Results and the trial record

AOD-9604 before and after: what the trials measured

By the Decadewise team Education only Last updated 2 August 2026 13 cited sources

Short answer

The before and after that exists is a line chart: the final trial plotted average weight change for placebo and three amounts of AOD-9604 across 24 weeks, and no group separated from placebo by a full kilogram at either measured endpoint.

The number everyone quotes: in an earlier 12-week trial the 1mg group changed -2.8kg against -0.8kg on placebo. On that trial's own pre-specified primary analysis the comparison scored p=0.10, against the 0.05 it needed.

The trial built to confirm it: 536 subjects across 16 sites, a dietitian-supervised diet and exercise plan, and amounts at and below 1mg for 24 weeks.

Its own result line reads "Not met": weight loss against placebo came in under 1kg in every group, in a trial powered to find 1.8kg.

What followed: the sponsor ended development of the drug for obesity on 21 February 2007.

What no trial has tested: the subcutaneous injection sold today. Every human exposure on record went into a vein or was swallowed.

On this page

Short answer The six trials The 12-week result The 24-week trial Why the two disagree What the photos are of Fat, not just weight Where it sits with regulators What is not settled The safety picture How this is sourced Related pages The disclaimer

The published human record

What was measured, and in how many people?

The whole weight-loss program sits in one paper: Stier, Vos and Kenley (Journal of Endocrinology and Metabolism, 2013) summarize six randomized, double-blind, placebo-controlled trials of AOD-9604, labeled METAOD001 through METAOD006. Two of the three authors are listed with the sponsor, Metabolic Pharmaceuticals Pty Ltd. Three later human trials exist under the molecule's second code, LAT8881, and none of them is about weight, which is why they sit in the sourcing section at the foot of this page rather than here.

Adding the six enrollment figures it prints gives 925 people: 15, 23, 17, 36, 300 and 534. That total is our arithmetic on the paper's numbers, not a figure the paper states.

What that paper is for, and what it leaves out: safety. It walks through IGF-1, fasting glucose, insulin, oral glucose tolerance, antibodies and adverse events trial by trial. It prints no weight result for any of the six, including the two whose stated objective it describes as assessing efficacy.

The six trials in the 2013 safety review

Trial	Phase and design	People	Route and amounts	Length
METAOD001	phase 1, dose escalation	15 healthy adult males, BMI 24 to 30	intravenous, 25 to 400mcg per kg	single doses, 7-day washouts
METAOD002	phase 2a, 4 by 4 Latin square	23 obese males, BMI 35 or above	intravenous, 25, 50 and 100mcg per kg	single doses, 7-day washouts
METAOD003	phase 2a, 4 by 4 Latin square	17 obese males, BMI 35 or above	oral capsules, 9, 27 and 54mg	single doses, 2-week washouts
METAOD004	phase 2a, dose escalation	36 obese males, BMI 30 or above	oral capsules, 9, 27 or 54mg daily	7 days, plus 7-day follow-up
METAOD005	phase 2b, parallel group	300 obese adults, BMI 35 or above	oral capsules, 1, 5, 10, 20 or 30mg daily	12 weeks
METAOD006	phase 2b, parallel group, 16 centers	534 enrolled, 502 randomized, BMI 30 to 45	oral tablets, 0.25, 0.5 or 1mg daily	4-week run-in, 24 weeks, 4-week follow-up

Every row comes from the 2013 review. Note what is absent from the whole column: an injection under the skin. The first two trials used a vein and the other four used the mouth.

So the two efficacy readouts live somewhere else entirely: in the sponsor's own market announcements to the Australian Securities Exchange, dated 13 December 2004 and 21 February 2007. Both are retrievable, both print their statistics, and neither went through peer review.

A 2026 peer-reviewed review reaches the same conclusion about where the evidence sits: it states that conclusions regarding the lack of clinical efficacy in obesity rest primarily on sponsor-reported phase 2 outcomes rather than peer-reviewed randomized efficacy trials, and describes AOD-9604 as a tier B/C agent, one of the compounds with only limited or indirect human data (Dominikowski, 2026).

What did the 12-week trial show?

