evipedia.ai
← Back

Humanin for Health & Longevity

Evidence Review created on 10/25/2026 using AI4L / Opus 5.5

Also known as: HN, HNG, S14G-Humanin, MT-RNR2 Peptide

Motivation

Humanin is a short chain of 24 amino acids that cells build from a small stretch of the genetic material held inside mitochondria, the compartments that supply most of a cell’s energy. It was identified in human brain tissue in 2001. Interest in it comes from its origin: mitochondria make it and release it into the blood, so it is studied as a signal that mitochondria send to the rest of the body.

Researchers have measured blood humanin across age groups, in people with heart disease, memory loss and diabetes, and in the children of people who lived past one hundred. Separate laboratory work has tested a longer-acting modified form in cells, worms and mice. No clinical trial has yet given humanin or any of its forms to people, so the human record rests on measurements of what the body produces on its own.

This review examines that evidence: what the human measurements and the animal work cover, how far each reaches, which harms have been raised, and what has never been tested in people.

Benefits - Risks - Protocol - Conclusion

This section lists high-level overviews of humanin from experts and journals that discuss the peptide, or the mitochondrial-derived peptide class it belongs to, in depth.

No content from the six priority experts and publications qualified. Searches of all six platforms (documented in the comment above) returned no humanin overview, only passing mentions: a single interview question on Lifespan.io, cited in Historical Context rather than listed here, a two-sentence research summary in a Lifespan.io roundup, and one name-check in a Chris Kresser podcast episode on injectable peptides.

Grokipedia

  • Humanin

    The dedicated article covers the mitochondrial reading frame, natural variants, receptor interactions and discovery history, and is the fastest orientation to how the peptide is encoded and signaled.

Examine

No Examine article on humanin exists. A direct search of examine.com returned no results for the term.

Examine covers ingredients sold as dietary supplements; humanin is not sold as one. It is an injectable research peptide with no approved or over-the-counter product, which is the likely reason for its absence.

ConsumerLab

No ConsumerLab article on humanin exists. A direct search of consumerlab.com returned no results for the term.

ConsumerLab tests consumer supplements, foods and household products. Humanin is sold only as a research-grade peptide and is not available as a consumer supplement, so no product exists for ConsumerLab to test.

Systematic Reviews

Two systematic reviews cover humanin; both pool mechanistic and observational work rather than trials of the peptide.

Fewer than five qualify because the field has produced almost no pooled evidence; the list is not padded. The principal risk raised in this review — tumor promotion and chemoresistance — has no systematic review or meta-analysis at all, so that side of the trade-off is unrepresented in the pooled literature.

Mechanism of Action

Humanin is encoded by a short open reading frame (a stretch of genetic code that can be read into a protein) inside the mitochondrial ribosomal RNA gene MT-RNR2, so mitochondria rather than the cell nucleus specify it (Hashimoto 2001). It acts in two places.

Inside the cell it binds Bax (a protein that perforates mitochondrial membranes to start programmed cell death) and prevents its movement from the cytosol to the mitochondrion, blocking cytochrome c release (Guo 2003); it also activates chaperone-mediated autophagy, the route by which cells deliver damaged proteins for disposal (Gong 2018). Outside the cell it behaves like a cytokine (an immune signaling protein), binding a complex of CNTFR-α/WSX-1/gp130 (three linked surface proteins of the interleukin-6 receptor family) (Hashimoto 2009) and separately FPRL1/FPR2 (formyl peptide receptor-like 1), a G-protein-coupled receptor that amyloid-beta also uses (Ying 2004). Binding activates STAT3, ERK1/2 and AKT signaling (three growth and survival relays), with different responses in old and young hippocampus (Kim 2016).

It is pharmacologically a peptide: cleared by peptidases rather than liver cytochrome P450 enzymes (the main drug-processing enzymes), with a short, analogue-dependent plasma half-life, detectable in liver but not brain or heart after injection in rats, and circulating bound to IGFBP-3 (the main carrier of insulin-like growth factor 1) (Chin 2013).

Competing readings exist: the same survival signal is read as protection of healthy tissue and as a lifeline for malignant and senescent cells (worn-out cells that no longer divide but keep signaling) (Mendelsohn 2018).

