Chapter 16: IGF-1, Myostatin Inhibitors, and the Muscle-Building Peptides

"The great tragedy of Science — the slaying of a beautiful hypothesis by an ugly fact." — Thomas Henry Huxley, "Biogenesis and Abiogenesis" (1870)

Overview

This chapter contains the cleanest example in the entire book of a mechanism that is completely, unambiguously correct — and a set of drugs built on it that did not work.

Myostatin is a brake on muscle growth. That is not a hypothesis. It was identified in the 1990s, and within a few years the natural experiments were lining up: mice engineered without it grew enormous, cattle breeds selectively bred for "double muscling" turned out to carry loss-of-function mutations in the same gene, a racing dog breed produced a heavily muscled variant from the same cause, and a case report described a human child with the same kind of mutation and the same striking bulk. Release the brake, get more muscle. It is one of the most satisfying mechanistic stories in modern physiology.

So companies built drugs to release the brake and ran them in people with muscular dystrophy and other muscle-wasting conditions, in real randomized placebo-controlled trials. And in program after program, across multiple companies and multiple molecular approaches, the same thing happened: lean mass went up, and what patients could actually do did not reliably follow. Muscle volume on imaging increased. Walking distance, timed function tests, and strength did not move as the mechanism promised. Programs were discontinued.

That result is the spine of this chapter — the most concrete demonstration you will get of a principle Chapter 2 stated abstractly and Chapters 14 and 15 applied to growth hormone: mechanism is necessary and never sufficient. Roughly nine in ten compounds entering human trials never reach approval, and this is what one of those failures looks like from the inside: not fraud, not sloppiness, not a bad idea, just a chain of reasoning with one unexamined link in it.

The unexamined link is the assumption that more muscle mass means more of what muscle is for. Section 16.6 takes that apart, and is the section to read if you read only one. Chapter 8 gave you the same lesson pointing the opposite way: on GLP-1 drugs, lean mass falls and measured physical function improves. Here, lean mass rises and function does not.

Along the way: IGF-1 and its narrow, real approval; the gray-market variant IGF-1 LR3; follistatin and the compound sold as follistatin-344; sarcopenia, the indication where a working drug would matter most; and the cancer question, which has to be stated carefully in both directions.

In this chapter, you will learn to:

• Describe how muscle hypertrophy actually happens, and why no peptide substitutes for loading
• Explain what mecasermin's approval proves and what it conspicuously does not
• Distinguish a developmental gene knockout from adult pharmacological blockade
• State the human trial pattern for myostatin-pathway inhibitors accurately and without exaggeration
• Separate mass from strength from function from independence, and rate claims accordingly
• Explain why "more lean mass" is intuitively compelling and evidentially weak
• Discuss IGF-1 and cancer risk without either dismissing it or overclaiming it

Learning Paths

💊 GLP-1 — read §16.6 and nothing else, but read it twice. It is the argument from Chapter 8 §8.8 running in reverse, and watching one principle produce opposite-looking results is what converts it from a memorized fact into a tool. 🏋️ Performance — this is your chapter. Read it in full, and read §16.5, §16.6, and §16.8 as one continuous argument; they only work together. 🔬 Science — read straight through. §16.3 and §16.4 are the molecular material; §16.9 connects to the growth-signaling epidemiology and Chapter 14's acromegaly discussion. 💄 Cosmetic — §16.6 transfers directly. Cosmetic peptide claims lean on surrogate measurements (collagen markers in a dish) exactly as this field leaned on lean mass; Chapter 30 collects the payoff. 🏥 Clinical — §16.7 is the one to read closely: patients ask about sarcopenia constantly, and the honest answer means explaining why the mass-based intuition is wrong. §16.9 is the risk conversation.

16.1 How muscle actually grows

Before any peptide, the physiology — because a startling number of muscle-peptide claims fall apart against facts that have nothing to do with peptides.

