Geroprotectors

The "longevity molecule" aisle — NMN, resveratrol, rapamycin, urolithin A, spermidine and the rest — is rich in mouse data and mechanism and consistently disappointing in human trials. The few agents with hard human outcomes are prescription drugs proven in people who are already ill, and the gap between promise and proof has widened, not narrowed.

For a healthy adult, skip the category: no supplement on the market has been shown to extend life at anything like the scale of exercise, Mediterranean-style eating or quitting smoking. If you want to experiment, do it with food — legumes, mushrooms, berries, walnuts, broccoli — and if you want a prescription agent, do it with a knowledgeable clinician, not a marketing site.

Why this category keeps disappointing

The pattern repeats: rich mechanism and animal data, then thin or null human trials. Five more years of research have on balance deflated most of these claims rather than confirmed them, and industry funding distorts the published literature significantly. Some agents still carry stronger signals than others — rapamycin, SGLT2 inhibitors and GLP-1 drugs well ahead of resveratrol or NMN at typical over-the-counter doses. For the mitochondrial and metabolic markers several of these are sold on, see Metabolic flexibility.

Rapamycin: the best candidate, and still an experiment

Rapamycin blocks mTOR, the cell's central nutrient-sensing switch. It reliably extends median lifespan in mice, by roughly 10–25% depending on dose and sex (females gain more), replicated across all three sites of the US National Institute on Aging's Interventions Testing Program.[1]

The human record is much thinner, and there is still no trial with hard outcomes in healthy adults. Related mTOR-blocking drugs (not rapamycin itself) modestly improved older adults' flu-vaccine response in small trials,[2] but a 1,000-person phase 3 trial found no reduction in respiratory illness.[3] A review of 19 human studies found gains in some immune, heart-and-vessel and skin measures but none overall in muscle, hormones or the brain; people who were already ill had more infections and higher cholesterol and triglycerides.[4]

PEARL, the largest and longest trial to date of weekly low-dose rapamycin in healthy adults, brought no more side effects than placebo over 48 weeks, though rapamycin users seemed to have more stomach upsets, but it missed its main goal, and its compounded drug delivered only about a third as much rapamycin as standard generic sirolimus,[5] so its reassuring safety record does not carry over to the 5–6 mg a week of standard sirolimus some clinicians prescribe off-label. The one placebo-controlled trial at that dose is not encouraging: in 40 people aged 65–85, rapamycin added nothing to an exercise programme, hinted at blunting its gains and, though the same share of each group (85%) had side effects, the rapamycin group logged more of them (99 against 63), including one pneumonia that may have been caused by the drug.[6]

Rapamycin is a licensed immunosuppressant. Its known side effects are mouth ulcers, raised blood fats, disturbed blood sugar and lowered immunity, and the weekly-pulse schedule is an attempt to avoid the immunosuppression, not a demonstrated way of doing so. It is not approved for longevity, and dosing for healthy adults is far from worked out. It remains experimental — reasonable only for informed adults working with a knowledgeable physician.

Metformin: a diabetes drug that may cost you your training

Metformin, a common, cheap starting drug for type 2 diabetes, switches on AMPK, the cell's energy-shortage signal and a counterweight to mTOR. Its longevity reputation rests on long-term trial follow-up and records data that do not hold up well as evidence about aging. In people with diabetes it is linked to lower death rates than older sulfonylurea drugs — a low bar, since sulfonylureas are themselves associated with harm. The claim that metformin users outlive people without diabetes comes from one heavily criticised 2014 UK records analysis,[7] and a Danish national study that tested it again found people starting metformin had higher death rates than matched people without diabetes.[8] Treat the survival claim as an association with a plausible confound. The trial meant to settle it in healthy people, TAME (Targeting Aging with Metformin), has never enrolled anyone and is far short of its funding.[9]

There is also a concrete reason for an active adult to avoid it. In two trials in older adults, metformin weakened the fitness and mitochondrial gains from aerobic training[10] and the muscle (though not the strength) gained from strength training.[11]

For healthy adults the balance favours not taking it off-label. Even for prediabetes, lifestyle comes first — aiming for 5–7% weight loss and at least 150 minutes a week of moderate activity — with metformin considered for those at highest risk.[12] In type 2 diabetes it remains a common starting drug, but people with heart disease, high cardiovascular risk or kidney disease are now steered to an SGLT2 inhibitor or a GLP-1 drug whatever their blood sugar.[13]

