Reduced adiposity in rodent models [3]
Consistent reduction in body fat in obese rodents with increased lipolysis and reduced lipogenesis, without GH-like growth effects.
Research information only. PrecisePep is an educational library. It is not medical advice and is not intended to assist, encourage or enable self-administration. The compounds documented here are research chemicals — most are not approved for human use. Read the full disclaimer.
PrecisePep/Peptide Library/AOD-9604
Metabolic & Adipose Regulation
hGH fragment 176-191 · AOD9604 · lipolytic fragment
The C-terminal fragment of growth hormone, developed to deliver GH-like fat loss without GH-like side effects — which it achieved by delivering, in the pivotal human trial, no fat loss either.
AOD-9604 is the rare research peptide with a clear human answer — and the answer was no. It is a 16-amino-acid analogue of the C-terminal region of human growth hormone — the hGH(177–191) fragment with an N-terminal tyrosine added, conventionally referred to as fragment 176–191 — the region associated with GH's lipolytic activity. The design premise was elegant: isolate the fat-burning part of the molecule and leave behind the growth-promoting part, giving lipolysis without IGF-1 elevation, insulin resistance or tissue growth.
The first half worked. AOD-9604 does not raise IGF-1 and does not produce GH-like growth effects. The second half did not. A phase 2b obesity trial run by Metabolic Pharmaceuticals failed to separate from placebo on weight loss, and development was discontinued.[1][2]
It persists in the research-chemical market largely because the pre-clinical rodent data was encouraging and the human data is not widely known. It has since been affirmed GRAS in the United States as an ingredient and appears in some cosmetic and joint-health contexts — a very different regulatory posture from being an approved fat-loss drug, which it is not.
Most compounds on this site sit in the gap between promising pre-clinical data and no human test. This one was tested and did not work for its intended purpose. That is a more informative result than the silence surrounding most of the library, and it is routinely omitted from vendor descriptions.
Lipolysis and lipogenesis modulation. In rodent adipose tissue AOD-9604 increases lipolysis and inhibits lipogenesis, reproducing the metabolic component of GH activity. The effect appears to involve beta-3 adrenergic receptor pathways rather than the GH receptor itself.
No GH receptor binding. The fragment does not bind the GH receptor, does not raise IGF-1, and does not produce the insulin resistance or tissue growth associated with GH — the entire point of the design.
Cartilage and joint effects. A later research direction with some pre-clinical support, and the basis for its appearance in joint-health formulations. Distinct from the fat-loss claim and separately unproven in humans.
Why rodent results may not translate. Rodent adipocyte beta-adrenergic physiology differs substantially from human. Compounds that drive lipolysis convincingly in mice failing in humans is a well-worn pattern in obesity pharmacology, and this is a textbook example of it.
Effects that have been reported. A study-tier entry means a result was published — not that the effect is established, reproducible, or transferable from a rodent to a person.
Highest confidenceReported in peer-reviewed literature, registered clinical trials, or regulatory filings. Note the model: much peptide literature is in vitro or rodent work, not human trial data.
Consistent reduction in body fat in obese rodents with increased lipolysis and reduced lipogenesis, without GH-like growth effects.
Confirmed in human studies — the design objective was met. AOD-9604 does not stimulate the GH/IGF-1 axis.
Clinical studies including a 12-week trial reported good tolerability with no significant adverse-event signal, and no effect on glucose tolerance.
The pivotal obesity trial did not demonstrate significant weight reduction against placebo, and clinical development was discontinued. This is the finding that matters most on this page.
Affirmed generally recognised as safe in the United States for certain ingredient uses — a safety designation for a food or cosmetic context, not an efficacy finding and not a drug approval.
Unproven — inference onlyExtrapolated from a known mechanism, receptor profile, or adjacent compound class. Biologically plausible but not directly demonstrated for this compound in this context.
The original premise, and the reason it is still sold. The mechanism is real in rodents; the human trial that tested the premise did not support it.
Pre-clinical support and the basis for joint-health formulations. No human efficacy data.