The headline the sponsor wrote, on 13 December 2004: AOD9604 Phase 2b Clinical Trial Successful. Under it: effectively induces weight loss, excellent tolerability.

What sat beneath that headline: 300 obese patients enrolled at five Australian sites, 283 starting treatment across six groups, once daily by mouth for 12 weeks, with everyone given the same general diet and exercise advice. Average starting weight was 122kg and average BMI 42.

The pre-specified primary analysis, printed in the announcement's own appendix: the 1mg group changed -2.8 plus or minus 0.7kg against -0.8 plus or minus 0.6kg on placebo, and the p-value for that comparison, by Dunnett's multiple comparisons procedure, was 0.10. The four higher amounts scored 0.84, 1.00, 0.73 and 0.44.

Weight change at week 12, primary analysis, from the sponsor's appendix

Group	Week 12 minus week 0	P against placebo
Placebo	-0.8 plus or minus 0.6kg	reference
1mg	-2.8 plus or minus 0.7kg	0.10
5mg	-1.6 plus or minus 0.6kg	0.84
10mg	-1.0 plus or minus 0.6kg	1.00
20mg	-1.7 plus or minus 0.6kg	0.73
30mg	-2.2 plus or minus 0.7kg	0.44

Adjusted means and standard errors from the general linear model, with the p-values the announcement prints beside them. No group cleared 0.05 on this analysis.

The word the appendix uses for the best group is "near significance": not significant. That is the sponsor's own description of its own primary comparison.

What was reported as significant instead: a secondary and confirmatory analysis, planned in advance, that fitted a rate of weight change per patient across every fortnightly measurement rather than comparing two timepoints. On that analysis the 1mg group returned p below 0.0001, the 30mg group 0.005 and the 20mg group 0.015.

The 2007 document confirms both halves of that from the outside: reviewing the earlier trial, it records that the response was bimodal for both genders, that the best group was 1mg, that it fell short of significance at p=0.1 where 0.05 was required, and that it reached significance in the female subgroup.

Two tests on one trial, and only the second one cleared its threshold. The company printed both, which is to its credit. A page that quotes 2.8kg and stops has taken the number and left the test behind.

The distribution behind that average is the interesting part: in the intention-to-treat set, 26 percent of the 1mg group lost 4kg or more against 8 percent on placebo, and 9 percent lost 8kg or more against none on placebo. In the higher groups the announcement says the effect was typically restricted to a few people, mainly men.

Which is where a genuine before and after would come from: a minority of participants inside one group, in a trial whose primary comparison did not separate from placebo. That is a real observation and it is not a result.

The confirmatory trial

What happened in the 24-week trial?

It had a name, the OPTIONS study, and an explicit job: to explore amounts at and below 1mg and to confirm the 2004 observation under conditions closer to a phase 3 trial, in line with the FDA guidelines the sponsor names.

How it was built: 536 subjects enrolled at 16 Australian sites, BMI 30 to 45, aged 18 to 65. A four-week single-blind placebo run-in started a dietitian-supervised diet and exercise plan, then randomization to placebo, 0.25mg, 0.5mg or 1mg once daily by mouth for 24 weeks, then four weeks of follow-up.

The primary endpoint, in the trial's own wording: statistically significant weight loss after 12 weeks of treatment for any one of the three amounts against placebo, plus safety and tolerability.

How large a difference it could see: the announcement states the trial was powered for an 80 percent chance of reaching significance if weight loss against placebo was 1.8kg.

The result, in the appendix, in two words: Not met.

The size of the miss: weight loss against placebo at both the 12-week primary and the 24-week secondary timepoint, after allowing for the diet and exercise plan, was less than 1kg in every group. Against a 1.8kg detection target, that is not a near miss.

And the secondary endpoints, which are the ones a before-and-after reader actually wants: waistline reduction over 24 weeks, body fat measured by whole-body DEXA scan, and changes in glucose control and lipid profiles. The appendix records the lot in three words: No relevant findings.

What the sponsor concluded from its own data: that AOD-9604 does not combine or synergize with successful use of an ongoing diet and exercise plan, but may show some effect with moderate weight loss effort in females.

