A Bayesian meta-analysis pooled 12 studies measuring the exact muscle-building spike supplement ads sell. The gap: 0.004 percentage points per hour, at the 0.35th percentile of normal variation.
99.65 percent of the variation in how strongly your muscles respond to protein comes from something other than whether the protein was plant or animal.
Every protein tub on the shelf sells the same promise: more muscle protein synthesis. The acute MPS spike, the burst of muscle-building activity that shows up higher after whey than after pea protein in short-term lab studies, is the scientific basis beneath every "fast-absorbing" claim you have ever read on a supplement label.
It is also the single most common argument a fitness creator can make for choosing animal protein over plant.
A team of researchers at the University of Lisbon, including Brad Schoenfeld, wanted to know the exact size of that advantage. Their Bayesian meta-analysis pooled 12 studies that directly compared plant and animal protein for the MPS spike.
Twenty-six direct comparisons, 321 participants, studies published from 2012 to 2024. Soy, pea, wheat, rice, potato, oat, and mycoprotein, each tested head-to-head against whey, casein, milk, beef, or egg.
The advantage of animal protein for muscle protein synthesis was 0.004 percentage points per hour.
That number needs context. The researchers placed it against the full range of MPS variation that occurs after any person consumes any protein source. The effect landed at the 0.35th percentile.
Meaning 99.65 percent of the variation in how strongly your muscles respond to protein comes from something other than whether the protein was plant or animal. Your sleep, your training stimulus, your total calorie intake, your stress — all of those move the MPS needle more than protein source does.
The 95 percent credible interval (the range where the real difference almost certainly lives) spans from −0.002 to 0.011. It crosses zero. Three-quarters of the 12 studies found no significant difference at all.
And the between-study noise, how much the result bounced from lab to lab running the same comparison, was 2.5 times larger than the signal itself. Pick two labs, run the same plant-versus-animal test, and the difference between the labs exceeds the advantage being measured.
There is a 90 percent probability the gap favors animal protein. But probability without magnitude is how supplement ads work.
Ninety percent sure of an advantage that sits at the 0.35th percentile of biological variation, in data rated as LOW certainty by the GRADE framework, with noise louder than the signal. That is not nothing. But it is very close.
For younger adults, the people most likely to be standing in front of the protein shelf making this decision, the gap was even smaller: 0.001, functionally zero.
Three out of four studies found no MPS difference between plant and animal protein. The gap that remains is smaller than the variation between labs running the same test, measured in a window that captures one-fifth of the day.
- Three-quarters of the 12 studies found no significant difference in muscle protein synthesis between plant and animal protein.
- The only age group with a meaningful gap (older adults) had a confound: every study gave the animal protein condition 28 to 50 percent more leucine. When one study equalized leucine, the gap vanished.
- Eleven of 12 studies measured MPS for five hours or less, capturing 21 percent of a 24-hour muscle-building cycle. Almost no data exists on what happens in the other 79 percent.
- A separate meta-analysis of 43 randomized controlled trials measured identical muscle mass gains from plant and animal protein over weeks to months of training. The negligible acute gap produces zero long-term difference.
The Experiment That Tested the Wrong Variable
If there is one group where protein source is supposed to matter most, it is older adults. The theory: aging muscles become more resistant to the building signal, requiring higher-quality protein to overcome that resistance. Three studies in this meta-analysis tested that theory directly, in men aged 71 to 72.
The result looked like a win for animal protein. Larger effect (0.012), higher probability (97.7 percent). At first glance, the one subgroup where the gap matters.
Then look at the leucine.
Yang gave the whey group 2.4 grams of leucine and the soy group 1.6 grams. Fifty percent more leucine in the animal condition. Gorissen gave the casein group 3.2 grams and the wheat group 2.5 grams, 28 percent more.
Those two studies produced the largest MPS gaps in the entire meta-analysis. Not because animal protein is better for older adults, but because the animal group got far more of the amino acid that starts the building process.
Then McKendry pushed both conditions above the leucine threshold. Whey at 5.4 grams, pea protein at 4.1 grams. MPS was similar. The gap disappeared when the leucine playing field was level.
The researchers noted that the leucine dose "exceeded the critical threshold for initiating MPS," which "conceivably could negate the beneficial effects of animal-based proteins." The experiment did not prove protein source matters for older adults. It proved leucine dose matters. And when one study controlled for it, the supposed protein-source advantage vanished.