Visceral fat worsens insulin resistance and androgen excess, so reducing it is a genuine PCOS strategy. The problem is upstream: the phase 2b trial designed to test whether AOD-9604 produces fat loss found no separation from placebo. A fat-loss mechanism that did not produce fat loss cannot deliver the downstream benefit.
Proposed on the basis of a distinct mechanism from incretins. If the compound does not produce fat loss alone in humans, additivity is a difficult claim to sustain.
The most charitable reading of a negative trial. Possible in principle; entirely unevidenced.
Uncontrolled reportsAggregated from research-community logs, forums, vendor literature and self-reported experience. Uncontrolled, unverified, subject to placebo, polypharmacy and product-purity confounds. Lowest evidentiary weight.
Reported, and difficult to separate from the diet and training changes typically made at the same time — which is precisely what a placebo-controlled trial exists to disentangle, and it found nothing.
Reported with some consistency and aligning with the cartilage research direction rather than the fat-loss one.
The most reliable community observation, and consistent with the human safety data. A compound with little effect tends to have few side effects.
Non-response reports are unusually prominent for this compound. That is what agreement between anecdote and trial data looks like.
Adverse effects, contraindications and unknowns. The grey-market tier is where under-reporting is worst: uncontrolled use produces no systematic safety surveillance at all, so absence of a reported harm is not evidence of safety.
Highest confidenceReported in peer-reviewed literature, registered clinical trials, or regulatory filings. Note the model: much peptide literature is in vitro or rodent work, not human trial data.
No significant adverse-event signal across clinical studies, and no effect on glucose tolerance or IGF-1.
Absence of the fluid retention, arthralgia, carpal tunnel symptoms and insulin resistance that accompany GH-axis stimulation.
Unproven — inference onlyExtrapolated from a known mechanism, receptor profile, or adjacent compound class. Biologically plausible but not directly demonstrated for this compound in this context.
Trials ran to twelve weeks. Multi-year use in healthy adults is uncharacterised, as for everything here.
The most substantive risk attached to a compound that does not appear to work: time, money and attention spent on it instead of on interventions that do.
A 16-residue fragment is simple to synthesise, which cuts both ways — easy to make correctly and easy to substitute.
Uncontrolled reportsAggregated from research-community logs, forums, vendor literature and self-reported experience. Uncontrolled, unverified, subject to placebo, polypharmacy and product-purity confounds. Lowest evidentiary weight.
Mild and infrequent.
The community adverse-effect column for this compound is close to empty.
Figures below are documented, not recommended. Study ranges come from published protocols in the stated model; grey-market ranges are what the research community reports using and carry no safety validation of any kind.
Highest confidenceReported in peer-reviewed literature, registered clinical trials, or regulatory filings. Note the model: much peptide literature is in vitro or rodent work, not human trial data.
Unproven — inference onlyExtrapolated from a known mechanism, receptor profile, or adjacent compound class. Biologically plausible but not directly demonstrated for this compound in this context.
Reasoned from GH physiology — insulin opposes lipolysis. Applies to the mechanism regardless of whether the mechanism produces a clinical effect.
Community convention aligning administration with a fasted, catecholamine-elevated window.
Uncontrolled reportsAggregated from research-community logs, forums, vendor literature and self-reported experience. Uncontrolled, unverified, subject to placebo, polypharmacy and product-purity confounds. Lowest evidentiary weight.
The standard community range, typically morning and fasted.
Split dosing on half-life grounds.
Typical course length before assessment.
| Approach | Dose | Frequency | Timing |
|---|---|---|---|
| Standard | 300 mcg | Once daily | Morning, fasted |
| Higher-end | 500 mcg | Once daily | Morning, fasted |
| Split | 250 mcg | Twice daily | Fasted windows |
Combination effects. Almost no peptide combination has been studied as a combination in humans; stack logic is overwhelmingly mechanistic inference and community practice. Each linked compound has its own page.
Highest confidenceReported in peer-reviewed literature, registered clinical trials, or regulatory filings. Note the model: much peptide literature is in vitro or rodent work, not human trial data.
