Digestery
Supplements › Spermidine

Does spermidine slow aging in people?

Updated Nov 13, 2026 · 52 papers read, mostly from abstracts

Under review. A person has not yet checked this short answer against the sources. Until they do, search engines are asked not to list this page. See the 5 open items

Short answer

Not shown in people.

Spermidine extends lifespan in yeast, worms, flies, and mice. No human study has tested whether it slows aging.

Confidence

Very low

What “very low” means

This page says what the research shows. It does not say whether you should take spermidine.

Where the evidence stands

Each dot is one paper. Steps near the top answer the question more directly. Select a dot to see the study.

↑ Stronger evidence

  1. People: real aging outcome

    Did people given spermidine live longer or get fewer age-related diseases?

    0

    Not done yet. This is the step that would answer the question.

  2. People: memory or function

    Does a supplement change how people perform?

    3

    2 benefit, 1 no difference

  3. People: lab tests or safety

    Does a supplement change lab values? Is it safe?

    6

    3 benefit, 3 no difference

  4. Diet surveys

    Do people who eat more spermidine do better? A link, not a cause.

    12

    5 benefit, 2 mixed, 3 no difference, 1 no direction, 1 harm

  5. Blood levels and genes

    Do people with more spermidine in the blood do better? A link, not a cause.

    12

    3 benefit, 1 mixed, 4 no direction, 4 harm

  6. Animals

    Does it work in yeast, worms, flies, mice, or rats?

    9

    7 benefit, 1 no difference, 1 harm

  7. Cells in a dish

    What does it do in a dish?

    1

    1 harm

↓ Weaker evidence

  • Points to benefit
  • Mixed
  • No difference
  • No direction
  • Points to harm

Not on the ladder: 4 trials of products that mix spermidine with other ingredients, 1 trial with no results yet, and 4 opinion or review papers. Counts are papers, not people. Some trials appear in more than one paper.

See every study

What we found

  • The largest trial gave 100 older adults spermidine or a placebo for a year. Their memory did not differ. Schwarz 2022
  • In diet surveys, people who ate more spermidine died less often in some studies but not in others. Surveys show a link, not a cause. Details
  • Supplements of 15 mg and 40 mg a day did not raise spermidine in the blood. Senekowitsch 2023, Keohane 2024
  • 5 of the 9 papers on trials of spermidine alone have authors with ties to The Longevity Labs, which sells spermidine supplements. Who funds this

What the research says

In animals: longer life, with exceptions

Spermidine extended lifespan in yeast, flies, and worms, and in mice by up to 25% in one study. It did not extend lifespan in middle-aged rats. In mice on a high-fat diet, it made heart scarring worse.

Studies: Eisenberg 2009, Eisenberg 2016, Yue 2017, Filfan 2020, Mühlfeld 2021

In diet surveys: a link in some, not in others

Two studies found that people who ate more spermidine died less often: one of 829 adults followed for 20 years, and one of 23,894 US adults. A Japanese study of 29,079 adults over 16 years found no link. A US study of 54,517 men found fewer deaths overall, but its authors say other factors may explain it.

Surveys cannot show cause. People who eat more foods rich in spermidine, such as vegetables, grains, beans, and nuts, may differ in other ways.

Studies: Kiechl 2018, Wu 2022, Nagata 2024, Wang 2026

In blood tests: the links point both ways

Higher blood spermidine went with fewer heart events in heart attack and kidney patients. In other groups, it went with more fractures, more strokes, and more signs of brain aging.

One reading is that blood spermidine is a sign of disease, not a cause. These studies cannot tell.

Studies: Yu 2022, Chen 2025, Kong 2021, Zheng 2022, Wortha 2023

In trials: small, short, and mostly lab markers

Nine papers report seven or eight trials that gave people spermidine alone. The largest had 100 people. None lasted more than 12 months, and none measured lifespan or an age-related disease.

The largest trial found no memory effect at 0.9 mg a day. Smaller studies reported gains in memory scores, vaccine responses, and lab markers.

Studies: Schwarz 2022, Wirth 2018, Pekar 2021, Alsaleh 2026, Bruno 2025

Do supplements reach the blood?

Two trials measured this. At 15 mg a day for 5 days, blood spermidine did not rise, but blood spermine did. At 40 mg a day for up to 28 days, blood polyamines did not change much. For comparison, the largest trial used 0.9 mg a day.