Muscle hypertrophy results from net protein accretion. Skeletal muscle is in continuous turnover: proteins are synthesized and proteins are broken down, all day, every day. Muscle protein synthesis and muscle protein breakdown both run constantly, and the muscle you have is the running balance between them. Over any given day the balance swings positive after eating and after training, and negative during fasting and rest. Growth is what happens when the positive swings, integrated over weeks and months, exceed the negative ones.

That framing has an underappreciated consequence. There is no separate "growth process" to switch on; there is only a balance to tilt, and it can be tilted from either side. A drug that reduced breakdown and a drug that increased synthesis would both produce accretion, and would produce different-looking tissue while doing it. Hold that thought; §16.6 collects on it.

Mechanical loading is the dominant signal. When a muscle fiber is subjected to tension, structural proteins deform and that deformation is converted into biochemical signaling — mechanotransduction. The downstream pathway most associated with it converges on mTOR, a gatekeeper for translation. Load, plus amino acids, plus adequate energy, produces synthesis. Remove load and the balance tips the other way within days: bed rest, immobilization, and spaceflight all produce rapid muscle loss in healthy people.

Satellite cells contribute nuclei. Muscle fibers are unusual cells: long, multinucleated syncytia in which each nucleus governs a finite surrounding volume of cytoplasm — the myonuclear domain. Growing a fiber substantially appears to require adding nuclei, supplied by resident stem cells called satellite cells that sit beneath the fiber's basal lamina, activate in response to loading and damage, proliferate, and fuse into the existing fiber. This matters because several compounds in this chapter act partly through satellite cell behavior, and because it explains why growth is slower and more structurally involved than "add protein, get bigger" suggests.

THE PROTEIN BALANCE — what actually determines muscle size

     SYNTHESIS  ────────────►│◄──────────── BREAKDOWN
                             │
   driven by:                │              driven by:
   · mechanical loading       │              · fasting
   · amino acid availability  │              · inflammation, illness
   · energy availability      │              · disuse, immobilization
   · anabolic signaling       │              · glucocorticoids
     (mTOR pathway)           │              · aging (partly)
                             │
                   NET BALANCE over weeks
                             │
        positive ──► hypertrophy    negative ──► atrophy

  Nothing in this diagram is a switch. It is a running total, and the
  single largest lever on the left-hand side is not chemical.

The diagram locates the peptides correctly. Every compound in this chapter operates on the anabolic-signaling term or the breakdown term. None replaces the loading term, and the loading term is the largest one.

Resistance training is the dominant intervention and no peptide substitutes for it. This is not a throat-clearing disclaimer; it is the most robustly evidenced statement in the chapter. Progressive resistance training reliably increases muscle mass and strength in essentially every population tested, including people in their eighties and nineties, people with chronic disease, and people who have never trained. The effect sizes are large and the literature is enormous. The compounds in this chapter are studied as adjuncts, or as therapies for disease states in which loading is impossible or insufficient. A drug that helps someone whose muscle is failing for a genetic reason no amount of training can fix is solving a different problem than a healthy adult who wants larger arms.

It also sets the comparison bar. Any performance claim here is implicitly a claim about adding something on top of training — much harder to demonstrate than an effect in an untrained or wasting population, and almost never what the supporting evidence measured.

One last piece of vocabulary, because it does real work in §16.3. Hypertrophy means existing fibers get bigger; hyperplasia means there are more of them. In adult humans hypertrophy is the dominant and probably near-exclusive mode — fiber number is largely set during development. Remember that when we get to the cattle.

16.2 IGF-1, IGF-1 LR3, and mecasermin's narrow approval

Chapter 14 left growth hormone at the point where its anabolic effects are largely mediated by something else. This is that something else.

IGF-1 — insulin-like growth factor 1 — is a 70-amino-acid single-chain polypeptide structurally related to proinsulin, which is where its name comes from and where one of its side effects comes from. It is produced largely in the liver in response to GH and mediates many of GH's anabolic effects. Muscle and other tissues also produce IGF-1 locally in response to loading, and local production may matter more for muscle than circulating levels do — a nuance that returns shortly.