Historical Context & Evolution

Humanin’s original purpose was neuroprotection, not longevity. In 2001 the Nishimoto laboratory at Keio University screened a library of the genes active in an unaffected brain region of an Alzheimer’s patient, looking for anything that stopped neurons dying from familial Alzheimer’s mutations and amyloid-beta; one clone did, and they named its peptide product Humanin (Hashimoto 2001). Structure-activity work the same year mapped the active domain and showed that replacing serine at position 14 with glycine yielded HNG, the more potent analogue used in nearly all later animal work (Hashimoto 2001).

Two findings moved it out of Alzheimer’s research. The Bax discovery gave it a general anti-apoptotic mechanism (Guo 2003), and clamp studies (the standard test of insulin action) showing that humanin infusion improved insulin action in rodents made it a metabolic candidate (Muzumdar 2009). Its longevity framing came later, when circulating humanin was tied to the growth hormone and insulin-like growth factor axis and to lifespan across mouse models (Lee 2014).

Scientific opinion has not settled. The discovery of MOTS-c and the small humanin-like peptides created a mitochondrial-derived peptide field with commercial backing — CohBar, co-founded by two of the field’s leading researchers, who are named as consultants and stockholders in their own papers — yet no humanin candidate reached a clinical trial, and CohBar began liquidating in 2023. Nir Barzilai has argued that an expired patent, not negative data, is what left humanin undeveloped (Lifespan.io interview with Nir Barzilai).

Expected Benefits

High 🟩 🟩 🟩

No benefit reaches High: no human clinical endpoint or validated clinical surrogate has ever been measured in people given humanin, because no trial has administered it, so no outcome class exists to replicate across research groups.

Medium 🟩 🟩

No benefit reaches Medium: there is no single human trial of administered humanin, and the observational cohorts measure the body’s own peptide rather than the intervention, which keeps that class of evidence indirect.

Low 🟩

Preserved Coronary Endothelial Function

Higher circulating humanin accompanies a better coronary blood-flow response to acetylcholine, and lower levels accompany coronary artery disease. The evidence is indirect: both studies measured the body’s own peptide, not administered humanin, and neither tested whether raising it changes vessel function (Widmer 2013; Cai 2022).

Magnitude: 0.9 ng/mL higher in patients with normal coronary endothelial function than in those with dysfunction (2.2 ± 1.5 vs 1.3 ± 1.1 ng/mL, p = 0.03); the paper reports no 95% CI (confidence interval, the range in which the true value probably lies).

Slower Cognitive Aging

A mitochondrial variant (rs2854128) at the end of the humanin reading frame tracks with lower circulating peptide and with an older cognitive age at the same chronological age. Indirect: genotype stands in for the body’s own output, and no humanin was given (Yen 2018).

Magnitude: about 2 years greater cognitive age in African American carriers of the alternate allele versus about 0.2 years in Caucasian American carriers; no treated comparison group exists and the paper reports no confidence interval.

Family History of Exceptional Longevity

Children of centenarians, who outlive age-matched peers, carry higher circulating humanin. The comparison is observational and indirect — it measures endogenous peptide, not supplementation — and cannot separate humanin from the rest of an inherited profile (Yen 2020).

Magnitude: direction only — higher in children of centenarians (n = 18) than in age-matched controls (n = 19), reported as significant; the literature reports no outcome figure for this comparison.

Exercise-Linked Glucose Handling ⚠️ Conflicted

Twelve weeks of resistance training raised muscle humanin in prediabetic men; serum levels tracked their glucose change (Gidlund 2016). Indirect: exercise was the intervention, not humanin. Acute resistance exercise moved nothing, and high-endurance athletes had lower levels (von Walden 2021; Alser 2022). Net reading: only sustained training raised muscle humanin.

Magnitude: muscle humanin 35% higher after 12 weeks in the resistance-training arm (84.7 to 111.8 pg/mL), with a significant group-by-time difference (p = 0.022) and no change in the walking or control arms; the trial reports no 95% CI.

Speculative 🟨

Lifespan Extension ⚠️ Conflicted

Humanin overexpression lengthened lifespan in Caenorhabditis elegans, requiring the stress-resistance gene daf-16, yet the same report found no lifespan difference in analogue-treated mice. Net reading: the signal is worm-only (Yen 2020).

