Walk into any gym and you will find a wall of tubs promising the same three molecules: leucine, isoleucine and valine. They are called the branched-chain amino acids, or BCAAs, because of a small kink in their carbon skeleton, and they have been marketed since the 1980s as the part of protein that actually builds muscle. The global market for them runs into the hundreds of millions of dollars a year.
Meanwhile, a laboratory in Madison, Wisconsin has spent the last decade systematically removing those same three molecules from the diets of mice to see what happens. What happens, repeatedly, is that the mice get leaner, healthier and in some cases live substantially longer.
On 24 July 2026, that lab published its clearest single-amino-acid result yet in Nature Aging.[1] It deserves to be read carefully, because the headline number is real and the caveat attached to it is large.
What the valine study actually did
Mariah Calubag and colleagues in Dudley Lamming's group at the University of Wisconsin–Madison took C57BL/6J mice and fed them a diet in which valine (C5H11NO2) was selectively reduced, starting in early life and continuing until death. Total calories and total protein stayed comparable. Only one of twenty amino acids was pulled down.
The metabolic results arrived early and applied to both sexes. Valine-restricted mice stayed leaner across their lives and held better glycaemic control. As they aged, both males and females scored lower on a frailty index, carried fewer senescent cells and developed fewer cancers.[1] That combination — less fat, better glucose handling, less frailty, less cancer — is close to the textbook description of what geroscientists mean by extended healthspan.
Then the survival curves diverged. Median lifespan in male mice increased by 23 percent. In females, it did not increase at all.
The asterisk: why only the males
Sex-specific longevity results are common in rodent aging research and are usually the first thing an honest write-up flags. Here the authors went looking for a mechanism. In the preprint version of the work, multi-tissue gene network analysis found that valine restriction produced a male-specific downregulation of PI3K–Akt signalling — a growth pathway repeatedly tied to longevity across species — and that this tracked with the lifespan effect.[2]
What makes it stranger is the direction of travel elsewhere. The obvious guess would be that starving cells of an amino acid quiets mTORC1, the nutrient sensor that rapamycin inhibits. It did not. mTORC1 signalling went up in both sexes.[1] Whatever valine restriction is doing, it is not a dietary imitation of rapamycin, and the field's default explanation for why low-protein diets work does not fit this result cleanly.
Worth noting too: the female mice were not left behind on health. They were leaner, less frail, less cancer-prone. They simply did not die later. That gap between feeling better and lasting longer shows up often enough in aging research to be worth taking seriously rather than smoothing over.
Three amino acids, three different answers
The valine paper only makes sense next to its siblings. BCAAs were treated as a unit for decades — supplements still sell them in a fixed 2:1:1 ratio — but restricting them one at a time has produced three distinct outcomes.
| Amino acid | Formula | What restriction did in mice |
|---|---|---|
| Isoleucine | C6H13NO2 | Extended lifespan in genetically diverse mice of both sexes, more strongly in males; improved metabolic health young and old[3] |
| Valine | C5H11NO2 | Improved healthspan in both sexes; extended median lifespan by 23% in males only[1] |
| Leucine | C6H13NO2 | No metabolic benefit in the comparison study; restricting it did not improve metabolic health the way isoleucine or valine did[4] |
That last row is the awkward one for supplement marketing, but not in the direction you might expect. Leucine is the amino acid the industry singles out as the muscle trigger — and it is also the one whose removal appears to do the least for metabolic health in these models. The two amino acids that matter for aging in mice are the two the fitness world treats as filler.
The lab has also started asking whether the distinction extends to the brain. In a 2025 preprint using the 3xTg Alzheimer's mouse model, restricting each BCAA separately produced different effects on disease pathology, and the effects were strongly sex-specific: isoleucine restriction helped short-term memory most in males, valine restriction most in females.[4] That work is not yet peer reviewed, and mouse models of Alzheimer's have a poor record of predicting human results. It belongs in the "interesting, unproven" column alongside the rest of the brain aging literature.
What any of this means for a human
Here is where the story usually gets oversold, so it is worth laying out the human evidence in order of strength.
