Most comparisons on this site put two compounds side by side that compete for the same job. This one is different. HGH 191AA and IGF-1 LR3 sit on the same axis at different heights — one a pituitary hormone, the other an engineered analog of that hormone's best-known downstream mediator. The vertical relationship invites a specific assumption: that acting at the lower tier is a more direct route to the same output.
The assumption has a name. It was the dominant model of growth regulation for four decades, and it was substantially revised in 1999 — the most useful thing to understand about this pair.
Two Tiers, Two Molecules
HGH 191AA is recombinant human growth hormone. The FDA label for one marketed somatropin product describes it as produced by recombinant DNA technology in Escherichia coli, comprising 191 amino acids with a molecular weight of about 22,000 daltons and a sequence identical to human growth hormone of pituitary origin. UniProt's somatotropin entry (P01241) records a 217-residue precursor with a signal peptide at 1–26, leaving a mature chain of 191 residues held by two disulfide bonds — so the product name uses mature-chain numbering, a coordinate-frame issue covered in the fragment class primer. The 191AA-versus-192AA question is a manufacturing artifact addressed in the synthesis routes piece.
IGF-1 LR3 is not IGF-1. Native IGF-1 is described on the mecasermin label as 70 amino acids in a single chain with three intramolecular disulfide bridges and a molecular weight of 7649 Da. LR3 is an 83-residue analog carrying a Glu3→Arg substitution and a 13-residue N-terminal extension, both aimed at reducing binding-protein affinity rather than improving receptor engagement. That engineering, and the circulated half-life claim that runs backwards from the only direct clearance measurement, are covered in the LR3 deep-dive.
So: a natural hormone reproduced faithfully, versus a deliberately altered version of a natural mediator.
The Hypothesis That Made Them Look Interchangeable
The tiered picture came out of mid-century work on a serum factor that stimulated cartilage sulfate incorporation and appeared to be produced under growth hormone control. In 1972, Daughaday and colleagues named it in a one-page Nature note, "Somatomedin: proposed designation for sulphation factor" (Nature 1972;235:107). The somatomedin hypothesis held that growth hormone's growth effects were mediated indirectly, by a hepatically produced circulating factor. If that were the whole story, the mediator tier would be a shortcut.
Two 1999 papers using Cre/loxP deletion of the igf1 gene in liver tested it directly. Yakar and colleagues reported a dramatic reduction in circulating IGF-I while body weight, body length and femoral length did not differ from wild-type littermates, concluding that the model "challenges the concept that circulating IGF-I is crucial for normal postnatal growth" and points to an autocrine/paracrine role instead (PNAS 1999;96:7324–7329). Sjögren and colleagues found serum IGF-I reduced by 75% with "no discernible effect on postnatal body growth" (PNAS 1999;96:7088–7092). Both confirmed the liver is the principal source of IGF-I in blood. Neither found the blood pool was what mattered for growth.
Running the other way, Green, Morikawa and Nixon had already proposed a dual effector theory in which growth hormone acts both directly and indirectly, the direct action promoting differentiation of precursor cells — and stated plainly that "IGF-I is not able to substitute for growth hormone in promoting this differentiation" (Differentiation 1985;29:195–198).
Read together: the tier model survives as anatomy; the substitution reading of it does not. These are mouse-genetic and cell-biology findings, rungs 1 and 2 of the evidence hierarchy — but they are what the whole framing rests on.
Two Receptors, Two Machines
The receptors are not variations on a theme. The growth hormone receptor is a class I cytokine receptor, and de Vos, Ultsch and Kossiakoff's crystal structure showed one hormone molecule engaging the extracellular domains of two receptor molecules (Science 1992;255:306–312). How that engagement transmits inward was revised much later: Brooks and colleagues described a ligand-induced transition from a parallel transmembrane helix pair to a left-handed crossover, separating the receptor's JAK2 binding motifs, removing each pseudokinase domain from its partner's kinase domain and permitting trans-activation (Science 2014;344:1249783). Signal proceeds through JAK2 to STAT transcription factors.
IGF-1 LR3 engages the IGF-1 receptor, a receptor tyrosine kinase, feeding PI3K/Akt and Ras/MAPK. Two caveats carry over from the LR3 post: IGF-1R forms hybrids with the insulin receptor, and the two share substantial homology, so ligand selectivity is not pathway selectivity.
That the axis can break below the receptor is instructive too: Kofoed and colleagues reported growth hormone insensitivity associated with a STAT5b mutation (NEJM 2003;349:1139–1147) — a defect in the transcription factor, not the hormone or its receptor. "Acting downstream" is not a single location.
The Hormone Carries a Pattern; the Mediator Does Not
Growth hormone is secreted in pulses, and the pattern is not decoration. Rampersaud, Connerney and Waxman reported that plasma GH profiles — pulsatile in males, persistent in females — regulate sex differences in hepatic STAT5 activation, and that the male rhythm of pulse-stimulated STAT5 activation "induces dynamic, repeated cycles of chromatin opening and closing at several thousand liver DHS." A single physiological replacement dose given to hypophysectomised male mice restored, within 30 minutes, chromatin accessibility at 83% of those dynamic male-biased sites (eLife 2023;12:RP91367).
