Humanin peptide: what the mouse and cell studies show, why nobody has given it to a person, and what the tumour study means for the vials
What 589 indexed publications and 5 randomized trials actually state about humanin, sentence by sentence, each tied to its source and labeled by the kind of evidence it is.
At a glance
What humanin is, and why the body makes more of it with age
Humanin is a peptide in the mitochondrial peptides class (mitochondria-derived cytoprotective peptide). Europe PMC indexes 589 publications naming it or a listed alias in a title or abstract, including 5 randomized controlled trials and 3 clinical trials of any design, as of . The strongest evidence tier in that literature is human clinical trials.
This page reports what those studies state, sentence by sentence, with each statement tied to its source and labeled by the kind of evidence it is. It does not recommend use, and it does not translate study doses into anything personal.
What the evidence level means
Publication counts show what is indexed, not how strong the evidence is. The ledger below separates human clinical records from preclinical, analytical, and in-vitro work.
Interpretation boundary: An indexed publication count is a discovery measure. It does not establish efficacy, safety, approval, or a personal treatment plan.
Humanin in two minutes
What it is. A small peptide encoded inside a mitochondrial ribosomal-RNA gene, found in 2001 in a search for what keeps neurons alive under Alzheimer's-type attack. It blocks the cell-death proteins BAX and BID and signals survival from outside the cell. The analogue HNG, a thousand times more potent, is what animal studies and vendors use.
What the research actually shows. 589 indexed publications. In mice and rats, humanin or HNG at 0.4 to 4 mg/kg protects neurons, heart, retina, kidney and testis from chemical and metabolic injury. In people, every study measured the body's own humanin: it rises after endurance exercise and with a Mediterranean diet, is highest in centenarians, and shifts in diabetes, kidney disease, heart failure and cancer. No human has been given humanin in a study.
The finding vendors leave out. In a 2020 mouse study of triple-negative breast cancer, humanin protected the tumour cells, blunted chemotherapy and accelerated growth and lung metastasis. An anti-apoptotic peptide protects whatever cells it reaches.
Status. Not approved, not in development; a research reagent. Caught by WADA's S0 rule.
Where the evidence is thinnest. All of it in people: dose, effect, safety, and whether a peptide that peaks in the oldest old is a cause of their longevity or a sign of their stress.
Editorial synthesis Written from general pharmacology, the development record and what is commonly reported; not a cited finding. Cited findings on this page carry their own tier.
How humanin works: a peptide encoded in mitochondrial DNA
Written from general pharmacology and the development record, and marked as editorial synthesis; the measurements behind it are quoted in the biomarker section.
- Editorial synthesis from general knowledge
- Editorial synthesis from general knowledge
- Editorial synthesis from general knowledge
Human studies: what was measured, since nothing was given
Every human study in the ledger measured humanin. None administered it. This table says what each one measured and after what.
| Study | Participants | What changed or was compared | Where humanin was measured | What it showed |
|---|---|---|---|---|
| von Walden 2021 | 30 healthy adults | One bout of endurance or resistance exercise | Plasma and muscle | Circulating humanin rose after endurance exercise, not resistance |
| Gidlund 2016 | 55 men with impaired glucose regulation | 12 weeks of resistance training or Nordic walking | Muscle biopsies and serum | Muscle humanin protein rose after resistance training |
| Delgado-Peraza 2023 | People with Alzheimer's disease | Exercise programme | Neuron-derived extracellular vesicles in blood | Vesicle humanin and BDNF rose with exercise |
| Xu 2019 | 158 people with vascular dementia | Hyperbaric oxygen, 12 weeks | Serum | Humanin measured as a marker alongside cognition scores |
| Vicinanza 2025 | 49 older adults with atrial fibrillation | Mediterranean-diet adherence | Plasma | Higher adherence, higher humanin and SHMOOSE |
| Bolignano 2026 | 83 haemodialysis patients, 24 months | Cardiovascular events | Plasma | U-shaped: low and high humanin both predicted events |
| Liao 2026 | 39 atrial-fibrillation patients, 39 controls | Disease vs health | Plasma and atrial tissue | Humanin and MOTS-c altered in AF; HNG then tested in mice |
| Mohamed 2026; Coradduzza 2025 | Breast and prostate cancer patients | Disease vs health | Serum; tissue | Raised in breast cancer serum; down-regulated in prostate lesions |
| Zhao 2023; Küçükali 2026; Rodríguez-Esparragón 2025 | Children with IBD; multiple sclerosis; other cohorts | Disease vs health | Serum; transcripts | Levels shift with chronic inflammation and disease |
| Nashine 2017; Nashine 2026 | Cells carrying mitochondria from macular-degeneration patients | HNG or a humanin fragment added in culture | Cultured cybrid cells | Protected the AMD mitochondria and reduced pro-apoptotic gene expression: a cell study, counted as human by the fetcher |
| Study | Model | Dose and route | Duration | Result |
|---|---|---|---|---|
| Köm Akipek 2026 | Rats given paroxetine | Humanin 1 mg/kg/day, subcutaneous infusion | Phase 1 dosing period | Reversed SSRI-induced sexual dysfunction and dopamine changes |
| Hekim 2026 | Diabetic mice | Humanin 4 mg/kg intraperitoneal | 15 days | Less testicular oxidative damage |
| Bulut 2026 | Diabetic mice | Humanin intraperitoneal, repeated | Not stated | Lower oxidative stress, inflammation, apoptosis; hormone balance restored |
| Lin 2026 | Rats with inherited retinal degeneration | HNG 0.4 or 4 mg/kg intraperitoneal, twice weekly | 1 or 4 weeks | Gene-expression changes in retina; functional results in full text |
| Abueid 2026 | Rats, renal ischaemia-reperfusion | S14G-humanin | Not stated | Antioxidant enzymes restored |
| Liao 2026 | Mice with angiotensin-induced atrial fibrillation | HNG or MOTS-c | Not stated | Less fibrosis and mitochondrial dysfunction |
| Elhusseiny 2026 | Human muscle cells | 10 μM HNG or MOTS-c | Culture | MOTS-c preserved myotube size against dexamethasone; HNG partially |
| Moreno Ayala 2020 | Mice with triple-negative breast cancer | Systemic humanin | Tumour growth period | Tumour apoptosis reduced, chemotherapy blunted, growth and lung metastasis accelerated |
Editorial synthesis Written from general pharmacology, the development record and what is commonly reported; not a cited finding. Cited findings on this page carry their own tier.
