Chapters:
00:00 Intro
00:11 Our findings
02:05 New meta-analysis: Protein benefits continue up to 3.1 g/kg per day?
02:30 1: The researchers did not control for energy intake and nutrient timing
05:21 2: The high protein intake recommendations are based on a handful of studies
06:23 3: The researchers did not present their actual data
06:56 What does the model say?
09:18 Muscle mass or lean body mass?
10:26 Conclusion
12:30 Outro
Transcript:
How much protein do you need to consume per day to maximize muscle growth and strength development? To answer this question, in 2017. I collaborated on a meta-analysis of the available literature on this topic. We found that a protein intake of 1.6g/kg maximized fat free mass gains. There were no further benefits of higher protein intakes. For my Freedom Unit loving friends, that’s 1.7g/lb. That’s total protein intake for total body weight. Our results were closely replicated by a 2022. meta-analysis finding that strength gains were maximized by a protein intake of 1.5g/kg, just 0.1g/kg difference with our results. I also collaborated on a study to see if there were any improvements in recovery of higher protein intakes. We compared protein intakes of 1.8 and 2.9g/kg per day. We found no differences between these protein intakes for muscle soreness, muscle damage or strength recovery.
All of these results converge with a scientific consensus that has been forming since the 80s. Research, especially by Tarnopolsky et al. in strength athletes and bodybuilders found that strength development, muscle growth, body composition, nitrogen balance, and muscle protein synthesis are all maximized by protein intakes as low as 0.7g/lb, so 1.6g/kg per day. In my Personal Training Course we’ve collected all the available studies on all the major topics in exercise and nutrition sciences. We have found 45 studies comparing different protein intakes that controlled for meal frequency and were generally comparable. Of these 45 studies none found benefits of more than 1.6g/kg per day, so 0.7g/lb, and exactly in line with our meta-analysis and the subsequent meta- analysis on strength development. While some bodybuilders and especially supplement companies will continue to tout benefits of protein intakes up to 1 g/lb or even higher intakes, there has been a relative consensus in evidence based fitness. And for many years there was peace. But recently the piece has been disturbed. A new meta-analysis is making the rounds, showing that, supposedly, the benefits of protein go up to 3.1g/kg. Almost twice as much as previously believed. So is it time for a paradigm shift, or are people crying wolf?
I think the reason that this meta analysis was not previously given much attention, because it simply doesn’t deserve much attention for multiple reasons. The first reason is that in this meta-analysis essentially all available studies comparing different protein intakes were compared. The nice thing is that you have a lot of data. The bad thing is that these data are confounded by numerous factors. The analysis did not control for energy intake, nutrient timing, meal frequency, or pretty much any factor. So it’s just a matter of lumping all available data together and seeing if higher protein intakes beats lower protein intakes. The problem is that you don’t know if it’s the higher protein intake or the fact that they have an extra meal among their workouts, or because they simply had higher energy intakes. Energy intake is a particularly big confounder because they had a lot of weight loss studies in there, even a lot of aggressive weight loss studies. As a result of that, presumably. protein intake in the control group actually decreased significantly. So we’re not just looking at what happens when you increase your protein intake, we’re looking at a context of what happens when you decrease your protein intake versus what happens when you increase your protein intake. Not controlling for energy intake is particularly problematic because in some of the studies, they actually co-administered carbohydrates with the protein. So again, we don’t know if it’s the additional carbohydrates, the energy intake in general or the additional protein that made a difference for their lean body mass gains.
To quote the authors: “Only 4 of the 105 articles included in this meta-analysis examined the dose response relationship under identical conditions. So basically, only 4 of the studies were actually really good comparisons, and the other 101 studies were confounded by at least one other factor. Now, the offers did make some attempt to control for the confounding effect of energy intake. Namely, they controlled for the effect of weight change. However, just throwing weight change as a covariate in a regression model an extremely rudimentary way to control for the effect of energy intake. For example, if you have a study that finds no benefits of protein intakes for lean body mass, but a lot of fat loss, it will register as a very positive effect on lean body mass due to the confounding effect of weight, because normally you would expect weight change to correlate with lean body mass change. So what the model essentially will do is penalize for weight change, but that means it will reward weight loss. And because a lot of the studies were weight loss studies, what you can get is a study in which protein intake actually has no effect on lean body mass, but because it has an effect on weight loss, maybe because of the satiating effect of higher protein intakes, you can actually get a study where protein intake did not increase lean body mass objectively, but the model will count it as if it did because there was greater fat loss in that study. And conversely, you can have a study where protein intake actually results in a lot of lean body mass gain, pure lean body mass gains, but that will not register because it will penalize for weight change and assume that some of the weight change was not lean body mass. So you get a lot of these confounding effects when you try to control for weight change and use that as a proxy to control for energy intake.
It’s much better to just compare studies that were actually equal and matched in calories, and then see if there is an effect of protein intake independent of the effect of energy intake. The second major limitation is that these super high protein intakes that are now commonly recommended by some influencers are based on a handful of studies. The graph goes up to 3.1 g/kg in terms of results for strength trainees, but the actual data in many cases don’t. The highest assigned protein intake, so it doesn’t mean you actually consumed it, was 240g, with only one other study group actually having a protein intake assigned above 200g. So if we extrapolate these results to individuals with higher body weights like me, I’m 200 pounds, then you would get way, way higher protein intakes. I would have to consume like 300g protein based on these results. You have to be very careful extrapolating these results of mostly individuals with lower body weights to strength trainees that are a lot more muscular.
