HGH Fragment 176-191: What the Studies Tested

Key Takeaways

Four molecules circulate under names that look interchangeable: native hGH 176-191, hGH 177-191, AOD-9604 and full-length human growth hormone.
AOD9401 is not a separate molecule, since the 2000 Journal of Molecular Endocrinology paper defines it as the synthetic peptide containing residues 177-191.
Two qualifying published studies directly tested the native hGH 176-191 sequence, in 1978 and 1982, and both are rodent studies.
Those two studies measured carbohydrate endpoints, including blood glucose, plasma insulin, glycogen synthase and pyruvate dehydrogenase, rather than lipolysis or fat mass.
In the 1978 screen, native hGH 176-191 produced a short-lived rise in blood glucose and a more sustained rise in plasma insulin.
Two of the six peptides screened in 1978, hGH 179-191 and hGH 180-191, were inert, so activity is not carried by the disulfide loop alone.
The hormone-sensitive lipase and acetyl-CoA carboxylase findings belong to AOD9401 and hGH 177-191, not to the native fragment.
The 2001 knockout study concluded that AOD-9604's lipolytic actions are not mediated directly through the beta-3 adrenergic receptor.
No qualifying human study of the unmodified native hGH 176-191 fragment was identified in the reviewed literature.
Native hGH 176-191 was not identified in the FDA 503A or 503B records reviewed for this article, while the 2026 WADA Prohibited List names it at section S2.2.3.

Four different molecules circulate under names that look almost identical. Native hGH 176-191. hGH 177-191, published in some papers as AOD9401. AOD-9604. And full-length human growth hormone, the parent of all three. Most pages written about the first one quote findings generated with the others. This article sorts the published record by the molecule each study actually tested, so a reader can see which result belongs to which peptide.

What the evidence establishes in one paragraph

Two qualifying published studies directly tested the native hGH 176-191 sequence. Both are rodent studies, one from 1978 and one from 1982, and both measured carbohydrate endpoints such as blood glucose, plasma insulin and enzyme activity ratios. The lipolysis findings, the adipose results and the entire human development record belong to hGH 177-191, to AOD9401 or to AOD-9604. No qualifying human study of the unmodified native hGH 176-191 fragment was identified in the reviewed literature.

Four molecules, and why the names keep collapsing

HGH Fragment 176-191 is the sixteen-residue C-terminal segment of human growth hormone, beginning at the native phenylalanine. Its molecular identity, analytical fields and documentation requirements sit on the H Frag product page and are not repeated here.

Designation	Residues	First residue	Relationship
hGH 176-191	16	Phe, native position 176	The native excised segment
hGH 177-191 / AOD9401	15	Leu	One residue shorter, the Monash working peptide
AOD-9604	16	Tyr, substituted at position 176	177-191 carrying tyrosine in place of the native phenylalanine
Full-length hGH	191	Phe at position 1	The parent hormone, used as a comparator in most of these studies

One naming quirk drives most of the confusion. Describe AOD-9604's tyrosine as replacing residue 176 and the compound reads as "176-191." Describe the same tyrosine as sitting in front of residue 177 and it reads as "177-191 plus tyrosine." Both descriptions are in circulation, including in regulatory documents, and both are accurate for one molecule. That is how an AOD-9604 result ends up filed under the fragment's name without anyone making an obvious mistake.

AOD9401 deserves its own sentence, because it is not a fifth compound. The 2000 paper in the Journal of Molecular Endocrinology defines AOD9401 as the lipolytic domain of human growth hormone containing residues 177-191. It is a programme designation for the 177-191 peptide.

**The gate used throughout this article.** A study counts as direct evidence for hGH 176-191 only when the tested article was the unmodified native sequence. Tyrosine at the first position disqualifies it. A fifteen-residue article disqualifies it. When a paper describes its own test article inconsistently, the study is recorded as unresolved rather than assigned. Every finding below is reported under the molecule that produced it.