This is mouse liver chromatin, not a human outcome. But it makes pattern a property of the signal at the hormone tier — extending the timing-as-variable thread that runs from pulsatile versus continuous GnRH in the hormonal class primer through the receptor pharmacology primer. A mediator delivered into circulation carries no such pattern; neither does exogenous hormone on an injection schedule rather than a hypothalamic one.
A review published 11 September 2026 records three once-weekly long-acting growth hormone products available in the EU — lonapegsomatropin, somatrogon and somapacitan — with trials demonstrating non-inferiority to daily GH on annualised height velocity, safety profiles consistent with daily GH, and a note that ongoing real-world studies will further clarify long-term safety and metabolic outcomes (Ann Endocrinol 2026, online ahead of print). The engineering direction at the hormone tier has been toward flatter, less frequent exposure; the chromatin work says pulse shape is read as information. Two literatures measuring different endpoints, and neither settles the other.
What the Approved Files Say, and Which Way Development Went
Here the comparison stops being interpretive. Recombinant somatropin carries a broad set of indications across several paediatric growth conditions and adult GH deficiency. Recombinant native IGF-1 — mecasermin, not LR3 — is approved for a narrow population: growth failure in paediatric patients with severe primary IGF-1 deficiency, defined on the label by height SDS ≤ −3.0, basal IGF-1 SDS ≤ −3.0 and normal or elevated growth hormone, or with GH gene deletion who have developed neutralising antibodies to GH.
Then the label states the point outright, under Limitations of Use: "INCRELEX is not a substitute to GH for approved GH indications." It adds that it is not indicated for secondary forms of IGF-1 deficiency, such as GH deficiency. The approved mediator is authorised precisely where the hormone cannot work, and explicitly not where it can.
The mediator's file carries its own liabilities, in label voice: hypoglycaemia including hypoglycaemic seizures, attributed to insulin-like effects; anaphylaxis; intracranial hypertension; lymphoid tissue hypertrophy; contraindications including closed epiphyses and malignant neoplasia. A 2026 review of the Global IGFD registry — 346 patients at an April 2025 cut-off, spanning Laron syndrome to milder phenotypes — reports height improvement in a majority and names hypoglycaemia as the most frequent targeted adverse event (Growth Horm IGF Res 2026;85:101719).
And the gradient across one axis, verified against the ClinicalTrials.gov v2 API today: somatropin returns 1,302 study records, mecasermin 40, and IGF-1 LR3 — under that name or "Long R3 IGF-I" — returns zero. Not zero results. Zero registrations: no human trials, no human PK, not even an intention on file.
LR3's modifications reduce binding-protein affinity so the analog distributes largely as free protein. It is worth seeing what programmes on the same tier chose instead. Tailor and colleagues fused IGF-1 to single-chain antibody fragments targeting matrilin-3, a cartilage matrix protein, to concentrate it at the growth plate. In mice made GH-resistant with pegvisomant, the fusion partially restored growth plate height without increasing kidney cell proliferation, and showed significantly reduced hypoglycaemic effect compared with injection of IGF-1 itself (Front Endocrinol 2025;15:1523931; several authors were employees of the sponsoring company with declared financial interests in the molecule). Their stated motivation is recombinant IGF-1's limitation set: injection burden, incomplete correction of the growth deficit, off-target effects.
One approach restrains systemic exposure and aims the mediator at a tissue. The other removes the system that restrains it. Answers to different questions — and only one has been asked in a registered trial.
What the Paperwork Cannot Settle
Both sit past the practical solid-phase ceiling and past the 40-residue line discussed in what counts as a peptide, so both are recombinant, and route sets the failure modes: host cell proteins, residual host DNA, endotoxin where the expression host is itself the source, and misfolding.
Folding is the dominant question for both, and a purity percentage cannot answer it. Misfolded and disulfide-scrambled species are isomers — identical mass, identical composition, invisible to the intact-mass spectrum on a typical certificate. Native IGF-1 and LR3 carry three disulfide bridges to hGH's two, meaning more pairings available to get wrong. Answering it needs non-reduced peptide mapping or another folding-sensitive method, as set out in MS/MS and peptide mapping. For a folded protein, an absent folding or activity line is a larger gap than an absent counterion line.
Measuring either molecule in a biological sample is separately method-dependent, with documented immunoassay-versus-mass-spectrometry discordance for IGF-1, covered in the blood assay primer. See /quality/ and the /library/.
Short FAQ
Is IGF-1 LR3 a more direct version of growth hormone? It acts lower on the same axis, but "more direct" implies the two produce the same output. The 1999 liver-deletion work and the dual effector theory argue against that, and the approved mediator's own label states it is not a substitute for GH in GH-approved indications.
Which has better human data? Ambiguous as asked. Somatropin and native rhIGF-1 both have extensive human files; IGF-1 LR3 has none. Its documented use is as a serum-free cell culture supplement — a real industrial laboratory role, and a different claim from organism-level activity.
Do the pulsatility findings mean weekly growth hormone is a mistake? No. Registered trials report non-inferiority on height velocity, and the chromatin work is mouse liver. The narrower point stands: pulse shape is part of what the hormone tier transmits, so "how much" and "on what schedule" are separate variables — and a mediator carries neither.
This article is educational and for the laboratory research community. Trulogic Labs products are sold for laboratory and research use only and are not for human consumption.