcome via mechanisms independent of homocysteine lowering. Elucidation of these mechanisms is important in the drive to develop effective treatments for prevention of CHD.

So is this an actual concern?

Dr. Phillips wrote: "In the final analysis – requirement or optimization – the argument is moot. Available evidence as well as hypothetical calculations aren't consistent with the tremendously elevated protein needs John says must exist."

Yes and no. For protein status or positive protein "balance" – perhaps. But isn't focusing on this one outcome variable a bit too isolationist for you as an inquisitive individual? After all, isn't there so much more to nutrition than focusing on one physiological outcome – protein turnover? Aren't athletes (even just physique athletes) after something much more than just an improvement in protein synthesis?

As mentioned earlier, athletes are after the balance between lean mass and fat mass. They're after optimal biochemical conditions for performance. And they're after recovery of every system, from muscular to neural to immune. In my estimation, focusing on the protein status response to protein intake alone is committing an error, perhaps the same error many sports nutrition people are making with their chronic focus on high carb diets.

Think about it, high carb diets evolved to be the recommendation for athletes because of the original studies showing that high carb diets lead to high muscle glycogen. And high muscle glycogen seemed to correlate with good performance (in endurance events, no less). So the sports nutrition community has based its entire recommendations on one thing – muscle glycogen. What about all the other stuff that supports training, immune function, etc.?

Now, with this as an analogy, are we doing the same thing if we stick to the argument that since aminos aren't limiting one potential body comp outcome – protein turnover – that we can ignore protein from there on out?

From this perspective, I'm not quite sure either of us can jump to any conclusions with firm protein intake recommendations for athletes until we answer a few important questions, such as:

1) Could altered intakes of protein impact immune function?

2) Could they impact body composition (i.e. the all important balance between fat mass and muscle mass, not just muscle mass in isolation)?

3) Could they impact the neurochemical/neurotransmitter balance in the body?

4) Could they impact hormones that lead to body comp changes?

5) Could they lead to alterations in physiological set points for cellular function?

6) Could they alter CNS function?

7) Could they offer other nutrients that support body comp changes?

8) Might individual differences govern the responses to each of the seven aforementioned questions?

Until these questions are answered, each of our conclusions about what's best for athletes, from a dietary perspective, are based only on a very small glimpse of what might be going on when we alter macronutrient intakes. So I guess this is where we reach the impasse.

Dr. Phillips wrote: "I'm sticking to my concluding statement in my review, since I see no true evidence to support any other recommendation otherwise; namely, that 12-15% of ones energy should come from protein."

I agree that the literature has little to offer in the way of comprehensive dietary suggestions for athletes seeking to find their best competitive body composition, immune function, CNS balance, etc. However, sometimes we have to operate in the void where literature is lacking. To operate within that void, I've done some pretty cool, albeit loosely controlled experiments on small sample sizes to come up with my own conclusions. And in this realm, I'm sticking to my recommendations – more protein.

Of course, this statement "more protein" is a gross oversimplification. Caveats include, macros can't be viewed in a vacuum; "more protein" must be matched by appropriate shifts in dietary carbohydrate and fat, and these suggestions may have behavioral implications in free living humans that can't be controlled.

(Of course, my original article didn't touch on much of this and that was, admittedly, a weakness, but such is the article format. If you tried to cover it all, it wouldn't be an article, it would be a book. But if you kept telling the story it wouldn't be a book any longer, but a volume of books. And so on. So you've gotta stop somewhere, leaving stuff out along the way, based on your judgment.)

One more caveat, Dr. Phillips. I'm sure that we can agree to disagree, but let's do so officially at ACSM where, regardless of its low protein content (usually 0.3 – 0.5%), you'd better let me buy you a beer! Thanks for the discussion and the forum to share these types of ideas.

One More Response from Dr. Phillips:

I do have to admit that there's a metabolic advantage to consumption of higher than normal protein: provision of muscle building blocks, BCAA, establishes good hormonal environment, better macronutrient mix for lean/fat mass gain, and Layman's work shows improved loss profile, too.

I suppose what it comes down to is where exactly this point is: 1g/kg/d, 1.5g/kg/d, 2g/kg/d, more? Surely at some point, the protein simply can't be shoved into the system and is oxidized and the N is turned into urea. Since N (i.e., ammonia) is toxic to most species then our enzymatic machinery would be elevated to deal with the excess protein/N.

Hence, once you've started eating "high" protein you'd better continue doing it, because suddenly downshifting to lower protein would result in degradation of that protein – higher peaks of protein synthesis maybe, but lower valleys when not eating or when consuming lower than your body is set up to process (this is theoretical but does have some experimental support).

Eventual consequences of high protein – Homosyteine (not likely too big of a deal), kidney (no), bone loss (no, in fact the opposite is likely true), association with fat (maybe, if protein is purely from animal sources), expense (it's all relative, but potentially yes, if highly purified supplemental forms are consumed). Anything else, I don't think so.

Your points, immunity (maybe), neurotransmitters (I don't know of any evidence), CNS function (likely relates to the previous, but again I don't know), there are many other theoretical/potential possibilities, but they've yet to be proven. I actually had a really good chat with Don Layman last week in Houston where he and I were giving talks to the International Livestock Congress (beef protein). He presented some terrific data and agreed that likely 1.4-1.5g/kg/d is a good intake when you're losing weight (in women). Men are likely similar. As for gaining weight, your points are well taken and many I can't argue with since no good data exist to support either.

If there's one area I think you and I (and Dr. Ivy also at University of Texas) can agree on, it's that immediate post-exercise nutrition has a very large impact on gains – we're just about to submit some very solid data on this issue. This isn't to say that later on isn't important, but the immediate 1-2 hours after exercise is important, particularly for resistance training athletes.

One More Reply from Dr. Berardi

I agree with your comments above on protein. Therefore, in the end, it appears that the only real negative to eating a lot of protein is the fact that you'd need to keep intake high always (unless you know when you'll want it low and ramp down slowly).

I've had to do this with my athletes for sure, especially athletes like cross country skiers who travel out to a glacier to train and live for a few weeks. Altitude, sleeping in tents, skiing all day, not much in the way of nutritional volume... gotta find some anti-catabolic strategies for them.

So really it just becomes a preference thing. However, all else being equal, with all the extra metabolic processing needed for protein, you do drive up energy expenditure without much risk, something that many athletes could benefit from (body comp wise).

What's Next?

After completing the debate, Drs. Berardi and Phillips shook hands and walked back to their neutral corners content with a great discussion. In fact, rumor has it they're even considering collaborating on a few research projects to get at the heart of some of these protein questions. See what can come of a healthy discussion!