Longevity Supplements: What the evidence actually shows.
Each supplement and drug is rated outcome by outcome from counted human evidence: randomized trials and meta-analyses indexed in PubMed, each linked with its finding. Animal and laboratory studies are listed separately and never decide a rating.
RatingsSupportedAt least one positive meta-analysis or positive large trial (1,000+ participants), and no counted negative results.MixedCounted evidence points both ways.Early or small trialsOnly small randomized trials or uncontrolled pilots, all positive.Not supported in trialsOnly null results among counted studies.Possible harm in trialsCounted studies show only null or harmful results, including at least one harmful result.No human trialsNo counted human trial evidence for this outcome.
Category
Rating
25 compounds
Prescription drugs
Metformin
Prescription
Prescription diabetes drug · Prescription drugs
Muscle growth with strength training (older adults)Possible harm in trials
HarmRandomized trial In healthy older adults doing resistance training, metformin blunted the gain in muscle size (MASTERS trial). Walton 2019, Aging Cell
Slowing ageing in peopleNo human trials
No human trial results yet.
Other papers (not counted: animal, laboratory, observational or review)
review The rationale for the TAME trial; effects on ageing in people are unproven. Barzilai 2016, Cell Metab
PubMed lists 3,450 human randomized trials and 1,010 meta-analyses mentioning Metformin for any use (searched 2026-10-02; the exact search is in the data file).
Immune function in older adultsEarly or small trials
BenefitRandomized trial In older volunteers, the rapalog RAD001 (everolimus) improved the response to flu vaccination by about 20%. Mannick 2014, Sci Transl Med
BenefitRandomized trial, 264 people In 264 older adults, six weeks of low-dose TORC1 inhibition was followed by fewer reported infections over the next year. Mannick 2018, Sci Transl Med
Slowing ageing in peopleNo human trials
No human trial results yet.
Other papers (not counted: animal, laboratory, observational or review)
animal Extended lifespan in mice, even when started late in life. Harrison 2009, Nature
PubMed lists 1,969 human randomized trials and 466 meta-analyses mentioning Rapamycin and rapalogs for any use (searched 2026-10-02; the exact search is in the data file).
PubMed lists 284 human randomized trials and 36 meta-analyses mentioning Acarbose for any use (searched 2026-10-02; the exact search is in the data file).
Heart attacks, strokes and cardiovascular deathSupported
BenefitRandomized trial, 3,297 people In 3,297 people with type 2 diabetes at high cardiovascular risk, semaglutide lowered cardiovascular death, heart attack or stroke (SUSTAIN-6). Marso 2016, N Engl J Med
BenefitRandomized trial, 17,604 people In 17,604 people with cardiovascular disease and overweight or obesity but no diabetes, semaglutide reduced major cardiovascular events (SELECT). Lincoff 2023, N Engl J Med
Body weightSupported
BenefitRandomized trial, 2,539 people In 2,539 adults with obesity, weekly tirzepatide produced substantial, sustained weight loss over 72 weeks (SURMOUNT-1). Jastreboff 2022, N Engl J Med
BenefitRandomized trial, 3,127 people In 3,127 adults with obesity, the daily pill orforglipron reduced weight by up to 11.2%, versus 2.1% with placebo, over 72 weeks. Wharton 2025, N Engl J Med
PubMed lists 1,255 human randomized trials and 722 meta-analyses mentioning GLP-1 drugs for any use (searched 2026-10-02; the exact search is in the data file).
Flavonoid; tested with the drug dasatinib · Senolytics
Physical function in lung fibrosis (with dasatinib)Early or small trials
BenefitUncontrolled pilot In a small open-label pilot in idiopathic pulmonary fibrosis, dasatinib plus quercetin was linked to better physical function. Justice 2019, EBioMedicine
Senescent cells in people (with dasatinib)Early or small trials
BenefitUncontrolled pilot, 9 people In 9 people with diabetic kidney disease, three days of dasatinib plus quercetin reduced senescent cells in fat tissue. Hickson 2019, EBioMedicine
Other papers (not counted: animal, laboratory, observational or review)
PubMed lists 232 human randomized trials and 81 meta-analyses mentioning Quercetin (with dasatinib) for any use (searched 2026-10-02; the exact search is in the data file).