IGF-1 acts on the IGF-1 receptor, a tyrosine kinase closely related to the insulin receptor, signaling through pathways you have met before: PI3K/Akt driving protein synthesis and cell survival, MAPK driving proliferation. The receptor promotes growth and inhibits apoptosis — cells exposed to IGF-1 grow and are less likely to die on schedule. Keep that second clause; §16.9 needs it.

🧬 The Molecule — IGF-1 and the binding proteins that are the whole story

Almost nothing you read about IGF-1 in the consumer space mentions the binding proteins, and the binding proteins are the physiology.

Free IGF-1 in blood has a half-life measured in minutes — but almost none of it is free. Upward of 99% travels bound to a family of six IGF binding proteins (IGFBPs), mostly IGFBP-3, which itself associates with the acid-labile subunit to form a large ternary complex. In that complex IGF-1's half-life stretches to hours, it cannot easily leave the circulation, and it is inert until released.

So the body maintains a large reservoir of an extremely potent growth factor held chemically on a leash, with local proteases determining where and when a little gets let off. The binding proteins are not packaging; they are a control system that buffers concentration, extends lifetime, restricts distribution, and localizes activity.

This is why "raise your IGF-1" means much less than it sounds like. Total IGF-1 is what a lab reports; free IGF-1 at a particular receptor in a particular tissue at a particular moment is what does anything, and the relationship between the two is governed by a six-protein regulatory system you did not adjust. It is also, as we are about to see, precisely the system the gray-market variant was engineered to defeat.

Mecasermin: the approval, and exactly how narrow it is

Mecasermin is recombinant human IGF-1, and it is approved for severe primary IGF-1 deficiency — a rare condition in which a child cannot produce IGF-1 despite normal or elevated GH, usually because the GH receptor or its downstream signaling is defective. These children are profoundly short. Giving them GH does nothing, because the problem is downstream of GH. Giving them IGF-1 works: they grow. It is a genuine, mechanism-confirming approval, establishing that IGF-1 is biologically active in humans and that replacing it in severe deficiency produces the predicted effect.

That narrow approval proves the mechanism and does not license performance claims. This is exactly the structure Chapter 14 taught with growth hormone, worth stating in the same words: replacement in deficiency is not enhancement in sufficiency. A child who makes no IGF-1 and a thirty-year-old with a normal level are not in the same experiment — the first is being restored to a physiological range, the second pushed above it. Nothing about the first result predicts the second, and the history of endocrinology is substantially a history of people assuming it does.

💊 In the Clinic — what mecasermin's label actually covers

Mecasermin's approved population is small: children with severe primary IGF-1 deficiency, and children with GH gene deletion who have developed neutralizing antibodies to GH. It is a growth indication in a pediatric rare disease — not a muscle indication, and not an adult indication.

The label's most prominent warning is hypoglycemia, and the reason is structural. IGF-1 resembles insulin closely enough that at sufficient concentration it engages insulin receptors, and the resulting drop in blood glucose can be severe; managing that risk is why the labeling addresses food intake around administration at all. Other labeled concerns include tonsillar and adenoid hypertrophy — a reminder that a general growth factor grows things you were not aiming at.

None of that is guidance. It is here because it is the only place in this chapter where a regulator has reviewed human IGF-1 safety data, and what a regulator thought worth warning about tells you what the molecule does in people. A compound whose approved use carries a serious hypoglycemia warning is not a benign compound at higher exposure in someone whose IGF-1 was normal to begin with.

