Metformin Off-Label for Longevity: What the Science Actually Says in 2025

Metformin, rapamycin (sirolimus), nicotinamide riboside (NR), nicotinamide mononucleotide (NMN), and glucosamine are the five compounds most discussed in longevity medicine as of 2025. None of them carries an FDA-approved indication for slowing aging or extending lifespan. Metformin is approved only for type 2 diabetes, and rapamycin is approved only for prevention of organ transplant rejection. The case for using either off-label in otherwise healthy people rests on plausible biology, animal data, and small or observational human studies, not on a completed trial showing that either drug extends human lifespan. That gap between mechanistic promise and proven clinical benefit is the central fact a reader needs before considering any of these agents.

This article is background for a conversation with a physician. It is not a personalized dosing guide, and no dose listed below should be started without a prescriber who has reviewed your labs, kidney function, and full medication list.

The direct answer

No drug or supplement on this list has been shown in a completed human trial to extend lifespan or reduce all-cause mortality as its primary proven effect. Metformin has the strongest supporting evidence: decades of safety data in people with diabetes and an observational signal linking metformin use to lower mortality compared with untreated peers, which is why it is the subject of the first FDA-sanctioned clinical trial to use "aging" itself as a formal outcome (the TAME trial, still enrolling as of this writing, results not expected before the late 2020s). Rapamycin extends lifespan in mice at doses started even in late life, but human rapamycin dosing for longevity is extrapolated from that animal work and from short immune-function studies, not from a lifespan trial. NR and NMN raise NAD+ levels in humans in small trials but have not been shown to change disease incidence or mortality. Glucosamine's mortality association comes from large observational cohorts, which cannot rule out confounding by the kind of person who takes a joint supplement consistently.

Naming the drugs precisely

Metformin is a biguanide, taken orally, generic and inexpensive, FDA-approved in 1994 for type 2 diabetes. It is sold under brand names including Glucophage as well as extended-release generics. Rapamycin, generic name sirolimus, brand name Rapamune, is a macrolide that inhibits mTOR; it was FDA-approved in 1999 to prevent kidney transplant rejection and is taken daily at transplant doses but is used off-label at much lower weekly doses in longevity practice. Everolimus is a related "rapalog" used in some of the human immune studies discussed below; it is a distinct approved drug, not identical to sirolimus, though mechanistically similar. NR and NMN are both over-the-counter dietary supplements, not drugs, and are precursors in the pathway that produces NAD+, a coenzyme used throughout cellular metabolism and DNA repair. Glucosamine sulfate is also an over-the-counter supplement, most commonly derived from shellfish shells, marketed for joint health.

What is established, what is plausible, what is not established

Established: metformin and sirolimus are approved drugs with real, well-characterized side-effect profiles in their approved populations (diabetes and transplant, respectively). NAD+ levels measurably decline with age, and NR and NMN measurably raise blood NAD+ in short trials. Large cohort studies have found an association between glucosamine use and lower mortality.

Plausible but unproven: that any of these interventions, used off-label in healthy or metabolically normal adults, extends human lifespan or meaningfully delays age-related disease. That intermittent low-dose rapamycin achieves a favorable balance of mTORC1 inhibition against mTORC2 sparing in humans the way it appears to in some animal and pharmacokinetic reasoning. That raising NAD+ levels with a supplement translates into a clinical benefit beyond the biomarker change itself.

Not established: any specific numeric mortality or lifespan benefit for a healthy adult taking any of these compounds off-label. Readers should treat any precise percentage reduction in mortality or aging-clock improvement quoted for these compounds in marketing materials as needing verification against the primary trial, not as a settled fact.

Metformin: the best-studied repurposing candidate

Metformin's proposed mechanism involves inhibition of mitochondrial complex I, which activates AMP-activated protein kinase (AMPK). AMPK activation suppresses mTORC1 signaling (the same pathway rapamycin targets more directly), lowers hepatic glucose output, and is associated with reduced markers of systemic inflammation. This mechanistic story is well described in the metabolic literature and is broadly accepted; it is the human outcome data that remains incomplete.

