Does metformin slow aging in people without diabetes?
Claim attributed to Longevity physicians and biohackers (TAME-trial advocates; Peter Attia historically, who later softened his stance) , Metformin became a flagship "geroprotector" in longevity circles largely on one confounded observational study (Bannister 2014) and the high-profile, NIH-discussed TAME trial. Attia publicly walked back his enthusiasm. This claim is distinct from metformin's genuine, well-established role as a first-line type 2 diabetes drug.
No completed trial has ever tested whether metformin slows aging in people without diabetes, the decisive study was designed but never launched. The mechanism is plausible, the best animal data are null, and there is a randomized signal that it may blunt the gains from exercise.
It lowers a blood-sugar drug's familiar markers in people who don't need it, and says nothing about whether they live longer, while one randomized trial hints it may dull the very exercise gains that do.
What it’s supposed to target
- AMPK
- Mitochondrial complex I
- Hepatic glucose output
- Insulin sensitivity
Metformin's best-described action is to mildly inhibit mitochondrial complex I, which raises the cell's AMP:ATP ratio and switches on AMPK, the master sensor of low energy. Activated AMPK tells the liver to make less glucose and tissues to take up more, which is how it lowers blood sugar. The longevity pitch extends that logic: AMPK activation overlaps with the response to caloric restriction and exercise (more autophagy, tidier mitochondria, lower mTOR), so the theory is that metformin flips some of the same pro-longevity switches.
In worms and some mouse strains metformin can extend lifespan, and large observational datasets suggested diabetics taking it sometimes outlived non-diabetic controls, the hint that launched the TAME trial. But the mechanism is messier than the slogan: the concentrations that activate AMPK in a dish exceed typical clinical levels, effects differ by tissue and strain, and metformin can blunt some of the gains from exercise. Cheap and plausible, but the human anti-aging claim still rests on a trial that has not yet reported.
Mechanism is theory, not proof. A plausible pathway explains why something might work, not whether it does. The verdict rests on the evidence below, not the elegance of the theory.
What would have to be true
Metformin's AMPK/mTOR/mitochondrial effects would have to translate into slower aging in humans, plausible mechanistically, but this link is untested in people.
That benefit would have to appear in people without diabetes, not just diabetics, a link never tested in any completed trial.
The benefit would have to outweigh harms, this link currently fails, given a randomized signal that metformin blunts exercise adaptation.
The supportive human data would have to be sound, a weak link, since the key study is observational, confounded, and (per secondary reports) industry-funded.
What the evidence actually shows
The decisive human trial has never been run
The only study designed to settle the question, TAME (Targeting Aging with Metformin), a roughly 3,000-person, ~6-year randomized controlled trial across 14 sites coordinated by Wake Forest, exists only on paper. The American Federation for Aging Research describes it as awaiting "visionary donors to help launch" the trial; it has not enrolled and has reported nothing. Aging is not a recognized regulatory indication and metformin is generic and cheap, so no conventional sponsor has an approval pathway to fund it. The result is a genuine evidence vacuum: the human anti-aging claim is untested, not refuted. The most-cited supportive datapoint, Bannister 2014, is a retrospective UK database study reporting that diabetics on metformin had survival comparable to or slightly better than matched non-diabetic controls, a finding its own authors caution cannot establish lifespan extension, confounded by healthy-user and indication bias.
The best independent data are null, or point to harm
Where rigorous, disinterested evidence exists, it does not favour the claim. In the NIH-funded NIA Interventions Testing Program (Strong 2016), metformin alone did not significantly extend lifespan in genetically heterogeneous mice (+7% median in males, P=0.35; no effect in females); only the metformin-plus-rapamycin combination extended life (+23%), and the paper notes that combination did not significantly beat rapamycin alone. More pointedly, a double-blind RCT in older adults (Konopka 2019, n=53, 12 weeks of aerobic training) found metformin abolished the exercise-induced rise in skeletal-muscle mitochondrial respiration and roughly halved the VO2max gain (attenuated ~50%, p=0.08). That trial was funded by Dairy Management and Dexcom, parties with no stake in making metformin look bad, which strengthens the unfavourable signal. Separately, the randomized HOME trial in diabetics found metformin lowered vitamin B12 by ~19% and raised the absolute risk of deficiency by ~7 percentage points.