Studies: Senekowitsch 2023, Keohane 2024

How the research grew

Animal studies came first. Human trials began in 2018. 24 of the 52 papers came out in 2024 to 2026.

Timeline of 52 spermidine and aging papers from 2009 to 2026, in five groups: human trials, diet surveys, blood and genes, animal and cell studies, and opinion. Animal studies start in 2009. The first human trial is in 2018. The largest finished trial, in 2022, found no memory benefit. A heart trial of 187 people has no results yet.
Each dot is one paper, colored by what it found. Numbered rings mark six key studies.

Doses tested

4 of 7 trials gave less spermidine than a typical diet already provides. The largest finished trial gave 0.9 mg a day.

Spermidine doses in 7 human trials, from 0.9 to 40 mg a day, compared with a typical diet of 8 to 13 mg a day. SmartAge, 0.9 mg, 100 people: no difference in memory. 1.2 mg: no change in health tests. 1.5 or 3.3 mg: markers improved at the higher dose, with no placebo. 6 mg: stronger vaccine response in weak responders. 15 mg: blood spermidine did not rise. 24 mg, POLYCAD: results not published. 40 mg: safe, blood levels changed little.
One row per trial, so the SmartAge brain-scan paper is not a separate row. Not shown: 2 trials with no dose in the abstract and 4 trials of mixed products.

Spermidine in food

Wheat germ, soybeans, aged cheddar, and mushrooms have the most. A typical diet gives about 8 mg a day in the US and 13 mg in the EU. Results differ a lot between studies of the same food.

Spermidine in 14 foods, in milligrams per 100 grams: wheat germ 35, soybeans 13 to 21, aged cheddar 20, mushrooms 8.9, green peas 4.7 to 6.5, broccoli 2.7 to 4.1, down to milk at 0.02 to 0.05. A second chart compares a typical diet of 8 to 13 mg a day with trial doses of 0.9 to 40 mg a day.
Each dot is one study's average. Values are for food as tested; cooking changes them a little.
See the numbers
Foodmg per 100 gCountry testedSource
Wheat germ35JapanMuñoz-Esparza 2019 (2,437 nmol/g)
Soybeans, dried21JapanAtiya Ali 2011: “Soybean, dried”, 207 mg/kg
13GermanyAtiya Ali 2011: “Soybean, dried”, 128 mg/kg
Cheddar, aged 1 year20UKAtiya Ali 2011: “Cheddar, matured (1 year)”, 199.5 mg/kg
Mushrooms8.9JapanAtiya Ali 2011: “Mushroom”, 88.6 mg/kg
Green peas6.5UKAtiya Ali 2011: “Green peas, cooked”, 65.2 mg/kg
4.7Czech RAtiya Ali 2011: “Green peas, frozen”, 46.6 mg/kg
Broccoli4.1NorwayAtiya Ali 2011: “Broccoli, fresh”, 41.3 mg/kg
3.3GermanyAtiya Ali 2011: “Broccoli, fresh”, 33.2 mg/kg
2.7NorwayAtiya Ali 2011: “Broccoli, cooked”, 27.3 mg/kg
Cauliflower3.1GermanyAtiya Ali 2011: “Cauliflower, fresh”, 31.2 mg/kg
2.8NorwayAtiya Ali 2011: “Cauliflower, fresh”, 28.3 mg/kg
2.6NorwayAtiya Ali 2011: “Cauliflower, cooked”, 26.2 mg/kg
2.5UKAtiya Ali 2011: “Cauliflower, fresh”, 24.8 mg/kg
Chickpeas2.9FranceAtiya Ali 2011: “Chickpea”, 28.8 mg/kg
Whole-grain bread2.4UKAtiya Ali 2011: “Whole grain”, 24.4 mg/kg
1.6FranceAtiya Ali 2011: “Whole grain”, 15.8 mg/kg
1.3NorwayAtiya Ali 2011: “Whole grain”, 13.1 mg/kg
Blue cheese2.4NorwayAtiya Ali 2011: “Blue, Norwegian”, 23.8 mg/kg
2.0JapanAtiya Ali 2011: “Blue, Japanese (1-5 months)”, 20.3 mg/kg
Potatoes, cooked2.4GermanyAtiya Ali 2011: “Potato, cooked”, 23.5 mg/kg
1.5UKAtiya Ali 2011: “Potato, cooked”, 15.2 mg/kg
1.1NorwayAtiya Ali 2011: “Potato, cooked”, 10.9 mg/kg
Beef, raw lean1.9UKAtiya Ali 2011: “Beef, raw, lean”, 19 mg/kg
0.55NorwayAtiya Ali 2011: “Beef, raw, lean”, 5.5 mg/kg
0.31SpainAtiya Ali 2011: “Beef, raw, lean”, 3.1 mg/kg
0.26JapanAtiya Ali 2011: “Beef, raw, lean”, 2.6 mg/kg
Apple0.25UKAtiya Ali 2011: “Apple”, 2.5 mg/kg
0.1JapanAtiya Ali 2011: “Apple”, 1 mg/kg
Milk0.05UKAtiya Ali 2011: “Milk, semi-skimmed”, 0.5 mg/kg
0.04UKAtiya Ali 2011: “Milk, full cream”, 0.4 mg/kg
0.02SpainAtiya Ali 2011: “Milk, unspecified”, 0.17 mg/kg