Fourteen months of analogue injections lowered collagen deposition, fibroblast proliferation and apoptosis in aging mouse hearts. Animal work only; no human cardiac imaging or biopsy data exist (Qin 2018).

Protection of Normal Tissue During Chemotherapy

In mice, analogues limited bortezomib-induced bone-growth impairment and chemotherapy-induced germ-cell apoptosis without blunting tumor kill. The basis is animal and tumor-graft work; no human oncology study has tested it (Eriksson 2014; Jia 2019).

Retinal Pigment Epithelium Protection

Humanin preserved mitochondrial function and limited senescence in cultured retinal pigment epithelial cells and cell lines carrying mitochondria from macular-degeneration donors. Cell culture only; no animal or human eye outcome reported (Sreekumar 2016; Nashine 2017).

Autophagy Induction in Skeletal Muscle

Humanin raised autophagy gene expression and reduced misfolded-protein accumulation in mouse muscle, and autophagy was required for its worm lifespan effect. Mechanistic and animal only; no human muscle data (Kim 2022).

Benefit-Modifying Factors

  • Mitochondrial haplogroup and natural variants: carriers of haplogroup N1b (a maternal-line mitochondrial DNA group) make humanin P3S, enriched among Ashkenazi centenarians carrying APOE4 (a cholesterol-transport gene variant that raises Alzheimer’s risk); a commentary adds that non-Ashkenazi carriers are untested (Miller 2024; Logan 2024).

  • The rs2854128 polymorphism: this mitochondrial single-nucleotide change sits at the end of the humanin reading frame, does not alter the peptide sequence, and accompanies roughly 14% lower circulating levels, with a stronger cognitive association in African American participants (Yen 2018).

  • Baseline humanin level: people with coronary artery disease or Alzheimer’s disease start from lower circulating levels, while type 2 diabetes showed no difference, so any headroom for benefit is largest in the first two groups (Conte 2021; Cai 2022).

  • Growth hormone axis activity: administering growth hormone or insulin-like growth factor 1 lowers circulating humanin, so high growth-axis activity, whether natural or from therapy, works against the proposed benefit (Lee 2014).

  • Sex: rodent pharmacokinetic and chemoprotection studies used male animals, and the ovarian and germ-cell work is sex-specific, so no sex comparison of response exists (Chin 2013; Wang 2021).

  • Pre-existing conditions: insulin resistance with polycystic ovary syndrome (a hormonal and metabolic disorder of the ovaries) showed lower follicular-fluid humanin, while sepsis-associated acute kidney injury in children showed higher serum levels, so direction of deviation is condition-specific (Wang 2021; Atreya 2024).

  • Age: hippocampal AKT and ERK1/2 signaling responded to analogue injection in old mice but not young, suggesting the signaling benefit may require an aged background (Kim 2016).

Potential Risks & Side Effects

High 🟥 🟥 🟥

No risk reaches High: no adverse event in a person given humanin has ever been documented, since the peptide has never been administered to people in a reported study, so no human harm endpoint exists to replicate.

Medium 🟥 🟥

No risk reaches Medium: no single human trial and no consistent comparison-group observational data describe harm from administered humanin; the only human harm signal comes from cohorts measuring the body’s own peptide, which is indirect.

Low 🟥

Excess Mortality and Cardiac Events at Extreme Circulating Levels

In chronic hemodialysis patients, both very low and very high circulating humanin accompanied more deaths and non-fatal cardiovascular events, a U-shaped association. Indirect: it measures endogenous levels in advanced kidney disease, not administered peptide (Bolignano 2024).

Magnitude: HR (hazard ratio, the relative rate of events between groups) 5.84 (95% CI 1.65–20.64) above 759.5 pg/mL and 4.25 (95% CI 1.52–11.86) below 450.7 pg/mL, versus the mid-range reference quartile, falling to 4.50 (95% CI 1.21–18.73) and 2.49 (95% CI 1.01–7.49) in the fully adjusted model; events occurred in 6 of 23 reference-quartile patients (26.1%), the only quartile for which the paper reports an absolute rate.

Contaminated, Mislabeled or Immunogenic Research-Grade Product

Humanin is sold only as a research chemical, and reviews of the unregulated injectable-peptide market document contamination and manufacturing impurities (Moiz 2026); the US Food and Drug Administration has flagged aggregation-related immunogenicity for other synthetic peptides (FDA Category 2 list). Indirect: this evidence covers the product class, not humanin.