That second clause is a subgroup and the document says so: females whose run-in weight loss was under 2kg, 149 subjects, a stratum defined before randomization but read after the primary endpoint had already failed. The announcement reports it as an examination of subgroups, not as a finding.

Then the program stopped. The announcement of 21 February 2007 states that the trial results do not support the commercial viability of the drug as a treatment for obesity, and that development of the drug for this condition is terminated.

One number in that document does not reconcile with itself, and it is worth saying out loud: the text says 536 subjects enrolled, the week-0 attendance table totals 502 across the four groups, and the weight chart above that table is labeled N=503. The 2013 review prints 534 enrolled and 502 randomized for the same trial. We report all of them rather than quietly choose one.

0.10

p-value for the best group in the 12-week trial's own primary analysis, against the 0.05 it needed

1.8kg

difference against placebo the 24-week trial was powered to detect

under 1kg

weight loss against placebo actually seen, in every group, at 12 and at 24 weeks

published human trials of subcutaneous AOD-9604, on the searches printed in the sourcing section

Metabolic Pharmaceuticals market announcements of 13 December 2004 and 21 February 2007, both retrieved from the Internet Archive on 2 August 2026; Stier, Vos and Kenley, J Endocrinol Metab 2013; our own PubMed, Europe PMC and ClinicalTrials.gov searches of 2 August 2026.

Why did the two trials disagree?

The first difference is what the placebo group was doing: in 2004 everyone got general diet and exercise advice and placebo lost 0.8kg over 12 weeks. In 2007 everyone got a dietitian-supervised plan for 28 weeks, and the announcement points at the high levels of weight loss seen in the placebo group as a candidate explanation, suggesting the drug effect may have been overwhelmed by an effective weight loss plan.

The second difference is which amounts were on the table: 2004 ran 1 to 30mg and found its best signal at the very bottom of that range, which is an unusual shape for a dose response. 2007 was designed to explore at and below that point, at 0.25, 0.5 and 1mg. If the 2004 low-end signal was noise, a trial built around it would find nothing, and nothing is what it found.

The third difference is which analysis was allowed to count: in 2004 the primary comparison missed and a secondary rate analysis cleared. In 2007 the primary endpoint was fixed in advance at 12 weeks, and it is that endpoint the appendix marks as not met.

What both trials agree on completely: tolerability. The 2007 announcement records that its preliminary analysis shows no evidence of any difference between placebo and any AOD-9604 group on safety or tolerability, and the 2013 review reaches the same conclusion across all six trials. A compound can be well tolerated and still not work, and this is what that looks like on paper.

What a cleared endpoint reads like, for contrast: the primary endpoint figures for the modern weight-loss drugs, and the papers behind them, are set out on our semaglutide guide. Reading the two pages next to each other is the quickest way to see the distance between a result and a near miss.

What are the before and after photos of?

Not the trials. We found no participant image and no body-composition photograph in either sponsor announcement or in the 2013 safety review, and the only visuals either announcement carries are plots of weight, lipids, glucose and IGF-1.

And not the same product either, which is the larger point. Every human exposure in the record went into a vein at 25 to 400mcg per kg as a 20-minute infusion, or was swallowed at 0.25 to 30mg a day. The 2026 review that tabulates both sides puts the trial figures next to what circulates online: 250 to 500mcg a day subcutaneously, split across two or three injections, on 8 to 12 week cycles (Dominikowski, 2026).

So the route moved and the amount moved with it. A milligram swallowed and a quarter of a milligram under the skin are different exposures of a peptide whose bioavailability differs by route, and the second of those has never been measured in a published human trial that we could retrieve.

Before that becomes a biology question it is an arithmetic one: 250mcg and 1mg are a factor of four apart and are written in two different units, which is exactly where the milligram and microgram confusion starts.

The other thing a photograph cannot show is what was in the vial. A trial tablet came with a manufacturing chain and release testing behind it. A research vial comes with a certificate whose limits are worth understanding before it is trusted, and reading a peptide certificate of analysis is a skill in itself. FDA names peptide-related impurities and active pharmaceutical ingredient characterization as open questions for this substance specifically.