Food values: Atiya Ali and colleagues, 2011, Table 1, given there in mg per kg. Wheat germ and diet averages: Muñoz-Esparza and colleagues, 2019, given there in nmol per g and µmol a day, converted with spermidine's molar mass of 145.25 g/mol.

The two papers that started this page

A reader submitted these two papers. Here is how each one holds up against the other research.

OpinionSubmitted by a reader

Spermidine as an anti-aging vitamin

Madeo and colleagues, Autophagy, 2019 (online October 2018). Read the paper

Main point. The body makes less spermidine with age, so it may need to come from food or supplements, like a vitamin. The paper has no new data.

What other studies found

Stake. Three of four authors hold stakes in companies in this field, including one that sells spermidine supplements.

Untested in people. Later trials have not confirmed it.

Cell studySubmitted by a reader

Polyamines and cancer cell growth

Suzuki and colleagues, Journal of Biological Chemistry, 2025. Read the paper

Main point. In cancer cells in a dish, polyamines, the family that includes spermidine, help the cells make a protein that supports their growth.

What studies in people found about cancer

  • Above-median polyamine intake went with 39% higher odds of new colon polyps. Vargas 2012
  • Polyamine intake did not raise colorectal cancer risk in 87,602 women. Vargas 2015
  • Spermidine intake was not linked to prostate cancer in 54,517 men. Wang 2026
  • Japanese women with the highest intake had a possibly higher cancer death rate. The range included no effect. Nagata 2024

Stake. None declared.

An open question. No study in people shows that spermidine causes cancer, and none rules it out.

Who funds this research

Nine papers report trials that gave people spermidine alone. The Longevity Labs sells spermidine supplements.

  • Stake: The Longevity Labs 3
  • Authors linked to a stake 2
  • Stake: a company 1
  • No disclosure in PubMed 1
  • No stake declared 2

The two papers with no declared stake found a gain in memory scores in a preliminary nursing-home study, and no rise in blood spermidine at 15 mg a day.

Among diet surveys, the first study to link spermidine with longer life also has co-authors with ties to The Longevity Labs. Kiechl 2018

Funding rule. A stake is always shown. Confidence is lowered only when every supporting study comes from a party with a stake. That is not the case here, so funding did not lower the confidence. "Authors linked to a stake" means the paper itself has no disclosure in PubMed, but co-authors disclosed a stake in other papers.

What to watch

Trial in progress: POLYCAD

187 adults over 65 with heart disease took 24 mg a day or a placebo for 48 weeks. The trial measures heart size, fitness, muscle, and inflammation. Its protocol expected it to finish in August 2026. No results were in PubMed on Nov 12, 2026. Thorup 2025

Email alerts are coming soon.

All 52 papers

Each paper shows what it tested, what it found, and who has a stake. Links go to PubMed.

  1. Trial: memory or functionSchwarz et al., 2022JAMA Network Open

    100 people · adults aged 60 to 90 with memory worries

    Tested0.9 mg a day from wheat germ for 12 months, vs placebo (the SmartAge trial)

    No difference on the main memory test. Extra analyses hinted at benefits for verbal memory and inflammation.

    No differenceStake: The Longevity LabsAbstract only

    Three authors hold equity in or advise The Longevity Labs. Another was its chief executive during the trial.