Magnitude: direction only — risk rises with unverified sourcing and with absent certificates of analysis, sterility testing or endotoxin testing; the literature reports no outcome figure for humanin products specifically.

Speculative 🟨

Tumor Promotion and Chemoresistance ⚠️ Conflicted

Humanin itself accelerated breast tumor growth and glioblastoma (brain cancer) invasion in mice; analogue work found no loss of tumor kill. Net reading: the animal-level hazard remains unresolved (Moreno Ayala 2020; Ha 2024; Eriksson 2014).

Amplified Senescent-Cell Inflammatory Signaling

In senescent human fibroblasts, humanin raised respiration and selected components of the senescence-associated secretory phenotype (the inflammatory output of worn-out cells) through JAK (a cytokine signaling relay). Cell culture only (Kim 2018).

Suppression of Growth and Fertility Signaling

Humanin-transgenic mice were shorter and lighter with smaller litters, and analogue injection lowered circulating insulin-like growth factor 1 in rodents. Animal only; no human reproductive data (Yen 2020; Chin 2013).

Fibril Formation by Humanin Variants

Humanin point substitutions, including the S14G analogue sequence, assembled into distinct amyloid-like fibrils in solution, raising a theoretical aggregation concern for injected peptide. Biophysical in-vitro work only; no tissue or human finding (Morris 2025).

Risk-Modifying Factors

  • Mitochondrial haplogroup and the rs2854128 variant: haplogroup U6a7a1a carriers natively make the potent S14G sequence and N1b carriers P3S, so added peptide stacks on an unlike native form; rs2854128 carriers start roughly 14% lower, widening the exposure jump (Logan 2024; Yen 2018).

  • FPRL1/FPR2 receptor expression: chemotherapy raised this receptor on glioblastoma cells, and blocking it removed the analogue’s protective effect on them, so tumors that express it heavily are the ones the pro-tumor signal would reach (Peña Agudelo 2023).

  • Baseline humanin level: the hemodialysis cohort’s harm signal sat above 759.5 pg/mL, so a high starting level is the condition under which adding more carries the clearest human-level concern (Bolignano 2024).

  • Analogue sequence chosen: single-residue substitutions, the S14G sequence of HNG among them, aggregate into different fibril forms in vitro, so which analogue a vial holds may change aggregation behavior (Morris 2025).

  • Sex: the chemoprotection and germ-cell work is male-specific, and ovarian and follicular-fluid work female-specific; no study has compared harm between sexes, so sex-based risk differences are unknown (Jia 2019; Wang 2021).

  • Active or prior cancer: humanin was found upregulated in triple-negative breast tumor tissue (a subtype that lacks the three usual treatment targets) compared with healthy donor mammary gland, and silencing it increased tumor-cell chemosensitivity in culture (Moreno Ayala 2020).

  • Advanced kidney disease: end-stage kidney disease on dialysis is the one human population in which an association between humanin levels and death has been measured, and it was adverse at both extremes (Bolignano 2024).

  • Age: cancer incidence and senescent-cell burden both rise with age, and the two main speculative hazards act on exactly those substrates, so the theoretical risk profile worsens at the older end of the target range (Kim 2018).

Key Interactions & Contraindications

Documented interactions:

  • Growth hormone and insulin-like growth factor 1 therapy: four weeks of growth hormone injections lowered plasma humanin about 20% in growth-hormone-deficient children, and analogue injection lowers insulin-like growth factor 1 in rodents, so each reduces the other’s expected benefit (Lee 2014; Chin 2013).

Theoretical interactions (inferred from mechanism, not studied in humans):

  • Cytotoxic chemotherapy (doxorubicin, cyclophosphamide, temozolomide, bortezomib): inferred from blockade of Bax-dependent programmed cell death, the route most of these agents use to kill tumor cells; animal results split between preserved and reduced tumor kill (Eriksson 2014; Peña Agudelo 2023).

  • Insulin and insulin secretagogues (drugs that make the pancreas release insulin: glipizide, glimepiride, repaglinide): inferred from additive glucose lowering, since central and peripheral humanin infusion improved insulin sensitivity in rodent clamp studies; the clinical consequence would be hypoglycemia (abnormally low blood sugar) (Muzumdar 2009).