Heart attacks and strokes, added to statinsNot supported in trials
No benefitRandomized trial, 3,414 people In 3,414 patients with cardiovascular disease and low LDL on statins, adding niacin gave no additional benefit (AIM-HIGH). Boden 2011, N Engl J Med
No benefitRandomized trial, 25,673 people In 25,673 adults with vascular disease, extended-release niacin with laropiprant did not reduce major vascular events and increased serious adverse events (HPS2-THRIVE). Landray 2014, N Engl J Med
Other papers (not counted: animal, laboratory, observational or review)
historical trial Coronary Drug Project (1975), an early niacin trial; PubMed has no abstract, so it is not counted here. 1975, JAMA
PubMed lists 497 human randomized trials and 79 meta-analyses mentioning Niacin (vitamin B3) for any use (searched 2026-10-02; the exact search is in the data file).
Berberine
Plant alkaloid · Heart and metabolism
Blood sugar in type 2 diabetesSupported
The trials are mostly small and of low methodological quality.
BenefitRandomized trial In a small pilot trial in type 2 diabetes, berberine lowered blood sugar and improved blood lipids. Yin 2008, Metabolism
BenefitMeta-analysis A meta-analysis found berberine lowered blood sugar and blood lipids in type 2 diabetes, but the trials were small and of low quality. Dong 2012, Evid Based Complement Alternat Med
BenefitMeta-analysis A second meta-analysis reported effects comparable to standard treatment on blood sugar, lipids and blood pressure, with no serious side effects. Lan 2015, J Ethnopharmacol
Cholesterol and triglyceridesSupported
The trials are mostly small and of low methodological quality.
BenefitMeta-analysis The second meta-analysis found effects on high cholesterol comparable to standard treatment. Lan 2015, J Ethnopharmacol
PubMed lists 110 human randomized trials and 49 meta-analyses mentioning Berberine for any use (searched 2026-10-02; the exact search is in the data file).
BenefitRandomized trial In people with diabetic nerve damage, five weeks of oral ALA improved symptoms; 600 mg a day gave the best balance of benefit and side effects. Ziegler 2006, Diabetes Care
Blood sugar in type 2 diabetesEarly or small trials
BenefitRandomized trial, 57 people In 57 people with type 2 diabetes, eight weeks of ALA lowered fasting and after-meal blood sugar. Ansar 2011, Saudi Med J
Other papers (not counted: animal, laboratory, observational or review)
PubMed lists 287 human randomized trials and 58 meta-analyses mentioning Alpha-lipoic acid for any use (searched 2026-10-02; the exact search is in the data file).
Omega-3 (fish oil)
EPA and DHA · Heart and metabolism
Heart attacks and cardiovascular events (fish-oil supplements)Mixed
No benefitRandomized trial, 15,480 people In 15,480 people with diabetes but no cardiovascular disease, 1 g of fish oil a day did not reduce serious vascular events (ASCEND). Bowman 2018, N Engl J Med
No benefitRandomized trial, 25,871 people In 25,871 generally healthy adults, fish oil did not reduce major cardiovascular events or cancer (VITAL). Manson 2019, N Engl J Med
BenefitMeta-analysis A meta-analysis of 13 trials with 127,477 people found small reductions in heart attacks (rate ratio 0.92) and cardiovascular death (0.93). Hu 2019, J Am Heart Assoc
Cardiovascular events with prescription EPA (high-risk patients)Supported
BenefitRandomized trial, 8,179 people In 8,179 statin-treated patients with raised triglycerides, prescription EPA (icosapent ethyl, 4 g a day) reduced ischaemic events, including cardiovascular death (REDUCE-IT). Bhatt 2019, N Engl J Med
PubMed lists 3,417 human randomized trials and 864 meta-analyses mentioning Omega-3 (fish oil) for any use (searched 2026-10-02; the exact search is in the data file).