  2. Trial: memory or functionWirth et al., 2018Cortex

    30 people · adults aged 60 to 80 with memory worries

    TestedSpermidine-rich plant extract for 3 months, vs placebo

    Memory scores were moderately better than placebo. A small pilot, and the uncertainty range touched zero.

    Points to benefitAuthors linked to a stakeAbstract only

    No disclosure in PubMed. Three co-authors (Madeo, Sigrist, Stekovic) disclosed stakes in The Longevity Labs in other papers.

  3. Trial: memory or functionPekar et al., 2021Wiener klinische Wochenschrift

    85 people · nursing-home residents aged 60 to 96

    TestedA higher vs a lower spermidine dose for 3 months

    People with mild dementia on the higher dose gained 2.23 points on a memory screening test (MMSE). The authors call these first, preliminary results.

    Points to benefitNo stake declaredAbstract only

    No competing interests. The authors funded the study themselves.

  4. Trial: markers or safetySchwarz et al., 2026Aging and Disease

    75 people · people from the SmartAge trial, with brain scans

    Tested0.9 mg a day for 12 months, vs placebo

    Small differences in blood flow and white matter in two brain regions. No change in brain size, and no link to memory. The authors call it exploratory.

    Points to benefitAuthors linked to a stakeAbstract only

    No disclosure in PubMed. Three co-authors (Madeo, Eisenberg, Sigrist) disclosed stakes in The Longevity Labs in the 2022 SmartAge paper.

  5. Trial: markers or safetyAlsaleh et al., 2026Aging Cell

    40 people · adults over 65 after a COVID-19 vaccine dose

    Tested6 mg a day for 13 weeks, vs placebo

    Well tolerated. In people who had responded poorly to the vaccine, antibody responses got stronger. A pilot study.

    Points to benefitStake: The Longevity LabsAbstract only

    Two authors consult for The Longevity Labs, which also helped fund the work. The paper says the company had no input into design or results.

  6. Trial: markers or safetyBruno et al., 2025Alternative Therapies in Health and Medicine

    12 people · healthy adults

    Tested1.5 or 3.3 mg a day from a rice germ extract for 56 days. No placebo group described.

    The higher dose raised two autophagy markers by 7.3% and 13.4% and lowered the inflammation marker CRP by 20.8%, compared with starting values.

    Points to benefitNo disclosure in PubMedAbstract only

    No disclosure in PubMed. The study tested a named commercial extract (Miricell, Nutraland USA).

  7. Trial: markers or safetySenekowitsch et al., 2023Nutrients

    12 people · healthy adults

    Tested15 mg a day for 5 days, vs placebo

    Blood spermidine did not rise. Blood spermine did. The authors say doses under 15 mg a day are unlikely to have short-term effects.

    No differenceNo stake declaredAbstract only

    No conflict of interest declared.

  8. Trial: markers or safetyKeohane et al., 2024Nutrition Research

    37 people · healthy men aged 50 to 70

    Tested40 mg a day of high-purity spermidine for up to 28 days, vs placebo

    No safety problems found. Blood and urine polyamine levels did not change much.

    No differenceStake: a companyAbstract only

    Two authors work for Chrysea Labs, which also funded the other authors.

  9. Trial: markers or safetySchwarz et al., 2018Aging

    30 people · adults aged 60 to 80 with memory worries, plus mice

    Tested1.2 mg a day from wheat germ for 3 months, vs placebo

    No differences in safety tests, weight, or self-rated health. Well tolerated.

    No differenceStake: The Longevity LabsAbstract only

    Four authors have a financial interest in The Longevity Labs.

  10. Trial: mixed productMatsumoto et al., 2019Nutrients

    44 people · healthy adults near the top of the healthy weight range

    TestedYogurt with a probiotic and arginine for 12 weeks, vs plain yogurt

    Blood vessel function improved, and blood spermidine rose. Not a spermidine supplement.

    Points to benefitStake: a companyAbstract only

    Two authors work for Kyodo Milk Industry, which funded the study and sells products with this probiotic.

  11. Trial: mixed productFélix et al., 2024Antioxidants

    41 people · adults aged 30 to 63

    TestedA blend of AM3, 0.6 mg spermidine, and hesperidin for 2 months, vs placebo

    Immune tests improved, and an immune-based age score fell by 11 years. The effect cannot be credited to spermidine alone.

    Points to benefitNo stake declaredAbstract only

    No conflicts of interest declared.