  • Glucocorticoids (steroid anti-inflammatory drugs: prednisone, dexamethasone): inferred from opposing actions on the Hedgehog pathway (a signaling route that drives bone lengthening), where humanin prevented glucocorticoid-induced impairment of bone growth in cultured growth-plate cartilage and rodents (Zaman 2019).

  • Senolytic combinations (dasatinib plus quercetin, fisetin): inferred from humanin’s cytoprotection of senescent cells, which could oppose drugs whose purpose is to kill them; the proposed consequence is reduced senolytic effect (Mendelsohn 2018).

  • High-dose zinc supplements: inferred from in-vitro chelation, where zinc bound humanin through its single cysteine with micromolar affinity; the proposed consequence is less free peptide reaching receptors, weakening the humanin effect (Armas 2006).

  • Glucose-lowering supplements (berberine, chromium, alpha-lipoic acid): inferred from the same additive glucose lowering as insulin above, since rodent clamp studies raised insulin sensitivity; the clinical consequence would be hypoglycemia (Muzumdar 2009).

No over-the-counter medication interaction is documented, and none is inferable from humanin’s targets: no common non-prescription drug class acts on its receptors or on Bax-dependent programmed cell death.

Populations who should avoid Humanin:

  • Anyone with active, recently treated or previously treated cancer: humanin was upregulated in triple-negative breast tumor tissue, and humanin administration accelerated tumor growth and lung metastasis in tumor-bearing mice (Moreno Ayala 2020).

  • Anyone receiving cytotoxic chemotherapy: inferred from the shared Bax-dependent cell-death route and from analogue-driven chemoresistance in glioblastoma models; no human data (Peña Agudelo 2023).

  • People with end-stage kidney disease on maintenance hemodialysis, particularly with humanin above 759.5 pg/mL: this threshold marked the high-risk arm of the U-shaped mortality and cardiovascular association (Bolignano 2024).

  • Pregnancy and lactation: inferred from reduced body length and litter size in humanin-transgenic mice and from suppression of the growth-factor axis; no human pregnancy data (Yen 2020).

  • People under 18 or still growing: inferred from humanin’s action on growth-plate cartilage and the Hedgehog pathway, where it altered bone growth in animal models; no human growth data (Zaman 2019).

Mitigating actions where one is known: separating high-dose zinc from any injection by several hours addresses the chelation concern; glucose monitoring addresses the additive glucose-lowering concern; no mitigating action is known for the chemotherapy or cancer concerns, where the inferred response is avoidance rather than adjustment.

Risk Mitigation Strategies

Parameters below without a citation reflect common practice rather than published evidence.

  • Oncology screening before any use: age-appropriate cancer screening and a personal and family cancer history address the tumor-promotion signal seen with humanin administration in tumor-bearing mice (Moreno Ayala 2020).

  • Certificate of analysis for every vial: identity and purity by mass spectrometry plus sterility and endotoxin results address contamination, impurity and mislabeling, the documented hazards of the unregulated injectable-peptide market (Moiz 2026).

  • Baseline and periodic humanin measurement: a starting level avoids pushing an already-high concentration higher, the condition associated with excess mortality above 759.5 pg/mL in dialysis patients (Bolignano 2024).

  • Glucose self-monitoring during the first weeks: fingerstick readings before and two hours after meals detect the additive glucose lowering inferred from rodent clamp studies, preventing symptomatic hypoglycemia (Muzumdar 2009).

  • Lowest workable dose and widest interval: animal regimens used twice-weekly injection rather than daily, which limits cumulative exposure to a peptide with no human safety record (Yen 2020).

  • Stop rule on any new mass or unexplained weight loss: discontinuation pending evaluation addresses the one hazard — tumor promotion — that would be irreversible if it occurred.

  • Aseptic injection technique: single-use needles, alcohol site preparation and refrigerated storage of reconstituted peptide address injection-site infection and abscess, the generic hazard of self-injected non-pharmaceutical product.

Therapeutic Protocol

Parameters below without a citation — timing, cycling and titration steps — reflect common practice rather than published evidence.