Studies, graded, and who paid
The pivotal TAME trial is designed but, as of 2026, has not enrolled or reported. The relevant human question is open.
In the NIA Interventions Testing Program, metformin alone did not significantly extend mouse lifespan (+7% males, P=0.35; no female effect).
A randomized trial found it blunted exercise-induced fitness and muscle adaptations; B12 depletion and GI effects are documented in diabetics.
| # | Study | Type | Size | Funding / COI | Key limitations |
|---|---|---|---|---|---|
| 1 | Bannister 2014, UK CPRD observational cohort | Retrospective observational cohort | ~78,241 metformin + ~12,222 sulphonylurea vs ~90,463 non-diabetic controls; 503,384 person-years | Funding unknown Often cited as AstraZeneca/Bristol-Myers Squibb funded; PubMed lists only 'Research Support, Non-U.S. Gov't' and the publisher disclosure is paywalled. Industry funding asserted on secondary sources only, not independently verified. | Observational; healthy-user, prevalent-user and indication bias; comparison structure cannot establish lifespan extension. Authors themselves caution against an anti-aging reading. |
| 2 | Strong 2016, NIA Interventions Testing Program (mice) | Randomized controlled animal lifespan study (3 sites) | Genetically heterogeneous mice; multi-cohort across Jackson Lab, U Michigan, UT Health San Antonio | Independent US National Institute on Aging (AG013319, AG022303/07/08, AG024824). No industry funding. | Animal model; dosing/timing differs from human protocols. Note: combo benefit vs rapamycin alone was not statistically significant (P=0.12), so 'attributable to rapamycin' is suggestive, not proven. |
| 3 | Konopka 2019, exercise-adaptation RCT | Double-blind randomized controlled trial | 53 older adults (~62 yrs): placebo n=26, metformin n=27; 12 weeks | Mixed Dairy Management Inc. and Dexcom Inc.; authors declared no conflicts. Not funded by metformin makers and funders had no stake in a negative result, strengthens credibility. | Small (n=53); the VO2max attenuation did not reach significance (p=0.08). Suggestive signal of harm, not definitive. |
| 4 | Mohammed 2021, critical review | Critical narrative review | N/A (review) | Funding unknown Funding not stated in accessed text; authors academic pharmacologists (Weill Cornell-Qatar / U Calgary). | Narrative synthesis, not primary data. Concludes the human lifespan-extension evidence in disease-free people 'remains controversial.' |
| 5 | AFAR, TAME trial (status) | Planned RCT (not yet enrolling) | Planned ~3,000 adults aged 65-79, ~6 years, 14 sites; not yet enrolled | Mixed Coordinated by AFAR; reliant on philanthropy and NIH negotiations. No results to bias; listed for status only. | Not launched; no data. Documents why high-quality human evidence is absent. |
| 6 | de Jager 2010 (HOME trial), B12 outcome | Randomized placebo-controlled trial | 390 patients; ~4.3 years | Independent Investigator-led HOME trial. Demonstrates harm, so no favourable-sponsor concern. | Conducted in diabetics; absolute risk-benefit in healthy non-diabetics is inferred. B12 fell ~19%; absolute deficiency risk rose ~7 percentage points. |
| 7 | Infante 2021, B12 safety review | Narrative/mini-systematic review | N/A (review) | Funding unknown Authors Infante M, Leoni M, Caprio M, Fabbri A; declared no competing interests; no funding stated. | Review, not primary data; corroborates GI intolerance and long-term B12 depletion but does not itself report the 19%/7pp figures (those are de Jager's). |
This is the recurring longevity pattern: a plausible mechanism and one confounded, much-cited human study standing in for the randomized trial that was never funded.
Unproven ≠ disproven
Unproven is not disproven. Metformin may yet help healthy people, but no completed trial has measured it, and 'Unproven' marks an open question, not a verdict of failure.
Where claim and evidence diverge
The gap between a blood-sugar drug with attractive cell-signalling effects and a proven human anti-aging therapy is exactly the trial (TAME) that has never enrolled a single participant.