The evidence ledger

Study	Exact form tested	Model	Design	Endpoint	Result	Direct 176-191 evidence
Ng and Bornstein 1978	Six peptides including native hGH 176-191	Normal rats	Single dose, 5 nmol/kg, in vivo	Blood glucose, plasma insulin, insulin sensitivity	172-191, 176-191, 177-191 and 178-191 produced a short-lived rise in blood glucose and a more sustained rise in plasma insulin. 179-191 and 180-191 were inert. Peptides containing 178-191 reduced insulin sensitivity	Yes
Ma and colleagues 1982	hGH 176-191 and hGH 177-191, side by side	Rat, in vivo and in vitro	Metabolite and enzyme activity study	Blood glucose, lactate, glycogen synthase and pyruvate dehydrogenase activity ratios	Transient rises in glucose and lactate. Reduced activity ratios for synthase in muscle, adipose and liver, and for the dehydrogenase in muscle and adipose but not liver, with total enzyme unchanged	Yes
Wu and Ng 1993	hGH 177-191	Rat adipose tissue, peptide-treated rats	In vivo treatment, ex vivo assay	Lipogenic activity, glycerol release	Antilipogenic action described as identical to that of intact growth hormone. No significant lipolytic effect was found, measured as glycerol release from epididymal fat pads	No
Wijaya and Ng 1993	The antilipogenic C-terminal fragment	Rat adipocytes	In vitro	Glucose transport	Reduced insulin-stimulated glucose transport	No
Natera and colleagues 1994	Synthetic hGH 177-191	Obese mice	Chronic in vivo	Cumulative body-weight gain, adipose tissue mass	Both reduced	No
Ng and colleagues 2000	AOD9401, defined in the paper as the 177-191 domain	Rat adipose tissue, obese Zucker rats	In vitro plus 20 days in vivo	Hormone-sensitive lipase, acetyl-CoA carboxylase, diacylglycerol release, weight gain, adipocyte diameter	Lipase stimulated, carboxylase inhibited, weight gain reduced, mean adipocyte diameter fell from 110 to 80 micrometres	No
Heffernan and colleagues 2000	AOD9401, oral	ob/ob mice for 30 days, human adipose tissue ex vivo	Controlled	Weight gain, lipogenic and lipolytic activity, energy expenditure, fat oxidation	Lower weight gain from day 16 with food intake unchanged, reduced lipogenesis, increased lipolysis	No
Ng and colleagues 2000, Hormone Research	AOD-9604	Obese Zucker rats	19 days oral, euglycaemic clamp	Weight gain, adipose lipolytic activity, insulin sensitivity	Weight gain 15.8 grams against 35.6 grams in controls, increased lipolytic activity, no adverse effect on insulin sensitivity	No
Heffernan and colleagues 2001, International Journal of Obesity	AOD-9604	ob/ob and lean mice	14 days by osmotic pump	Weight, energy expenditure, fat and glucose oxidation, plasma glucose, insulin, glycerol	Growth hormone and AOD-9604 both reduced weight gain in obese mice, with increased fat oxidation and raised plasma glycerol	No
Heffernan and colleagues 2001, Endocrinology	AOD-9604	Obese mice and beta-3 adrenergic receptor knockout mice	14 days	Weight, fat mass, receptor RNA, lipolysis, energy expenditure	Chronic effects on weight and lipolysis were absent in knockouts, while acute energy expenditure and fat oxidation persisted. The authors concluded the lipolytic actions are not mediated directly through the receptor	No
Metabolic Pharmaceuticals human programme, as summarised in FDA's Pharmacy Compounding Advisory Committee briefing of 4 December 2024	AOD-9604	Adults with obesity	Six randomised, double-blind, placebo-controlled trials	Safety, IGF-1, glucose tolerance, weight	Safety findings favourable, the pivotal efficacy trial did not separate from placebo, development halted in 2007	No
Habibullah and colleagues 2022	Not the native sequence. The title names hGH fragment 176-191, but the paper gives its sequence as YLRIVQCRSVEGSCGF, which is the tyrosine form, while describing it as carrying a tyrosine to phenylalanine substitution at the last position	MCF-7 cells	In vitro	Anti-proliferative activity of loaded chitosan nanoparticles	Dual-loaded nanoparticles more anti-proliferative than doxorubicin alone	No

What the two native-fragment studies actually measured

Glucose, not fat.

The 1978 work was a structure-activity screen. Six synthetic C-terminal peptides went into normal rats at a single 5 nmol/kg dose, and the native 176-191 sequence landed in the active group. Its measured effect was a short-lived rise in blood glucose with a longer rise in plasma insulin. Two of the six peptides did nothing at all, which tells you the activity depends on residues at the front of the window rather than on the internal disulfide loop by itself.