  • No established human regimen: no dose, route or schedule for humanin has been studied in people; nothing below is a validated human protocol, and the only quantitative anchors are animal regimens.

  • Animal regimen most often cited: the analogue HNG at 4 mg/kg intraperitoneally (into the abdominal cavity) twice weekly in middle-aged mice, continued for 6 to 14 months (Yen 2018; Yen 2020; Qin 2018).

  • Lower animal regimen: 1 µg per mouse twice weekly during a two-week chemotherapy cycle, the schedule used for the bone-growth protection result (Eriksson 2014).

  • Competing approach — raising endogenous output: acute endurance exercise raised plasma humanin and resistance training raised muscle humanin, an alternative to injection with human rather than animal evidence (von Walden 2021; Gidlund 2016).

  • Route in gray-market practice: subcutaneous injection of reconstituted freeze-dried (lyophilized) powder, chosen for convenience; animal work used the abdominal-cavity route, which is not used in people.

  • Who popularized each approach: the University of Southern California gerontology laboratory and CohBar drove the exogenous-analogue approach; the exercise-physiology groups in Stockholm and Auckland drove the endogenous approach (Woodhead 2021).

  • Best time of day: no human or animal study compared dosing times; practice follows convenience, typically morning, with no evidence behind it.

  • Half-life: short and analogue-dependent in rodents — longer for the non-IGFBP-3-binding analogue HNGF6A than for HNG, and longer in rats than mice; no human value exists (Chin 2013).

  • Single versus split dosing: every animal regimen used a single injection twice weekly, so no split-dose comparison exists; the short rodent half-life argues against assuming sustained exposure.

  • Genetic polymorphisms: haplogroup U6a7a1a carriers naturally produce the potent S14G sequence and haplogroup N1b carriers the P3S variant, which could change what an added analogue contributes (noted in the Logan 2024 commentary).

  • Sex-based differences: rodent pharmacokinetics was characterized in males and the longevity dosing work in females, with no head-to-head comparison, so no sex-specific dose can be stated (Chin 2013).

  • Age: hippocampal signaling responded to the analogue in old mice but not young, and the animal regimens began in middle age, so the modeled population is older rather than younger (Kim 2016).

  • Baseline biomarkers: lower starting humanin, as seen in coronary artery disease and Alzheimer’s cohorts, defines the group with room to move; high baseline defines the group with least (Cai 2022).

  • Pre-existing conditions: cancer history, chemotherapy, dialysis-dependent kidney disease and pregnancy are the conditions that change the calculation most, as covered in Key Interactions & Contraindications.

Discontinuation & Cycling

  • Lifelong versus short-term: neither is established. The longevity animal work ran continuously from middle age to death, while the chemoprotection work ran for a single two-week cycle (Yen 2020; Eriksson 2014).

  • Withdrawal effects: none reported in any species. Transgenic mice carried lifelong elevation without a described rebound, and no study has measured what happens when administration stops (Yen 2020).

  • Tapering: no tapering protocol exists and none is suggested by the mechanism, since the peptide is cleared quickly in rodents rather than accumulating (Chin 2013).

  • Cycling for sustained efficacy: untested. No study compared continuous with intermittent administration, so whether receptor signaling desensitizes over months is unknown (Kim 2016).

  • Reason to stop: the stop triggers that appear in practice are a new mass, unexplained weight loss, a cancer diagnosis or starting chemotherapy — all tied to the tumor-promotion signal rather than to withdrawal concerns.

Sourcing and Quality

  • No pharmaceutical-grade product exists: humanin is not an approved drug in any jurisdiction and does not appear on FDA’s section 503A bulk drug substances list, so no legally compounded preparation is available.

  • Not on FDA’s safety-risk list either: humanin is absent from the Category 2 list of nominated bulk substances with identified safety risks, which means it was never nominated for compounding rather than that it was cleared.

  • What to look for in a research peptide: identity and purity by high-performance liquid chromatography and mass spectrometry above 95%, lot-specific certificate of analysis, sterility and bacterial endotoxin results, and freeze-dried rather than pre-dissolved supply.

  • Name confusion to check: vials may hold native humanin, the analogue HNG (S14G), the non-binding analogue HNGF6A, or colivelin, a different hybrid peptide; the animal evidence differs by molecule (Chin 2013).

