Metformin is a cheap, decades-old diabetes pill, and it has become one of the most discussed "longevity" prescriptions. Its label covers one thing: blood sugar control in type 2 diabetes [1]. Anyone taking it to slow aging is using it off-label. The interest is real, and so are the caveats. Below is every major study in the case, sorted by what kind of evidence it is.
The short answer: unproven in people without diabetes
Metformin lowers blood sugar and is labeled for adults and children 10 and older with type 2 diabetes [1]. FDA has never approved it for aging, longevity or weight loss. That isn't unusual, since aging isn't an approved indication for any drug. But it means no regulator has weighed the longevity claim.
What does exist splits into three piles: animal studies, observational studies of people who already had diabetes, and randomized trials that measured something other than lifespan. None of them answers the question a healthy 50-year-old is asking.
| Study | Who | Design and size | Main result | Can it show metformin extends life? |
|---|---|---|---|---|
| UKPDS 34 (1998) | Overweight adults with new type 2 diabetes | Randomized trial, 753 people, median 10.7 years | 36% lower all-cause death vs diet-based care | No. Everyone had diabetes |
| Diabetes Prevention Program (2002) | Adults with prediabetes | Randomized trial, 3,234 people, 2.8 years | 31% less new diabetes than placebo; lifestyle change did better (58%) | No. The outcome was diabetes, not death |
| Bannister (2014) | UK patients starting metformin vs matched people without diabetes | Observational, 78,241 vs 90,463 | Controls' adjusted median survival 15% shorter | No. Observational, and later reversed |
| Stevenson-Hoare (2023) | Welsh patients on metformin vs matched controls | Observational, 129,140 on metformin, up to 20 years | Early edge reversed after 5 years; shorter survival overall | No, and it argues against |
| Shadyab (2025) | Women 60+ with new type 2 diabetes | Target-trial emulation, 438 matched women | 30% lower risk of death before 90 vs sulfonylureas | No. Compared with another drug, not placebo |
| Konopka (2019) | Healthy adults around 62 | Randomized trial, 53 people, 12 weeks of aerobic training | Smaller gains in fitness and insulin sensitivity on metformin | No. Shows a possible cost |
| Walton, MASTERS (2019) | Healthy adults 65+ | Randomized trial, 94 people, 14 weeks of strength training | Less lean mass and thigh muscle gained on metformin | No. Shows a possible cost |
| TAME | Older adults (as proposed) | Proposed in 2016 | No results | It was designed to; not listed on ClinicalTrials.gov under that name |
Sources are linked in the sections below. Observational studies show associations, not cause and effect.
Researchers got interested because of mice and diabetes data
The mouse story is mixed too. In a 2013 study, male mice given metformin from middle age lived longer and stayed healthier on several measures; the authors linked this to higher activity of AMPK, an energy-sensing enzyme, and less oxidative damage and inflammation [10]. But when the National Institute on Aging's testing program ran metformin alone at that dose in genetically mixed mice at three sites, it didn't significantly extend lifespan [11].
The human interest came from diabetes. Metformin users seemed to do better than expected on several age-related diseases, which is why researchers proposed a trial called TAME, Targeting Aging with Metformin, in 2016 [9].
The diabetes study that started the buzz was later reversed
The headline finding came from UK health records in 2014. Researchers followed 78,241 people with type 2 diabetes who started on metformin and 90,463 matched people without diabetes. After adjusting for other factors, the people without diabetes had a median survival time 15% shorter than the metformin group [4]. People on sulfonylureas, another diabetes drug, did far worse than both.
That sounded like metformin might do more than treat diabetes. Then a larger Welsh study looked over a longer window. Among 129,140 metformin users followed for up to 20 years, metformin patients had shorter survival than matched controls without diabetes. Within the first three years they did better, but that edge reversed after five years [5].
Both studies are observational, so they can't separate the drug from the people who get it. Doctors tend to prescribe metformin to patients healthy enough to take it, and that alone can make users look better than average.
Older women with diabetes: a 2025 finding against another drug
A 2025 analysis of the Women's Health Initiative asked whether starting metformin, rather than a sulfonylurea, made women with new type 2 diabetes more likely to reach 90. Among 438 matched women aged 60 and older, the rate of death before 90 was 3.7 per 100 person-years on metformin against 5.0 on sulfonylureas, a 30% lower adjusted risk (hazard ratio 0.70) [6].
The authors are careful about what that means. The comparison was with another diabetes drug, not with placebo, and the observational design leaves room for confounding, so "causality cannot be inferred" [6]. Sulfonylureas also looked harmful in the 2014 UK data, so part of the gap may be the comparison drug.
