AOD-9604: Research, Mechanism, and What the Evidence Actually Shows

What it is

Human growth hormone is a long protein with many jobs: it grows tissue, raises blood sugar, enlarges organs at high doses, and breaks down fat. Researchers at Monash University in Australia noticed that the fat-burning part seemed to live in one short stretch near the end of the molecule, amino acids 176 to 191. They cut that stretch out, added one amino acid to stabilize it, and called it AOD-9604: Anti-Obesity Drug 9604.

The idea was to keep the fat signal and throw away everything else. A company called Metabolic Pharmaceuticals took it into human trials in the 2000s, and that is where the story turns.

What it is studied for

Fat loss, first and mainly. More recently, cartilage repair, based on lab and animal work. Along the way it collected an unusually complete safety file, which is what earned it the FDA’s “Generally Recognized as Safe” status as a food ingredient in 2010. That designation says the compound appears safe to eat. It says nothing about whether it works.

What the evidence actually shows

In humans

The trial that matters was a 24-week, randomized, placebo-controlled study of oral AOD-9604 in roughly 300 obese adults at several doses. The company reported that it missed its primary endpoint: weight loss was not meaningfully better than placebo at most doses, with a small, statistically significant drop in fat mass in one dose group. That was not enough to keep funding a drug, and development stopped. The results were announced by the company rather than published in full; the paper that did come out of the program covers safety and tolerability across the human studies, and on that front the news was good (Stier et al., 2013).

That is the tension with this peptide. The largest human test found no useful weight loss. Vendors and forums lean on the mechanism and the mouse data and skip past that.

In animals

The mouse data is what everyone quotes, and it is striking. Obese mice given AOD-9604 lost about half their body fat in 19 days without eating less or losing muscle. Mice bred without a particular fat-cell receptor (beta-3 adrenergic) barely responded, which pinned down part of the mechanism (Heffernan et al., 2001). Obese rats lost fat too, and isolated human fat cells broke down fat faster in the dish.

Then the same molecule went into people and did very little. AOD-9604 is a textbook case of the gap between animals and humans; see why most peptide evidence is preclinical.

From user reports

People injecting it describe slow fat loss over a cycle, usually alongside diet and exercise that would produce fat loss on their own, and sometimes less joint pain. Some inject near the belly hoping to target that fat; the peptide enters the bloodstream regardless of where the needle goes, and no study supports spot reduction.

How it works, as far as anyone knows

AOD-9604 does not bind the growth hormone receptor and does not raise IGF-1, which is the point of the design (Ng & Bornstein, 2000). In fat cells it appears to act through beta-3 adrenergic receptors, the switches that tell a fat cell to release its stores; to raise the activity of hormone-sensitive lipase, the enzyme that does the actual cutting of stored fat; and to slow the creation of new fat. All of that comes from mice and cell cultures.

Because it skips the growth hormone receptor, it also skips the effects that make full growth hormone a problem: insulin resistance, water retention, carpal tunnel, organ growth. In every study that looked, blood sugar and insulin did not move.

The cartilage lead

In lab dishes and animals, AOD-9604 pushes cartilage cells to make more of the proteins cartilage is built from. Some Australian clinics inject it directly into arthritic joints on that basis. No controlled human trial of that use has been published.

How it is used in studies

The human trials used oral doses. Users inject it: 250 to 300 micrograms a day under the skin, in the morning on an empty stomach, for cycles of up to 12 weeks, on the theory that low insulin helps the fat signal. A 5 mg vial mixed with 2 mL of bacteriostatic water gives 300 micrograms in 0.12 mL, kept refrigerated and used within about four weeks. None of this dosing comes from a trial. See the storage and reconstitution guide.

Side effects and unknowns

The safety story is the best part. In trials, side effects were about the same as placebo: some injection-site redness, an occasional headache, rare mild nausea. The toxicology behind the GRAS status found no cancer signal, no genetic damage, no effect on reproduction, and no effect on blood sugar.

Two caveats. The GRAS finding covers eating it as a food ingredient, not injecting it for months; nobody has studied long-term injection. And the cartilage use is a new route with no safety data at all. See the peptide safety guide.

How it compares

• Semaglutide and tirzepatide: 15 to 21% body-weight loss in trials of thousands. Not in the same category.
• Tesamorelin: FDA approved, about 18% less visceral fat in Phase III, with growth hormone side effects to go with it.
• Ipamorelin and CJC-1295: raise growth hormone itself, with broader effects and the blood-sugar and IGF-1 concerns AOD-9604 avoids.
• AOD-9604: the lowest risk in this group and the weakest human result.

The fat loss guide puts them side by side.

Legal status

In the United States, AOD-9604 is GRAS as a food ingredient and not approved as a drug for anything. The FDA has moved to restrict it from compounding. It is sold as a research chemical. WADA bans it as a growth hormone fragment, in and out of competition.

Bottom line

AOD-9604 is a clever design with a clean safety record and a human trial that did not work. Mice lost half their fat; people lost roughly nothing. If someone recommends it, the honest pitch is “very unlikely to hurt you,” not “will make you lean.”

References

1. Stier H, et al. (2013). Safety and tolerability of the hexadecapeptide AOD9604 in humans. J Endocrinol Invest. PubMed
2. Heffernan MA, et al. (2001). The effects of human GH and its lipolytic fragment (AOD9604) on lipid metabolism following chronic treatment in obese mice and beta(3)-AR knock-out mice. Endocrinology. PubMed
3. Ng FM, Bornstein J. (2000). Hyperglycemic action of synthetic C-terminal fragment of human growth hormone. Am J Physiol. PubMed

See also

• Peptide
  CJC-1295: What the Evidence Says About This GHRH Analog

  CJC-1295 is a longer-lasting version of GHRH, the signal that tells the pituitary to release growth hormone. The DAC form raised GH and IGF-1 for days in one small trial; the no-DAC form (Mod GRF 1-29) has no human trial of its own.

• Peptide
  Ipamorelin: What the Evidence Says About This Selective GH Secretagogue

  Ipamorelin is a five-amino-acid peptide that mimics ghrelin to release a pulse of growth hormone, and unlike its predecessors does it without raising cortisol, prolactin, or hunger. Human data: one dosing study and one abandoned surgery trial.

• Peptide
  Semaglutide: What the Clinical Evidence Shows About This GLP-1 Agonist

  Semaglutide is a lab-made, week-long version of the gut hormone GLP-1, sold as Ozempic, Wegovy, and Rybelsus. It has the largest trials of any peptide on this site: about 15% weight loss over 68 weeks and 20% fewer major heart events.