Comparison
Longevity supplements and drugs 2026: what is proven in humans
21 substances from metformin to NMN: what is approved, what has been shown in humans and where the data stop.
6 min read updated November 15, 2026
The short answer
The best-studied are approved medicines such as metformin, acarbose, rapamycin and canagliflozin: they are well supported for their actual diseases, and they extended lifespan in mice. Among supplements, NMN and NR reliably raise NAD+ levels. That any of these slows ageing or extends life in humans has not been shown so far, and resveratrol did not hold up in trials. The best human evidence is for unspectacular things such as exercise, muscle strength and sleep (as of 15 November 2026).
All substances at a glance
| Substance | Status | Shown in humans | BK-Score |
|---|---|---|---|
| Approved medicines, used off-label for longevity | |||
| Metformindiabetes drug | prescription (type 2 diabetes); longevity use off-label | well established for diabetes; ageing benefit in healthy people open, the TAME trial is planned | Well supported, heavily overhyped Human evidence 8/10 |
| Acarbose (Glucobay)diabetes drug | prescription (type 2 diabetes); longevity use off-label | established for diabetes; longer life so far only in mice (ITP) | Well supported, heavily overhyped Human evidence 8/10 |
| Rapamycin (sirolimus)mTOR inhibitor | prescription (transplantation); longevity use off-label | strongest life extension in mammals (mice); only small studies in humans | Thin human evidence Human evidence 5/10 |
| Canagliflozin (Invokana)SGLT2 inhibitor | prescription (type 2 diabetes); longevity use off-label | heart and kidney protection in patients; longer life in male mice (ITP) | Well supported, heavily overhyped Human evidence 9/10 |
| Dasatinib + quercetinsenolytic combination | dasatinib prescription-only (leukaemia); not approved as a senolytic | mechanism shown in tissue samples; benefit in one of three controlled trials | Thin human evidence Human evidence 5/10 |
| Supplements and foods | |||
| Nicotinamide riboside (NR)NAD+ precursor | EU: authorised as a novel food | reliably raises NAD+ in blood; no clinical benefit shown | Supported, with caveats Human evidence 6/10 |
| NMNNAD+ precursor | EU: safety assessed positively in 2026, authorisation pending | raises NAD+ in blood; no clinical benefit shown | Thin human evidence Human evidence 5/10 |
| Urolithin Agut metabolite | EU: novel food application filed | small, manufacturer-funded trials on mitochondria and muscle | Thin human evidence Human evidence 5/10 |
| Spermidinepolyamine | authorised as wheat germ extract (novel food) | only one pilot study in humans | Thin human evidence Human evidence 5/10 |
| Taurineamino sulfonic acid | allowed as a food ingredient | blood pressure slightly lowered; the longevity claim was corrected in 2025 | Supported, with caveats Human evidence 6/10 |
| Ergothioneinemushroom antioxidant | EU: authorised as a novel food since 2017 | small, short trials without hard endpoints | Thin human evidence Human evidence 5/10 |
| GlyNAC (glycine + NAC)glutathione building blocks | glycine freely available; NAC is a medicine in Germany | one small trial from a single research group | Hype far ahead of evidence Human evidence 3/10 |
| Fisetinflavonoid, senolytic candidate | not marketable in isolated form in the EU | some inflammation markers lowered; senolytic effect not shown | Thin human evidence Human evidence 4/10 |
| Resveratrolpolyphenol | available as a supplement | longevity benefit not confirmed, trials mostly negative | Studied – no benefit shown Human evidence 4/10 |
| Pterostilbeneresveratrol relative | EU: not authorised | in one larger trial blood pressure fell and LDL rose | Thin human evidence Human evidence 4/10 |
| Ca-AKGcitric acid cycle molecule | available as a supplement | hardly any human data; no life extension in the independent mouse test | Hype far ahead of evidence Human evidence 2/10 |
| Apigeninflavone from chamomile | not marketable in isolated form in the EU | human trial completed, results unpublished | Tested in humans, results unpublished Human evidence 1/10 |
| Infusions, sprays and research substances | |||
| NAD+ infusioninfusion | offered in clinics; not approved for anti-ageing | one positive trial in heart failure; no controlled trial for energy or anti-ageing | Hype far ahead of evidence Human evidence 3/10 |
| NAD+ nasal spray and injectionNAD+ directly | not an approved medicine | no efficacy trials | Hype far ahead of evidence Human evidence 2/10 |
| 17α-estradioloestrogen isomer | sold as a scalp solution; not approved for oral use | longer life in male mice; oral use not studied in humans | Hype far ahead of evidence Human evidence 2/10 |
| SLU-PP-332"exercise mimetic" | not approved anywhere, preclinical only | mouse data only | Not studied in humans Human evidence 0/10 |
How to read the table
- Mice are not humans. Many longevity promises come from the US Interventions Testing Program (ITP), in which mice lived longer. No completed trial has yet tested whether this works the same in humans.
- A lab value is not a benefit. NMN and NR reliably raise NAD+ in blood. Whether that makes you age more healthily is a different question, and it is open.
- The BK-Score rates the state of knowledge. "Well supported, heavily overhyped" for metformin means: the diabetes data are solid, but the anti-ageing promise goes far beyond them.
- Medicines belong in medical hands. For drugs used off-label we give no amounts.
What has been shown to work
The strongest human data exist for things that cost little or nothing: regular exercise and strength training, enough sleep, not smoking, little alcohol and a plant-based diet, plus treating high blood pressure and raised blood lipids. A starting point is Longevity: Where do I start?.
Head-to-head comparisons
NMN vs NR · NMN vs NAD+ infusion · Metformin vs rapamycin · Resveratrol vs pterostilbene · Urolithin A vs spermidine · Fisetin vs quercetin
Frequently asked questions
Which longevity supplement has the best evidence?
None has been shown to extend life in humans. The most solid lab data exist for NR and NMN, which reliably raise NAD+ levels, and for urolithin A with small randomised trials on mitochondria and muscle. A benefit for health or lifespan has not been shown for any of them.
Does NMN do anything?
NMN raises NAD+ levels in the blood. Whether this leads to healthier ageing has not been shown. EFSA assessed beta-NMN in 2026 for safety only, not for efficacy; it is not yet authorised in the EU.
Is metformin an anti-ageing drug?
Metformin is a well-studied diabetes drug. Whether it slows ageing in healthy people is what the planned TAME trial is meant to find out; there are no results yet. For healthy people it is prescription-only and off-label.
What has been shown to help you live longer?
The best human evidence is for lifestyle factors: regular exercise, muscle strength, enough sleep, not smoking and a plant-based diet, plus treating high blood pressure and raised blood lipids.
Information only, not medical advice and no usage or dosage recommendation. Prescription and unapproved drugs belong in medical hands.