There is a close structural parallel elsewhere in this library: a fragment sold under its own name while every human number attached to it belongs to the parent protein, which is the situation TB-500 sits in. AOD-9604 is the better-documented case, because here the fragment itself did reach people, twice, at scale, and the results were published by the company that paid for them.

Did anyone measure fat and not just weight?

Twice, and both readings are already on this page. Neither supports a body-recomposition claim.

2004, by computed tomography of the abdomen: abdominal fat area changed -35 plus or minus 15 square centimeters in the 1mg group against -23 plus or minus 15 on placebo, and abdominal circumference -1.3 plus or minus 0.8cm against +0.6 plus or minus 0.8cm. The announcement itself calls these methods inherently more variable than weight, and the standard errors it prints are as large as the differences.

2007, by whole-body DEXA: body fat reduction was a named secondary endpoint in a trial nearly twice the size, running twice as long, with the better instrument. The result recorded for the secondary endpoints is No relevant findings.

That is the honest ranking of the two measurements: the weaker method in the smaller trial produced a suggestive number with error bars swallowing it, and the stronger method in the larger trial produced nothing. When two measurements of the same thing disagree, the better-powered one is not automatically right, but it is the one that has to be explained away.

For contrast inside the same hormone axis: the same 2026 review puts tesamorelin at the top of its evidence tiers, supported by randomized controlled trials within an FDA-approved clinical indication. A different molecule with a different mechanism, and a picture of what this kind of evidence looks like when it exists.

Where does AOD-9604 sit with regulators?

In the United States, on a compounding list, not an approval list. FDA's page on bulk drug substances for use in compounding that may present significant safety risks carries AOD-9604 in the table headed "Bulk drug substances nominated but withdrawn". That page was current as of 22 April 2026 when we retrieved it.

What the agency's own entry says, in three parts: that compounded drugs containing AOD-9604 may pose significant risk for immunogenicity for certain routes of administration and may have complexities with regard to peptide-related impurities and active pharmaceutical ingredient characterization; that FDA has identified no, or only limited, safety-related information and therefore lacks sufficient information to know whether the drug would cause harm when given to humans; and that FDA has also identified serious adverse events that may be associated with AOD-9604, though causality is not clear.

That third clause deserves to be read slowly. It is not a finding of harm and it is not a clearance. It is a regulator saying reports exist and it cannot attribute them, which is a different and more uncomfortable statement than either.

There is a claim in the literature that points the other way, and it needs its label: the 2013 safety review states that AOD-9604 was determined by a qualified panel of experts to be Generally Recognized As Safe under conditions of intended use in foods. That is an industry self-determination process described by authors two of whom work for the sponsor. It is not an FDA approval, and it is not what the agency's current page says.

In sport it is named outright. The 2026 World Anti-Doping Agency Prohibited List, in force since 1 January 2026, files growth hormone fragments at S2.2.3 and gives the examples as AOD-9604 and hGH 176-191. Class S2 is marked prohibited at all times, in and out of competition.

Anti-doping is also where most of the recent literature comes from. A 2015 method paper validated a urine test with a limit of detection of 50 pg/mL, identified six potential metabolites, and noted that the peptide was available on several internet websites and had recently been identified in confiscated vials in the United States (Cox, 2015). Detection work is a large share of what has been published about this peptide since the trials ended, which is a pattern it shares with the other compounds sold beside growth hormone.

What the record does not settle

Five gaps, and every one of them sits between the trials and the product:

The injected route. On the six trials in the 2013 review, on the three later trials registered under the molecule's second code LAT8881, and on the database searches printed below, no human trial has given this compound under the skin. Everything human went into a vein or into the mouth.
Anything past 24 weeks. The longest human exposure in the record is the 24-week OPTIONS study, and its treatment period ended there.
A usable body-composition result. The whole-body DEXA endpoint returned nothing, and the 2004 computed tomography readings carry standard errors as large as the differences they describe.
Joints and tissue repair. The claims made for AOD-9604 outside weight rest on animal work, including intra-articular injection with or without hyaluronic acid in a rabbit osteoarthritis model (Kwon & Park, 2015). We found no human trial of that use.
What a research vial holds. FDA names impurities and ingredient characterization as open for this substance, and a vial bought online has no trial supply chain behind it.