📊 Evidence Rating

Claim: Mecasermin (recombinant IGF-1) improves growth in children with severe primary IGF-1 deficiency. Rating: ✅ Strong clinical evidence (assessed 2026) Reason: It is an approved therapy for that indication, supported by human trial data in the target population, with a characterized safety profile including a known hypoglycemia risk. What would change it: Post-marketing evidence that long-term outcomes in treated children are worse than the growth benefit justifies, or a safety signal serious enough to withdraw the indication. Note that this rating is narrow by construction — it says nothing about IGF-1 in anyone who is not IGF-1 deficient.

IGF-1 LR3: engineered to defeat the leash

IGF-1 LR3 — often written "Long R3 IGF-1" — is a modified IGF-1 variant carrying a thirteen-residue N-terminal extension and a substitution at position 3. Both changes serve one purpose: reduced binding-protein affinity and extended activity. It slips the leash. Less of it is captured by IGFBPs, more of it stays free, and its activity persists far longer than native IGF-1's.

The provenance is genuinely informative. Variants of this kind were developed as cell-culture reagents, used in bioprocessing to stimulate growth of mammalian cells in industrial fermenters, where a longer-acting, binding-protein-resistant growth factor is cheaper than the native one. It is a laboratory tool that migrated into vials sold to humans, without anyone running the studies that would normally sit between those two states.

It is not approved anywhere, and is known by a code rather than a generic name — which, per Chapter 1 §1.8, is itself a fact about regulatory history. The international naming authority assigns stem names to compounds in serious clinical development. Decades on, this one has none.

Now put that design intent next to the molecule box above. The binding-protein system buffers concentration, limits distribution, and localizes activity; IGF-1 LR3's entire rationale is to escape it. That is the deliberate removal of a regulatory brake on a growth-signaling pathway, sustained over time, in a person whose own IGF-1 production is presumably normal. The long-term consequences have not been studied in humans, and §16.9 explains why that unstudied question is not a comfortable one.

And the nuance promised earlier: animal work suggests locally produced, load-responsive muscle IGF-1 is more closely tied to hypertrophy than systemic levels are. Held loosely, as animal evidence — but it means even the mechanistic case for injected systemic IGF-1 is weaker than it first appears.

📊 Evidence Rating

Claim: IGF-1 or IGF-1 LR3 increases muscle mass, strength, or athletic performance in healthy adults. Rating: ❌ Hype outpaces evidence (assessed 2026) Reason: There are no completed randomized human outcome trials of IGF-1 or IGF-1 LR3 for muscle growth or performance in healthy adults; neither is approved for any such use; and the risk of chronically elevating a growth-signaling axis in a healthy person is unquantified rather than established as low. What would change it: A completed, adequately powered, randomized, placebo-controlled trial in healthy trained adults reporting both mass and function, with a safety follow-up long enough to say something about the growth-signaling question. Nothing short of a controlled trial resolves this, because the confounding in the uncontrolled setting (§16.8) is total.

16.3 Myostatin: the brake, and the animals born without it

Now the other half of the chapter, and the more interesting biology.

Myostatin, also called GDF-8, is a negative regulator of muscle growth. A TGF-β superfamily member, produced by skeletal muscle itself, secreted, and acting back on muscle — an autocrine and paracrine brake. It binds the activin receptor type IIB (ActRIIB), which recruits a type I receptor, which phosphorylates Smad2 and Smad3, which move to the nucleus and suppress the transcriptional program of growth. In parallel it damps the Akt/mTOR signaling that drives synthesis.

The architecture is the interesting part. Muscle makes a molecule whose job is to stop muscle from getting bigger: growth is not merely switched on when conditions allow, it is actively and continuously restrained.

🧬 The Molecule — myostatin (GDF-8) and why it has a family

Myostatin does not act alone, and that fact complicates every drug ever built against it.

It belongs to the TGF-β superfamily, a large group of secreted signaling proteins sharing receptors and downstream machinery. Its closest relatives include activin A and GDF-11, and both signal through the same ActRIIB receptor myostatin uses. They are not redundant — activin A is involved in reproductive signaling, inflammation, and fibrosis — but they overlap at the receptor.