The drugs with real outcome data — in people who are already ill

SGLT2 inhibitors (empagliflozin, dapagliflozin) make the kidneys pass glucose into the urine. They reduce deaths from any cause in large trials of people with type 2 diabetes and heart disease (5.7% died against 8.3%, about a third fewer),[14] heart failure with reduced pumping (about 17% fewer deaths)[15] and chronic kidney disease (4.7% against 6.8%).[16] Pooled across five heart-failure trials the drop in deaths is smaller, about 8%; the clearer benefit there is fewer hospital stays.[17] They are approved only for those three conditions, cause more urinary-tract and genital yeast infections and, rarely, diabetic ketoacidosis, and have no safety or benefit data in healthy adults. The first placebo-controlled trial in healthy older people is still recruiting.[18] As a longevity drug for the healthy, the evidence is insufficient.

GLP-1 drugs (semaglutide, tirzepatide) also have real outcome data, moderate and growing, though thinner than SGLT2's. In people with heart disease and overweight or obesity but no diabetes, semaglutide cut heart attacks, strokes and cardiovascular deaths combined by 20%,[19] and it produces durable double-digit weight loss. Tirzepatide's evidence is narrower: in people with obesity and one kind of heart failure it cut heart-failure events, but not cardiovascular deaths.[20]

Semaglutide also slowed an epigenetic aging clock in a randomised trial, by about 2.3 clock-years per calendar year over 32 weeks in 84 adults with HIV-related fat redistribution.[21] That was a small after-the-fact analysis in an unusual group, scored by the company that sells the clock, and omega-3 supplements had already nudged such clocks slightly in an earlier trial[22] — suggestive, not settled.

The same drugs cause meaningful muscle and bone loss unless you eat enough protein and train. With overweight and cardiometabolic risk, they are worth discussing seriously with a doctor; for a healthy normal-weight adult chasing longevity, off-label "microdosing" is unproven and carries real compounding-pharmacy risks. Full treatment in GLP-1 receptor agonists.

NAD+ boosters: the blood level goes up, nothing else does

NMN and NR reliably raise blood NAD+ at 250–1,000 mg a day. The benefit that is supposed to follow keeps failing to appear. One small trial found better muscle insulin sensitivity in postmenopausal women with prediabetes on 250 mg of NMN,[23] but pooled, eight trials in 342 adults found NMN moved no blood-sugar, insulin or blood-fat measure,[24] trials in people over 60 found no gain in muscle or physical function from either,[25] and adding NR to a walking programme in people with high blood pressure lowered it no more than walking alone.[26]

There is also an open safety question. NR and NMN, like niacin, raise blood levels of 2-PY and 4-PY, the end products of NAD+ breakdown. In observational studies of heart patients, people with higher levels had more heart attacks, strokes and deaths over three years, and 4-PY inflamed blood-vessel walls in mice.[27] Whether supplement-driven rises cause harm is unknown. There are no long-term human safety data, and a theoretical worry about feeding an undetected cancer has not been settled either way.

NMN has been legal to sell as a US supplement again since September 2025, which says nothing about whether it works.[28] Save the money; the wider vitamin B3 picture is in B vitamins.

Interesting molecules, thin human evidence

Urolithin A is made by gut bacteria from the ellagitannins in pomegranate, berries and walnuts, and switches on mitophagy, the recycling of damaged mitochondria. It has the best randomised data of any novel longevity compound — most of it paid for by its maker — but the largest trial, in 88 overweight middle-aged adults, raised hamstring (but not thigh-front) strength while missing its main goal, peak power,[29] and a review of five trials could reach no firm conclusion.[30] Roughly 1 in 10 people make none from food.[31] No trial has yet shown it helps older adults, including those losing muscle, where resistance training and enough protein come first; for most people, pomegranate and walnuts are the cheaper bet.

Spermidine. In an observational study of one town's residents, people in the top third of dietary intake had 39% lower death rates than those in the bottom third.[32] But the largest human trial, twelve months in 100 older adults, was negative on its main cognitive measure,[33] and 15 mg a day for five days did not raise blood spermidine at all.[34] Get it from legumes, mushrooms, aged cheese, wheat germ and natto rather than pills.

