Metformin: The $4 Longevity Pill

The Cheapest Longevity Drug Is Sitting in Your Pharmacy

Metformin costs a few dollars a month as a generic. It has been prescribed for type 2 diabetes since the 1950s in Europe and since 1995 in the United States. And for about a decade, it has been the quiet favorite of the longevity crowd: the drug you can get through a telehealth visit, the one podcast hosts say they take, the one that supposedly makes you biologically younger while the rest of the supplement aisle fights over mouse data.

The narrative is tidy. People with diabetes who take metformin appear to live longer than people without diabetes. Metformin touches the same cellular energy pathways that caloric restriction does. A large trial to test it as an anti-aging drug has been designed and endorsed by the people who study aging for a living.^1,12 What could go wrong?

Quite a lot, as it turns out, and most of it is in the data the narrative skips. The headline survival finding comes from a study design with a known bias. The mouse data are weaker than the popular telling. The two randomized trials that put metformin into healthy older adults alongside exercise found it blunted the gains. And the big trial that was supposed to settle the question, as far as I can establish from public reporting, has not enrolled a single participant.

Why a Diabetes Drug Looks Like an Aging Drug

Metformin lowers blood sugar mainly by reducing how much glucose the liver makes, and it improves how muscle responds to insulin. To do this, it partially inhibits complex I, an enzyme in the mitochondria, the cell's power plants. That mild energy stress activates AMPK, a cellular fuel gauge that tells the cell to conserve energy and burn fat, and it dampens mTOR, a growth-signaling pathway that is heavily implicated in aging biology.¹²

That is why researchers noticed it. AMPK activation and lower mTOR signaling are the same neighborhood as caloric restriction and rapamycin, the two interventions with the most convincing lifespan effects in animals. Metformin also lowers insulin and insulin-like growth factor signaling, reduces some markers of inflammation, and alters the gut microbiome. Each of these is a plausible route to slower aging.

Plausible, though, is the operative word. The same mitochondrial inhibition that looks beneficial in a diabetic liver may be harmful in an older adult's skeletal muscle, which depends on mitochondrial function to adapt to exercise. I will come back to that, because it is the most important caveat in this entire issue.

The Study That Launched a Thousand Podcasts

The finding that made metformin a longevity candidate came from a 2014 cohort study of UK primary care records. Bannister and colleagues compared roughly 78,000 people: diabetics on metformin alone, diabetics on a sulfonylurea alone, and matched people without diabetes. The metformin group survived about 15 percent better than the non-diabetic controls. The sulfonylurea group did far worse.²

A "too good to be true" result like this one usually is. Suissa and Azoulay showed in 2012 that observational studies of metformin were riddled with time-related biases, chiefly immortal time bias, in which the way patients are sorted into "users" guarantees they survive long enough to be counted as users.³ A later meta-analysis by Campbell and colleagues pooled the observational literature and still found lower mortality with metformin, but it was built from the same kind of studies, and the authors acknowledged the limits of that design.⁴

The honest reading: metformin looks better than the alternatives people with diabetes had at the time. That tells you something about sulfonylureas and about who stays on metformin. It tells you very little about whether metformin adds years to a healthy 55-year-old.

Mice Were Not Quite So Impressed

In 2013, Martin-Montalvo and colleagues reported that a low dose of metformin in the diet extended the mean lifespan of male mice of one inbred strain by around 6 percent, while a tenfold higher dose was toxic and shortened it.⁵ That is a narrow window, and an inbred strain is a narrow test.

The National Institute on Aging's Interventions Testing Program is the field's gold standard for this question. It tests compounds in genetically diverse mice across three independent sites. When it tested metformin alone, at 0.1 percent of the diet, lifespan did not significantly change. Metformin combined with rapamycin did extend lifespan, which suggests the rapamycin was doing the heavy lifting.⁶ Rapamycin, by contrast, has repeatedly extended lifespan in this program.

What Randomized Trials Actually Show

There is a robust randomized literature for metformin. It is just not about aging. In the Diabetes Prevention Program, 3,234 adults with prediabetes were randomized to placebo, lifestyle change, or metformin. Metformin cut the incidence of type 2 diabetes by 31 percent over about three years. Lifestyle change cut it by 58 percent.⁷ The long-term follow-up kept a 17 percent reduction at 15 years.⁸

That is a real, durable effect, and it is the reason metformin is endorsed for people at high risk of diabetes. It is also a different claim from "slows aging," and the gap matters. Preventing diabetes in people who are headed for it is a disease-prevention benefit. The longevity claim is that metformin helps people who are metabolically healthy.

The newest piece of randomized data points the same way. The MeMeMe trial, reported in Diabetes Care in late 2024, randomized 1,442 adults aged 50 to 79 with metabolic syndrome to metformin 1,700 mg a day or placebo, with and without a Mediterranean diet program, for up to five years. Metformin substantially reduced the development of type 2 diabetes. It did not produce a significant difference in cancer or cardiovascular events.⁹ If metformin were quietly slowing aging in a group this metabolically at risk, you would expect at least a flicker in those outcomes. There was none.

The Trials That Should Worry the Biohackers

This is the part of the story that rarely makes the podcast. If you are an older adult who wants to stay strong, the single best-supported intervention you have is resistance and aerobic exercise. Two randomized, placebo-controlled trials tested what happens when you add metformin to exercise in older people. Both found that metformin got in the way.

Muscle mass and mitochondrial capacity are two of the strongest predictors of how well people age. A drug that interferes with both during exercise, in the very population it is meant to help, is a serious complication of the longevity story. It does not prove metformin shortens healthspan. It proves the simple version, "it hits AMPK so it must be good," is wrong at least some of the time.

A recently posted conference abstract reports that a two-year trial of metformin in adults 65 and older with glucose intolerance slowed progression of a frailty index compared with placebo.¹³ I could not access the full data, so I am treating it as a preliminary signal worth watching, not as evidence. Frailty indices are also soft outcomes, and a population with glucose intolerance is not a healthy population.

TAME: A Great Idea With No Sponsor

The Targeting Aging with Metformin trial was designed by Nir Barzilai and colleagues to test metformin in roughly 3,000 adults aged 65 to 79 across about 14 centers, with a composite endpoint of first occurrence of cardiovascular disease, cancer, dementia, or death.¹² The design was meant to get the FDA to treat aging-related decline as an indication. It is a smart trial.

Public reporting in 2026 describes it as designed and regulatory-ready but unfunded, with no participants enrolled and estimates of its cost in the tens of millions of dollars. Nobody owns the patent on metformin, so nobody profits from a result. That is a structural reason the evidence is missing, and it is not a reason to assume the answer is yes.

Cheap Is Not the Same as Free of Consequences

Metformin has decades of safety data in diabetics, which is genuinely reassuring. But the longevity use means long-term exposure in people who do not have the disease the drug was approved for, and some of the known problems deserve attention.

There is also a supply-side issue. Between 2019 and 2020, regulators recalled batches of metformin, mostly extended-release products, after finding the probable carcinogen N-nitrosodimethylamine (NDMA) above acceptable limits. Contamination of this kind was a manufacturing problem, not a property of the drug, but it is a reminder that "cheap generic" varies by supplier.¹⁵

Metformin for longevity is off-label, and in the United States it requires a prescription. Longevity clinics and telehealth services sell it, usually bundled with other things. If someone is selling you a metformin "protocol" for a monthly fee, notice what you are paying for, since the drug itself costs almost nothing.

A Good Drug in Search of the Right Question