Four years later the follow-up study put 176-191 and 177-191 side by side and traced a mechanism. Both peptides pushed glycogen synthase and pyruvate dehydrogenase from their active forms into their inactive forms, without changing how much enzyme was present. Blood glucose and lactate rose transiently. The authors read that as direct action on the target tissues.

Both papers came out of a research programme studying growth hormone's diabetogenic action, meaning its tendency to raise blood sugar. That is worth holding next to the claim, common on pages selling this peptide, that the fragment has no effect on glucose. The only qualifying primary evidence points the other way, in rats.

Where the lipolysis findings actually come from

Every lipolysis result in the ledger belongs to 177-191, to AOD9401 or to AOD-9604.

The hormone-sensitive lipase and acetyl-CoA carboxylase findings, quoted across the market as the fragment's mechanism, were generated with AOD9401 in isolated rat adipose tissue. The oral ob/ob study used the same peptide. The Zucker rat weight figures came from AOD-9604. So did the entire human programme, and that record sits in what AOD-9604 research shows about the fragment distinction rather than here.

None of it was produced with the native sequence. No direct study identified here establishes that replacing the native phenylalanine with tyrosine leaves these endpoints unchanged. That experiment is the gap.

Lipolysis, fat mass and weight change are not the same finding

The ledger keeps its endpoint column narrow on purpose, because the field tends to compress a chain of separate measurements into one claim.

Endpoint	What it measures	Where it appears in the ledger
Antilipogenesis	Reduced synthesis of new lipid	Wu and Ng 1993, 177-191
Lipolysis and fatty-acid release	Breakdown of stored triglyceride, with glycerol as the index	AOD9401 and AOD-9604 studies
Adipocyte effects	Cell-level change such as mean diameter	Ng 2000, AOD9401
Fat-mass change	Change in adipose tissue mass	Natera 1994 and the AOD-9604 mouse work
Weight change	Whole-body mass, including non-fat compartments	Zucker and ob/ob studies, plus the human trials
Metabolic outcomes	Glucose, insulin, enzyme activity ratios	The two native-fragment studies

An enzyme stimulated in isolated tissue is not fat removed from an animal. Slower weight gain in a genetically obese rodent is not weight loss in a person. Each step along that chain needs its own experiment, and for the native fragment none of those experiments has been published.

The beta-3 adrenergic claim, traced back

Search results repeat one mechanism for this fragment more than any other. It acts through the beta-3 adrenergic receptor, a receptor found mainly on fat cells, and through hormone-sensitive lipase.

Split that claim in half and both halves move. The lipase work is AOD9401's. The beta-3 material comes from the 2001 knockout study, which used AOD-9604, and the authors concluded that the lipolytic actions are not mediated directly through the receptor. Chronic effects disappeared in animals lacking it, yet acute energy expenditure and fat oxidation carried on. The paper reads as a limit on the mechanism rather than proof of it.

So the popular version reverses a primary source, then applies the reversed version to a molecule that source never tested. Both moves are invisible if you only check that the citation exists.

A page asserting beta-3 adrenergic lipolysis for native hGH 176-191 is making two unsupported moves at once. The citation behind it is usually real. The attribution is what fails, which is why this error survives a quick reference check.

Where the native fragment sits in U.S. regulation

Native hGH 176-191 was not identified in the FDA 503A or 503B nomination and category records reviewed for this article. No interim category, no withdrawn-nomination entry, no advisory-committee history.

The U.S. compounding actions found in those records concern AOD-9604. It entered Category 2 in September 2023, came out in September 2024 after the nominations were withdrawn, and went before the Pharmacy Compounding Advisory Committee on 4 December 2024, which voted against inclusion. FDA also notes that it identified serious adverse events that may be associated with AOD-9604, with causality unclear. How that framework operates is covered in the 2026 bulks-list explainer.

Neither molecule is an approved drug. Research-use-only status describes labelling and intended use. It is not a finding on safety, and it is not a finding on efficacy.

One authority does name the fragment directly. The 2026 World Anti-Doping Agency Prohibited List, effective 1 January 2026, names hGH 176-191 at section S2.2.3, growth hormone analogues and fragments, alongside AOD-9604, prohibited at all times.