  12. Trial: mixed productRhodes et al., 2024Nutrition Research

    5 people · healthy young men

    TestedOne dose of a mix of spermidine and three other compounds

    Blood taken after the dose reduced inflammation signals in lab cells.

    Points to benefitNo disclosure in PubMedAbstract only

    No disclosure in PubMed.

  13. Trial: mixed productGrant et al., 2026Scientific Reports

    42 people · adults

    TestedA mix of spermidine and three other compounds (Mimio) for 8 weeks, vs placebo

    More people reported less appetite at mealtimes (91% vs 47%).

    Points to benefitStake: a companyAbstract only

    Funded by Mimio Health. One author works there.

  14. Trial: no results yetThorup et al., 2025Trials

    187 people · adults over 65 with heart disease

    Tested24 mg a day for 48 weeks, vs placebo (the POLYCAD trial)

    Results not yet published. The protocol expected the trial to finish in August 2026.

    No directionNo stake declaredAbstract only

    No conflicts declared. A supplement firm (DoNotAge) donated the capsules.

  15. Diet surveyKiechl et al., 2018American Journal of Clinical Nutrition

    829 people · adults aged 45 to 84, followed 20 years

    TestedSpermidine in the diet, from repeated food questionnaires

    Each step up in intake (one standard deviation) went with a 26% lower death rate (HR 0.74). A second study group showed the same pattern.

    Points to benefitAuthors linked to a stakeAbstract only

    No disclosure in PubMed. Three co-authors (Madeo, Eisenberg, Stekovic) disclosed stakes in The Longevity Labs in other 2018 papers.

  16. Diet surveyWu et al., 2022Frontiers in Public Health

    23,894 people · US adults (NHANES 2003 to 2014)

    TestedSpermidine in the diet

    The top quarter of intake had 30% lower death from any cause (HR 0.70) than the bottom quarter.

    Points to benefitNo stake declaredAbstract only

    No commercial or financial relationships declared.

  17. Diet surveyNagata et al., 2024British Journal of Nutrition

    29,079 people · Japanese adults, followed 16 years

    TestedSpermidine in the diet, from a food questionnaire

    No link with death from any cause. A possible higher cancer death rate in women (HR 1.38, range 0.99 to 1.93).

    No differenceNo disclosure in PubMedAbstract only

    No disclosure in PubMed.

  18. Diet surveyWang et al., 2026Frontiers in Nutrition

    54,517 people · US men in a cancer screening trial

    TestedSpermidine in the diet, from a food questionnaire

    No link with prostate cancer or prostate cancer death. A link with fewer deaths overall, which the authors say may come from other factors.

    No differenceNo stake declaredAbstract only

    No commercial or financial relationships declared.

  19. Diet surveyQian et al., 2024Nutrients

    77,092 people · UK adults aged 60 and over, followed 12 years

    TestedSpermidine in the diet

    The fourth of five intake groups had lower dementia risk (HR 0.68). The pattern was a curve, not a steady trend.

    MixedNo stake declaredAbstract only

    No support from any organization declared.

  20. Diet surveyHan et al., 2024Nutrients

    184,732 people · UK adults, followed 11.5 years

    TestedAll polyamines in the diet, not spermidine alone

    Every group above the lowest intake had lower death risk. The lowest risk was in the second-lowest group (HR 0.82), so more was not better.

    MixedNo stake declaredAbstract only

    No conflicts of interest declared.

  21. Diet surveyMa et al., 2025Journal of Affective Disorders

    2,674 people · US adults aged 60 and over (NHANES 2011 to 2014)

    TestedSpermidine in the diet

    The top intake group had lower odds of poor scores on all four thinking tests. Measured at one point in time.

    Points to benefitNo stake declaredAbstract only

    No conflicts of interest declared.

  22. Diet surveyLópez-González et al., 2026European Journal of Nutrition

    2,664 people · older adults with metabolic syndrome, followed 1 year

    TestedChanges in dietary spermidine

    The group with the highest and rising intake had better liver fat scores and HbA1c (down 0.14%). Its BMI also fell.

    Points to benefitNo stake declaredAbstract only

    No spermidine-company tie declared. Some authors report food-industry funding.

  23. Diet surveyThorup et al., 2026European Heart Journal Open

    187 people · older adults with heart disease

    TestedSpermidine in the diet

    Each doubling of intake went with higher fitness (peak oxygen uptake up 2.6 mL/kg/min). Measured at one point in time.