What did the studies find?
Results in mice, rats and cells, plus human studies that measured the body's own humanin after exercise, diet or disease. No study has given humanin or its analogue HNG to a person. One row per primary study in this record's ledger. Fields are read from each paper's indexing and abstract. "Analytical method" marks papers about detecting or measuring the compound rather than about its effects.
| Study | Design | n | Species | Dose | Route | Duration | Reported outcome | Tier |
|---|---|---|---|---|---|---|---|---|
| Hosseini 2026 | Randomized controlled trial (humanin measured, not given) | 12 | Humans | — | — | — | The WBRT group showed significant increases in gray matter, subcortical gray matter, cerebellar gray matter volume, and cerebrum white matter compared to the CON group (p < 0.05). | Human clinical trial |
| Vicinanza 2025 | Clinical trial (humanin measured, not given) | 49 | Humans | — | — | — | Patients with high adherence exhibited higher plasma levels of SHMOOSE ( p = 0.046) and Humanin ( p = 0.045). | Human clinical trial |
| Basereh 2025 | Randomized controlled trial (humanin measured, not given) | 15 | Humans | — | — | — | Both CT and AST enhanced antioxidant defenses and reduced IF, with CT + S showing the best effects. | Human clinical trial |
| Delgado-Peraza 2023 | Randomized controlled trial (humanin measured, not given) | — | Humans | — | — | — | NDEV levels of proBDNF, BDNF, and humanin increased in the exercise group, especially in APOE ε4 carriers, but remained unchanged in the control group. | Human clinical trial |
| von 2021 | Randomized controlled trial (humanin measured, not given) | 10 | Humans | — | — | — | Circulating levels of HN were significantly elevated by acute EE but not RE. | Human clinical trial |
| Xu 2019 | Clinical trial (humanin measured, not given) | 158 | Humans | — | — | 5 days; 12 weeks | There was no significant difference in MMSE scores and serum Humanin levels between the two groups before treatment ( p > 0.05). | Human clinical trial |
| Shen 2019 | Clinical trial (humanin measured, not given) | — | Humans | — | — | — | Results- In the screening stage, 1012 differentially methylated CpG sites annotated in 672 genes were found to be significantly associated with large-artery atherosclerotic stroke (mean methylation difference >5%, P… | Human clinical trial |
| Nashine 2017 | Clinical trial (humanin measured, not given) | — | Humans | — | — | — | In AMD cybrids, HNG protected the AMD mitochondria, reduced pro-apoptosis gene and protein levels, upregulated gp130 (a component of the HN receptor complex), and increased the protection against amyloid-β-induced… | Human clinical trial |
| Gidlund 2016 | Randomized controlled trial (humanin measured, not given) | 55 | Humans | — | — | 12 weeks | There was a significant correlation between humanin levels in serum and the improvements in the 2 h glucose loading test in the resistance training group. | Human clinical trial |
| Shen 2026 | Human study (humanin measured, not given) | — | Humans | — | — | — | DHQ adjunctive therapy reduced serum circulating mitochondrial DNA levels in patients with CLF, increased serum SOD2 and Humanin expression, inhibited activation of the NLRP3-mediated pyroptotic pathway, improved… | Observational, human |
| Bolignano 2026 | Multicentre study (humanin measured, not given) | 33 | Humans | — | — | — | Humanin displayed a curvilinear association with CVMM, with both low ( 778 pg/mL) levels linked to an increased risk. | Observational, human |
| Nashine 2026 | Human study (humanin measured, not given) | — | Humans | — | — | — | HNF 14 treatment was associated with improved metabolic activity and reduced cytotoxicity in AMD cybrids, with minimal effects in normal cybrids. | Observational, human |
| Mohamed 2026 | Human study (humanin measured, not given) | — | Humans | — | — | — | Serum humanin concentrations were significantly elevated in breast cancer patients compared with healthy controls (p Conclusion These findings suggest that humanin could serve as a promising biomarker for breast cancer… | Observational, human |
| Küçükali 2026 | Human study (humanin measured, not given) | — | Humans | 6 ng | — | — | Humanin-like 3 levels did not correlate with demographic and clinical variables of MS and NMOSD. | Observational, human |
| Rodríguez-Esparragón 2025 | Human study (humanin measured, not given) | — | Humans | — | — | — | HN and MOTS-c transcript levels differed significantly among study groups, whereas plasma protein concentrations did not discriminate between AD and MCI. | Observational, human |