CoQ10, ubiquinone, ubiquinol · Heart and metabolism
Blood pressureMixed
BenefitMeta-analysis A 2007 meta-analysis of older trials reported blood-pressure reductions in people with hypertension. Rosenfeldt 2007, J Hum Hypertens
No benefitMeta-analysis A 2016 Cochrane review, pooling the two trials at acceptable risk of bias, found no clinically significant effect on blood pressure. Ho 2016, Cochrane Database Syst Rev
Heart failure outcomesEarly or small trials
BenefitRandomized trial, 420 people In 420 people with chronic heart failure, CoQ10 reduced major adverse cardiovascular events (Q-SYMBIO). Mortensen 2014, JACC Heart Fail
Muscle symptoms on statinsMixed
No benefitMeta-analysis A 2015 meta-analysis found no significant benefit for statin-related muscle problems. Banach 2015, Mayo Clin Proc
BenefitMeta-analysis A 2018 meta-analysis of 12 trials (575 patients) found less muscle pain, weakness, cramps and tiredness. Qu 2018, J Am Heart Assoc
PubMed lists 426 human randomized trials and 116 meta-analyses mentioning Coenzyme Q10 for any use (searched 2026-10-02; the exact search is in the data file).
Green tea extract (EGCG)
Epigallocatechin gallate · Heart and metabolism
CholesterolSupported
Very few long-term trials exist.
BenefitMeta-analysis A meta-analysis found green tea drinks or extracts modestly lowered total and LDL cholesterol, with no effect on HDL. Zheng 2011, Am J Clin Nutr
BenefitMeta-analysis A Cochrane review found favourable effects on cardiovascular risk factors but very few long-term trials. Hartley 2013, Cochrane Database Syst Rev
Other papers (not counted: animal, laboratory, observational or review)
observational Japanese cohort: green tea drinking was linked to lower all-cause and cardiovascular death (an association). Kuriyama 2006, JAMA
safety review Derived a safe intake of 338 mg EGCG a day for tea preparations taken as a solid bolus dose. Hu 2018, Regul Toxicol Pharmacol
PubMed lists 543 human randomized trials and 171 meta-analyses mentioning Green tea extract (EGCG) for any use (searched 2026-10-02; the exact search is in the data file).
Muscle and mitochondria
Creatine
Creatine monohydrate · Muscle and mitochondria
Muscle strength and lean mass, with strength training (older adults)Supported
BenefitMeta-analysis In older adults doing resistance training, creatine added about 1.37 kg of lean mass and improved upper- and lower-body strength. Chilibeck 2017, Open Access J Sports Med
MemorySupported
The benefit was concentrated in older adults (66 to 76), and results varied between trials.
BenefitMeta-analysis A 2023 meta-analysis found creatine improved memory, mainly in adults aged 66 to 76. Prokopidis 2023, Nutr Rev
Other papers (not counted: animal, laboratory, observational or review)
PubMed lists 408 human randomized trials and 42 meta-analyses mentioning Creatine for any use (searched 2026-10-02; the exact search is in the data file).
Urolithin A
Gut-microbe metabolite of ellagitannins · Muscle and mitochondria
Muscle enduranceNot supported in trials
No benefitRandomized trial, 66 people In 66 older adults, urolithin A was safe, but its primary endpoints (walking distance and hand-muscle energy production) did not significantly improve. Liu 2022, JAMA Netw Open
Markers of mitochondrial healthEarly or small trials
BenefitUncontrolled pilot In a first-in-human trial in older adults, urolithin A was safe and changed blood and muscle markers of mitochondrial health over four weeks. Andreux 2019, Nat Metab
Other papers (not counted: animal, laboratory, observational or review)
animal Induced mitophagy and extended lifespan in worms; improved exercise capacity in rodents. Ryu 2016, Nat Med
PubMed lists 25 human randomized trials and 0 meta-analyses mentioning Urolithin A for any use (searched 2026-10-02; the exact search is in the data file).
