Humanin: Evidence, Mechanism, Dosing & Legal Status

Humanin (HN) is a 24-amino-acid peptide and the first identified member of the mitochondrial-derived peptide (MDP) family — encoded within the 16S rRNA region of the mitochondrial genome and discovered in 2001 while screening for factors that protect neurons from amyloid-β toxicity.12 It is marketed in longevity and biohacking circles as a neuroprotective, anti-aging peptide. Its mechanism is real and well-studied; its proof in humans does not exist. This monograph separates the two.

This article is informational and editorial content for research and educational purposes only. It is not medical advice, not a protocol to follow, and not a sourcing guide. Humanin is not an FDA-approved drug; it is sold as a "research chemical not for human use" and is prohibited in sport. Dosing figures are reported strictly as seen in the published (almost entirely animal) literature for completeness — not as recommendations. Consult a licensed clinician before any health decision.

What is humanin and how does it work?

Humanin is translated from a short open reading frame within the mitochondrial 16S rRNA region, making it the founding member of the MDP family that also includes MOTS-c and the SHLPs.118 Substituting glycine for serine at position 14 yields S14G-humanin, or "HNG," which is roughly 1,000-fold more potent than wild-type humanin and is the form used in most experiments.51 Key residues are well mapped: F6 and K21 govern IGFBP-3 binding, C8 is required for BAX/tBID binding, and a cluster including L9–L11, P19 and V20 governs the secretion needed for cytoprotection.2

The mechanism — all of it preclinical — runs through three convergent cytoprotective pathways. First, humanin binds insulin-like growth-factor-binding protein-3 (IGFBP-3) at its C-terminal domain, interfering with importin-β binding and suppressing IGFBP-3-mediated, caspase-dependent apoptosis; this was the original 2003 mechanistic finding and the link the field used to predict effects on glucose homeostasis via the IGF/insulin axis.2 Second, humanin binds the pro-apoptotic Bcl-2-family proteins BAX, tBID and BimEL, preventing BAX translocation to mitochondria and blocking the intrinsic apoptotic cascade — an effect specific to BAX-dependent death.2 Third, it signals through a trimeric receptor (CNTFR-α / gp130 / WSX-1) to activate JAK2/STAT3 and drive AKT and ERK1/2; notably, in aged but not young mouse hippocampus, humanin increased AKT and ERK1/2 phosphorylation, an age-dependent in-vivo effect.3 Native humanin's circulating half-life is only about 30 minutes, which is why the longer-acting, far more potent HNG analog is used experimentally; as a peptide it has negligible oral bioavailability, and no human pharmacokinetic data exist.124

What is the evidence by indication?

The orientation point is decisive: there are no human randomized controlled trials, no Phase 1/2 interventional trials, and no completed clinical endpoints for exogenous humanin or HNG in any indication — a fact you can verify against the empty interventional record on registries such as ClinicalTrials.gov.24 Every efficacy claim below is preclinical (Grade C) unless it is an observational human biomarker correlation (association, not proof of benefit).

Neuroprotection is humanin's founding role — rescue of neurons from amyloid-β toxicity in vitro, with HNG ameliorating cerebral infarction and suppressing inflammatory cytokines in a mouse middle-cerebral-artery-occlusion stroke model, and preventing age-related cognitive decline in middle-aged mice.94 The human signal is purely observational: Alzheimer's patients have lower CSF humanin (a tiny sample), and the mitochondrial SNP rs2854128 associates with roughly 15% lower circulating humanin and accelerated cognitive aging in about 16,000 older adults, with a more pronounced effect in African-American ancestry.14 That is correlation, not evidence that giving humanin treats dementia.

The longevity narrative deserves a careful read. Transgenic humanin-overexpressing C. elegans lived about 7.3% longer in a daf-16/FOXO-dependent manner, humanin declines with age across species, and circulating levels are higher and more sustained in children of centenarians than in children of non-centenarians.119 But these are associations in tiny samples — and crucially, in the very mouse-healthspan experiment that drives the longevity story, midlife HNG (4 mg/kg IP twice weekly for 14 months) did not significantly extend lifespan at that dose.1 Metabolic and cardioprotective effects are similarly preclinical: chronic HNG reduced visceral fat, increased lean mass and lowered IGF-I in mice, and S14G-humanin protected the mouse heart against pressure-overload and streptozotocin-induced dysfunction.678 Anti-inflammatory effects in asthma, gout and retinal models are all animal or cell-based.101112

What doses appear in the literature?

