AOD-9604 and adiponectin: the fat-loss mechanism that doesn't involve growth hormone

Most peptides on the fat-loss menu route through the growth-hormone axis. Sermorelin, tesamorelin, CJC-1295, ipamorelin — all four are growth-hormone secretagogues, releasing more endogenous GH on the assumption that elevated IGF-1 then drives body-composition changes. The trade-off baked into that route is the IGF-1 elevation itself: long-term high IGF-1 levels track with concerning health signals, from insulin resistance to certain cancer risks.

AOD-9604 was designed to bypass that trade-off. The compound is the C-terminal lipolytic fragment of human growth hormone, with the growth-promoting region clipped off. The conceit is fat loss without the GH-axis exposure. Whether that conceit holds in humans is a question the published trial record has only partly answered.

The endocrine half of adipose tissue

Fat is not just storage. It is the body’s largest endocrine organ, secreting a family of hormones called adipokines that regulate insulin sensitivity, inflammation, appetite, and energy expenditure. Adiponectin is the most-studied of them — its plasma level falls as body fat rises, and lower adiponectin tracks with insulin resistance, metabolic syndrome, and cardiovascular risk. Leptin signals satiety from filled fat depots. TNF-α and IL-6 link adipose tissue to systemic inflammation. The full list is longer.

Any fat-loss intervention worth taking seriously has to engage that endocrine network honestly. Growth hormone does, by binding receptors that include adipocytes among their many targets and shifting body composition through a constellation of downstream effects — some lipolytic, some growth-promoting, some both. AOD-9604’s design idea was that the lipolytic half could be separated from the growth-promoting half if you used the right structural fragment.

What the structural conceit actually is

Human growth hormone is a 191-residue protein. The C-terminal region — roughly the last 15-20 amino acids — has long been associated with the lipolytic and antilipogenic activities of the full hormone in animal models, distinct from the growth-promoting region elsewhere on the molecule. AOD-9604 takes that lipolytic region (specifically hGH residues 177-191) and adds an N-terminal tyrosine for synthetic stability, giving a 16-amino-acid fragment.

Heffernan and colleagues’ 2001 paper in International Journal of Obesity established the conceit. In obese mice, AOD-9604 reduced body weight gain through increased fat oxidation and reduced lipogenesis, without the IGF-1-mediated growth-promoting signal of full hGH. The lipolytic activity retained; the growth-promoting activity dropped. Mechanistically, that is a clean story — the right region of the right hormone, isolated and stabilised, doing one of its parent’s jobs without the rest.

The reason the cell-level story matters: it would make AOD-9604 the only peptide on the fat-loss menu that arguably acts on the adipocyte directly rather than upstream through GH/IGF-1. That positions it differently in conversations about adipose endocrinology — including whether it might modulate adipokine output (adiponectin among them) in ways the GH secretagogues do not. The published evidence on AOD-9604’s specific effects on circulating adiponectin in humans is, again, thin.

The Phase IIb that did not work

By the early 2000s the Australian sponsor Metabolic Pharmaceuticals was running AOD-9604 through the clinical pipeline for obesity. The 2004 Wilding pipeline update in Current Opinion in Investigational Drugs summarises the Phase IIa work in obese adults — short-term weight reduction, the kind of early-phase signal that justifies moving into a larger pivotal trial.

The pivotal Phase IIb trial that followed did not meet its primary efficacy endpoint. The peptide showed numerical weight reduction at some doses but not the statistically significant separation from placebo that would have justified Phase III investment. Metabolic Pharmaceuticals discontinued the obesity development program. The molecule has not been registered for a Phase III trial in any indication in the twenty years since.

That is the load-bearing fact about AOD-9604’s clinical file. The mouse-model mechanism story (Heffernan 2001) and the Phase IIa early signal (Wilding 2004) are real findings, but the trial that asked whether they translated to a clinically meaningful effect at scale answered with a null primary endpoint. The development arc — early signal, null pivotal, sponsor walks away — is the same shape as ipamorelin’s, and the conclusion the regulatory system reached for both molecules was the same.