Senolytics. Dasatinib plus quercetin, a protocol to clear worn-out "senescent" cells, reduced their markers in small pilots, but the first randomised trial, in postmenopausal women, missed its main goal.[35] Dasatinib is a prescription leukaemia drug; this is not a home regimen. Fisetin, marketed as an elite senolytic, rests on lab data: in the only placebo-controlled trial with posted results, in 74 people with knee osteoarthritis, it did no better than placebo on pain, function or inflammation;[36] in a bone study in older women it did not lower a bone-loss marker compared with no treatment;[37] and in the Interventions Testing Program it did not extend mouse lifespan in either sex.[38] Quercetin alone lowers systolic blood pressure by roughly 2–3 mmHg, with very-low-to-moderate certainty.[39]

Astaxanthin extended median lifespan about 12% in male mice in the same testing program, with no significant effect in females.[40] Human data stop at skin and metabolic markers.

Taurine extended median mouse lifespan by 10–12% in both sexes,[41] but its central premise — that blood taurine falls with age — failed in humans, monkeys and mice.[42] The 2025 study did not test supplements, so the mouse result stands; there is no human longevity trial.

Calcium alpha-ketoglutarate extended remaining lifespan in female mice started in late middle age by about 10–17%, in a study whose authors held shares in the company selling it.[43] The headline human claim — an average 8-year drop in epigenetic age in 42 customers — is uncontrolled, unblinded and unreplicated, and these clocks are noisy in individuals.[44] A 120-person placebo-controlled trial has finished collecting data, but its results are not yet published.[45]

Sulforaphane, from broccoli sprouts, switches on the cell's own antioxidant defences, but a prediabetes trial missed its blood-sugar target and no trial has looked at aging.[46] Eat the broccoli.

Old promises that didn't hold

Resveratrol is the field's cautionary tale. It is almost entirely broken down on its first pass through the gut and liver, so only traces reach the blood.[47] Two decades of trials are largely null,[48] the most positive one found small bone gains as a secondary outcome with maker funding,[49] and it failed to extend lifespan in the mouse testing program where rapamycin succeeded.[50] Red wine contains too little to matter. A diet rich in varied polyphenols is the better bet.

GlyNAC (glycine plus N-acetylcysteine) rests mostly on one lab's pilot trials — 12 people per arm in the main randomised one.[51] The only placebo-controlled trial from another group, run by Nestlé in 114 healthy older adults, found no rise in glutathione overall, though at a lower dose for only two weeks.[52]

Berberine at 500 mg two or three times a day genuinely lowers blood sugar and cholesterol in the short term, comparably to metformin, cutting LDL cholesterol by about 15 mg/dL. But it has no diabetes-prevention, cardiovascular-outcome or longevity trials, whatever the "natural metformin" marketing says.

Low-dose lithium orotate. Places with more lithium in their drinking water tend to have lower suicide rates, but the studies disagree a lot and publication bias may inflate the link.[53] For dementia the picture is mixed, with a Scottish cohort of 37,597 people finding no protection, although lithium levels in its water were very low.[54] Small trials in people who already had memory problems point to slower decline,[55] and a 2025 paper found brain lithium depleted in mild cognitive impairment and Alzheimer's, and low-dose lithium orotate reversed the disease's pathology in mice.[56] No trial has tested lithium in healthy adults, or used the orotate form in people. The longevity case is a hypothesis.

What to actually do

  1. Skip the category. Sleep, exercise, diet and the core supplements capture most of the realistic benefit at a fraction of the cost and risk.
  2. If you want to experiment, use food. Spermidine from legumes, mushrooms and aged cheese; polyphenols from berries, olive oil, dark chocolate and tea.
  3. If you want a prescription geroprotector, see a clinician who knows the field. Rapamycin, SGLT2 inhibitors and GLP-1 drugs carry the strongest signals — the last two with outcome trials only in people with disease — and all have real side effects.

The online "longevity stack" of NMN, resveratrol, urolithin A, spermidine, lithium orotate and ten other compounds has no support from human randomised trials.