This finding comes from three studies in men aged 71 to 72, in a meta-analysis that has not yet been peer-reviewed. The concern about protein quality with age is real. The evidence behind it tested the wrong variable.
The post-workout window, the exact moment the supplement industry prices protein source into the product, showed zero moderating effect on the MPS gap. The effect size was 0.004 with exercise and 0.004 without. The moment when protein source is supposed to matter most showed the same negligible difference as rest.
Twenty-One Percent of the Day
Even granting the tiny MPS gap, even setting aside the leucine confound, there is a deeper problem with the entire measurement.
Eleven of 12 studies measured muscle protein synthesis for five hours or less. One study extended to 24 hours. One.
Five hours is 21 percent of a 24-hour muscle-building cycle. The whey protein spike, the fast burst of building activity that looks impressive on a graph, is a measured reality in the opening act. But plant proteins release amino acids more slowly. Their contribution to the building process may extend into hours that almost nobody measured.
Whey spikes MPS within the first hour. The question is whether the first five hours capture the full picture or just the first act.
The study’s own authors warned that acute MPS measures are "not necessarily indicative of long-term changes in muscle mass." The researchers who pooled this data are telling you the measurement their meta-analysis captured does not reliably predict whether your muscles actually grow over weeks and months.
The MPS spike is the scientific basis for "fast-absorbing" and "anabolic window" marketing. It is technically real. The gap it measures, 0.004 at the 0.35th percentile, was captured in a window that covers one-fifth of the process. That is the distance between the finding and the narrative printed on the supplement tub.
The concern about protein quality with age is real. The evidence behind it tested the wrong variable.
What This Study Does Not Know
This is a preprint, posted on SportRxiv. The 12 individual studies it pools have been peer-reviewed and published. The synthesis combining them has not.
Eleven of 12 studies included only men. One study included 8 men and 2 women out of roughly 321 total participants. Whether the near-zero MPS gap holds for women is an open question from this data.
Every study tested isolated protein supplements, not whole-food meals. The researchers noted that whole-plant proteins contain compounds, including fibre, protease inhibitors, phytates, lectins, and polyphenols, that can reduce digestibility. Pea protein isolate in a lab and a plate of lentils at dinner are not the same thing. The equivalence finding is about supplements.
The overall certainty of evidence is LOW by the GRADE framework, driven by serious concerns about study-to-study inconsistency and the width of the credible interval. The direction is clear. The gap is centered near zero. The precision will sharpen as larger trials arrive.
The supplement tub says ‘supports muscle protein synthesis.’ This meta-analysis says that support is indistinguishable from the background noise of normal biological variation.
Why the Mechanism Matches the Outcome
If the acute MPS gap is 0.004, at the 0.35th percentile of variation, measured in a five-hour window, confounded by leucine in the only population where it appeared meaningful, and rated as low certainty, then there is essentially nothing to accumulate into a long-term difference.
The mechanism matches the outcome. A separate meta-analysis of 43 randomized controlled trials found that plant and animal protein produce functionally identical changes in muscle mass over weeks to months of training.
The acute gap is negligible. The long-term gap is zero. One explains the other.
What you can take from this page: every time someone cites the MPS spike as evidence that whey is superior, you now know the size of that advantage (0.004), the share of variation it explains (0.35 percent), the fraction of the day it was measured in (21 percent), and the certainty of the evidence behind it (low).
The supplement tub says "supports muscle protein synthesis." This meta-analysis says that support is indistinguishable from the background noise of normal biological variation.
That is not a debunking. It is a measurement. And the measurement does not support the marketing.
If protein source sits at the 0.35th percentile of MPS variation, the question shifts: what occupies the other 99.65 percent?
The variables that actually move the muscle-building needle, according to this data and the broader exercise science literature, are total daily protein intake (hitting adequate grams matters far more than where those grams come from), training stimulus (progressive overload is the primary driver of muscle protein synthesis), sleep quality and duration (recovery is when the building happens), and overall caloric balance (a surplus supports muscle gain regardless of protein source). Protein source is last in line.
For protein purchasing, this meta-analysis suggests the decision can be made on taste, cost, digestive comfort, or environmental preference. The MPS data does not give animal protein an advantage worth optimizing for, particularly in adults under 40 where the gap was 0.001. The variable worth optimizing is total daily protein, not where it comes from.
What this means for you
For younger adults, the MPS gap between plant and animal protein was 0.001 percentage points per hour, with only a 65 percent probability that animal protein was even slightly better. That probability is barely above a coin flip.