| Coradduzza 2025 | Human study (humanin measured, not given) | — | Humans | — | — | — | Results showed significant downregulation of Humanin and GAS5 in both PL and PCa compared to BPH, supporting their role in early disease transition. | Observational, human |
| Zhao 2023 | Human study (humanin measured, not given) | 40 | Humans | — | — | — | Serum humanin levels were significantly decreased in children with IBD compared to healthy controls. | Observational, human |
| Abueid 2026 | Animal study | 48 | Rats | — | — | — | Glutathione level and superoxide dismutase activity, which were diminished in the I/R group, were significantly restored following HNG administration. | Animal, preclinical |
| Köm 2026 | Animal study | 10 | Rats | 1 mg/kg/day | oral, subcutaneous | — | Humanin co-administration significantly reversed these adverse effects, restoring ejaculatory frequency, copulatory efficiency, sperm parameters, and hormonal balance. | Animal, preclinical |
| Lin 2026 | Animal study | — | Rats | — | — | 4 weeks | The results showed that high dose HNG at 4 weeks after first injection (WAFI) was associated with the largest change in gene expression in the RPE and retina of treated animals, altering expression of genes involved in… | Animal, preclinical |
| Jang 2026 | Animal study | — | — | — | — | — | — | Animal, preclinical |
| Hekim 2026 | Animal study | 10 | Mice | 4 mg/kg | intraperitoneal | 15 days | Humanin treatment significantly increased TAS and GSH levels while reducing TOS levels compared to the STZ group (P < 0.05). | Animal, preclinical |
| Bulut 2026 | Animal study | 10 | Mice | — | — | — | STZ-induced diabetes significantly increased oxidative stress markers, pro-inflammatory cytokines, and apoptotic activity while reducing antioxidant defenses and anti-inflammatory cytokines compared with controls. | Animal, preclinical |
| Gemeda 2026 | Animal study | — | Cattle | — | — | — | Humanin levels in seminal plasma were significantly higher (p < 0.05) in normozoospermic individuals than in other groups. | Animal, preclinical |
| Bulut 2026 | Animal study | 10 | Mice | — | intraperitoneal | — | STZ-induced diabetes markedly disrupted metabolic hormone balance, as indicated by decreased leptin and irisin levels and increased asprosin concentrations. | Animal, preclinical |
Doses reported in studies
Each entry is a sentence quoted from the study's abstract, with the study's design, sample size and year in front of it. Comparator doses in the same sentence are not attributed to humanin.
- Source
pmid-42480246 - Source
pmid-42438323 - Source
pmid-42471450 - Source
pmid-41732124
Every humanin and HNG dose in the literature, in one table
Every figure here was given to a rodent or a culture. No human dose exists.
| Compound | Species | Route | Dose | Schedule |
|---|---|---|---|---|
| Humanin | Rat | Subcutaneous infusion | 1 mg/kg/day | Continuous, study phase |
| Humanin | Mouse | Intraperitoneal | 4 mg/kg | Daily, 15 days |
| HNG | Rat | Intraperitoneal | 0.4 mg/kg (low) or 4 mg/kg (high) | Twice weekly, 1 to 4 weeks |
| HNG (cited, not in ledger) | Mouse, aged females | Intraperitoneal | 4 mg/kg | Twice weekly, 6 months |
| HNG or MOTS-c | Human muscle cells | Culture | 10 μM | Co-treatment with dexamethasone |
| HNG or humanin fragment | Human AMD cybrid cells | Culture | Not stated in the abstract | Culture |
Figures for this compound circulate on forums and vendor pages. None has been tested in a human study, so none is reproduced here; the table holds every dose a study actually gave. The community's figures are the mouse doses scaled by body weight, a conversion that ignores the fact that the mouse studies dosed for weeks in animals with an induced disease, not for years in healthy people.
Editorial synthesis Written from general pharmacology, the development record and what is commonly reported; not a cited finding. Cited findings on this page carry their own tier.
Side effects: nothing recorded in people, and a tumour signal in mice
No human has received humanin in a study, so there is no side-effect record. What exists is the biology: humanin blocks the cell-death machinery, and in a 2020 mouse study of breast cancer it protected the tumour and sped its spread. Events and their frequency as each study reported them, with the denominator where the abstract gives one.
- Editorial synthesis from general knowledge · primary document to be added to the ledger
- Source
pmid-41518474 - Source
pmid-41849628
Who humanin is discussed for, and the cautions that recur
Studied: rodents with induced injury to brain, heart, retina, kidney or testis; cells from macular-degeneration patients. Measured in: healthy exercisers, prediabetic men, people with dementia, dialysis patients, cancer patients, centenarians. Given to: no human. Community: longevity enthusiasts, usually with MOTS-c and SS-31.