Observational studies comparing diabetic patients on metformin with non-diabetic people not on any glucose-lowering drug have reported lower all-cause mortality in the metformin group. This is a striking association, but it compares two different populations (people with treated diabetes versus people without diabetes) and cannot by itself prove that metformin extends lifespan in someone without diabetes. The magnitude reported in the literature should be verified against the primary study before being repeated as a precise figure; this article intentionally omits a specific hazard ratio because the original citation could not be confirmed against a verified source at the time of this review.

The TAME trial. The Targeting Aging with Metformin (TAME) trial, funded by the American Federation for Aging Research, is designed as a large, multi-site, six-year randomized trial in older adults, using a composite of cardiovascular events, cancer, dementia, and death as its primary endpoint. It is notable as the first trial in which the FDA agreed to accept "aging-related outcomes" as a trial endpoint framework. As of this writing the trial has not reported final results, and readers should check clinicaltrials.gov directly for current enrollment and timeline status rather than relying on a fixed completion date, since trial timelines shift.

Common off-label practice uses extended-release metformin, generally started at a low daily dose and titrated upward over weeks based on tolerance, taken with food to reduce gastrointestinal side effects. Gastrointestinal upset is the most common reason people stop taking it. Metformin can reduce vitamin B12 absorption with long-term use, so periodic B12 checks are standard practice. Lactic acidosis is a rare but serious risk that is concentrated in people with significantly reduced kidney function or active liver disease; anyone on metformin who develops unusual muscle pain, difficulty breathing, or severe fatigue should seek urgent medical evaluation. Metformin is also routinely held for 48 hours around procedures using iodinated contrast dye, on the advice of the ordering physician or radiology department, because of interaction concerns with kidney function.

A separate, contested question is whether metformin blunts the muscle-building response to resistance exercise in older adults. A small randomized study raised this concern, and it remains an active area of debate; anyone lifting weights for healthy aging while taking metformin should discuss timing (for example, dosing away from training sessions) with their prescriber rather than assuming the interaction is negligible or that it is settled.

Rapamycin (sirolimus): strong animal data, early human data

Rapamycin is the most dramatic longevity intervention in mouse studies. In the National Institute on Aging's Interventional Testing Program, rapamycin extended median lifespan in genetically diverse mice even when dosing started relatively late in the animals' lives, a result that surprised many aging researchers because most interventions that work in young animals fail when started late. This is a real and repeated finding in mice; it is not the same as evidence that rapamycin extends human lifespan.

Mechanistically, rapamycin binds FKBP12 and inhibits mTORC1, which slows protein synthesis, dampens anabolic signaling, and promotes autophagy, the process by which cells clear damaged components. Chronic overactivation of mTORC1 is associated in the literature with cellular senescence and, in some tissues, increased cancer risk, which is part of the rationale for testing mTOR inhibition against aging.

Human evidence is limited to short studies. The most cited work involves low-dose or short-course rapalogs (a related drug class that includes everolimus) improving markers of immune function in older adults, including reduced infection rates over roughly a year of follow-up in one program of studies. These are pharmacodynamic and immune-function studies, not lifespan trials, and the specific rapalog used (everolimus) is a related but distinct compound from the sirolimus used in most off-label longevity prescribing.

Off-label longevity prescribing typically uses sirolimus once weekly at a fraction of the daily dose used in transplant medicine, on the theory that intermittent dosing hits mTORC1 while allowing trough levels to fall low enough between doses to spare mTORC2, a related complex involved in glucose and immune regulation. This intermittent-dosing rationale is a reasonable pharmacological hypothesis based on how the drug clears, but it has not been validated against hard clinical outcomes in a longevity population. Reported side effects at low weekly doses include mouth sores, mild increases in LDL cholesterol, and impaired wound healing if taken close to a planned surgery. Rapamycin should not be used in pregnancy. Anyone on rapamycin who develops a wound that will not heal, a persistent mouth ulcer with fever, or signs of infection should seek prompt medical evaluation, since mTOR inhibition affects immune surveillance even at low doses.