Blood BCAA levels track with metabolic disease. This is solid and old. A 2009 Duke metabolomics study found a BCAA-related signature that separated obese from lean people and correlated with insulin resistance; feeding rats a high-fat diet supplemented with BCAAs made them as insulin resistant as high-fat-fed controls despite eating less and gaining less weight, and the effect was reversed by rapamycin.[5] Since then, elevated circulating BCAAs have been linked to insulin resistance, metabolic syndrome and later type 2 diabetes across multiple populations.[6] A 14-year cohort of 3,421 older adults published in 2026 found serum BCAAs statistically mediated 19–38 percent of the association between body composition and new-onset diabetes.[7]
But that same 2026 cohort contains a detail that should slow everyone down: after adjustment, dietary protein intake showed no significant correlation with serum BCAA levels.[7] High blood BCAAs in humans look substantially like a readout of impaired metabolism — a broken disposal system — rather than a straightforward consequence of what someone ate for lunch. Lowering the intake does not automatically lower the marker.
Lowering BCAA intake has been tested in humans, briefly. The same Wisconsin group ran a randomised controlled trial of a moderately protein-restricted diet in 2016 and found improved markers of metabolic health, alongside mouse work showing a BCAA-specific diet reproduced those benefits through different pathways.[8] That is a genuine human signal — but it measured blood markers over weeks, not lifespan over decades, and it restricted protein rather than a single amino acid.
Eating fewer amino acids has a documented downside in older adults. The Golestan Cohort Study followed 47,337 Iranian adults for a median of 15 years and recorded 9,231 deaths. The relationship between amino acid intake and mortality was non-linear and, crucially, age-dependent: people in the lowest fifth of essential amino acid intake had a 16 percent higher risk of death overall (HR 1.16, 95% CI 1.07–1.26), and while high intake looked unfavourable in the under-65s, it was low intake that carried increased mortality in adults over 65.[9]
That is the same fork we described in protein after 40: the age at which you run the experiment changes the answer. A mouse restricted from four weeks old is not a model of a 68-year-old deciding whether to skip the protein shake.
So should you throw out the BCAA tub?
Probably — but for a duller reason than longevity.
If you already eat adequate protein, isolated BCAA powder is redundant. Whey, eggs, chicken, fish, soy and legumes all deliver leucine, isoleucine and valine in quantities that dwarf a 5-gram scoop, packaged with the other sixteen amino acids muscle actually needs to be built. The case for BCAA supplementation was always thin for people eating enough protein, and no result in this article changes that either way.
What the mouse data does not support is the inverse move — deliberately cutting protein to lower your valine. Nobody has shown that works in a human, the one long-term human dataset on low amino acid intake points to harm in older adults, and age-related muscle loss is one of the best-documented predictors of losing independence. Trading a proven risk for an unproven benefit is a bad trade.
The realistic near-term translation is pharmaceutical, not dietary. Selectively removing one amino acid from a human diet is not practical, which is why the authors' own framing is that "interventions that mimic" valine restriction may have translational potential.[1] That means a drug targeting BCAA metabolism — the branched-chain ketoacid dehydrogenase complex is the obvious candidate — and it means the same long wait that every other entry on our list of drugs that may slow aging is currently serving.
The honest summary
A 23 percent lifespan extension in male mice from removing one amino acid is a striking finding and a real one. It also applies to one sex, one inbred mouse strain, one lifelong laboratory diet and a signalling pathway whose behaviour surprised the researchers who measured it. Two of the three BCAAs now have lifespan data; the third does not; the mechanism does not match the field's leading theory; and the human evidence is a mix of biomarkers that may be effects rather than causes and one large cohort suggesting older adults should eat more amino acids, not fewer.
None of that makes the work less interesting. It makes it a research programme rather than a diet plan. For further reading on where this sits in the broader picture, see our overview of protein restriction and longevity and our assessment of which longevity supplements have human evidence behind them.
Common questions
Are BCAA supplements bad for longevity?
No human trial has tested BCAA supplements against lifespan, so nobody can answer this with data. What exists is a chain of indirect evidence: mice fed less valine or less isoleucine live longer, and people with higher blood BCAA levels are more likely to be insulin resistant. Neither establishes that a scoop of BCAA powder shortens a human life. The stronger point is that the supplement is redundant if you already eat enough protein.
Should I cut protein to reduce my valine intake?
No, and the mouse studies do not suggest it. Those animals ate normal total protein with one amino acid selectively lowered — something ordinary food cannot do. The Golestan cohort found the lowest fifth of essential amino acid intake carried a 16 percent higher mortality risk, concentrated in adults over 65.[9] Cutting protein risks a documented harm to chase a hypothetical benefit.
Which foods are highest in valine?
Valine is in essentially every protein food, which is precisely why selective restriction is a laboratory diet rather than a menu. It is most concentrated in whey and dairy protein, eggs, beef, chicken, fish, soy and legumes. Plant proteins carry a somewhat lower proportion of branched-chain amino acids than whey — a difference of degree, not a switch you can flip.