No human study exists, so the cautions come from the biology:
- Cancer, present or past. Humanin blocks apoptosis, the process the body uses to remove damaged and malignant cells; in mice it protected breast-cancer cells, blunted chemotherapy and accelerated metastasis, and its levels are raised in breast-cancer patients. This is the strongest caution on the page and the one vendors omit.
- Reading high levels as good. Humanin is highest in centenarians and in the sickest dialysis patients alike; the field reads it as a stress signal. Adding more to a healthy person is untested in every direction.
- Chemotherapy. The mouse study found humanin protected tumour cells from chemotherapy specifically.
- Pregnancy, children, interactions. No data of any kind.
- Product identity. Humanin and HNG are different molecules with a thousandfold potency difference; a vial's label decides which the buyer gets, and no study has checked.
- Tested athletes. Unapproved substances are prohibited under WADA S0 at all times.
Editorial synthesis Written from general pharmacology, the development record and what is commonly reported; not a cited finding. Cited findings on this page carry their own tier.
Biomarkers measured in studies
Laboratory and clinical measures the studies report tracking, including which hormones did and did not change.
-
Brain structural changes were assessed using magnetic resonance imaging (MRI), and blood samples were analyzed for Humanin (HN), Fibroblast Growth Factor 21 (FGF21), Growth Differentiation Factor 15 (GDF-15), Brain-Derived Neurotrophic Factor (BDNF), and Insulin-like Growth Factor 1 (IGF-1) as well as oxidative and inflammation biomarkers.
Sourcepmid-41975304· quoted verbatim from the abstract -
To investigate Humanin and SHMOOSE (Small Human Mitochondrial ORF Over SErine tRNA), as potential mitochondrial biomarkers of Med-Diet adherence and their associations with markers of oxidative stress.
Sourcepmid-41883858· quoted verbatim from the abstract -
This study examined whether combined aerobic and resistance training (CT) and astaxanthin (AST) supplementation synergistically improve oxidant and inflammatory status as well as metabolic indices in T2DM, focusing on the mediatory role of Humanin (HN) and microRNAs (miRNA-122, miRNA-126-3p, and miRNA-146a).
Sourcepmid-41249265· quoted verbatim from the abstract -
Integration of Humanin significantly enhances the predictive accuracy of CVMM risk models in HD patients, supporting its potential role as an additive biomarker in this high-risk population.
Sourcepmid-41849628· quoted verbatim from the abstract -
Humanin, a mitochondrial-derived peptide (MDP) with reported cytoprotective properties, has been implicated in cancer biology and may play a role in breast cancer pathogenesis.
Sourcepmid-41518474· quoted verbatim from the abstract -
Our aim was to determine whether serum levels of humanin-like 3 (encoded by the MTRNR2L3 gene) may serve as a diagnostic biomarker for differentiation of NMOSD from relapsing remitting multiple sclerosis (RRMS) presenting with clinical features reminiscent of NMOSD.
Sourcepmid-41777515· quoted verbatim from the abstract
Reported interactions
Sentences in which the compound is studied alongside another agent. Read the quotation: the word interaction can also be statistical.
-
Humanin co-administration significantly reversed these adverse effects, restoring ejaculatory frequency, copulatory efficiency, sperm parameters, and hormonal balance.
Sourcepmid-42480246· quoted verbatim from the abstract
Reported timelines
Onset, peak and duration figures as each study reported them.
- Source
pmid-27923980 - Source
pmid-42471450
What is measured over time, and what is not
The human time course is the body's own peptide responding to stimuli: circulating humanin rose within 30 minutes to 3 hours of a single endurance bout; muscle humanin rose over 12 weeks of resistance training; blood levels predicted cardiovascular events over 24 months of dialysis follow-up, at both extremes. The animal time course is weeks: 15 days of daily injection in diabetic mice, 4 weeks of twice-weekly HNG in rats, 6 months in the cited aged-mouse study. Nothing has been measured after giving humanin to a person at any time point, so 'how long does it take to work' has no human answer and 'what does it do over years' has none in any species.
Editorial synthesis Written from general pharmacology, the development record and what is commonly reported; not a cited finding. Cited findings on this page carry their own tier.
Study durations
Treatment and follow-up periods as stated in each abstract.
- Source
pmid-42438323
Routes: intraperitoneal and subcutaneous in rodents, nothing in people
Intraperitoneal injection in most rodent studies, a subcutaneous infusion pump in one rat study, and culture medium. No study has given humanin to a person by any route. Routes of administration named in each study.
- administration by oral, subcutaneous route reported
In Phase 1, fifty rats were randomized into control, sham, paroxetine (20 mg/kg/day via oral gavage), humanin (1 mg/kg/day via subcutaneous infusion), and paroxetine + humanin groups (n = 10/group) to evaluate sexual behavior, sperm quality, serum hormones, and dopaminergic activity within the nucleus accumbens (NAc) and medial preoptic area (MPOA).
Sourcepmid-42480246· quoted verbatim from the abstract - administration by intraperitoneal route reported
Humanin (4 mg/kg) was administered intraperitoneally for 15 days following diabetes induction.