No study has combined AOD-9604 with another compound.
Unproven — inference onlyExtrapolated from a known mechanism, receptor profile, or adjacent compound class. Biologically plausible but not directly demonstrated for this compound in this context.
Proposed as a distinct mechanism alongside appetite suppression. The premise requires AOD-9604 to have a fat-loss effect, which the human trial did not find.
Sometimes proposed as complementary. Redundant if the mechanism is a subset of GH activity, and the secretagogue delivers the whole molecule.
The cartilage research direction is the more defensible pairing rationale, and it is pre-clinical.
Uncontrolled reportsAggregated from research-community logs, forums, vendor literature and self-reported experience. Uncontrolled, unverified, subject to placebo, polypharmacy and product-purity confounds. Lowest evidentiary weight.
Common, and the effect is not separable from the incretin doing the work alongside it.
Whole schedules — dosing, titration, cycle length, and off-time — as documented in trials, as proposed from mechanism, and as circulated in the research community. Reproduced for study and comparison only.
Highest confidenceReported in peer-reviewed literature, registered clinical trials, or regulatory filings. Note the model: much peptide literature is in vitro or rodent work, not human trial data.
Randomised placebo-controlled, twelve weeks, weight as primary endpoint. It did not separate from placebo.
Unproven — inference onlyExtrapolated from a known mechanism, receptor profile, or adjacent compound class. Biologically plausible but not directly demonstrated for this compound in this context.
This is the one compound on this site where a properly designed human trial asked the question everyone is asking and returned a negative answer. That should carry more weight than a rodent lipolysis result.
Uncontrolled reportsAggregated from research-community logs, forums, vendor literature and self-reported experience. Uncontrolled, unverified, subject to placebo, polypharmacy and product-purity confounds. Lowest evidentiary weight.
The general community pattern.
It does not appear in either documented protocol.
The visceral-adiposity rationale for PCOS is coherent. It also depends on a fat-loss effect that a placebo-controlled human trial specifically looked for and did not find.
AOD-9604 is not approved as a drug anywhere and has never been studied in PCOS. More importantly for this section: its phase 2b obesity trial failed to separate from placebo on weight loss, and clinical development was discontinued.
PCOS has treatments that work. Metformin is first-line for the metabolic phenotype, inexpensive, with decades of data. Combined hormonal contraceptives address hyperandrogenism and cycle regulation; letrozole is first-line for ovulation induction; inositol has reasonable evidence and an excellent safety profile; GLP-1 receptor agonists now have randomised evidence in this population. Weight loss of 5–10% alone restores ovulation in a meaningful proportion.
Against that, a compound with no human data is not an alternative — it is a substitution of something unmeasured for something measured. The comparison worth making is not “peptide versus nothing”.
Pathophysiology
Visceral adipose tissue worsens insulin resistance and contributes to androgen excess, and central adiposity is common in PCOS. Reducing it improves insulin sensitivity and, through that, the hormonal picture. The reasoning is sound and is exactly why weight loss is a first-line PCOS intervention.
Mechanistic rationale
AOD-9604 is the lipolytic fragment of growth hormone, proposed to reduce fat without the hyperglycaemia of full-length GH — attractive in a population where insulin resistance is the problem. The difficulty is upstream of PCOS entirely: the human trial designed to test whether it produces fat loss found it did not beat placebo. A fat-loss mechanism that did not produce fat loss in people cannot mitigate insulin resistance through fat loss.
Community reports
Reports of mild fat loss exist and are confounded by concurrent diet and training. Non-response is unusually prominent for this compound — which is what agreement between anecdote and trial data looks like.
Components carrying the argument: Lipolysis without GH glycaemic effects
Pathophysiology
Most compounds in this library sit in the gap between promising pre-clinical data and no human test. This one was tested in a placebo-controlled trial and did not work for its intended purpose.
Mechanistic rationale
That makes it more informative than the silence around most of the library, not less. A negative trial is evidence. Building a PCOS rationale on top of an effect that a trial looked for and did not find requires explaining why it would appear in this population when it did not in the one studied — and no such explanation has been offered.