A compound that was well tolerated and did not beat placebo is still a compound that did not beat placebo.

And one thing the record settles more clearly than most: this is not a case of thin evidence or a program that ran out of money before it could be tested. AOD-9604 was tested properly, twice, by the people who wanted it to work, and the second test was designed by them to be predictive of a phase 3 result. The sponsor said so in the same announcement that closed the program.

The safety picture, from the trials

What the trial record holds: six randomized, double-blind, placebo-controlled trials, no anti-AOD9604 antibodies detected in any patient assayed, no withdrawal or serious adverse event that occurred related to intake of AOD-9604, and no statistically significant effect on IGF-1, fasting glucose, insulin or oral glucose tolerance, which was the entire point of using a fragment rather than growth hormone (Stier, 2013). Adverse events did occur, and in quantity: 88.9 percent of subjects in the 12-week trial and 78.7 percent in the 24-week trial reported at least one, distributed similarly across placebo and every active group. Five patients in the 12-week trial reported a serious event: three in the 20mg group, a basal cell carcinoma, a moderate lipoma and a squamous cell carcinoma, one breast cancer in the 5mg group and one malignant melanoma in the 10mg group, none of which the investigator considered possibly, probably or definitely related to study medication, and 18 patients in the 24-week trial reported at least one serious event, spread evenly across groups. What none of that covers is the product people inject: a different route, a quarter to a half of the largest amount the final trial tested, from a supply chain with no release testing behind it, and an FDA entry stating the agency has seen serious adverse events it cannot attribute. A clean file on pharmaceutical-grade tablets is not a clean file on an unlicensed vial.

How this page is sourced

Which document is behind which number: the trial designs, the adverse-event figures and the metabolic readouts come from the 2013 peer-reviewed safety review; every weight, fat and statistical figure comes from the sponsor's two market announcements, read directly rather than through anyone's summary; the route and reported-use comparison and the evidence tiering come from a 2026 peer-reviewed narrative review; the animal work, the detection assay and the two regulatory positions come from the remaining references.

Thirteen sources: eight peer-reviewed papers with DOIs and PubMed IDs, two sponsor market announcements retrieved from the Internet Archive, two regulator documents, and one journal record with no full text available to us. The two announcements are labeled as sponsor announcements throughout, because that is what they are: they were not peer reviewed, and the company that wrote them had a commercial interest in both outcomes.

Why the announcements are cited at all, given that: because they are the only place the efficacy numbers exist. The 2026 review states the same thing in its own words, that conclusions here rest on sponsor-reported outcomes rather than peer-reviewed efficacy trials, and we would rather print that limitation than launder the figures through a secondary source. Both documents were retrieved on 2 August 2026 from the Internet Archive's capture of the sponsor's own website, HTTP 200, and both carry their original dates and appendices intact.

How the absence claims on this page were tested, because an absence claim is a claim about a whole literature: on 2 August 2026 we searched PubMed for AOD9604 across all fields, which returns 23 records, and applied PubMed's own Clinical Trial publication-type filter to that same query, which returns 0. Europe PMC, searched the same day for the quoted phrase "AOD9604", returns 24 records, and for "AOD-9604" returns 26. ClinicalTrials.gov, queried the same day on AOD9604 and on AOD-9604, returns 0 studies for each; the query hGH 176-191 returns one record, NCT03755791, which is a liver cancer trial matched on a fragment of the string. All six obesity trials were run in Australia by an Australian sponsor and none appears in that registry. We did not search the Australian registry: its search endpoint returned HTTP 403 to an automated request on the same date, which is a bot challenge rather than an empty result.