Sourcepmid-42438323· quoted verbatim from the abstract - administration by intraperitoneal route reported
Humanin was administered intraperitoneally for 15 consecutive days.
Sourcepmid-42346352· quoted verbatim from the abstract
Weight-normalized doses, as published
Per-kilogram figures exactly as each study published them, for the species it studied.
- Source
pmid-42438323
Weight-normalized figures above are reproduced exactly as each study published them, for the species it studied. Dose does not scale linearly with body mass between species: metabolic rate, clearance and receptor density differ, and regulatory guidance uses allometric, not proportional, conversion. A mg/kg figure from a rat study is a fact about that rat study and nothing else.
What people report outside the literature
Forum reports describe subcutaneous injection of humanin or HNG for longevity, energy and 'mitochondrial health'. No study has given either to a person; the doses people use are scaled from mouse experiments. What circulates in user communities and clinic marketing, reported so a reader knows the figures and their origin. None of it has been tested in a trial, and none of it is a suggestion.
- Editorial synthesis from general knowledge
Reconstitution mathematics
Lyophilised peptide is supplied as a mass in a vial and dissolved in a stated volume of diluent. Two figures follow from that and nothing else:
- Concentration = mass in vial ÷ diluent volume. A 5 mg vial in 2 mL is 2.5 mg/mL, or 2,500 mcg/mL.
- Volume for a given mass = that mass ÷ concentration. 250 mcg at 2,500 mcg/mL is 0.1 mL. On a U-100 insulin syringe, where 1 mL is 100 units, that is 10 units.
Study doses on this page are quoted as published, in the units the authors used. Converting between mcg, mg and IU, or between a mass and a syringe volume, is arithmetic; deciding a mass is not, and this page does not do it. The reconstitution calculator performs the conversion with the formula shown.
Reading a certificate of analysis
A certificate of analysis is a laboratory report on one batch. Reading it means checking four things in order.
- Identity. Mass spectrometry should report an observed molecular mass matching the peptide's theoretical mass within the instrument's stated tolerance. A certificate with no mass spectrometry, or one that lists only the expected value, has not confirmed identity.
- Purity. HPLC purity is the area of the main peak as a share of all peaks, at a stated wavelength. It says nothing about what the other peaks are, and it is not a measure of quantity.
- Quantity. Net peptide content, by amino acid analysis or UV, is the fraction of the vial's mass that is peptide rather than counter-ions and water. A 5 mg vial at 80% net content holds 4 mg of peptide, and every concentration figure above changes accordingly.
- Provenance. The batch number on the certificate should match the vial, the laboratory should be named and independent of the seller, and the test date should be recent.
This page rates no supplier and links to none. It describes how to read the document.
Equipment described in studies
Published human studies of injectable peptides report the same small set of materials: single-use insulin syringes with fixed needles, most often U-100 with 29 to 31 gauge needles; bacteriostatic water or sterile water for injection as diluent, with the choice stated in the methods; alcohol swabs; and refrigerated storage of reconstituted solution. Animal studies more often report intraperitoneal or intragastric administration and state the vehicle used. Where a study in this record's ledger names its materials or vehicle, that sentence appears under storage or routes above; where it does not, nothing is assumed here.
Storage and handling
Humanin and HNG are laboratory reagents with supplier handling data: lyophilised, stored frozen, reconstituted fresh; both aggregate into fibrils in solution, a property the mechanism papers study on purpose. That is laboratory guidance; vials sold for injection carry unverified contents.
Editorial synthesis Written from general pharmacology, the development record and what is commonly reported; not a cited finding. Cited findings on this page carry their own tier.
Common mistakes in how humanin is discussed
- Counting the human studies as trials of humanin. All seventeen in the ledger measured the body's own peptide. None gave it.
- Reading centenarians' high levels as the cause of their age. The field reads it as a mitochondrial stress signal; the dialysis data, where high levels predicted harm, fit that reading.
- Omitting the tumour study. An anti-apoptotic peptide protecting cancer cells in mice is the predictable result, and it ranks second on Google for the compound's name.
- Scaling mouse doses to people. The rodent studies dosed diseased animals for weeks; the conversion produces a number with no study behind it.
- Treating humanin and HNG as one product. HNG is a thousand times more potent in cell assays; which one a vial contains is unverified.
- Selling the mitochondrial trio as a set. SS-31 is an approved drug, MOTS-c has small human studies, humanin has none; no study has combined any two in a person.
Editorial synthesis Written from general pharmacology, the development record and what is commonly reported; not a cited finding. Cited findings on this page carry their own tier.