The consequence is a specificity problem that shaped the whole therapeutic field. You can block myostatin at three levels: the ligand (an antibody binding myostatin), the receptor (a decoy that soaks up everything ActRIIB would have bound, or an anti-receptor antibody), or the intracellular signal. The narrower you go, the more selective you are and the weaker the effect tends to be; the broader you go, the larger the effect and the more of the family you silence along with your target.

That trade-off is a property of the biology, not an oversight, and it is why compounds in this space differ so much in effect size and side-effect profile despite being described in the popular press as "myostatin inhibitors" as though they were one thing.

The natural knockouts

The reason myostatin captured everyone's imagination is that nature had already run the experiment, repeatedly, in the open.

Mice. The gene was identified in the 1990s, and mice engineered to lack it developed dramatically increased muscle mass — the "mighty mouse" images that circulated widely — from both larger fibers and, in the mouse, more of them.

Cattle. Two European breeds, the Belgian Blue and the Piedmontese, had been bred for generations for a trait stockmen called "double muscling." When the myostatin gene was examined, both carried loss-of-function mutations. Human breeders had, without knowing it, selected for a myostatin knockout.

Dogs. A myostatin mutation was identified in whippets, a racing breed. Dogs with two copies show a heavily muscled phenotype quite unlike a normal whippet; more interestingly, dogs with one copy were reported to be overrepresented among the fastest racers — partial loss of function tracking with athletic performance in a species bred and timed for exactly that.

Humans. A case report published in 2004 described a child with loss-of-function mutations in the myostatin gene showing muscle bulk far outside the normal range from infancy. A single case report — and, in its way, decisive: the pathway operates in humans as it does in everything else.

This is real and striking. Nobody here will tell you the mechanism is fake. It is one of the best-supported gene-to-phenotype stories in mammalian physiology, and had you shown it to a room of physiologists in 1998 and asked whether a myostatin-blocking drug would build muscle in people, the room would have said yes — and would have been reasoning well.

⚠️ Hype Check — "the double-muscled cattle prove myostatin inhibitors work"

The claim, in its usual form:

"Look at a Belgian Blue. Look at the mighty mouse. Myostatin is the brake on muscle growth, it's proven, and blocking it is the closest thing to a cheat code that exists."

What's true in it. All of the biology. The gene does what they say, the animals are as described, the human case report exists, and the pathway is conserved. The underlying facts are not wrong.

Where it fails. Three places, and the third is the one nobody sees coming.

First: a developmental knockout is not adult blockade. The Belgian Blue was never exposed to myostatin. It developed from a single cell in its absence, through the entire period when fiber number is set, and its extra muscle is substantially hyperplasia — more fibers — which adult humans do not add. Blocking myostatin in a fully formed adult is a different intervention on a different substrate at a different life stage. The animals tell you the pathway is real. They do not tell you what a drug does to a grown man.

Second: the animals come with a bill. Double-muscled cattle breeds are associated with substantial calving difficulty — Belgian Blue calves are frequently delivered by cesarean section as routine husbandry — along with reduced fertility, sensitivity to heat and stress, and tendon and joint problems. And the heavily muscled homozygous whippets are not the ones that win races; the heterozygotes are. A phenotype can be spectacular and still be, on net, costly.

Third, and this is the one: in myostatin-null animals, force per unit of muscle cross-section has been reported to be reduced. The muscle is bigger and, gram for gram, weaker; the increase in strength is smaller than the increase in size. That finding is the seed of everything in §16.5 and §16.6 — the dissociation between mass and function was visible in the animal models before a single human trial was run, if you were looking for it. Most people were looking at the photographs.

Verdict: the mechanism is proven and the extrapolation is not. A textbook case of mechanism substituting for evidence — and unusually, we no longer have to speculate about how the extrapolation would fare, because it was tested (§16.5).

16.4 Follistatin and follistatin-344

If myostatin is the brake, follistatin is the body's own brake-release.