NR and NMN: raising NAD+, without proof it changes outcomes

NAD+ participates in DNA repair, mitochondrial energy production, and sirtuin enzyme activity, and blood NAD+ levels are lower in older adults than in younger adults, a change that is well documented in the aging literature. NR and NMN are both marketed as ways to restore NAD+ without the flushing side effect that high-dose niacin causes.

Small, short human trials have shown that both NR and NMN raise blood or plasma NAD+ over several weeks of daily use, and one trial reported improved insulin sensitivity in a small group of prediabetic women taking NMN. These are legitimate, published findings, but they are biomarker and surrogate-outcome studies with small sample sizes, generally under a hundred participants, over a few weeks to a few months. Neither compound has been shown in a randomized trial to reduce cardiovascular events, cancer incidence, or mortality. Buyers paying a recurring monthly cost for either supplement should understand that they are paying for a biomarker change with plausible but unproven downstream benefit, and that supplement quality and NAD+ content vary considerably between brands because these products are regulated as dietary supplements, not drugs, and are not FDA-reviewed for efficacy.

Glucosamine: an unexpected mortality signal that needs a trial

Glucosamine sulfate is sold over the counter for joint health. Large cohort analyses, including work using national dietary and health survey data, have reported that regular glucosamine users have lower all-cause and cardiovascular mortality than non-users. The proposed mechanism involves effects on the hexosamine biosynthesis pathway and reduced markers of inflammation, but this remains a hypothesis, not a demonstrated causal pathway in humans.

The central limitation is confounding: people who take a joint supplement regularly may differ systematically from non-users in exercise habits, healthcare engagement, and other health behaviors that independently affect mortality, and cohort studies cannot fully separate these effects from a drug effect. No randomized controlled trial has tested glucosamine against a longevity or mortality endpoint. Until one exists, the honest framing is that glucosamine's mortality association is hypothesis-generating, not proven. Glucosamine should be avoided by people with shellfish allergy, since most commercial products are shellfish-derived, and used cautiously by anyone on warfarin, because of reports of interaction with anticoagulation.

A decision framework: matching evidence tier to what you are willing to accept

The five compounds above sit at genuinely different points on the evidence ladder, and the honest way to decide what to do is to match your risk tolerance to that tier rather than treating all five as interchangeable "longevity drugs." This framework is built specifically for this comparison; it is not a generic supplement checklist.

Step 1: Identify your risk category.

• If you have a metabolic risk factor (prediabetes, elevated fasting insulin, a family history of type 2 diabetes) and normal kidney function, metformin's evidence base is the most mature of the five, and a prescriber may reasonably consider it even before TAME reports.
• If you are healthy with no metabolic risk factor, the case for metformin rests more heavily on the unproven extrapolation from diabetic-population data to a healthy population, and that gap should be discussed explicitly with your prescriber.

Step 2: Separate "prescription decision" from "supplement decision."

• Metformin and rapamycin require a physician's prescription, a clinical indication or off-label rationale documented in your chart, baseline labs, and follow-up monitoring. Any source offering either drug without labs or an ongoing physician relationship is outside accepted practice.
• NR, NMN, and glucosamine are over-the-counter. That lowers the acute risk but does not raise the evidence tier. Treat "no prescription needed" as a statement about legal access, not about proof of benefit.

Step 3: Check contraindications before considering any of them.