Humanin vs MOTS-c, SS-31 and the other mitochondrial peptides
| Compound | What it is | Human evidence | Status |
|---|---|---|---|
| Humanin / HNG | 24-amino-acid peptide encoded in mitochondrial 16S rRNA; anti-apoptotic | None as a treatment; seventeen biomarker studies | Research reagent; not in development |
| MOTS-c | 16-amino-acid peptide encoded in mitochondrial 12S rRNA; metabolic; 277 indexed publications | 4 randomized trials; 13 human studies in its ledger | Not approved; WADA S4.4 |
| SS-31 (elamipretide) | Synthetic tetrapeptide binding cardiolipin; 493 indexed publications | 16 randomized trials; approved for Barth syndrome 2025 | Prescription medicine (Forzinity) |
| SHLP2, SHLP3, SHMOOSE | Other small peptides encoded in mitochondrial DNA | Biomarker studies only | Research reagents |
| NAD+ precursors (NMN, NR) | Oral supplements raising NAD+ | Small trials of NAD+ levels | Dietary supplements; no record on this site |
The 'mitochondrial stack' sold online puts a research reagent with no human data beside a compound with small human studies and an approved orphan drug, and treats them as one class of evidence. They are one class of origin. Only SS-31 has been given to people in controlled trials, and no study has combined any two of the three.
Editorial synthesis Written from general pharmacology, the development record and what is commonly reported; not a cited finding. Cited findings on this page carry their own tier.
Regulatory status: a research reagent with no development programme
Registry facts only. Jurisdiction-level status, compounding categories and sport prohibitions need their own cited documents before they appear here.
- Editorial synthesis from general knowledge · primary document to be added to the ledger
Open questions and limitations
What is genuinely unknown, stated plainly. A missing field is not evidence that the outcome is safe, effective, or established.
- No human has been given humanin or HNG in a study; every human row in the ledger measured endogenous levels.
- The HNG healthspan mouse study is cited editorially and is not in the ledger. The 2020 tumour-progression study (Moreno Ayala, pmid-32444831) and the 2001 discovery paper (Hashimoto, pmid-11371646) were added to it on 29 September 2026.
- No pharmacokinetic study of humanin or HNG in any species is indexed.
Questions people ask
What is HNG (humanin-G)?
HNG is humanin with glycine replacing serine at position 14 (S14G), an analogue roughly a thousand times more potent than the natural peptide in cell-protection assays. It is the form used in most animal studies, including the retinal-degeneration rat work and the mouse healthspan study, and the form usually sold as a research chemical. Like humanin itself, it has never been given to a person in a study.
Does humanin help Alzheimer's disease?
Humanin was discovered in 2001 in a screen for genes that protected neurons from Alzheimer's-type insults, and it protects cultured neurons and mouse models from amyloid toxicity. In people, an exercise study found neuron-derived vesicles carrying more humanin after training in Alzheimer's patients, and a dementia trial measured serum humanin as a marker. No one with Alzheimer's has been given humanin in a study.
What is the half-life of humanin?
Not measured in people. Humanin is a 24-amino-acid peptide degraded quickly in plasma; the analogue HNG was made partly for stability, and animal studies dose it every few days to twice a week. No human pharmacokinetic study exists, and no abstract in the ledger states a figure.
Does humanin extend lifespan?
In mice, one study gave old females the analogue HNG twice a week for six months and found better cognition and metabolic markers, not a longer life; the study is cited editorially and is not in the ledger. In people, humanin levels rise with age and are highest in centenarians, which the field reads as a stress response rather than a cause of longevity. No human has been given humanin to test the idea.
What is humanin?
Humanin is a 24-amino-acid peptide encoded, unusually, in mitochondrial DNA, within the 16S ribosomal RNA gene. Discovered in 2001 in a search for genes that protected neurons from Alzheimer's-type damage, it blocks the cell-death machinery (BAX and BID) and signals through a cell-surface receptor complex. Its blood levels rise with age and after exercise. It has been given to mice and rats, never to people.
What is humanin used for?
In research, as a tool for studying cell survival, mitochondrial stress and ageing, and as a candidate neuroprotectant and cardioprotectant in animal models. In people it is measured as a biomarker. Outside the literature it circulates for longevity and 'mitochondrial health', a use with no human study behind it.
What diseases is humanin linked to?
As a biomarker, its levels shift in Alzheimer's disease, type 2 diabetes, kidney disease, heart failure, atrial fibrillation, inflammatory bowel disease in children, and several cancers, where it is often raised. As an intervention in animals it has protected heart, brain, retina, kidney and testis. As a possible driver, a 2020 mouse study found it protected breast-cancer cells and sped tumour spread.
Has humanin been tested in humans?
Not as a treatment. All seventeen human studies in the ledger measured the body's own humanin, in blood, muscle or neuron-derived vesicles, after exercise, diet or hyperbaric oxygen or as a disease marker. No person has been given humanin or HNG in a published study, and none is registered.
Is there a humanin dose?
Not for people. In animals, humanin was given at 1 mg/kg a day by subcutaneous infusion in rats and 4 mg/kg intraperitoneally in mice; HNG at 0.4 or 4 mg/kg twice a week in rats and, in a cited mouse study, 4 mg/kg twice weekly for six months. The figures that circulate are those doses scaled by body weight.
What is the difference between humanin and HNG?
HNG is humanin with glycine in place of serine at position 14, about a thousand times more potent in cell-protection assays and more stable. Most animal studies use HNG, and so do most vendors. Neither has been given to a person in a study.
Does humanin extend lifespan?
No study has shown that in any species. A cited mouse study gave old females HNG twice weekly for six months and found better cognition and metabolic markers, not a longer life. In people, humanin levels rise with age and peak in centenarians, which the field reads as a stress-response signal rather than a cause of longevity.