The money trail
The flagship supportive study (Bannister 2014) is widely reported as funded by diabetes-drug makers AstraZeneca and Bristol-Myers Squibb, though we could not verify this from the paper itself (publisher disclosure paywalled), so we flag rather than assert it.
The independent, publicly funded evidence cuts the other way: the NIH-funded mouse study is null for metformin alone, and the dairy/device-funded exercise RCT, whose funders had no stake in a bad result, found a signal of harm.
The decisive TAME trial has stalled for years over money; because aging is not a regulatory indication and metformin is generic, no commercial sponsor has reason to fund it, which is the structural reason the human evidence is still missing.
The honest read
For people without diabetes, the anti-aging claim is unproven: the decisive trial has not run, the supportive human data are confounded, the best animal data are null for metformin alone, and there is a randomized signal it may blunt exercise gains. None of this touches metformin's separate, well-supported use in type 2 diabetes.
What would change this verdict
A completed, adequately powered RCT (such as TAME) showing metformin delays age-related disease or mortality in non-diabetic adults.
Replicated randomized evidence that metformin does not blunt exercise adaptation, neutralising the current harm signal in active people.
Sources
- Bannister CA, Holden SE, Jenkins-Jones S, et al. Can people with type 2 diabetes live longer than those without? A comparison of mortality in people initiated with metformin or sulphonylurea monotherapy and matched, non-diabetic controls. Diabetes Obes Metab. 2014;16(11):1165-1173.
- Strong R, Miller RA, Antebi A, et al. Longer lifespan in male mice treated with a weakly estrogenic agonist, an antioxidant, an alpha-glucosidase inhibitor or a Nrf2-inducer. Aging Cell. 2016;15(5):872-884. (NIA Interventions Testing Program)
- Konopka AR, Laurin JL, Schoenberg HM, et al. Metformin inhibits mitochondrial adaptations to aerobic exercise training in older adults. Aging Cell. 2019;18(1):e12880.
- Mohammed I, Hollenberg MD, Ding H, Triggle CR. A Critical Review of the Evidence That Metformin Is a Putative Anti-Aging Drug That Enhances Healthspan and Extends Lifespan. Front Endocrinol. 2021;12:718942.
- American Federation for Aging Research (AFAR). TAME, Targeting Aging with Metformin (trial design; Barzilai et al.).
- de Jager J, Kooy A, Lehert P, et al. Long term treatment with metformin in patients with type 2 diabetes and risk of vitamin B-12 deficiency: randomised placebo controlled trial (HOME trial). BMJ. 2010;340:c2181.
- Infante M, Leoni M, Caprio M, Fabbri A. Long-term metformin therapy and vitamin B12 deficiency: An association to bear in mind. World J Diabetes. 2021;12(7):916-931.
- Attia P. A recent metformin study casts doubts on longevity indications (commentary on the Keys et al. 2022 replication study).
People also ask
- Does metformin slow aging in people without diabetes?
- This is unproven. No completed randomized trial has ever tested metformin for anti-aging in non-diabetic humans; the pivotal TAME trial was designed but, as of 2026, has not enrolled or reported. The human question remains genuinely open.
- Does metformin extend lifespan in animals?
- Not reliably on its own. In the NIA Interventions Testing Program, metformin alone did not significantly extend mouse lifespan, showing only a non-significant 7% gain in males and no effect in females. The best animal data are essentially null.
- Can metformin interfere with exercise benefits?
- There is a randomized signal that it can. A trial found metformin blunted exercise-induced fitness and muscle adaptations. B12 depletion and gastrointestinal effects are also documented in people with diabetes, so it is not consequence-free for healthy, active people.
- Why has the TAME metformin trial never run?
- Money and regulation. Aging is not a recognized regulatory indication and metformin is generic, so no commercial sponsor has reason to fund it. The trial has stalled for years, which is the structural reason the human anti-aging evidence is still missing.
Compare head to head: Rapamycin vs metformin: which has the stronger longevity case?
Caveat is journalism, not medical advice. We check public claims against published evidence; we don’t diagnose, treat, or tell you what to take.