Dashnaiv Peptides
Compounds·mitochondrial peptides·mitochondria-derived cytoprotective peptide

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.

Human clinical trial Reviewed 49 sources Updated

At a glance

Evidence availability
Human clinical evidence
Strongest tier in the indexed literature
Indexed publications
589
Europe PMC, title or abstract, 2026-09-14
Human studies in ledger
0
0 randomized; study count, not efficacy proof
Approval
none found
no registered development programme
Routes reported
intraperitoneal, oral, subcutaneous
from studies in this ledger
Studied in
Cattle, Humans, Mice, Rats
25 primary studies in the evidence table
Reviewed by Adam Mirando, PharmD, on . What changed

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.

Human clinical studies in this ledger00 randomized · 0 other human clinical
Evidence strengthNot yet assessedSeparate from evidence availability
Efficacy signalNot yet assessedRead the study result, not the tier alone

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.

  • Humanin is a 24-amino-acid peptide (21 when made inside mitochondria) encoded in an unexpected place: within the mitochondrial 16S ribosomal RNA gene, MT-RNR2. It was found in 2001 by Hashimoto and colleagues screening a brain library from an Alzheimer's patient for sequences that kept neurons alive under amyloid attack, and named for the hope that it would restore 'humanity' to people with dementia. It is the founding member of the mitochondrial-derived peptides, a family that includes MOTS-c and the SHLPs, which are challenging the idea that mitochondrial DNA encodes only respiratory-chain machinery.Editorial synthesis
    Editorial synthesis from general knowledge
  • It protects cells two ways. Inside the cell it binds the pro-apoptotic proteins BAX and BID and sequesters them into fibrils, stopping them from puncturing the mitochondrial membrane; outside, it signals through a receptor complex (CNTFR, WSX-1 and gp130) and through formyl-peptide receptors to activate survival pathways. The analogue HNG, with a single serine-to-glycine change, is about a thousand times more potent in these assays and is what most animal studies use. In animals the effect is broad: protection of neurons, heart, retina, kidney, testis and muscle under various insults.Editorial synthesis
    Editorial synthesis from general knowledge
  • The human data are about levels, not treatment. Circulating humanin rises with age and is highest in centenarians; it rises after endurance exercise and with Mediterranean-diet adherence; it is altered in diabetes, kidney disease, heart failure and cancer. The field's reading is that humanin is a stress-response signal that mitochondria send when under pressure, so higher levels mark a body coping rather than a body thriving. Whether adding more from outside helps or harms a person has never been tested, and the tumour study is the reason 'helps' is not the default assumption.Editorial synthesis
    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.

Human studies of humanin in the ledger, all observational with respect to the peptide.
StudyParticipantsWhat changed or was comparedWhere humanin was measuredWhat it showed
von Walden 202130 healthy adultsOne bout of endurance or resistance exercisePlasma and muscleCirculating humanin rose after endurance exercise, not resistance
Gidlund 201655 men with impaired glucose regulation12 weeks of resistance training or Nordic walkingMuscle biopsies and serumMuscle humanin protein rose after resistance training
Delgado-Peraza 2023People with Alzheimer's diseaseExercise programmeNeuron-derived extracellular vesicles in bloodVesicle humanin and BDNF rose with exercise
Xu 2019158 people with vascular dementiaHyperbaric oxygen, 12 weeksSerumHumanin measured as a marker alongside cognition scores
Vicinanza 202549 older adults with atrial fibrillationMediterranean-diet adherencePlasmaHigher adherence, higher humanin and SHMOOSE
Bolignano 202683 haemodialysis patients, 24 monthsCardiovascular eventsPlasmaU-shaped: low and high humanin both predicted events
Liao 202639 atrial-fibrillation patients, 39 controlsDisease vs healthPlasma and atrial tissueHumanin and MOTS-c altered in AF; HNG then tested in mice
Mohamed 2026; Coradduzza 2025Breast and prostate cancer patientsDisease vs healthSerum; tissueRaised in breast cancer serum; down-regulated in prostate lesions
Zhao 2023; Küçükali 2026; Rodríguez-Esparragón 2025Children with IBD; multiple sclerosis; other cohortsDisease vs healthSerum; transcriptsLevels shift with chronic inflammation and disease
Nashine 2017; Nashine 2026Cells carrying mitochondria from macular-degeneration patientsHNG or a humanin fragment added in cultureCultured cybrid cellsProtected the AMD mitochondria and reduced pro-apoptotic gene expression: a cell study, counted as human by the fetcher
Animal and cell studies that gave humanin or HNG.
StudyModelDose and routeDurationResult
Köm Akipek 2026Rats given paroxetineHumanin 1 mg/kg/day, subcutaneous infusionPhase 1 dosing periodReversed SSRI-induced sexual dysfunction and dopamine changes
Hekim 2026Diabetic miceHumanin 4 mg/kg intraperitoneal15 daysLess testicular oxidative damage
Bulut 2026Diabetic miceHumanin intraperitoneal, repeatedNot statedLower oxidative stress, inflammation, apoptosis; hormone balance restored
Lin 2026Rats with inherited retinal degenerationHNG 0.4 or 4 mg/kg intraperitoneal, twice weekly1 or 4 weeksGene-expression changes in retina; functional results in full text
Abueid 2026Rats, renal ischaemia-reperfusionS14G-humaninNot statedAntioxidant enzymes restored
Liao 2026Mice with angiotensin-induced atrial fibrillationHNG or MOTS-cNot statedLess fibrosis and mitochondrial dysfunction
Elhusseiny 2026Human muscle cells10 μM HNG or MOTS-cCultureMOTS-c preserved myotube size against dexamethasone; HNG partially
Moreno Ayala 2020Mice with triple-negative breast cancerSystemic humaninTumour growth periodTumour 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.