This means the cellular argument for choosing whey over pea protein does not exist in your age group. Nine studies with 19 direct comparisons produced a gap so close to zero that optimizing protein source for MPS is like optimizing shoe color for a marathon time. The research suggests focusing your attention on total daily protein intake and training consistency instead.
The MPS gap was larger in older adults (0.012), but the three studies behind that number all gave the animal protein condition substantially more leucine, the amino acid that triggers the building process. The one study that equalized leucine found similar MPS regardless of protein source.
The practical signal from this data: ensure each protein serving delivers adequate leucine (roughly 3 to 4 grams or more), whether from animal or plant sources. The variable that mattered in these older-adult studies was leucine dose, not protein type. This finding comes from only three studies in men aged 71 to 72, so the evidence base is thin and the conclusion is preliminary.
For isolated plant protein supplements (pea isolate, soy isolate, wheat protein), this meta-analysis confirms the MPS response is nearly identical to animal protein. The gap is 0.004, well within the noise between labs running the same test.
The caveat: every study used protein isolates, not whole foods. The researchers noted that whole-plant proteins contain compounds (fibre, phytates, lectins, polyphenols) that can reduce digestibility. Your protein shake and your lentil curry are not the same thing for MPS purposes. The supplement data is solid. The whole-food question remains open.
Before you change anything
Who this applies to: healthy men aged 18 to 72 consuming isolated protein supplements. Eleven of 12 studies included only men. One study included 8 men and 2 women out of roughly 321 total participants. Whether the near-zero MPS gap holds for women is an open question from this data. All participants were healthy, with no clinical conditions. Most were untrained.
What the measurement captures: this is a preprint posted on SportRxiv. The 12 individual studies it pools have been peer-reviewed. The synthesis combining them has not. The data measures acute muscle protein synthesis (a cellular snapshot lasting 5 hours or less in most studies), not whether muscles actually grow over weeks and months of training. The researchers themselves warned that acute MPS is "not necessarily indicative of long-term changes in muscle mass."
How confident the researchers are: GRADE rated the certainty of evidence as LOW. That rating reflects two specific concerns: the results varied substantially from study to study (the noise between labs was 2.5 times the signal), and the credible interval was wide enough to include zero. LOW certainty means the direction is likely correct (the gap is near zero) but the precise magnitude could shift as larger trials arrive.
If the acute MPS gap between plant and animal protein is this small, what actually happens to muscle mass over weeks and months of training? A separate meta-analysis of 43 randomized controlled trials measured exactly that, tracking real muscle growth across 8 to 16 weeks. And for older adults concerned about the leucine confound: another study in this cluster tested whether adding leucine to plant protein closes the gap entirely.
What This Study Found
All findings from this paper, in plain language.
- Animal protein produced a 0.004 percentage point per hour higher MPS response than plant protein, a gap at the 0.35th percentile of normal variation with low certainty of evidence.
- Nine of 12 studies found no significant MPS difference between plant and animal protein. Three found lower MPS with plant protein.
- In younger adults under 30, the MPS gap was 0.001, functionally zero, with only 65 percent probability that animal protein was slightly better.
- In older adults over 65, the MPS gap was larger at 0.012, with 97.7 percent probability favoring animal protein, but based on only three studies in men aged 71 to 72.
- All three older-adult studies gave the animal protein condition 28 to 50 percent more leucine, confounding protein source with leucine dose. When one study equalized leucine, MPS was similar.
- Resistance exercise did not change the gap: the MPS effect size was 0.004 both with and without exercise, based on 4 studies with exercise and 11 without.
- Eleven of 12 studies measured MPS for five hours or less, capturing roughly 21 percent of a full day. Only one study extended to 24 hours.
- The between-study noise (how much results varied from lab to lab) was 2.5 times larger than the effect itself, contributing to low certainty of evidence.
- Study quality was moderate to strong: 4 studies rated strong, 8 moderate, none weak. Eleven of 12 were double-blind randomized controlled trials.
- Seven types of plant protein were tested (soy, pea, wheat, rice, potato, oat, mycoprotein) against four animal types (milk, whey, chicken, casein) across 26 direct comparisons.
- McKendry found that at a 50-gram dose with both conditions above the leucine threshold (5.4 vs 4.1 grams), MPS was similar in older men regardless of protein source.
- Whole-plant proteins contain fibre, phytates, lectins, and polyphenols that reduce digestibility. All 12 studies used isolated supplements, not whole foods.