Why searching the compound's own name is not enough, which an earlier version of this page learned the hard way: the molecule carries a second code, LAT8881, the name a later developer registered its trials under. Two peer-reviewed papers state the identity outright, one calling it "LAT8881 (AOD9604)" and the other "LAT8881 (formerly identified as AOD9604)" alongside the sequence H-YLRIVQCRSVEGSCGF-OH (Huerta, 2023; Harpur, 2023). Searched on that code on 2 August 2026, ClinicalTrials.gov returns 3 completed trials, all sponsored by Lateral Pharma Pty Ltd and none of them about weight: NCT03865953, oral, phase 2, neuropathic pain, 53 participants, completed 3 May 2020; NCT04153409, oral, phase 2, acute migraine, 21 participants, completed 10 April 2020; and NCT05298306, intravenous, phase 1, lumbar radicular pain, 26 participants, completed 16 June 2023. A search on the name a compound used to have finds the trials it used to run, and nothing after.

What that does and does not establish: it is a statement about three databases, eight query strings and one date. It is not an assertion about every trial ever run on earth. What holds is narrow and checkable: on those queries, on that date, no clinical trial gave this molecule under the skin, under either of its names. Nine trials of it turn up in total, the six the original sponsor ran for weight and the three Lateral Pharma registered for pain, and every one of them went into a vein or into the mouth.

The standard: the full sourcing and citation-verification standard for this site, including who checks a page before it ships, lives on the methodology page.

Last reviewed: 2 August 2026.