What are the side effects of humanin?
None recorded, because no human has received it in a study. The biological caution is that humanin blocks apoptosis: in a 2020 mouse study of triple-negative breast cancer, systemic humanin protected tumour cells, blunted chemotherapy and accelerated growth and lung metastasis, and its levels are raised in breast-cancer patients.
Is humanin FDA approved?
No. It is not approved anywhere, has no development programme or drug code, and is sold as a research reagent and, by other vendors, as a research chemical for injection.
Does humanin help Alzheimer's disease?
It was discovered in 2001 protecting neurons from Alzheimer's-type damage in culture and protects mouse models, which is where its name comes from. In people, exercise raised humanin in neuron-derived vesicles of Alzheimer's patients, and a dementia trial measured it as a marker. No one with Alzheimer's has been given humanin.
Why do centenarians have high humanin?
Levels rise with age and are highest in the oldest old, and within them correlate inversely with grip strength and insulin sensitivity, which points to humanin as a signal mitochondria send under stress rather than a marker of health. High levels also predicted cardiovascular events in dialysis patients. It is measured in these people, not given to them.
Does exercise increase humanin?
Yes. A single bout of endurance cycling raised circulating humanin within hours in healthy adults (resistance exercise did not), twelve weeks of resistance training raised muscle humanin in prediabetic men, and exercise raised humanin in neuron-derived vesicles in Alzheimer's patients. That is the body making its own; nothing was injected.
Is humanin the same as MOTS-c?
No. Both are mitochondrial-derived peptides, encoded in mitochondrial ribosomal-RNA genes, but humanin (16S rRNA) is anti-apoptotic and MOTS-c (12S rRNA) acts on metabolism through AMPK. MOTS-c has small human studies and mouse exercise data; humanin has no human treatment study. They are sold together on the strength of their shared origin.
Is humanin banned in sport?
It is not named on the WADA Prohibited List, but as a substance with no regulatory approval it falls under the list's S0 category, prohibited at all times.
Which species has humanin been studied in?
Given to mice and rats, in models of diabetes, retinal degeneration, kidney injury, SSRI side effects, atrial fibrillation and cancer; measured in humans, cattle and cells. No primate or human dosing study exists.
Sources
Full citations. Every claim above links to one of these by its id.
- 1
- 2
- 3Humanin peptide ameliorates reproductive dysfunction and restores neuroendocrine mechanisms in SSRI-treated male rats.
pmid-42480246· · peer-reviewed - 4Neuroprotective effect of intraperitoneal Humanin-G in retinal degeneration of Royal College of Surgeons rats.
pmid-42471450· · peer-reviewed - 5
- 6Humanin ameliorates diabetes-induced testicular damage in a streptozotocin-induced mouse model.
pmid-42438323· · peer-reviewed - 7Repeated Humanin Treatment Attenuates Oxidative Stress, Inflammation, and Apoptosis in Diabetic Cardiac Tissue.
pmid-42450608· · peer-reviewed - 8Network Topology and Interactomic Analysis Reveal the Regulatory Framework of the Humanin Protein Family (MTRNR2Lx Class).
pmid-42509775· · peer-reviewed - 9Humanin Mitigates Aβ-Induced Retinal Pigment Epithelium Injury via AMPK-Beclin1-Dependent Mitophagy.
pmid-42333946· · peer-reviewed - 10Humanin analogue promotes metabolic reprogramming to protect the ischemic heart
doi-10-64898-2026-06-16-732776· · preprint - 11Humanin as a Molecular Indicator of Semen Quality and Cryotolerance in Crossbred Cattle.
pmid-42223320· · peer-reviewed - 12
Show the remaining 37 sources
- 13Humanin and MOTS-c Attenuate Atrial Fibrillation by Suppressing Fibrosis and Mitochondrial Dysfunction.
pmid-42193373· · peer-reviewed - 14
- 15Humanin as an evolutionarily tuned mitochondrial peptide: Insights from mammalian oxidative stress diversity.
pmid-41864362· · peer-reviewed - 16Circulating Humanin Improves the Prognostic Accuracy of Cardiovascular Risk Models in Chronic Hemodialysis Patients.
pmid-41849628· · peer-reviewed - 17
- 18Characterization of the Effects of a Humanin Fragment Peptide (HNF 14 ) in Age-Related Macular Degeneration.
pmid-41827105· · peer-reviewed - 19Mitochondrial-derived peptides MOTS-c and humanin attenuate dexamethasone-induced atrophy in human skeletal muscle cells.
pmid-41732124· · peer-reviewed - 20Diagnostic potential of serum humanin in breast cancer among the Egyptian population.
pmid-41518474· · peer-reviewed - 21Humanin improves bone health in a glucocorticoid-treated mouse model of Duchenne muscular dystrophy.
pmid-41550496· · peer-reviewed - 22
- 23Insights into the Biomarker Potential of Humanin and Mots-c Expression and Telomere Length in Alzheimer's Disease.
pmid-41303353· · peer-reviewed - 24Exercise Improves Cardiac Dysfunction in D-Galactose-Treated Rats by Regulation of IGF-1-Humanin Pathway.