    Points to benefitNo stake declaredAbstract only

    No spermidine-company tie declared.

  24. Diet surveyMa et al., 2025Journal of Nutrition, Health and Aging

    Number not stated · UK Biobank adults

    TestedSpermidine in the diet and a biological age score

    No abstract in PubMed, so the result was not read.

    No directionNo stake declaredAbstract only

    No competing interests declared.

  25. Diet surveyVargas et al., 2012American Journal of Clinical Nutrition

    1,164 people · adults with past colon polyps, from two trials

    TestedAll polyamines in the diet

    Above-median intake went with 39% higher odds of new colon polyps (OR 1.39).

    Points to harmNo disclosure in PubMedAbstract only

    No disclosure in PubMed.

  26. Diet surveyVargas et al., 2015American Journal of Clinical Nutrition

    87,602 people · US women after menopause (Women's Health Initiative)

    TestedAll polyamines in the diet

    No higher risk of colorectal cancer. Some intake groups had slightly lower risk.

    No differenceNo disclosure in PubMedAbstract only

    No disclosure in PubMed.

  27. Blood level or genesKong et al., 2021Journal of Clinical Endocrinology & Metabolism

    1,504 people · Korean adults, followed 9 years

    TestedSpermidine in the blood

    Higher blood spermidine went with more fragility fractures (HR 1.35 per 1 μM).

    Points to harmNo disclosure in PubMedAbstract only

    No disclosure in PubMed.

  28. Blood level or genesZheng et al., 2022Frontiers in Nutrition

    Number not stated · adults in a community study

    TestedSpermidine in the blood

    The top blood spermidine group had about five times the stroke risk of the bottom group (HR 5.02, range 1.58 to 16.02).

    Points to harmNo stake declaredAbstract only

    No commercial or financial relationships declared.

  29. Blood level or genesWortha et al., 2023Alzheimer's & Dementia

    659 people · adults aged 21 to 81 (Study of Health in Pomerania)

    TestedSpermidine in the blood

    Higher blood spermidine went with a smaller memory center (hippocampus) and other signs of brain aging.

    Points to harmNo stake declaredAbstract only

    Some authors report drug-industry fees. None is tied to spermidine.

  30. Blood level or genesSternberg et al., 2022Journal of Clinical Neuroscience

    76 people · 43 people with Alzheimer's disease, 12 with mild memory loss, 21 healthy

    TestedSpermidine in the blood

    Blood spermidine was higher in people with mild memory loss than in healthy people.

    Points to harmNo disclosure in PubMedAbstract only

    No disclosure in PubMed.

  31. Blood level or genesXu et al., 2022Frontiers in Aging Neuroscience

    3,774 people · adults over 35

    TestedSpermidine in the blood

    Middle and high levels went with about 20% lower odds of mild memory loss. The link was a curve, not a straight line.

    MixedNo stake declaredAbstract only

    No commercial or financial relationships declared.

  32. Blood level or genesYu et al., 2022Nutrients

    377 people · heart attack patients, followed about 1 year

    TestedSpermidine in the blood

    Higher levels went with fewer major heart events (HR 0.566 for the top third vs the bottom third).

    Points to benefitNo stake declaredAbstract only

    No conflict of interest declared.

  33. Blood level or genesChen et al., 2025Cardiorenal Medicine

    297 people · kidney disease patients, followed 6 years

    TestedSpermidine in the blood

    Higher levels went with fewer heart events (HR 0.077). The link with death faded after accounting for kidney function.

    Points to benefitNo disclosure in PubMedAbstract only

    No disclosure in PubMed.

  34. Blood level or genesWang et al., 2025Hereditas

    Number not stated · genetic data

    TestedGenes that set blood spermidine (Mendelian randomization)

    Genes for lower spermidine went with higher coronary heart disease risk.

    Points to benefitNo stake declaredAbstract only

    No competing interests declared.

  35. Blood level or genesPucciarelli et al., 2012Rejuvenation Research

    78 people · three age groups, including 26 people aged 90 to 106

    TestedPolyamines in the blood

    Total polyamines were lower in older groups. People aged 60 to 80 had the lowest spermidine.

    No directionNo disclosure in PubMedAbstract only

    No disclosure in PubMed.

  36. Blood level or genesPekar et al., 2020Wiener klinische Wochenschrift

    80 people · nursing-home residents aged 60 to 96

    TestedSpermidine in the blood

    Blood spermidine was linked to memory test scores. The abstract does not say in which direction.