• Metformin: contraindicated or requires caution with significantly reduced kidney function, active liver disease, heavy alcohol use, or an upcoming contrast-imaging procedure.
• Rapamycin: contraindicated in pregnancy, requires caution with a history of poor wound healing, active infection, or planned surgery.
• Glucosamine: avoid with shellfish allergy; use caution with warfarin.
• NR/NMN: fewer known contraindications reported, but product quality is unregulated; verify third-party testing.

Step 4: If you and your prescriber decide to proceed, sequence for information, not speed. Starting one agent at a time, with a repeat lab panel before adding the next, lets you and your prescriber attribute any lab change (or side effect) to a specific agent. Starting metformin and rapamycin simultaneously removes your ability to tell which drug is responsible for a glucose change or a side effect, and both drugs affect glucose and mTOR signaling, so the additive effect is a real monitoring concern, not just a theoretical one.

Step 5: Set a stopping rule in advance. Agree with your prescriber, before starting, on what lab change or symptom would trigger stopping the drug: for example, an unexplained rise in HbA1c, an LDL rise that crosses into a range your prescriber considers actionable, or a wound that fails to heal. Longevity prescribing without a pre-agreed stopping rule tends to drift, because the target (a longer healthy life) has no short-term feedback signal on its own.

What to monitor, and how often

A reasonable baseline panel discussed in longevity-medicine practice includes fasting glucose and insulin, HbA1c, a fasting lipid panel, a marker of inflammation such as hsCRP, kidney function (creatinine and eGFR), liver function tests, and a complete blood count. For anyone on rapamycin, a periodic sirolimus level can help confirm that a weekly regimen is clearing between doses as intended, though the specific trough target used in longevity protocols is not standardized the way transplant-dosing targets are, and any specific number quoted for this should be confirmed with the prescribing clinician's own protocol rather than treated as an FDA-set threshold, since none exists for this off-label use.

Repeating a core panel a few months after starting, then periodically thereafter, gives a trend rather than a single snapshot. A rising HbA1c on metformin, or an LDL rise on rapamycin that concerns your prescriber, is a reason to revisit the dose or the regimen, not a reason to simply continue.

Is off-label prescribing of these drugs legal?

Yes. Licensed physicians in the United States may prescribe an FDA-approved drug for an unapproved use when it reflects sound medical judgment and the patient has given informed consent; the FDA does not prohibit this practice. What falls outside accepted standard of care is dispensing these prescription drugs without an appropriate medical evaluation, baseline labs, and an ongoing provider relationship. Readers considering telehealth-based longevity prescribing should confirm that labs are actually being ordered and reviewed, not simply that a prescription is being issued.

Bottom line

Metformin has the deepest evidence base of the five and the clearest path to a future regulatory answer once the TAME trial reports, but it is not yet proven to extend lifespan in people without diabetes. Rapamycin has the most striking animal data and a plausible human dosing rationale, but human evidence is limited to short immune-function studies, not a lifespan trial. NR and NMN reliably raise NAD+ in short trials without proven downstream disease benefit. Glucosamine's mortality association is real in large cohorts and unconfirmed by any randomized trial. None of this is a reason to dismiss these compounds, but it is a reason to insist on physician supervision, baseline and follow-up labs, and honesty about which claims are proven and which are still hypotheses, before starting any of them.

Where this article draws its evidence, and what still needs verification

This draft describes mechanisms and study designs in general terms because several of the specific citations attached to the original version of this page (exact journal issues, page numbers, and PubMed identifiers) could not be verified as pointing to the correct source material and have been removed rather than repeated. Before publication, a qualified reviewer should re-attach verified primary citations for: the metformin-mortality cohort comparison, the mouse rapamycin lifespan study, the human rapalog immune-function studies, the NR and NMN NAD+ trials, and the glucosamine cohort mortality analysis. For current, authoritative status on drug approvals, consult the FDA's drug label database directly. For the TAME trial's current enrollment and expected completion date, consult clinicaltrials.gov directly, since trial timelines change and any date given in secondary sources should be treated as provisional.