What a researcher can take from this

The native fragment has been synthesised and tested in rodents, with measured effects on blood glucose, plasma insulin and two enzyme activity ratios. In the later literature it appears mostly as a reference or comparator material. At least one report carrying its name in the title describes a test article that does not match the native sequence, which is the same attribution problem this article documents, showing up inside the primary literature rather than outside it.

And no qualifying human study of the unmodified native hGH 176-191 fragment was identified in the reviewed literature. Not a failed trial. None at all.

Got Questions?

Frequently Asked Questions

Two qualifying published studies directly tested the native hGH 176-191 sequence. Ng and Bornstein reported one in the American Journal of Physiology in 1978, and Ma and colleagues reported the other in Biochimica et Biophysica Acta in 1982. Both are rodent studies of carbohydrate metabolism, and everything else commonly cited for this fragment tested hGH 177-191, AOD9401 or AOD-9604.

Blood glucose, plasma insulin and insulin sensitivity. Six synthetic C-terminal peptides were given to normal rats at 5 nmol/kg. Four of them, including 176-191, produced a short-lived rise in blood glucose and a more sustained rise in plasma insulin, while two shorter peptides were inert. Peptides containing the 178-191 sequence reduced insulin sensitivity in insulin tolerance tests.

No qualifying human study of the unmodified native fragment was identified in the reviewed literature. The human data usually cited alongside it were generated with AOD-9604, a different molecule with its own registry number and its own evidence base.

Yes. The 2000 paper in the Journal of Molecular Endocrinology defines AOD9401 as the lipolytic domain of growth hormone containing residues 177-191. It is a programme designation for that peptide rather than a separate compound, which matters because the hormone-sensitive lipase and acetyl-CoA carboxylase findings attributed across the market to 176-191 were generated under that name.

From a 2001 study in Endocrinology using obese mice and beta-3 adrenergic receptor knockout mice, conducted with AOD-9604. The authors concluded that the lipolytic actions are not mediated directly through that receptor, because acute energy expenditure and fat oxidation persisted in animals lacking it. The version circulating online reverses that conclusion and applies it to a different molecule.

No direct study identified here establishes that replacing the native phenylalanine with tyrosine leaves these endpoints unchanged. The two compounds hold separate registry numbers and separate evidence bases, so a result generated with one is not a result for the other.

Because each one is a different measurement from a different experiment. Antilipogenesis is reduced lipid synthesis, lipolysis is triglyceride breakdown, fat-mass change is a tissue-level outcome, and weight change includes non-fat compartments. A result at one level does not establish a result at the next, and none of these has been measured for the native fragment.

In the two qualifying rodent studies of the native sequence, blood glucose rose transiently, alongside raised plasma insulin and lactate and reduced activity ratios for glycogen synthase and pyruvate dehydrogenase. Both studies came from research into growth hormone's diabetogenic action. No human data on this question were identified.

It was not identified in the FDA 503A or 503B nomination and category records reviewed for this article, with no interim category, no withdrawn-nomination entry and no advisory-committee history. The compounding record in those materials concerns AOD-9604, which entered Category 2 in September 2023, was removed in September 2024 after the nominations were withdrawn, and was reviewed by the Pharmacy Compounding Advisory Committee in December 2024, which voted against inclusion.

Yes. The 2026 World Anti-Doping Agency Prohibited List, effective 1 January 2026, names hGH 176-191 at section S2.2.3, growth hormone analogues and fragments, together with AOD-9604. The prohibition applies at all times and attaches to the compound rather than to any particular product format.

Because the references are real while the attributions are not. A citation to a 2000 or 2001 paper looks solid until the test article is checked, at which point the peptide turns out to be AOD9401 or AOD-9604. The failure mode is attribution rather than fabrication, which is why it survives a casual reference check.

A study whose methods state that the peptide tested was the unmodified native sequence beginning at phenylalanine, reporting a defined endpoint in a defined model. Where a paper never states the sequence of its test article, or describes it inconsistently, the study cannot be assigned to this molecule regardless of the name in its title. The same principle governs [reading a certificate of analysis](/how-to-read-peptide-certificate-of-analysis), where the document has to say what was actually measured.

References