    No directionNo stake declaredAbstract only

    No competing interests declared.

  37. Blood level or genesKaminsky et al., 2024Aging Cell

    Number not stated · twins

    TestedSpermidine in red blood cells

    Red cell spermidine is 82% inherited. The link to long life is indirect, through red cell size.

    No directionNo stake declaredAbstract only

    No competing interests declared.

  38. Blood level or genesThorup et al., 2026Clinical Nutrition

    192 people · older adults with heart disease

    TestedSpermidine in the diet vs in blood and muscle

    Doubling intake went with 18% higher blood levels. Muscle levels showed no link.

    No directionNo stake declaredAbstract only

    No conflicts declared.

  39. AnimalEisenberg et al., 2009Nature Cell Biology

    yeast, flies, worms; human cells in a dish

    TestedSpermidine

    Markedly longer lifespan in yeast, flies, and worms.

    Points to benefitNo disclosure in PubMedAbstract only

    No disclosure in PubMed. Several authors later held equity in The Longevity Labs, founded in 2016.

  40. AnimalEisenberg et al., 2016Nature Medicine

    mice and rats; diet data in people

    TestedSpermidine in food or water

    Longer lifespan in mice and a healthier aging heart.

    Points to benefitStake: The Longevity LabsAbstract only

    Five authors hold equity in The Longevity Labs.

  41. AnimalYue et al., 2017Cancer Research

    mice

    TestedLifelong spermidine

    Up to 25% longer life, and less liver scarring and liver cancer after chemical injury.

    Points to benefitNo disclosure in PubMedAbstract only

    No disclosure in PubMed.

  42. AnimalSoda et al., 2009Experimental Gerontology

    mice

    TestedFood high in spermine and spermidine

    Higher survival on the high-polyamine food.

    Points to benefitNo disclosure in PubMedAbstract only

    No disclosure in PubMed.

  43. AnimalFilfan et al., 2020GeroScience

    middle-aged male rats

    TestedSpermidine in drinking water

    No longer life. Some health measures improved, including less brain inflammation.

    No differenceNo stake declaredAbstract only

    No conflict of interest declared.

  44. AnimalMühlfeld et al., 2021Histochemistry and Cell Biology

    mice on a high-fat diet

    TestedSpermidine in drinking water for 30 weeks

    Spermidine made heart scarring and fat build-up worse. Exercise did not.

    Points to harmNo stake declaredAbstract only

    No conflict of interest declared.

  45. AnimalSchroeder et al., 2021Cell Reports

    mice and flies; diet data in people

    TestedSpermidine in food

    Old mice learned better.

    Points to benefitStake: The Longevity LabsAbstract only

    Authors hold equity in The Longevity Labs. Two are inventors on a related pending patent.

  46. AnimalHofer et al., 2024Nature Cell Biology

    yeast, worms, flies, mice; blood from fasting volunteers

    TestedBlocking the body's own spermidine during fasting

    Fasting raised spermidine. Blocking that rise removed fasting's lifespan benefit in animals.

    Points to benefitStake: The Longevity LabsAbstract only

    Authors hold equity in The Longevity Labs or co-founded Samsara Therapeutics.

  47. AnimalHofer et al., 2024Autophagy

    yeast, worms, flies, mice

    TestedBlocking spermidine during fasting or rapamycin

    The same pattern for fasting and for the drug rapamycin. May repeat data from the paper above.

    Points to benefitStake: The Longevity LabsAbstract only

    One author holds equity in The Longevity Labs and Samsara Therapeutics.

  48. CellSuzuki et al., 2025Journal of Biological Chemistry

    human cancer cells in a dish

    TestedPolyamines, the family that includes spermidine

    Polyamines help cancer cells make a protein that supports their growth (eIF5A2).

    Points to harmNo stake declaredRead in fullSubmitted by a reader

    No conflicts of interest declared.

  49. Opinion or reviewMadeo et al., 2019Autophagy (online October 2018)

    Argues that spermidine acts like a vitamin in old age. No new data.

    No weight on the ladderStake: The Longevity LabsRead in fullSubmitted by a reader

    From the full text: three of four authors co-founded Samsara Therapeutics, and two hold equity in The Longevity Labs.

  50. Opinion or reviewMadeo et al., 2018Science

    Review of spermidine in health and disease.