25 primary studies in this record's ledger, strongest evidence first. A dash means the abstract did not state the value.
StudyDesignnSpeciesDoseRouteDurationReported outcomeTier
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) — Humans6 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 Rats1 mg/kg/dayoral, 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 Mice4 mg/kgintraperitoneal15 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.

  • Rats received humanin at 1 mg/kg a day by subcutaneous infusion alongside paroxetine.Animal, preclinical
    Source pmid-42480246
  • Diabetic mice received humanin 4 mg/kg intraperitoneally for 15 days.Animal, preclinical
    Source pmid-42438323
  • Rats with retinal degeneration received the analogue HNG at 0.4 or 4 mg/kg intraperitoneally twice a week for one or four weeks.Animal, preclinical
    Source pmid-42471450
  • Human muscle cells in culture were treated with 10 micromolar HNG or MOTS-c against dexamethasone.Mechanistic, in vitro
    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.

Doses of humanin and HNG administered in studies.
CompoundSpeciesRouteDoseSchedule
HumaninRatSubcutaneous infusion1 mg/kg/dayContinuous, study phase
HumaninMouseIntraperitoneal4 mg/kgDaily, 15 days
HNGRatIntraperitoneal0.4 mg/kg (low) or 4 mg/kg (high)Twice weekly, 1 to 4 weeks
HNG (cited, not in ledger)Mouse, aged femalesIntraperitoneal4 mg/kgTwice weekly, 6 months
HNG or MOTS-cHuman muscle cellsCulture10 μMCo-treatment with dexamethasone
HNG or humanin fragmentHuman AMD cybrid cellsCultureNot stated in the abstractCulture

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.

  • No human has received humanin or HNG in a study, so no side effect has ever been recorded in a person. The one safety signal in the literature comes from mice: a 2020 study of triple-negative breast cancer found humanin and its receptors expressed in human tumours, that added humanin protected cancer cells from apoptosis, and that systemic humanin in tumour-bearing mice reduced tumour cell death, blunted chemotherapy's anti-tumour and anti-metastatic effect, and accelerated tumour growth and lung metastasis. That is what an anti-apoptotic peptide would be expected to do to a tumour, and it is not in the ledger yet; it ranks second on Google for the compound's name.Editorial synthesis
    Editorial synthesis from general knowledge · primary document to be added to the ledger
  • Humanin levels are raised in breast-cancer patients' serum, which the authors propose as a diagnostic marker; it is also consistent with the tumour-protection biology.Observational, human
    Source pmid-41518474
  • In dialysis patients, both low and high humanin levels predicted cardiovascular events, a U-shaped relationship.Observational, human
    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.

  • Randomized controlled trial humans n = 12 2026 Human clinical trial
    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.
    Source pmid-41975304 · quoted verbatim from the abstract
  • Clinical trial humans n = 49 2025 Human clinical trial
    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.
    Source pmid-41883858 · quoted verbatim from the abstract
  • Randomized controlled trial humans n = 15 2025 Human clinical trial
    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).
    Source pmid-41249265 · quoted verbatim from the abstract
  • Multicentre study humans n = 33 2026 Observational, human
    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.
    Source pmid-41849628 · quoted verbatim from the abstract
  • Human study humans 2026 Observational, human
    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.
    Source pmid-41518474 · quoted verbatim from the abstract
  • Human study humans 2026 Observational, human
    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.
    Source pmid-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.

  • Animal study rats n = 10 2026 Animal, preclinical
    Humanin co-administration significantly reversed these adverse effects, restoring ejaculatory frequency, copulatory efficiency, sperm parameters, and hormonal balance.
    Source pmid-42480246 · quoted verbatim from the abstract

Reported timelines

Onset, peak and duration figures as each study reported them.

  • Muscle humanin protein rose after 12 weeks of resistance training in men with impaired glucose regulation.Human clinical trial
    Source pmid-27923980
  • In rats, four weeks of high-dose HNG produced the largest gene-expression changes in the retina.Animal, preclinical
    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.

  • Fifteen days of daily injection in diabetic mice; one or four weeks of twice-weekly HNG in rats.Animal, preclinical
    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.

