AOD-9604: what the molecule is, what the animal studies and the six human trials measured, and what its record does and does not include

What AOD-9604 is: a modified C-terminal fragment of human growth hormone#

Human growth hormone (hGH) is a 191-amino-acid protein. AOD-9604 is a synthetic peptide of 16 amino acids copied from its last stretch, residues 177 to 191, with one change: a tyrosine at the N-terminal end, in the position occupied by phenylalanine-176 in the native hormone. The developer's 2013 safety paper writes the molecule as Tyr-hGH177-191, gives the sequence YLRIVQCRSVEGSCGF, and says the added tyrosine stabilizes the peptide ^[5]. Because it spans positions 176 to 191 of the parent protein it is often called "hGH fragment 176-191" or a modified form of it; the WADA Prohibited List names AOD-9604 and hGH 176-191 side by side ^[13]. The name is a development code from Metabolic Pharmaceuticals; the compound is also written AOD9604 or AOD 9604.

Identifiers for the free peptide, as recorded on the AOD-9604 product page from PubChem: molecular formula C78H123N23O23S2, molecular weight about 1,815.1 g/mol, CAS registry number 221231-10-3, PubChem CID 71300630. The two cysteines in the sequence correspond to those that form the C-terminal disulfide loop of native hGH; the papers opened for this article do not state how they are connected in the synthetic peptide, so a COA's mass spectrometry result is the practical check (see the final section).

The research idea came from Frank Ng's group in the Department of Biochemistry and Molecular Biology at Monash University in Melbourne, which studied fragments of the hGH C-terminus as a "lipolytic domain" of the hormone ^[1][2][3]. Commercial development was carried by Metabolic Pharmaceuticals Pty Ltd of Port Melbourne, later a subsidiary of the ASX-listed Calzada Limited, which funded all six human trials ^[5][8].

How AOD-9604 differs from the other GH-axis peptides#

Most peptides discussed alongside AOD-9604 act upstream of growth hormone. Sermorelin and tesamorelin are analogues of growth hormone-releasing hormone, and ipamorelin is a growth hormone secretagogue; each was studied for its ability to make the pituitary release GH, which raises IGF-1. AOD-9604 is a piece of growth hormone itself, and the Monash hypothesis was that this piece carried the hormone's effect on fat metabolism without its effects on growth and glucose. In the 2001 International Journal of Obesity paper, AOD-9604 did not compete with hGH for binding to the hGH receptor on BaF-BO3 cells and did not stimulate their proliferation, whereas hGH did ^[2], and the human safety paper reported no change in serum IGF-1 ^[5].

The animal studies and what each measured#

The preclinical literature is small and comes almost entirely from the Monash group between 2000 and 2001, with one later rabbit study funded by the developer. An earlier Monash paper on the unmodified fragment AOD-9401, given orally to ob/ob mice for 30 days, reported less weight gain with no change in food intake and higher adipose lipolysis ^[4]. The table names each model and what was measured; our guide to reading published peptide research explains why the model matters.

Read together, the Monash papers describe a consistent pattern in obese rodents: less weight gain, more adipose lipolysis and fat oxidation, and, unlike intact growth hormone, no worsening of glucose handling or insulin sensitivity ^[1][2][3]. Chronic effects disappeared in mice lacking the beta3-adrenergic receptor, yet acute AOD-9604 still raised fat oxidation in those animals, so the authors concluded the action is not mediated directly through beta3-AR even though both compounds raise its expression ^[3]. Two cautions: the authors were the compound's inventors, with little independent replication since, and ob/ob mice and Zucker rats are genetic models of obesity.

After the obesity program ended, the developer turned to cartilage. Calzada announced the rabbit study to the ASX in February 2013 as its sponsor, alongside in-vitro cartilage-cell work in Toronto and patent filings ^[10][11]. It is one animal study, in a small journal, funded by the patent holder; no human cartilage trial has been published.

The six human trials and the 2007 termination#

Between 2001 and 2007 Metabolic Pharmaceuticals ran six randomized, double-blind, placebo-controlled trials, METAOD001 to METAOD006, in roughly 900 adults; the company's later statements say 925 ^[5][8]. The first was a phase 1 study in 15 healthy men; METAOD002 to METAOD004 were small phase 2a studies in obese men, the first intravenous and the later two oral. The two phase 2b obesity trials were oral: METAOD005 randomized 300 obese adults to five groups and placebo for 12 weeks, and METAOD006 randomized 502 to three groups and placebo for 24 weeks ^[5].

The only peer-reviewed publication from the program is the 2013 safety and tolerability paper by Stier, Vos and Kenley ^[5]. It pools the safety data from all six trials and reports no effect on serum IGF-1 relative to placebo, no change in oral glucose tolerance, and no anti-AOD-9604 antibodies in the participants assayed. Serious adverse events were distributed similarly across groups and judged unrelated, although the full text records one diarrhea event rated "possibly related". The disclosures matter: Metabolic funded the trials, one author held an interest in Calzada, one was its former medical director, and the third advised it on European novel-food applications ^[5]. The paper reports no body-weight outcome from either phase 2b trial, although both were designed to measure it.

