Rapamycin vs metformin: the comparison stopped being close

Which one has better animal evidence?

Strong human evidence The 2025 Aging Cell meta-analysis pooled 167 studies across vertebrate species and reached an unusually clean conclusion: dietary restriction and rapamycin both increased lifespan across species and in both sexes, and metformin did not.^[1] Independent expert commentary at the time emphasized the same split rather than softening it.^[2]

That result did not come from nowhere. Rapamycin extended mouse lifespan in the Interventions Testing Program even when started at 600 days of age, and the effect replicated across sites and strains.^{[3] [4]} The same rigorous program has repeatedly failed to produce an equivalent metformin result, which is a meaningful negative given how specifically that program exists to catch irreproducible claims.

Which one has better human evidence?

Metformin, and it is not close, provided you are careful about what the evidence is for. It has been prescribed since the 1950s, is on every essential medicines list, and its safety profile over decades of continuous use is genuinely known. That is a category of knowledge rapamycin simply does not have at longevity doses.

What metformin does not have is a longevity result. Its geroprotective case rests largely on observational analyses suggesting people with diabetes taking it had mortality comparable to non-diabetic controls, an intriguing finding with obvious confounding problems.^[5] The trial designed to test it properly has been discussed for years without running.^[6] Rapamycin's human file is thinner still: a systematic review of human use found real biological effects and no lifespan endpoint anywhere.^[7]

Nobody's going to make enough money off of rapamycin to do the clinical trials that really should be done.

They do not work the same way

The two are often introduced together as calorie restriction mimetics, which obscures how differently they act. Rapamycin binds FKBP12 and inhibits mTORC1 directly and potently. Metformin works upstream, largely by activating AMPK and inhibiting mitochondrial complex I, which can lower mTOR signaling as a downstream consequence.

Directness matters here. One drug switches off the pathway the field cares about; the other nudges a network that eventually touches it. Our explainer on how mTOR inhibition actually works covers the first mechanism in detail.

The metformin finding nobody mentions

Early / mixed evidence A randomized trial in older adults found that metformin blunted the improvement in insulin sensitivity and muscle mitochondrial adaptation produced by exercise training.^[8] The people taking it got less out of the same workouts than the people on placebo.

For someone with type 2 diabetes that trade is usually still worth making. For a metabolically healthy person taking it speculatively for longevity while also training, it is an argument that deserves to be on the table, and it very rarely is.

Rapamycin vs metformin on risk

Their risk profiles are not comparable in kind. Metformin's problems are well mapped: gastrointestinal upset that causes many people to stop, vitamin B12 depletion over years, and lactic acidosis as a rare event largely confined to significant kidney impairment. Everything on that list is known, monitorable and reversible.

Rapamycin is an immunosuppressant with a dose-dependent adverse event profile drawn mostly from transplant medicine, plus an unquantified long-term risk in healthy people. We go through that separation in what actually changes at longevity doses. The honest summary is that metformin's risks are small and known while rapamycin's are larger and partly unknown.

Can you take them together?

People do, usually on the reasoning that hitting two nutrient-sensing pathways is better than one. No trial has tested the combination against either drug alone for any aging-related outcome, and the two push glucose handling in opposite directions, which makes the metabolic result of combining them genuinely unpredictable without monitoring.

Stacking on a hypothesis is not the same as stacking on evidence. Our page on what the metformin longevity evidence shows sets out that side of the case in full.

How the comparison should be read

They answer different questions. If you are asking which compound has the strongest evidence of slowing aging in a living organism, the answer is rapamycin and the 2025 meta-analysis made that clearer, not murkier. If you are asking which compound you can take for twenty years with a known safety profile, the answer is metformin, and that is not a small thing.

The fact that something fails in mice doesn't mean it's going to fail in people.

Neither answer is the one people actually want, which is a compound proven to extend healthy human life. Rapamycin vs metformin is a comparison between an animal result without human confirmation and a human safety record without a longevity result. Our main evidence guide lays out what would need to change for that to stop being true.

Frequently asked questions

What is better than rapamycin?

For lifespan extension in animals, nothing tested has beaten it, and a 2025 meta-analysis across vertebrates found only dietary restriction performing comparably. For human safety evidence, almost everything is better than it, including metformin, because it has decades of use in millions of people.

Can rapamycin increase longevity?

In animals, consistently. It extends lifespan in yeast, worms, flies and mice, including mice given it late in life. In humans nobody knows, because no trial has ever used lifespan or age-related disease incidence as an endpoint.

Do people on metformin look younger?

There is no evidence for that. The observational finding that generated the interest was about mortality in diabetics compared with matched non-diabetics, not appearance. Nothing in the trial literature measures how anyone looks.

Can I take metformin for 20 years?

Many people with type 2 diabetes have. That is precisely its advantage in this comparison: the long-term safety picture is genuinely known, including the main real risk, which is vitamin B12 depletion that goes unnoticed unless someone checks.

What happens when you take metformin for years?

Most people tolerate it, with gastrointestinal upset the usual reason for stopping early. Over years, B12 levels can fall enough to matter, which is why periodic testing is standard. Lactic acidosis is the rare serious risk and is largely confined to people with significant kidney impairment.

Which one has better animal evidence?

Rapamycin, decisively. The 2025 meta-analysis of 167 studies found it extended lifespan across vertebrate species in both sexes, while metformin did not produce a consistent effect. The rigorous mouse testing program has also repeatedly failed to show a metformin lifespan benefit.

Which one has better human evidence?

Metformin, by a wide margin, but not for longevity. It has enormous human data as a diabetes drug and essentially no controlled human data as a geroprotector. The trial that would test that has been proposed for years and has not run.

Do they work by the same mechanism?

No. Rapamycin inhibits mTORC1 directly by binding FKBP12. Metformin acts largely upstream through AMPK activation and mitochondrial complex I inhibition, which can reduce mTOR signaling indirectly. Calling them both mTOR inhibitors flattens a real difference in directness and potency.

Does metformin blunt exercise benefits?

A randomized trial in older adults found that metformin attenuated the improvement in insulin sensitivity and muscle mitochondrial adaptation from exercise training. That is a genuine trade-off for anyone taking it while training, and it is rarely mentioned in longevity coverage.

Can you take rapamycin and metformin together?

Some people do, and there is no trial testing the combination for any longevity outcome. Both affect glucose handling in opposite directions, which makes the combined effect on metabolism hard to predict without monitoring.

Which one is easier to get?

Metformin, without contest. It is cheap, widely prescribed for diabetes and prediabetes, and familiar to every physician. Rapamycin requires a prescriber willing to write it off-label for an unapproved purpose.

Is metformin useless for longevity then?

Not established would be more accurate than useless. Its case rests on observational data and disease-risk reduction in people with metabolic dysfunction, which is a real benefit for that group. What it lacks is evidence of slowing aging in people who are metabolically healthy.

Why do they get compared at all?

Because both are cheap, approved, well-characterized drugs that appeared in longevity research at the same time, and because both were framed as calorie restriction mimetics. The framing turned out to fit one of them considerably better than the other.

What would settle the question?

A properly powered human trial with an age-related disease endpoint, of the kind proposed for metformin and now beginning for rapamycin in specific populations. Until one reports, both cases rest on animals and inference.