Reported strictly as information, not a protocol — and unlike most peptides, humanin has no human dose at all. There is no established human dose, route or schedule; the figures below are animal experimental doses, reported for completeness.24 The route in animals is parenteral, almost always intraperitoneal injection in rodents, because the peptide structure precludes meaningful oral absorption.1 The pivotal mouse healthspan study used HNG at 4 mg/kg IP twice weekly for 14 months, starting at 18 months of age in female C57BL/6N mice.1 Most research uses the HNG (S14G) analog rather than native humanin, because native humanin's roughly 30-minute half-life and 1,000-fold lower potency make it impractical.15 No clinical reconstitution standard exists because there is no clinical product; material sold online is unregulated research chemical of unverified identity and purity.21

How safe is humanin?

There are no human safety data at all — with zero interventional human exposure, there is no characterized human adverse-event profile, no PK, and no drug-interaction data.24 The principal theoretical risk, and the one with the most animal support, is tumor-promotion and chemoresistance. Humanin's core anti-apoptotic, pro-survival biology is exactly what aggressive tumors exploit: exogenous humanin protected triple-negative breast cancer cells from apoptosis, promoted tumor progression, stimulated spontaneous lung metastases and impaired chemotherapy's anti-metastatic effect, while silencing humanin did the opposite.14 In glioblastoma, humanin drives progression and angiogenesis via the integrin-αV/TGF-β axis and facilitates chemoresistance, with silencing re-sensitizing cells to chemotherapy.1516 The findings are mixed, but the pro-tumor signal is real and unresolved. Paradoxically, humanin also protects normal tissues from chemotherapy toxicity in animals — the same property that can shield tumors.13

Two further cautions matter. Humanin-overexpressing mice were about 12% shorter and 10% lighter with roughly 46% smaller litters, pointing to growth and fertility trade-offs from chronic IGF-I-axis suppression.1 And in hemodialysis patients, both abnormally low and abnormally high circulating humanin associated with higher mortality and cardiovascular risk — a U-shaped relationship that argues against assuming "more humanin is better."17 By extrapolation, theoretical contraindications include active or prior malignancy, pregnancy and lactation, and pediatric use; none are formally established because no human use is sanctioned. As with all research-chemical peptides, product-quality risk — unverified composition, sterility and endotoxin status — is a concrete additional hazard.21

What is the FDA and WADA status in 2026?

Humanin is not an FDA-approved drug — no NDA or BLA for any indication. It is not an approved compounding bulk substance: it does not appear on the FDA 503A bulk-substances list, and there is no 503B outsourcing-facility status. It is distributed only as a research chemical labeled "for research use only — not for human consumption," and selling or using it for human administration is outside FDA's lawful framework, particularly as the FDA has tightened oversight of research peptides generally through 2025–2026.2123 It is not a DEA-controlled substance.

For athletes the picture is unambiguous despite humanin not being individually named on the 2026 Prohibited List. Category S0 (Non-Approved Substances) is a catch-all that prohibits any pharmacological substance not approved by any governmental health authority for human therapeutic use — explicitly including pre-clinical or clinical-development drugs and research-chemical peptides. Humanin and HNG therefore fall under S0 and are prohibited at all times, in and out of competition.2022 Legitimate laboratory research remains lawful under proper oversight; the prohibitions concern human therapeutic administration.21

Bottom line. Humanin pairs a coherent, multi-pathway preclinical mechanism and genuinely interesting human longevity correlations with a near-total absence of human proof — not a single interventional trial, no human efficacy, no PK, no safety profile. From a root-cause, mitochondrial-health lens the more defensible reading is that humanin is today best viewed as a biomarker of mitochondrial and metabolic resilience and a target for understanding healthy aging, with exogenous administration remaining experimental, unapproved and not advisable outside controlled research — especially for anyone with cancer history, during pregnancy, or for athletes (WADA S0). Verdict: promising mechanism, real safety questions, zero clinical validation — graded C. Regulatory facts here are current as of June 2026 and should be re-verified as FDA peptide oversight evolves.