Randomized trials show diabetes benefits, not slower aging
The strongest metformin evidence is in diabetes itself. In UKPDS 34, 753 overweight adults with newly diagnosed type 2 diabetes were randomized to metformin or conventional, mostly diet-based care for a median 10.7 years. Metformin cut all-cause death by 36% [2]. Everyone in that trial had diabetes, so it tells you about treating diabetes, not about aging.
In the Diabetes Prevention Program, 3,234 adults with prediabetes got placebo, metformin or an intensive lifestyle program. Over about 2.8 years, metformin cut new diabetes by 31% compared with placebo. The lifestyle program cut it by 58%, and the researchers called it significantly more effective than metformin [3].
Metformin blunted exercise gains in two trials of healthy older adults
This is the result that matters most to the people most interested in metformin: healthy, active adults. In a 12-week trial, 53 adults around age 62 did aerobic training on metformin or placebo. Metformin weakened the rise in insulin sensitivity and in VO2 max, a standard fitness measure, and it canceled the gain in muscle mitochondrial function [7].
The MASTERS trial tested strength training. Ninety-four adults 65 and older trained for 14 weeks on metformin or placebo. The placebo group gained more lean body mass and more thigh muscle than the metformin group [8]. Those are the exact gains people exercise for in later life.
Responses varied a lot from person to person in both trials. Still, if you already exercise, these trials are the closest thing to evidence about metformin in someone like you, and they point to a cost.
TAME, the trial meant to settle it, has no results
TAME was proposed in 2016 as a trial of metformin against age-related disease in older adults [9]. Right now, no trial by that name is listed on ClinicalTrials.gov [12], and no results have been published. Until it or something like it reports, the central question stays open.
Smaller studies exist. MILES, a metformin study in older adults, enrolled 16 people and finished in 2017; it measured biological markers, not lifespan [13]. A trial recruiting in 2026 compares semaglutide, tirzepatide and metformin on a DNA-methylation measure of biological age over 24 weeks in adults with obesity [14]. Neither can show that anyone lives longer.
Side effects, and who should not take it
Metformin carries a boxed warning for lactic acidosis, a rare buildup of acid in the blood that can be fatal. The label updated that warning in March 2026. Risk rises with kidney problems, age 65 and older, contrast imaging, surgery, heavy drinking and liver disease [1].
- Kidney function: metformin is contraindicated when eGFR, a kidney filtration measure, is below 30, and starting it between 30 and 45 isn't recommended.
- Stomach effects: diarrhea, nausea or vomiting, gas, weakness, indigestion, abdominal discomfort and headache are the most common side effects, each in more than 5% of people.
- Vitamin B12: metformin may lower B12 levels; the label advises checking B12 every two to three years.
- Low blood sugar: the risk rises when it's combined with insulin or drugs that raise insulin.
Those are label facts for people with diabetes [1]. There's no label guidance for healthy people taking it for aging, which is one more reason the decision belongs with a prescriber who knows your kidneys and your training habits.
Metformin vs rapamycin: neither has human proof of longer life
People often weigh metformin against rapamycin, the other off-label longevity prescription. Neither has human evidence of longer life. Rapamycin's strongest data are mouse lifespan studies; metformin's are observational diabetes cohorts and mixed mouse results. One mouse study found metformin plus rapamycin extended lifespan, though it compared against earlier rapamycin-only studies rather than running both side by side [11]. See our rapamycin guide for its evidence, and how people get a rapamycin prescription for access.
GLP-1 drugs have also joined the longevity pitch. The guide to microdosing GLP-1s looks at that claim the same way.
How people get metformin, and what it costs
Metformin is a generic prescription drug. For diabetes, insurance usually covers it. For longevity, you need a clinician willing to prescribe it off-label, either your own doctor or a telehealth program. Several programs list it; prices below are each program's monthly price for metformin alone. More detail is in our metformin guide.
| Provider | Product | Monthly cost |
|---|---|---|
| bmiMDPartner | Generic metformin | $55/moMonthly, or a 3-month prepay |
| PeterMDPartner | Generic metformin | $60/moPer 30 days; 60- and 90-day options |
| HealthRXPartner | Prescription metformin | $134/moMonthly; lower with prepay |
| Found | Generic metformin with membership | $162/moEvery 28 days; longer terms cost less |
Alphabetical. Prices from each provider's own site. Shed sells a three-drug combination that includes metformin; it isn't metformin alone, so it isn't listed here.
For a scored view of these programs, see longevity prescription programs, scored. A full guide to getting a metformin prescription online covers the access side in more depth.