AOD-9604: What the Research Actually Shows

Search AOD-9604 and you will find two versions of the same compound. One is a fat-loss peptide with impressive animal data. The other is a discontinued obesity drug that missed its endpoint in a 500-person trial. Both are true, and the gap between them is the most useful thing to understand about it.

What AOD-9604 actually is

AOD-9604 is a synthetic copy of the tail end of human growth hormone. Specifically it is residues 177 to 191 of hGH, with one extra tyrosine stuck on the front to make synthesis practical. That gives a 16-residue peptide of roughly 1.8 kilodaltons.

The design logic was surgical. Growth hormone does several unrelated things: it drives longitudinal growth and IGF-1 production through the growth hormone receptor, and separately it promotes fat breakdown and suppresses fat storage. Metabolic Pharmaceuticals, an Australian company, asked whether the lipolytic function could be separated from the growth function by isolating the region responsible for it. AOD-9604 was the answer to that question.

In vitro the separation held up. The fragment showed no growth hormone receptor binding and no receptor-mediated proliferative activity. It did not raise blood glucose the way intact growth hormone does.

Three names, one peptide

This trips people up constantly, so it is worth clearing up. These all refer to the same molecule:

• AOD-9604 — the development code, where AOD stands for anti-obesity drug
• hGH fragment 176-191 — the most common vendor label
• hGH(177-191) — the technically accurate sequence description

The 176 versus 177 discrepancy is the added tyrosine. It sits at the N-terminal position, so counting it pushes the numbering back by one. If a supplier lists both as separate products, that is a red flag about the supplier, not a difference in the peptide.

The rodent data that built the reputation

The foundational study is Ng and colleagues, published in Hormone Research in 2000, using obese Zucker rats. The design was straightforward: 500 micrograms per kilogram, given orally, daily, for 19 days.

Body weight gain in the treated animals was 15.8 grams against 35.6 grams in controls — a reduction of more than half. Adipose tissue taken from treated animals showed increased lipolytic activity, which is the mechanism doing the work rather than an appetite effect.

The result that mattered most was a negative one. Measured by euglycaemic clamp, chronic treatment produced no adverse effect on insulin sensitivity. Chronic intact growth hormone in the same model does impair it. That contrast is the entire reason the compound was interesting: the metabolic upside without the metabolic cost.

Follow-up work in obese mice pointed the same direction from subcutaneous dosing — increased fat oxidation, increased markers of lipolysis, attenuated weight gain, and no hyperglycaemia.

What AOD-9604 did in humans

It went into a 24-week study that enrolled 534 participants, 502 of whom were randomised, testing it as an oral anti-obesity drug. It failed to demonstrate a significant benefit on the primary weight-loss endpoint. Development for obesity stopped there.

There is a second caveat that deserves equal weight. The conclusions about AOD-9604 in humans rest primarily on sponsor-reported phase 2 programme outcomes rather than peer-reviewed randomised efficacy trials. So the honest summary is not “it was proven not to work” but something slightly weaker and more uncomfortable: the trial that was run did not show benefit, and the full data were never published to a standard that allows independent scrutiny.

What follows from that is simple. There are no robust human data supporting meaningful visceral fat reduction or broader metabolic benefit. Claims about body recomposition or athletic performance are extrapolation, not evidence.

Why the rat and human results diverged

This pattern is common enough in metabolic research that it is worth understanding rather than treating as a scandal. Several things differ between the two settings:

• The model. Obese Zucker rats carry a leptin receptor mutation. Their obesity has a specific genetic driver that human obesity mostly does not.
• The endpoint. The rodent work measured lipolytic activity in excised tissue and weight gain in caged animals on fixed diets. The human study measured weight change in free-living adults over six months.
• The dose and route. 500 micrograms per kilogram daily is a large dose, scaled to a small animal, given orally.
• Compensation. Humans who mobilise slightly more fat can eat slightly more without noticing. Caged rodents on measured feed cannot.

A compound can genuinely increase lipolysis and still produce no weight change, because weight is an energy balance outcome and lipolysis is one input into it.

AOD-9604 and the joint repair claims

After the obesity programme ended, AOD-9604 was repositioned toward osteoarthritis and cartilage. You will see this cited as if it were established. It is not.

The evidence there is animal work — principally an intra-articular rabbit osteoarthritis model, tested with and without hyaluronic acid, showing improved cartilage-related outcomes. That is a legitimate preclinical finding and it is also the whole of it. There is no convincing human clinical trial evidence for joint repair or tissue regeneration.

AOD-9604 regulatory status

AOD-9604 is not approved by any major regulatory agency for any indication. It appears on the FDA Category 2 Bulk Drug Substances list, the category for substances that present significant safety risks in compounding. It is not a dietary supplement ingredient and it is not a medicine.

Material sold under this name, including the 10mg vial we stock, is a laboratory reference material. It is not for human or veterinary use.

Handling AOD-9604 in the lab

Supplied lyophilized and vacuum sealed. Store the sealed vial at minus 20 degrees Celsius away from light. The dry cake tolerates transit temperatures comfortably — lyophilized peptides are stable for weeks at ambient, which is why shipping without cold chain is normal and not a quality signal either way.

Reconstitute with bacteriostatic water, running the stream down the glass rather than onto the cake, and swirl instead of shaking. Our reconstitution calculator handles the volume arithmetic for any target concentration. If you want to check what your certificate of analysis is actually telling you, we wrote a guide to reading a peptide COA.

Frequently asked questions

Is AOD-9604 the same as HGH fragment 176-191?
Yes. Same peptide, different label. The numbering difference is an added N-terminal tyrosine.

Does AOD-9604 raise IGF-1?
In the published models it did not. It showed no growth hormone receptor binding, which is the pathway through which intact growth hormone drives IGF-1.

Is AOD-9604 an approved drug anywhere?
No. It failed its obesity programme and holds no approval in any major jurisdiction.

Why is the animal data so much stronger than the human data?
Different model, different endpoint, different timescale, and humans compensate behaviourally in ways caged animals on measured feed cannot.

What purity should a vial show?
Look for HPLC purity stated against a named method and a mass spectrometry identity result matching the expected mass. A certificate without a chromatogram is a claim, not a measurement.

References

Ng FM, Sun J, Sharma L, Libinaka R, Jiang WJ, Gianello R. Metabolic studies of a synthetic lipolytic domain (AOD9604) of human growth hormone. Horm Res. 2000. PMID 11146367

Increase of fat oxidation and weight loss in obese mice caused by chronic treatment with human growth hormone or a modified C-terminal fragment. PMID 11673763

The emerging landscape of performance-enhancing peptides modulating the GH-IGF1 axis: bridging the gap between clinical evidence and patient self-administration. PMC13322892