Who should be careful

This is the one category on the site where the agents are active enough to hurt you, and several are prescription drugs used off-label. Anyone on blood thinners, immunosuppressants or diabetes drugs, or with a history of cancer, should treat the whole category as prescription territory.

  • Drug interactions run through CYP3A4, the liver enzyme that clears several statins, many calcium-channel blockers, several blood thinners and most immunosuppressants. Rapamycin, berberine and grapefruit all act on or through it. Stacking them is how people get unexpected drug levels; have a pharmacist review anything taken alongside prescription medicine.
  • Rapamycin. It is cleared by CYP3A4, so grapefruit and many common drugs can change its levels. Not in pregnancy or during an active infection; no live vaccines while on it; it slows wound healing, so stop before planned surgery. Check blood fats, fasting glucose or HbA1c (a longer-term blood-sugar marker) and a full blood count periodically.
  • Metformin. Can cause vitamin B12 deficiency with long-term use (check it periodically) and stomach upset, and blunts training gains. Not for people with significant kidney impairment.
  • Berberine. In healthy men it raised blood levels of a CYP3A4 test drug about 40% and also inhibited CYP2D6 and CYP2C9;[57] it may also block the drug pump P-glycoprotein. Interactions are plausible with statins, cyclosporine, dabigatran, many antidepressants, metoprolol, warfarin and many blood-pressure drugs. Combined with metformin, a sulfonylurea or insulin it risks low blood sugar. Stomach side effects affect 10–35% of users. Not in pregnancy.
  • Quercetin. 500 mg a day raised pravastatin levels by about a quarter,[58] its effect on CYP3A4 is inconsistent, and there is one case report of a stronger warfarin effect. Check with a pharmacist if you take statins, blood thinners or transplant drugs.
  • Dasatinib is a chemotherapy drug, with the bleeding, low-blood-count and fluid-around-the-lung risks that implies. It has no place in unsupervised use.
  • Lithium. No trial has tested the long-term safety of over-the-counter lithium, and products vary widely; in a survey of 211 people taking supplement lithium, side effects and withdrawal symptoms were more common than expected.[59] Lithium orotate most likely breaks down into ordinary lithium in the stomach,[60] which at prescription doses can harm the kidneys and thyroid. Talk to a doctor first if you take anti-inflammatory painkillers (NSAIDs), ACE inhibitors or diuretics, which can raise lithium levels, or have kidney or thyroid disease.
  • Niacin, NR and NMN share the same breakdown products, so the 2-PY/4-PY question above applies to all three.

Further reading

  • López-Otín C et al. Hallmarks of aging: An expanding universe. Cell 2023.[61]
  • Mannick JB, Lamming DW. Targeting the biology of aging with mTOR inhibitors. Nature Aging 2023.[62]
  • Roark KM, Iffland PH. Rapamycin for longevity: the pros, the cons, and future perspectives. Front Aging 2025.[63]
  • Moel M et al. Influence of rapamycin on safety and healthspan metrics after one year: PEARL trial results. Aging 2025.[64]
  • Stanfield B et al. Exercise and Weekly Sirolimus (Rapamycin) in Older Adults: RAPA-EX-01 Randomised, Double-Blind, Placebo-Controlled Trial. J Cachexia Sarcopenia Muscle 2026.[65]
  • Mannick JB et al. TORC1 inhibition enhances immune function and reduces infections in the elderly. Sci Transl Med 2018.[66]
  • Yoshino M et al. Nicotinamide mononucleotide increases muscle insulin sensitivity in prediabetic women. Science 2021.[67]
  • Ferrell M et al. A terminal metabolite of niacin promotes vascular inflammation and contributes to cardiovascular disease risk. Nat Med 2024.[68]
  • Singh A et al. Urolithin A improves muscle strength, exercise performance, and biomarkers of mitochondrial health in a randomized trial in middle-aged adults. Cell Rep Med 2022.[69]
  • Schwarz C et al. SmartAge: Spermidine and Cognition. JAMA Network Open 2022.[70]
  • Aron L et al. Lithium deficiency and the onset of Alzheimer's disease. Nature 2025.[71]
  • Asadi Shahmirzadi A et al. Alpha-ketoglutarate, an endogenous metabolite, extends lifespan and compresses morbidity in aging mice. Cell Metab 2020.[72]
— § —