1
Stier H, Vos E, Kenley D. Safety and tolerability of the hexadecapeptide AOD9604 in humans. J Endocrinol Metab. 2013;3(1-2):7-15. doi:10.4021/jem157w. Open access, full text retrieved 2 August 2026, HTTP 200. Not indexed in PubMed, so it carries no PMID. Authors Vos and Kenley are listed with Metabolic Pharmaceuticals Pty Ltd, the sponsor of all six trials. A 2026 review cites this paper under the DOI 10.4021/jem.v3i1-2.157, which does not resolve; the resolving DOI is the one printed here. jofem.org, article record.
2
Metabolic Pharmaceuticals Limited. AOD9604 Phase 2b Clinical Trial Successful. Market announcement, 13 December 2004, with a ten-page technical appendix carrying the weight, fat, lipid, glucose and IGF-1 tables quoted on this page. Sponsor announcement, not peer reviewed. Retrieved 2 August 2026 from the Internet Archive capture of the company website, HTTP 200. web.archive.org, 2004 announcement (PDF).
3
Metabolic Pharmaceuticals Limited. Metabolic's obesity drug: Phase 2B clinical trial results. Market announcement, Melbourne, 21 February 2007, five pages, including Appendix 1 on the OPTIONS study design and Appendix 2 on its results, where the primary endpoint outcome is recorded as "Not met" and the secondary endpoints as "No relevant findings". Sponsor announcement, not peer reviewed. Retrieved 2 August 2026 from the Internet Archive capture of the company website, HTTP 200. web.archive.org, 2007 OPTIONS announcement (PDF).
4
Dominikowski A, Rékoś Z, Olejarz M, Szczepanek-Parulska E, Domin R, Ruchała M. The emerging landscape of performance-enhancing peptides modulating GH-IGF1 axis: bridging the gap between clinical evidence and patient self-administration. Front Endocrinol (Lausanne). 2026;17:1822475. doi:10.3389/fendo.2026.1822475. PMID 42395176. Section 3.4.2 covers AOD-9604; the dosing table contrasts the trial routes with the amounts reported on bodybuilding forums. Full text read via PubMed Central, PMC13322892, retrieved 2 August 2026.
5
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;53(6):274-278. doi:10.1159/000053183. PMID 11146367. Despite its title, this is not a human study. The animals are obese Zucker rats, at an oral 500mcg per kg of body weight for 19 days, and the reported outcome is a reduction of over 50 percent in body weight gain against control, 15.8 grams against 35.6 grams.
6
Heffernan MA, Thorburn AW, Fam B, et al. Increase of fat oxidation and weight loss in obese mice caused by chronic treatment with human growth hormone or a modified C-terminal fragment. Int J Obes Relat Metab Disord. 2001;25(10):1442-1449. doi:10.1038/sj.ijo.0801740. PMID 11673763. Mice, not people.
7
Kwon DR, Park GY. Effect of intra-articular injection of AOD9604 with or without hyaluronic acid in rabbit osteoarthritis model. Ann Clin Lab Sci. 2015;45(4):426-432. PMID 26275694. No DOI is registered for this article; the PubMed record is the locator. Rabbits, not people.
8
Cox HD, Smeal SJ, Hughes CM, Cox JE, Eichner D. Detection and in vitro metabolism of AOD9604. Drug Test Anal. 2015;7(1):31-38. doi:10.1002/dta.1715. PMID 25208511. Source of the structural description used on this page, the C-terminal fragment of human growth hormone from amino acids 177 to 191 with an added N-terminal tyrosine, and of the 50 pg/mL urine limit of detection.
9
U.S. Food and Drug Administration. Certain Bulk Drug Substances for Use in Compounding that May Present Significant Safety Risks, table headed "Bulk drug substances nominated but withdrawn", entry "AOD-9604". Page content current as of 22 April 2026. Retrieved 2 August 2026, HTTP 200. fda.gov, bulk drug substances that may present significant safety risks.
10
World Anti-Doping Agency. World Anti-Doping Code International Standard, Prohibited List 2026, in force 1 January 2026. Class S2, peptide hormones, growth factors, related substances and mimetics, marked prohibited at all times, in and out of competition; subsection S2.2.3, growth hormone, its analogues and fragments, gives "growth hormone fragments, e.g. AOD-9604 and hGH 176-191". Retrieved 2 August 2026, HTTP 200, 382,029 bytes. wada-ama.org, 2026 Prohibited List (PDF).
11
Mendias CL, Awan TM. Safety and efficacy of approved and unapproved peptide therapies for musculoskeletal injuries and athletic performance. Sports Med. 2026. Online ahead of print, 12 April 2026. doi:10.1007/s40279-026-02437-0. PMID 41966639. Listed here because it is the most recent peer-reviewed review to cover AOD-9604 by name among unapproved peptides marketed direct to patients. Only its abstract was retrievable to us, so nothing beyond the abstract is claimed from it on this page.
12
Huerta MÁ, Garcia MM, García-Parra B, Serrano-Afonso A, Paniagua N. Investigational drugs for the treatment of postherpetic neuralgia: systematic review of randomized controlled trials. Int J Mol Sci. 2023;24(16):12987. Source for the identity statement "LAT8881 (AOD9604) is a synthetic C-terminal fragment of human growth hormone (GH) developed by Lateral Pharma", and for NCT03865953 as a phase 2 trial of oral LAT8881. Full text read via PubMed Central, PMC10455720, retrieved 2 August 2026. doi:10.3390/ijms241612987. PMID 37629168.
13
Harpur CM, West AC, Le Page MA, et al. Naturally derived cytokine peptides limit virus replication and severe disease during influenza A virus infection. Clin Transl Immunology. 2023;12(3):e1443. Source for "LAT8881 (formerly identified as AOD9604)" and for the sequence H-YLRIVQCRSVEGSCGF-OH, a 16-amino-acid synthetic form of the C-terminal fragment of human growth hormone with an added N-terminal tyrosine and a disulphide bond. Full text read via PubMed Central, PMC10034483, retrieved 2 August 2026. doi:10.1002/cti2.1443. PMID 36969366.

Around this page: five neighboring pages on the growth hormone axis, on what a vial holds, and on the arithmetic between a label and a syringe.

Ipamorelin guide: a secretagogue that raises the hormone this fragment was cut from, and what its own trial record covers.
CJC-1295 guide: the other half of the growth hormone axis stack, and the same question about what has been measured in people.
What an HPLC purity number misses: the gap between a high purity percentage and knowing which molecule the vial contains.
How injections were given in the trials: what the published routes were, compound by compound.
Tirzepatide against semaglutide: the head-to-head weight trial data, for a sense of scale on what a positive result looks like.

Next in the series: one endpoint gets checked against its own trial each week in The Decadewise briefing.

The disclaimer

Every page is reviewed by medical professionals before it ships, and written with longtime biohackers who were doing this before it was a trend. Reviewed still does not mean prescribed: nothing here is medical advice. It is research, trial data, and reported use, with the numbers intact so you can check them. For decisions about your body, see a doctor who can look at your labs.