The 2015 Kwon & Park rabbit osteoarthritis model shows AOD-9604 has off-label-indication mechanism interest beyond obesity — weekly intra-articular dosing reduced cartilage damage scores in a 32-rabbit collagenase-induced OA model. That work is animal-only, in a different indication than the development program ran on, and the human translation of the result has not been registered as a controlled trial.

The GLP-1 era reopened the question

For roughly fifteen years after the Metabolic Pharmaceuticals discontinuation, AOD-9604 sat largely dormant in the consumer market. The GLP-1 weight-loss boom changed that. Aggressive semaglutide and tirzepatide-driven weight loss reliably produces meaningful muscle and lean-tissue loss alongside fat reduction — the published GLP-1 body-composition literature documents a non-trivial proportion of total weight lost as non-fat tissue, varying by protocol, dose, and the patient’s baseline body composition. The market for an adjunct that protects lean mass during weight loss is now structurally larger than it was when AOD-9604 was developed.

The 2026 Arora review in J Clin Med places AOD-9604 among the candidate adjuncts in that conversation — pharmacologic strategies for preserving lean body mass during GLP-1-driven weight loss. The review is a survey of the candidate landscape, not a trial. The case for AOD-9604 specifically as a lean-mass-sparing adjunct in the GLP-1 era is mechanistically plausible and trial-empty. The Phase IIb null result was on weight loss alone in obese adults, not on lean-mass preservation during GLP-1-driven loss in mixed populations.

Mendias and Awan’s 2026 Sports Medicine survey places AOD-9604 in the broader grey-market peptide-evidence landscape — non-approved peptides operating outside FDA-recognised drug-application frameworks with scarce controlled human safety and efficacy data.

The regulatory file is closed for now

AOD-9604 is not on either the July 23-24, 2026 FDA PCAC docket or the February 2027 second-meeting wave. The molecule was referred to PCAC in September 2024 after removal from the Section 503A Category 2 list, evaluated at the December 4, 2024 PCAC meeting, and the committee voted against inclusion on the 503A bulk drug substances list. FDA’s stated reasoning followed the standard pattern for the four peptides reviewed in October-December 2024 (ipamorelin, CJC-1295, AOD-9604, thymosin α-1): insufficient human safety data, mechanistic concerns about unintended endocrine effects, and lack of reproducible efficacy data outside small or open-label trials. We covered how PCAC reviews work in What is the PCAC.

The PCAC vote does not mean AOD-9604 cannot be compounded under state pharmacy authority; it does mean the molecule is not on the active path toward inclusion on the federal 503A bulk substances list that would normalise compounding under enforcement-discretion arrangements. The compounded-supply route exists but is narrower than for peptides under active PCAC review.

What would actually re-open the AOD-9604 file?

For the obesity indication, a Phase IIb trial run on modern endpoints — body composition by DEXA, visceral adipose tissue by CT or MRI, not just weight on a scale. The original 2000s trials read mostly on weight loss; the published null primary endpoint may have missed a body-composition effect that better imaging would have detected. A modern trial would also need a longer duration than the original program’s typical 12-week window.

For the GLP-1-adjunct lean-mass-sparing indication, a controlled trial in adults on stable GLP-1 therapy randomised to AOD-9604 or placebo, with DEXA-measured lean-mass change as the primary endpoint over six to twelve months. The market case is clear; the trial evidence supporting AOD-9604 specifically for that case does not yet exist.

A published human pharmacokinetic study would help. Twenty-plus years after the Phase IIa program, the PK profile of AOD-9604 in humans — by injection, by route, dose-response, half-life — is still not characterised in the published literature in the way an approved drug’s would be.

Independent replication of the Phase IIa positive signal outside the original sponsor’s program would establish whether the early result was a real effect that the larger trial happened to miss, or a phase-noise positive that the pivotal trial correctly identified as null.

None of those trials is currently scheduled. The GLP-1 era has opened a new market case for the molecule without producing the controlled human evidence to support it.

Wolverine Health is being built around the version where a supervising physician has the actual trial readouts in front of them before recommending AOD-9604, the dispensing pharmacy is US-licensed, and the prescribing decision is anchored in a controlled trial in the relevant population — not the mechanism plausibility a 2001 mouse paper established. Join the waitlist for a note the day any of the trials above actually reads out.