BenefitRandomized trial In a small trial in older adults with mild cognitive impairment, lion's mane improved cognitive scores. Mori 2009, Phytother Res
BenefitRandomized trial A small trial reported improved cognitive function with lion's mane. Saitsu 2019, Biomed Res
BenefitRandomized trial In people with mild Alzheimer's disease, an erinacine A-enriched extract was linked to higher cognitive and daily-function scores than placebo over 49 weeks. Li 2020, Front Aging Neurosci
PubMed lists 5 human randomized trials and 0 meta-analyses mentioning Lion's mane for any use (searched 2026-10-02; the exact search is in the data file).
Bacopa
Bacopa monnieri · Brain
Memory and attentionSupported
Effects are small, and the meta-analysis calls for a large head-to-head trial.
BenefitRandomized trial In healthy older Australians, bacopa improved memory, with digestive side effects such as cramps and nausea. Morgan 2010, J Altern Complement Med
BenefitMeta-analysis A meta-analysis found bacopa may improve cognition, especially speed of attention, but called for larger trials. Kongkeaw 2014, J Ethnopharmacol
PubMed lists 27 human randomized trials and 2 meta-analyses mentioning Bacopa for any use (searched 2026-10-02; the exact search is in the data file).
Ashwagandha
Withania somnifera · Brain
Stress and anxietyEarly or small trials
Rare but serious liver injury has been reported.
BenefitRandomized trial, 64 people In 64 adults with chronic stress, a root extract improved stress resistance and self-rated quality of life. Chandrasekhar 2012, Indian J Psychol Med
BenefitRandomized trial In a small trial in healthy adults, an aqueous root extract reduced stress and anxiety and improved sleep quality. Salve 2019, Cureus
Other papers (not counted: animal, laboratory, observational or review)
case series Liver-injury case series: ashwagandha can cause cholestatic hepatitis, which can be dangerous in people with existing liver disease. Philips 2023, Hepatol Commun
PubMed lists 59 human randomized trials and 17 meta-analyses mentioning Ashwagandha for any use (searched 2026-10-02; the exact search is in the data file).
Vitamins and minerals
Magnesium
Essential mineral · Vitamins and minerals
Blood pressureSupported
BenefitMeta-analysis A meta-analysis of 34 trials (2,028 people) found magnesium lowered blood pressure slightly, by about 2 mm Hg. Zhang 2016, Hypertension
Other papers (not counted: animal, laboratory, observational or review)
observational Meta-analysis of cohort studies: higher dietary magnesium was linked to lower risk of stroke, heart failure and diabetes (an association, not trial evidence). Fang 2016, BMC Med
PubMed lists 359 human randomized trials and 75 meta-analyses mentioning Magnesium for any use (searched 2026-10-02; the exact search is in the data file).
Vitamin D3
Cholecalciferol · Vitamins and minerals
Preventing cancer and heart disease (generally healthy adults)Not supported in trials
No benefitRandomized trial, 25,871 people In 25,871 generally healthy adults, vitamin D did not lower cancer or cardiovascular events (VITAL). Manson 2019, N Engl J Med
No benefitRandomized trial, 5,110 people In 5,110 adults, monthly high-dose vitamin D for up to four years did not prevent cancer (ViDA, post hoc analysis). Scragg 2018, JAMA Oncol
Other papers (not counted: animal, laboratory, observational or review)
guideline 2024 Endocrine Society guideline: suggests vitamin D for specific groups (ages 1-18, over 75, pregnancy, high-risk prediabetes), not routine testing. Demay 2024, J Clin Endocrinol Metab
PubMed lists 1,892 human randomized trials and 726 meta-analyses mentioning Vitamin D3 for any use (searched 2026-10-02; the exact search is in the data file).
Zinc
Essential mineral · Vitamins and minerals
Infections in older adultsEarly or small trials
BenefitRandomized trial In a small trial in older adults, zinc lowered the rate of infections. Prasad 2007, Am J Clin Nutr
Other papers (not counted: animal, laboratory, observational or review)
PubMed lists 724 human randomized trials and 147 meta-analyses mentioning Zinc for any use (searched 2026-10-02; the exact search is in the data file).
Longevity candidates
Taurine
Amino sulfonic acid · Longevity candidates
Healthy ageing in peopleNo human trials
No human trial results yet.