  • Animal study rats n = 10 2026 Animal, preclinical
    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).
    Source pmid-42480246 · quoted verbatim from the abstract
  • Animal study mice n = 10 2026 Animal, preclinical
    administration by intraperitoneal route reported
    Humanin (4 mg/kg) was administered intraperitoneally for 15 days following diabetes induction.
    Source pmid-42438323 · quoted verbatim from the abstract
  • Animal study mice n = 10 2026 Animal, preclinical
    administration by intraperitoneal route reported
    Humanin was administered intraperitoneally for 15 consecutive days.
    Source pmid-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.

  • 4 mg/kg intraperitoneal in mice and rats; 0.4 mg/kg as the low dose in rats; 1 mg/kg/day by subcutaneous infusion in rats.Animal, preclinical
    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.

  • Humanin, usually as HNG, circulates for longevity, energy, 'mitochondrial health' and neuroprotection, injected subcutaneously two or three times a week, often in a set with MOTS-c and SS-31 and sometimes with NAD+ precursors. Reports describe more energy and clearer thinking, or nothing, and few complaints; users scale their doses from the mouse studies by body weight. None of this comes from a human study, because none exists. 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 above holds every dose a study actually gave to an animal.Editorial synthesis
    Editorial synthesis from general knowledge
Method note · shared across compounds

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.

Method note · shared across compounds

Reading a certificate of analysis

A certificate of analysis is a laboratory report on one batch. Reading it means checking four things in order.

  1. 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.
  2. 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.
  3. 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.
  4. 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.

Method note · shared across compounds

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

  1. Counting the human studies as trials of humanin. All seventeen in the ledger measured the body's own peptide. None gave it.
  2. 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.
  3. 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.
  4. Scaling mouse doses to people. The rodent studies dosed diseased animals for weeks; the conversion produces a number with no study behind it.
  5. Treating humanin and HNG as one product. HNG is a thousand times more potent in cell assays; which one a vial contains is unverified.
  6. 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

Humanin beside the mitochondrial peptides it is sold with, from their own records.
CompoundWhat it isHuman evidenceStatus
Humanin / HNG24-amino-acid peptide encoded in mitochondrial 16S rRNA; anti-apoptoticNone as a treatment; seventeen biomarker studiesResearch reagent; not in development
MOTS-c16-amino-acid peptide encoded in mitochondrial 12S rRNA; metabolic; 277 indexed publications4 randomized trials; 13 human studies in its ledgerNot approved; WADA S4.4
SS-31 (elamipretide)Synthetic tetrapeptide binding cardiolipin; 493 indexed publications16 randomized trials; approved for Barth syndrome 2025Prescription medicine (Forzinity)
SHLP2, SHLP3, SHMOOSEOther small peptides encoded in mitochondrial DNABiomarker studies onlyResearch reagents
NAD+ precursors (NMN, NR)Oral supplements raising NAD+Small trials of NAD+ levelsDietary 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.

Other compounds in its class

Same class in the registry: MOTS-c, SS-31. Each row shows what that compound's own record states; nothing is inferred across rows.

3 compounds in the mitochondrial peptides class. Each row states what that compound's own record says; nothing is inferred across rows.
CompoundTierPublicationsRCTsRecord
Humanin Human clinical trial 589 5 draft
MOTS-c Human clinical trial 277 4 draft
SS-31 Approved label 493 16 draft

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.

  • Not approved anywhere and not in clinical development: no company has registered a trial of humanin or HNG in people, and it has no drug-development code. It is sold as a research reagent by peptide-synthesis companies and as a research chemical for injection by others. Not named on the WADA Prohibited List; as an unapproved substance it falls under S0. Not known to have been nominated to FDA's compounding lists.Editorial synthesis
    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. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10
    Humanin analogue promotes metabolic reprogramming to protect the ischemic heart
    doi-10-64898-2026-06-16-732776 · · preprint
  11. 11
  12. 12
Show the remaining 37 sources
  1. 13
  2. 14
  3. 15
  4. 16
  5. 17
  6. 18
  7. 19
  8. 20
  9. 21
  10. 22
  11. 23
  12. 24
  13. 25
  14. 26
  15. 27
  16. 28
  17. 29
  18. 30
  19. 31
  20. 32
  21. 33
  22. 34
  23. 35
  24. 36
  25. 37
  26. 38
  27. 39
  28. 40
  29. 41
  30. 42
  31. 43
  32. 44
  33. 45
  34. 46
  35. 47
  36. 48
  37. 49

Reference card

Reference card · generated /compounds/humanin
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.

  1. · Reviewed by Adam Mirando, PharmD, on 2026-09-29.
  2. · 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.
  3. · 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.
  4. · 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.
  5. · Claims drafted extractively from 46 ledger sources by scripts/draft_claims.py: 27 claims, 25 evidence-table rows. Status researched -> draft.
  6. · Claims drafted extractively from 46 ledger sources by scripts/draft_claims.py: 33 claims, 25 evidence-table rows. Status researched -> draft.
  7. · Evidence fetched from Europe PMC and ChEMBL by scripts/fetch_evidence.py; claims not yet written.