  • Reputable sources: established peptide manufacturers such as Bachem and Phoenix Pharmaceuticals supply humanin as a laboratory reagent with analytical documentation; no consumer brand and no third-party-tested supplement product exists, and compounding pharmacies cannot legally supply it.

  • Storage and reconstitution: freeze-dried peptide is stored frozen and protected from light; once reconstituted in bacteriostatic water it is refrigerated and short-dated, since peptide degradation and microbial growth are the practical quality failures.

Practical Considerations

  • Time to effect: unknown in people. Animal benefits appeared over weeks to months — six months of injections before a balance difference in mice and ten before a memory difference — so no short-term readout is expected (Yen 2018).

  • Common pitfall — conflating peptides: humanin and MOTS-c are different molecules with different evidence; the metabolic meta-analysis most often cited for “mitochondrial-derived peptides” pools MOTS-c data only (Zhou 2024).

  • Common pitfall — scaling animal doses: the 4 mg/kg mouse dose is not a human dose; rodent half-life is short, species-dependent and measured by research assay, with no human pharmacokinetic study to convert from (Chin 2013).

  • Common pitfall — assuming oral or brain delivery: humanin is a peptide digested when swallowed, and after injection in rats it was undetectable in brain and heart, which is why nanoparticle and intranasal carriers are still research subjects.

  • Regulatory status: no approval anywhere, no legal compounding pathway on FDA’s section 503A bulks list, and supply labeled research-use-only; purchase and self-injection sit outside medical regulation.

  • Cost and accessibility: research-grade vials are inexpensive but come without pharmaceutical assurance, no clinician can prescribe it, and no insurer or health system covers it — so no payer has a financial stake in favoring or blocking it relative to alternatives.

Interaction with Foundational Habits

  • Sleep: no interaction is documented in either direction. No human or animal study measured sleep architecture, sleepiness or circadian timing with humanin, and the receptor pathways involved have no established sleep role, so the direction is best described as unknown rather than none.

  • Nutrition: indirect and positive. Higher adherence to a Mediterranean dietary pattern accompanied higher circulating humanin in older adults with atrial fibrillation, with olive oil, fish and legume intake correlating individually; the association was cross-sectional, so diet-driven causation is unproven (Vicinanza 2025).

  • Exercise: direct but conflicting. Acute endurance exercise raised plasma humanin and acute resistance exercise did not; twelve weeks of resistance training raised muscle humanin; yet high-endurance professional athletes had lower levels than low-to-moderate-endurance athletes (von Walden 2021; Alser 2022).

  • Stress management: no direct human interaction is documented. The mechanistic link runs the other way — humanin is released as part of the mitochondrial stress response — and the nearest evidence is animal work on glucocorticoid-induced bone-growth impairment (Zaman 2019).

Monitoring Protocol & Defining Success

Baseline testing establishes where a person starts on the two measures the peptide is expected to move — circulating humanin and the growth-factor axis it suppresses in rodents — and on the two safety measures tied to its inferred hazards, blood glucose and cancer surveillance. Because no human trial exists, there is no validated target and no published schedule; every biomarker below is read against a person’s own baseline rather than a treatment goal.

Ongoing monitoring in practice follows a simple cadence: glucose self-monitoring daily for the first two weeks, then a blood panel at 4 weeks, 12 weeks, and every 6 months thereafter, with age-appropriate cancer screening kept on its normal schedule. Success is defined narrowly: no deterioration in the safety markers, with any movement in the expected-effect markers read as biology rather than benefit, since no outcome has been shown to follow from it.