Other papers (not counted: animal, laboratory, observational or review)
animal Taurine supplementation increased healthspan in worms, rodents and monkeys and lifespan in worms and rodents; the authors call for human trials. Singh 2023, Science
animal Taurine depletion accelerated muscle ageing in mice. Ito 2014, PLoS One
PubMed lists 227 human randomized trials and 41 meta-analyses mentioning Taurine for any use (searched 2026-10-02; the exact search is in the data file).
Resveratrol
Polyphenol in grapes and red wine · Longevity candidates
Exercise benefits (older men)Possible harm in trials
HarmRandomized trial In older men doing exercise training, resveratrol blunted the cardiovascular benefits of exercise. Gliemann 2013, J Physiol
Other papers (not counted: animal, laboratory, observational or review)
biochemical Found that resveratrol's apparent activation of sirtuins depended on the test substrate, with no effect on Sir2 activity in yeast. Kaeberlein 2005, J Biol Chem
animal Improved health and survival of mice on a high-calorie diet. Baur 2006, Nature
PubMed lists 263 human randomized trials and 115 meta-analyses mentioning Resveratrol for any use (searched 2026-10-02; the exact search is in the data file).
No benefitRandomized trial, 100 people In 100 older adults with subjective memory decline, longer-term spermidine did not improve memory or biomarkers (SmartAge). Schwarz 2022, JAMA Netw Open
Other papers (not counted: animal, laboratory, observational or review)
animal Extended lifespan and protected the heart in mice; in people, higher dietary spermidine was linked to lower blood pressure and heart disease (an association). Eisenberg 2016, Nat Med
PubMed lists 41 human randomized trials and 2 meta-analyses mentioning Spermidine for any use (searched 2026-10-02; the exact search is in the data file).
No compounds match. Try a broader search or clear the filters.
How the counting works
For each compound we list the outcomes it is promoted for and the human studies that tested them. A study is counted only if it is a randomized trial or a meta-analysis or systematic review in people, indexed in PubMed with an abstract. Each listed finding is checked against an exact phrase in that abstract, and so is each participant count. Uncontrolled human pilots are shown but cannot decide a rating. Animal, cell, observational and review papers appear under “Other papers” and are never counted.
How ratings are set
Ratings are computed from the counted studies by fixed rules, not chosen by hand:
Supported: At least one positive meta-analysis or positive large trial (1,000+ participants), and no counted negative results.
Mixed: Counted evidence points both ways.
Early or small trials: Only small randomized trials or uncontrolled pilots, all positive.
Not supported in trials: Only null results among counted studies.
Possible harm in trials: Counted studies show only null or harmful results, including at least one harmful result.
No human trials: No counted human trial evidence for this outcome.
Download the data
The dataset is free to download. The CSV has one row per counted study; outcomes with no counted study have one row with the study fields blank. The JSON also holds the context papers, the exact abstract phrases each finding was checked against, and the PubMed searches behind the volume counts.
We list the key studies for each outcome, not every trial ever run. The PubMed totals on each card show the overall volume of research for any use, which is not a measure of quality.
A rating applies only to the outcome it names. No supplement has been shown in trials to slow human ageing itself.
“Supported” can rest on small effects; the findings and caveats say how large.
Medical disclaimer: This database is for education only and is not medical advice. Supplements can interact with medicines; talk to a clinician before starting or stopping anything. Prescription drugs are described from their trials and require a prescriber.
Method, limits and how to read your result
How it works
Each supplement or drug is listed with the specific outcomes it has been tested for, and each outcome gets a verdict based on human trials: supported, mixed, early or small trials, no human trials, not supported in trials, or possible harm. Every verdict links to its sources on PubMed.
Limits
Verdicts summarise trial evidence for typical adults and can change as new research appears. A supported verdict for one outcome does not mean the product helps with aging in general, and product quality varies between brands. Some popular supplements have failed large trials: antioxidant pills had no effect on mortality in a Cochrane review (Cochrane).
How to read your result
Read the verdict for the outcome you care about rather than the product as a whole, and check the possible-harm and interaction notes. Discuss any supplement with your doctor or pharmacist if you take prescription medicines.