What happened to those outcomes has to be pieced together from company communications, not journals. In December 2004 Metabolic issued a press release on the 12-week trial, reporting a difference in body weight between the lowest-exposure group and placebo that the company described favourably ^[6]; those figures have never appeared in a peer-reviewed paper and are not repeated here. The 24-week trial read out in early 2007. The audited 2007 annual report of Circadian Technologies, Metabolic's major shareholder, records that on 21 February 2007 Metabolic announced that the phase 2b data "did not support further clinical development" and that "the program had been terminated" ^[7]. A 2013 Fairfax Media report, citing the 2007 stock-exchange announcement, described the 24-week change as small and quoted Calzada: "AOD-9604 was not successful in human trials aimed at obesity" ^[15].

GRAS: what the developer announced, and what GRAS does and does not mean#

In June 2012 Calzada told the ASX that an expert scientific panel had reached a conditional GRAS determination for AOD-9604 as a food ingredient in the United States, conditional on the unpublished safety studies appearing in a peer-reviewed journal; its chief executive called the status "self-affirmed" ^[8]. The 2013 safety paper satisfied that condition and states that a GRAS panel judged the peptide "Generally Recognized As Safe" under its intended uses in foods ^[5]. The company repeated the self-affirmed wording in February 2013 ^[10] and in August 2014 announced the designation as final, with a view to licensing the compound to food and supplement companies ^[8].

GRAS is a food-law concept. A substance added to food needs FDA premarket approval unless it is "generally recognized, among qualified experts, to be safe under the conditions of its intended use" ^[9]. A company can reach that conclusion itself through a panel of qualified experts. Telling FDA is optional: "any person" may submit a GRAS notice, and FDA responds that it does not question the basis for the conclusion, that the notice lacks a sufficient basis, or, at the notifier's request, that it has stopped evaluating ^[9]. A self-affirmed determination that is never submitted does not appear in FDA's GRAS Notice Inventory and has not been reviewed by the agency. Some vendor pages cite an FDA GRAS notice number for AOD-9604; we could not open an inventory record under that number, so this page does not state that FDA received or responded to a notice. The primary sources support only the company's own expert-panel determination for a defined food use ^[8].

• GRAS does not mean approved. It is not a drug approval and not a finding of effectiveness.
• GRAS is tied to the intended use: a food-ingredient use at a stated intake level ^[8], not other routes, amounts or purposes.
• GRAS does not cover research-grade material; a food-grade ingredient and a lyophilized research peptide are different products made under different controls.

Regulatory position in the United States#

AOD-9604 was nominated for FDA's list of bulk drug substances that may be used in compounding under section 503A, and FDA placed it in category 2 of its interim policy, for nominated substances that may present significant safety risks. The nomination was later withdrawn, so on FDA's page, current as of April 22, 2026, AOD-9604 sits in the table of substances "nominated but withdrawn" rather than in the active category 2 list ^[12]. FDA's stated concerns are a significant risk of immunogenicity for certain routes of administration, complexities with peptide-related impurities and API characterization, no or only limited safety information, and serious adverse events that may be associated with the compound, with causality unclear ^[12]. There is no approved drug and no FDA determination of safety or effectiveness for any purpose.

Anti-doping status#

The 2026 WADA Prohibited List, in effect from January 1, 2026, names AOD-9604 in section S2.2.3, "Growth hormone (GH), its analogues and fragments", with hGH 176-191 beside it. Section S2 is prohibited at all times, in and out of competition ^[13]. The listing is not new: in 2014 anti-doping chemists published a validated urine detection method for AOD-9604 with a limit of 50 pg/mL and noted that the peptide had been found in confiscated vials in the United States ^[14]. The compound drew public attention in Australia in 2013 when a professional football club's supplements program was investigated over its use ^[15].

What has not been studied#

• Any human efficacy outcome in a peer-reviewed journal. The two phase 2b trials measured body weight; neither dataset was published, and the sponsor described the 24-week trial as not supporting further development ^[5][7].
• Independent human trials. All six were funded by the developer, and the only publication has sponsor-affiliated authors ^[5].
• Anything beyond 24 weeks, the longest exposure in the record ^[5].
• Any route in people other than oral or intravenous ^[5]; FDA's immunogenicity concern points to other routes ^[12].
• Cartilage or joint outcomes in people; the evidence is one sponsor-funded rabbit study and company-announced cell work ^[10][11].
• Replication of the rodent mechanism outside the originating laboratory ^[2][3].
• Stability of the research-grade material; our storage guide covers the general principles.

What an AOD-9604 Certificate of Analysis has to show#

Renuvia's AOD-9604 lot has not yet been reported by Freedom Diagnostics Testing; the COA library shows it as pending, and no purity or content figure is stated here until a certificate is published. When one is issued, from Renuvia or any supplier, check these points:

1. Identity by mass spectrometry. The observed mass should match the 16-residue tyrosine-modified fragment, about 1,815 Da for the free peptide, not an unmodified 176-191 or 177-191 fragment, which would differ by the tyrosine-for-phenylalanine change (16 Da) or by a whole residue. A report that distinguishes the disulfide form from the reduced form, a 2 Da difference, tells you more than one that does not.
2. Purity by HPLC, with the method stated. A short disulfide peptide can show adjacent peaks from the reduced form, oxidation and deletion sequences; the report should quantify them. Our guide to purity report misreadings explains the common traps.
3. Measured content in milligrams, separate from purity. Purity says how clean the material is, not how much is in the vial.
4. A lot number that matches the vial, and a named laboratory with a report number you can check; our guide to verifying that a COA is genuine walks through that.
5. Supplier-provided certificates labelled as such, as they are in the Renuvia library.

The guides to reading a Certificate of Analysis and to HPLC and mass spectrometry cover each section of a report in more detail.