Biomarker Optimal Functional Range Why Measure It? Context/Notes
Plasma humanin No established target; track change from the individual’s own baseline Expected-change marker: the direct output of administration Measured only by research ELISA (enzyme-linked immunosorbent assay, an antibody-based protein test); assays are not standardized between laboratories and no clinical test is offered
Insulin-like growth factor 1 (IGF-1) Age- and sex-specific laboratory reference interval (standard reference range) Expected-change marker: analogue injection lowered this and its binding protein in rodents Fasting not required; report must state the age band used, as intervals narrow with age; same-laboratory repeat testing avoids assay drift
Fasting plasma glucose 70–99 mg/dL (standard reference range) Safety check: a low result would stop or reduce use, given additive glucose lowering inferred from rodent studies Requires 8–12 hours fasting; best paired with fasting insulin on the same draw; capillary self-monitoring before and two hours after meals covers the first weeks
High-sensitivity C-reactive protein (hs-CRP) Below 1.0 mg/L, the low-risk category defined by Ridker 2003 Expected-change marker: analogue injection lowered circulating inflammatory markers in aged mice Conventional laboratory reference interval runs to about 3.0 mg/L; source for the functional range is Ridker 2003; avoid testing within two weeks of infection or hard training
Prostate-specific antigen (PSA), in men No established target; track change from the individual’s own baseline Safety check: a rise would stop use, given the tumor-promotion signal in animal models Published PSA figures are screening cutoffs rather than reference intervals, so the velocity of change matters more than any single value; avoid testing after ejaculation or cycling

Qualitative markers tracked alongside the panel:

  • Exercise capacity and recovery between sessions
  • Cognitive clarity and short-term memory in daily tasks
  • Energy stability across the day, especially before meals
  • Injection-site appearance, swelling or persistent redness
  • Any new lump, mass or unexplained weight change

Emerging Research

  • Perioperative mitochondrial-peptide measurement in kidney transplantation: NCT07678073, 68 participants, non-randomized, primary endpoint the change in humanin and MOTS-c; listed as recruiting, its registered completion date of October 2026 passed with no results posted. A positive result would support humanin as a stress-response marker; a null result would weaken the claim that it tracks tissue protection.

  • Exercise and mitochondrial peptides in chronic kidney disease: NCT07438002, 50 participants, randomized, humanin an exploratory outcome; listed as recruiting, its registered completion date of September 2026 passed with no results posted. A positive result would strengthen the exercise route to raising endogenous peptide; a null result would weaken the Low-graded exercise-linked benefit above.

  • Humanin as an early marker of kidney injury after heart transplantation: NCT06125249, 60 participants, observational; the registered completion date of June 2024 has passed and the record’s status is unknown with no results posted.

  • Plasma humanin in acute kidney injury: NCT06105229, 60 participants, observational; the registered completion date of June 2024 has passed, the status is unknown and no results have been posted.

  • Humanin isoforms and complications after cardiac surgery: NCT03431844 completed in August 2019 with 106 participants; no results are posted on the registry, and the group’s published output addressed genetic variants instead (Eltermaa 2019).

  • Future direction that could strengthen the case — a first human dose study: no pharmacokinetic or safety study in people has been run, so the rodent-only kinetics remain the field’s hardest gap (Chin 2013).

  • Future direction that could strengthen the case — variant-targeted use: whether the P3S variant’s apparent advantage extends beyond Ashkenazi carriers of APOE4 to the millions of other haplogroup N1b carriers is the open question raised in a commentary on that work (Miller 2024; Logan 2024).

  • Future direction that could weaken the case — resolving the tumor signal: the pro-tumor and chemoresistance findings have not been reconciled with the chemoprotection findings, and either outcome of that work would move the risk grade (Moreno Ayala 2020; Peña Agudelo 2023).

  • Future direction that could weaken the case — assay standardization: humanin is quantified only by in-house ELISAs with no clinical standard, so a standardized assay could shift or invalidate existing level-based associations (Miller 2020).

Conclusion

Humanin is a small peptide that mitochondria themselves encode and release into the blood. Its appeal to people who track their own biology is easy to see: it is made by the body, it has been measured in the long-lived and in the chronically ill, and it has protected cells and animals against a wide range of harmful stresses. The gap between that appeal and the evidence is the central finding of this review. Nobody has ever given humanin to a person in a reported study. Everything human in the record is a measurement of what the body makes on its own, which can show company but not cause.

The strongest signals are therefore modest ones: better blood-vessel function and a longer-lived family history go with higher natural levels, and exercise raises those levels. The main concerns run in the opposite direction from the appeal, because the same ability to keep stressed cells alive has made tumors grow faster and resist treatment in animals, and has amplified inflammatory output from worn-out cells in culture. Two laboratories that produced much of the favorable work had a financial stake in a company built on these peptides, a conflict of interest that stands beside the favorable record without erasing it. For now, humanin remains a well-grounded idea with an empty human column.

Top - Benefits - Risks - Protocol