    No weight on the ladderAuthors linked to a stakeAbstract only

    No disclosure in PubMed. Co-authors Madeo and Eisenberg disclosed stakes in The Longevity Labs in other papers.

  51. Opinion or reviewMadeo et al., 2020Annual Review of Nutrition

    Review of spermidine in food and health.

    No weight on the ladderAuthors linked to a stakeAbstract only

    No disclosure in PubMed. Co-authors Madeo and Eisenberg disclosed stakes in The Longevity Labs in other papers.

  52. Opinion or reviewAngelucci et al., 2026Degenerative Neurological and Neuromuscular Disease

    Review on Alzheimer's disease: the evidence is not enough to support using spermidine.

    No weight on the ladderNo stake declaredAbstract only

    Authors report company stakes. None is tied to spermidine.

How the short answer is set

Fixed rules turn the studies into the short answer, so the same studies always give the same answer. The rules start from GRADE, a standard system for rating evidence, in plain language. On the live site, the method has its own page.

ConfidenceWhat it means
HighWe are confident the answer is right.
ModerateThe answer is probably right, but it could change.
LowThe answer may well change.
Very low (this page)Any answer is very uncertain.
Not tested in peopleOnly cell or animal studies exist.

Why very low here

  • No trial measured lifespan or an age-related disease.
  • Diet surveys disagree, and surveys cannot show cause.
  • Blood studies point both ways.
  • Most trials are small or short.

Hard limit used. With no human trial of an aging outcome, the short answer cannot say that spermidine works in people.

How this page was made

Label

Search and study selection
Claude Fable 5.2
First read of all 52 papers
Claude Sonnet
Independent second read
Claude Haiku
Disagreements settled by
Claude Opus 5.5
Page written by
Claude Opus 5.5
Charts
Drawn by code from the study records and 2 food tables; food values copied by Claude Opus 5.5 and checked against the tables
Source check by a person
Not yet. Required before search engines may list this page.
Expert review
No
Papers read in full
2 of 52; the rest from PubMed abstracts
Method
v0.1, created Nov 12, 2026, 0 corrections
409numbers in the two reads checked against the abstracts. None was wrong.
42 / 52papers where both reads agreed on all five checked fields.
13disagreements in 10 papers, all settled against the abstract.

Search. PubMed, searched Nov 12, 2026. 629 records mention spermidine together with aging, longevity, lifespan, senescence, age-related, or mortality. From these and related searches, 52 papers were chosen: human trials of spermidine on memory, aging markers, safety, or blood levels; a selection of human surveys and blood studies of aging-related outcomes; the main animal lifespan studies; reviews; and the two submitted papers. Two trials on other topics were left out (hair growth, and salivary glands after radiotherapy). The search found no systematic review of spermidine and aging.

Checks. The second read did not see the first. When they disagreed, the top model read the abstract and sided with the first read 8 times and the second read 5 times. One caught error: the first read counted 432,682 patient records as spermidine measurements; they were records of red blood cell size. Kaminsky 2024

Review queue

These items wait for a person. Search engines are asked not to list the page until the first one is done.

  1. OpenThe short answer is new. It waits for a source check by a person.
  2. OpenTwo 2018 papers may report the same 30-person trial: Wirth 2018 and Schwarz 2018. Both count on the ladder for now.
  3. OpenHofer 2024 may repeat data from Hofer 2024. Both count for now.
  4. OpenFive papers have no disclosure in PubMed, but co-authors disclosed stakes elsewhere. Their full texts were not checked: Kiechl 2018, Wirth 2018, Schwarz 2026, Madeo 2018, Madeo 2020.
  5. OpenRandom sample for a person's check: Kong 2021, Félix 2024, Senekowitsch 2023.

Page history

DateChangeReason
Nov 13, 2026New design: larger text, a phone-first layout, and color-blind-safe colors. Charts redrawn, with phone versions.Easier reading for older readers and on phones. No study data changed.
Nov 13, 2026Added three charts: how the research grew, doses tested, and spermidine in food. Added a food table from 2 published sources.Charts let readers see the research at a glance. No study data changed.
Nov 12, 2026Page created from 52 papers.A reader submitted two papers: Madeo 2019 and Suzuki 2025.

Flag an error

A flag must point to a source: the passage in the paper, or a link to a study this page missed. A flag starts a new check. Votes never change a page.