Community reports
The GRAS affirmation frequently cited in marketing is a food-and-cosmetic ingredient safety designation. It is not an efficacy finding and not a drug approval.
Components carrying the argument: Applies to any AOD-9604 fat-loss claim
| Question | Position |
|---|---|
| Studied in PCOS? | No |
| Is the visceral-fat premise sound? | Yes — adiposity drives insulin resistance and androgens |
| Does AOD-9604 produce fat loss in humans? | Not in the phase 2b trial — no separation from placebo |
| Development status | Discontinued for obesity |
| Does IGF-1 rise? | No — the design objective was met |
| Evidenced route to the same goal | Weight loss by any means that works |
PCOS has treatments that work. Metformin is first-line for the metabolic phenotype, inexpensive, with decades of data. Combined hormonal contraceptives address hyperandrogenism and cycle regulation; letrozole is first-line for ovulation induction; inositol has reasonable evidence and an excellent safety profile; GLP-1 receptor agonists now have randomised evidence in this population. Weight loss of 5–10% alone restores ovulation in a meaningful proportion.
Against that, a compound with no human data is not an alternative — it is a substitution of something unmeasured for something measured. The comparison worth making is not “peptide versus nothing”.
Physical-chemistry reference for laboratory handling of the lyophilised material. Reproduced so that stability and concentration mathematics can be studied — see the calculator for the arithmetic and the reconstitution guide for sterile technique.
Supplied lyophilised, commonly 10 mg per vial. Reconstituted with bacteriostatic water. At 300 mcg daily a 10 mg vial is about a month.
Lyophilised: refrigerated or frozen. Reconstituted: 2–8 °C.
Commonly 10 mg lyophilised vials.
A 16-residue fragment with ordinary peptide stability characteristics.
Regulatory status changes and differs by country. This is a summary for orientation, not legal advice — check your own jurisdiction.
Not approved as a drug anywhere. Clinical development for obesity was discontinued after the phase 2b programme. It holds a GRAS affirmation in the United States for certain ingredient uses, which is a food-and-cosmetic safety designation rather than a therapeutic approval — a distinction frequently blurred in marketing that cites "FDA GRAS status" as though it were an endorsement.
On 4 December 2024 the FDA's Pharmacy Compounding Advisory Committee voted 0–12 against including AOD-9604 on the Section 503A Bulks List. The committee cited a lack of evidence for clinical effectiveness and safety in the uses under review.
This was an earlier and separate round from the July 2026 meeting, and it is the part of the story that the coverage of that meeting almost entirely omits. Across two sittings in October and December 2024, PCAC considered seven substances — including ipamorelin, kisspeptin-10, AOD-9604, CJC-1295 and thymosin alpha-1 — and rejected every one of them. In July 2026 the same committee recommended six of the seven it looked at. What changed between the two rounds was the FDA's April 2026 restructuring of the Category 2 list and the set of substances nominated, not the arrival of new efficacy trials for anything on this page.
A negative advisory vote is not a ban and is not binding. It does mean there is no current 503A compounding pathway for this substance, and that a committee reviewing the evidence was not persuaded by it.
Regulatory and trial claims re-checked against primary sources on 16 August 2026. Checked: PCAC rejected it 0-12 on 4 December 2024; no 503A pathway. Approvals, trial readouts and compounding decisions move faster than anything else on this page — a date here means someone looked, not that nothing has changed since.
AOD-9604 is a growth hormone fragment and falls within the scope of WADA S2 (peptide hormones, growth factors and mimetics). Treat as prohibited.
Links resolve to PubMed records, trial registries, publisher pages or primary documents. Where a body of work rather than a single paper is cited, the link opens a PubMed query so the full result set — including newer papers than this page — can be reviewed. Verify every claim against the primary source before relying on it.
For educational and research reference only. Nothing on this site is medical advice, a recommendation, or an instruction to administer any substance to a human being. Every figure on this page is a record of what has been reported, not a recommendation. Full disclaimer.