If you actually look at the nutrition surveys the highest reported protein intake, so this is what they actually reported in there, like, self-reported data was only 161g. Moreover, of the handful of studies that actually had these high protein intakes, whether assigned or actually consumed, which we don’t really know, 2 of them had study durations, just 2 to 4 weeks long. The third reason I’m incredibly skeptical of these results is that the researchers did not present their actual data. Yes. They actually did not present their data. The graph you see is just the theoretical graph of the model. It’s what the model expects the relationship to be between protein intake and lean body mass, or fat free mass. Any good study should include the actual effect sizes of the data so you can see if it makes sense and the lines actually follow through the effect sizes. Not including the effect sizes and just showing the final output of the model is a lot like saying: “Just trust me, bro”. To illustrate what this model says and why I think you should be skeptical of it, here are a few things that this model says.
For 1: protein intake had a greater effect on the lean body mass gains of not exercising individuals than of exercising individuals. That, of course, doesn’t make any sense. We know from a lot of research and theory that protein intake is more important for people that lift.
Second: A completely ludicrous result is that in this analysis untrained individuals supposedly benefited from protein intakes as high as 3.1 g/kg per day. Not just trained, but also untrained individuals. Now, you could maybe make the case for trained individuals, but there’s just no way in hell that this is true for untrained individuals, people that don’t even lift. And the model indeed found that when you controlled for weight change, the effect on untrained individuals dropped to 1.3 g/kg per day. This was called an inflection point. Afterwards, there was still a significant effect in the analysis, like a significantly positive relationship for trained individuals, for people that lift, but they did not do an actual breakpoint analysis. So all the analysis essentially shows is that for trained individuals there are benefits above 1.3 g/kg per day. We know that this is true. The benefits go up to 1.5 or 1.6, depending on which other meta-analysis you look at. Inherently, this meta-analysis actually doesn’t support that you have to have protein intakes up to 3.1 g/kg per day. Strictly speaking, the analysis just shows that there are positive effects or that at least in the literature, there is a positive association between higher protein intakes and greater lean body mass gains of protein intakes higher than 1.3 g/kg per day. And I agree with this. But what we would like to know is a break point analysis, which is what we did in our meta-analysis, which tests when the relationship essentially becomes statistically insignificant. And we found that this was at 1.6 g/kg per day.
Another clearly nonsensical outcome of the model was that the relationship between the amount of supplemental protein and lean body mass gains made no sense. Lean body mass gains, as you can see in the graph here, actually went down a little bit up to supplemental protein intakes, up to about 0.5 g/kg per day. Only after that point they increase. So this would say that if you supplement a little bit of protein it’s going to hurt your gains, and then if you start supplementing a lot of protein it’s going to be good for your gains. Obviously that makes no sense. This graph should actually be the other way around. It should be a diminishing returns curve where we know that supplementing a little bit of protein is, if your protein intakes not sufficient yet, very positive. But then if you keep going up to progressively higher dosages there should be diminishing returns.
Just trusting the researchers that they did the analysis well without reporting a break point analysis, without reporting R^2 estimates, without reporting how well the model actually fit the data, or showing graphs of how well their model fit the data is also made a little bit harder by the fact that the researchers confused muscle mass and lean body mass in the title of their paper. In the title they say they are studying muscle mass, but then they are actually studying lean body mass in their results. This is not just a semantic issue because they also studied people that don’t lift. In people that don’t lift you can certainly not assume that lean body mass equals muscle mass.
So overall, I highly applaud the researchers for the work they did. I’m grateful that they did this work, but this analysis, honestly, just wasn’t very good. And the researchers probably realized this because these researchers that did this analysis were actually the same researchers, it’s the same research group that did the analysis that I mentioned previously on strength development where they adopted other methods and they excluded studies that were weight loss trials, and there they found that for strength development the benefits plateau at a protein intake of 1.5 g/kg. Almost exactly in line with all the other research.
If you want to trust these researchers, well, their newer meta- analysis is actually one that supports 1.5 g/kg per day. So based on the totality of research, to be extra safe, I typically add a standard deviation of error to the estimate of what is already the maximum protein intake you need for maximum gains. And then you get a protein intake of 1.8 g/kg per day or 0.8 g/lb. This protein intake is enough for the vast majority of trainees, regardless of whether you are a powerlifter or a bodybuilder, because the research has studied strength development and muscle growth. Moreover, based on the current research, and I already did a video on this that you can find here, there is also no further benefits of higher protein intakes in energy deficit. Now, there is some newer research on this that I may include in a updated video, but based on the current research, there is still no strong support for higher protein intakes, even in trainees that are cutting, that are an energy deficit.
There’s also no benefit of higher protein intakes for satiety. Protein benefits for satiety also ceases to be significant at around the same protein intake, 1.6 g/kg per day, because of protein leverage. I discuss that topic in more detail, as well as the research I co-authored on this in this video. There are some exceptions though, for people that actually need higher protein intakes. That includes some vegans, depending on the protein quality of their diet, people with muscle memory, they probably need a little bit more protein because they can grow muscle faster, concurrent athletes, so athletes that do endurance training and strength training with very high volumes, they might also need a little bit extra protein, although most research still doesn’t find those benefits because some of the studies actually included concurrent athletes.
And finally, by far the biggest exception is people on gear. If you are on anabolic steroids, you’ll most likely need significantly higher protein intakes. That’s a completely different topic, though. For this video, take home message: For the vast majority of trainees, if you consume 1.8 or even if you don’t like high protein intakes, 1.6 g/kg per day, 0.7 g/lb, that’s total protein intake based on your total body weight, not lean body mass, that’s not even necessary, you are covered. You’re going to get pretty much maximum strength development, optimum body recomposition and good satiety. And I think this is a happy message because protein is expensive, and having to consume higher protein intakes just takes away from the variety that you can get in your diet. So the less protein we need to consume for maximum gains, the better.
Alright. I hope this helps you optimize your diet, save some money, and if you like this type of evidence based fitness content, I’d be honored if you like and subscribe.