pmid-41050739· · peer-reviewed - 25
- 26Diagnostic relevance of Humanin, GAS5 and miR-21/miR-103 in prostate disease risk stratification.
pmid-40768089· · peer-reviewed - 27Humanin variants aggregate to produce different fibril morphologies.
pmid-40543583· · peer-reviewed - 28[Gly14]-Humanin ameliorates high glucose-induced endothelial senescence via SIRT6.
pmid-39730568· · peer-reviewed - 29Mitochondrial-derived peptides, HNG and SHLP3, protect cochlear hair cells against gentamicin.
pmid-39433756· · peer-reviewed - 30
- 31Humanin Treatment Protects Against Venetoclax-Induced Bone Growth Retardation in Ex Vivo Cultured Rat Bones.
pmid-38328478· · peer-reviewed - 32A novel link between chronic inflammation and humanin regulation in children.
pmid-38322155· · peer-reviewed - 33Neuron-derived extracellular vesicles in blood reveal effects of exercise in Alzheimer's disease.
pmid-37730689· · peer-reviewed - 34Acute endurance exercise stimulates circulating levels of mitochondrial-derived peptides in humans.
pmid-34351816· · peer-reviewed - 35Inducible fold-switching as a mechanism to fibrillate pro-apoptotic BCL-2 proteins.
pmid-33764501· · peer-reviewed - 36Humanin selectively prevents the activation of pro-apoptotic protein BID by sequestering it into fibers.
pmid-33106313· · peer-reviewed - 37Humanin Promotes Tumor Progression in Experimental Triple Negative Breast Cancer.
pmid-32444831· · peer-reviewed - 38Senescence in the pathogenesis of age-related macular degeneration.
pmid-31897543· · peer-reviewed - 39
- 40Protective Effect of Hyperbaric Oxygen Therapy on Cognitive Function in Patients with Vascular Dementia.
pmid-31134827· · peer-reviewed - 41Epigenome-Wide Association Study Indicates Hypomethylation of MTRNR2L8 in Large-Artery Atherosclerosis Stroke.
pmid-31084332· · peer-reviewed - 42
- 43Humanin skeletal muscle protein levels increase after resistance training in men with impaired glucose metabolism.
pmid-27923980· · peer-reviewed - 44MOTS-c: A novel mitochondrial-derived peptide regulating muscle and fat metabolism.
pmid-27216708· · peer-reviewed - 45Distinct signaling cascades elicited by different formyl peptide receptor 2 (FPR2) agonists.
pmid-23549262· · peer-reviewed - 46The emerging role of the mitochondrial-derived peptide humanin in stress resistance.
pmid-23239898· · peer-reviewed - 47Mapping the specific cytoprotective interaction of humanin with the pro-apoptotic protein bid.
pmid-17927731· · peer-reviewed - 48Formyl peptide receptors: a promiscuous subfamily of G protein-coupled receptors controlling immune responses.
pmid-17084101· · peer-reviewed - 49A rescue factor abolishing neuronal cell death by a wide spectrum of familial Alzheimer's disease genes and Abeta.
pmid-11371646· · peer-reviewed
Reference card
- Compound
- Humanin, mitochondrial peptides
- Evidence tier
- Human clinical trial
- Indexed publications
- 589 · 5 RCTs · 3 other clinical trials
- Approval
- no registered development programme found
- Routes reported
- intraperitoneal, oral, subcutaneous
- Reviewed
- Adam Mirando, PharmD,
Study figures are as published in each source and are not recommendations. Print or save this card with its date.
Last reviewed and what changed
Newest first. These are the record's own revision dates, and the same dates feed the sitemap.
- · Reviewed by Adam Mirando, PharmD, on 2026-09-29.
- · Two papers the page cited editorially are now in the ledger: the 2020 triple-negative breast cancer study showing humanin promotes tumour progression (pmid-32444831) and the 2001 paper that identified humanin as a rescue factor (pmid-11371646). The table no longer marks the first as absent, and the open question narrows to the HNG healthspan study.
- · Misattribution check: reported_timelines (pmid-34351816): the study measured the body's own humanin after exercise and gave none, so it reports no onset or duration of effect; study_durations (pmid-34351816): the quoted interval is a blood-sampling schedule in a study that administered nothing. The sources stay in the ledger; only the claims that put another agent's result under this compound were removed.
- · Written under the sequencing rule after research/intents/humanin.json: guide (8 sections incl. a measured-versus-given human table and an animal-dose table), FAQ to 12 plus 9 from the map, animal dose, duration, timeline and adverse-event claims from the abstracts, the 2020 tumour finding as editorial with pending_source, mechanism, reported-use, regulatory; seventeen human rows relabelled 'measured, not given'; misdrafted interactions removed; intent-driven H1 and title.
- · Claims drafted extractively from 46 ledger sources by scripts/draft_claims.py: 27 claims, 25 evidence-table rows. Status researched -> draft.
- · Claims drafted extractively from 46 ledger sources by scripts/draft_claims.py: 33 claims, 25 evidence-table rows. Status researched -> draft.
- · Evidence fetched from Europe PMC and ChEMBL by scripts/fetch_evidence.py; claims not yet written.