Follistatin binds and inhibits myostatin and related family members. It is an endogenous secreted glycoprotein — named, confusingly for our purposes, for its original discovery as an inhibitor of follicle-stimulating hormone secretion. It binds activin A with very high affinity, and also binds myostatin, GDF-11, and several bone morphogenetic proteins, sequestering them from their receptors.

Two things about that sentence should already bother you.

First, follistatin is not selective. It is a broad-spectrum neutralizer of a whole branch of the TGF-β family. Blocking myostatin with follistatin is like silencing a phone by cutting the building's power. It works, and it does other things.

Second, follistatin is not really a peptide. The isoforms run to a few hundred amino acids and roughly 35–40 kDa — well into protein territory on Chapter 1's size spectrum — and, critically, it is glycosylated. Sugar chains attached after translation are part of what the molecule is, influencing folding, stability, clearance, and binding. A protein of that size with post-translational modifications cannot be made by the solid-phase chemical synthesis that makes peptides; it requires expression in living cells with the right machinery.

Follistatin-344 is the name attached to a gray-market compound. The "344" refers to the 344-residue precursor isoform, which is processed into the shorter forms found in circulation and bound to cell surfaces. It is not approved anywhere, for anything.

🩺 Safety and Risk — the identity problem with "follistatin-344," stated plainly

This callout is not about follistatin's biology. It is about a prior question that has to be settled first and usually cannot be.

Follistatin is a glycosylated protein of roughly 35–40 kDa. Producing it authentically requires expression in a mammalian or comparable cell system, purification, and analytical characterization — the same infrastructure a biologic manufacturer uses, because it is a biologic. That is expensive, technically demanding, and entirely unlike running a peptide synthesizer.

So when a vial is offered as "follistatin-344," the first question is not whether follistatin builds muscle. It is what is in the vial, and the honest answer available to a purchaser is: no one outside the seller knows. No assurance of identity, of correct glycosylation, of concentration, of sterility or endotoxin content. Programs that have analytically tested gray-market peptide products have repeatedly found contents that did not match the label — wrong substance, wrong amount, or nothing detectable. Chapters 19 and 34 cover that literature.

There is a further wrinkle specific to this compound. A large glycosylated protein, made wrong, is exactly the kind of molecule likely to provoke an immune response — and antibodies raised against an administered version could in principle cross-react with a person's own follistatin. That is a theoretical risk, stated as one: theoretical, with no monitoring, no data, and no way for an individual to detect it.

The point generalizes. When a product's biology is a large glycosylated protein and its market is unregulated vials, the biology is not the binding constraint. Decisions here belong with a clinician who can see your history, not with a vendor's product page.

The gene therapy trials, and why they are a different intervention

Some approaches to raising follistatin have been pursued via gene therapy in trials. Investigators have delivered a follistatin gene using an adeno-associated viral vector injected directly into muscle, in early-phase studies in muscular dystrophies and inclusion body myositis — small, largely open-label, reported as encouraging by their investigators and preliminary by everyone else.

They are worth mentioning because they are frequently cited in defense of the injectable gray-market product, and they do not support it at all. A one-time viral vector delivering a gene to a specific muscle, producing protein continuously from within the tissue, is a fundamentally different intervention from injecting a protein of uncertain provenance systemically. The trials are evidence about gene therapy; citing them for the vial is a category error, and one of the most common moves in this corner of the market.

16.5 The muscular dystrophy trials: what happened when the mechanism met patients

Here is the part of the story that most people who talk about myostatin have never heard.

The therapeutic logic deserves to be stated at full strength before we watch it fail. Duchenne muscular dystrophy is a genetic disease in which muscle progressively degenerates and is replaced by fat and fibrous tissue; boys typically lose ambulation in adolescence and die young. Related dystrophies and other wasting conditions follow similar arcs. There is a brake on muscle growth. Release it in a person whose muscle is disappearing and you should slow or partially reverse the loss — an enormous advance across a whole class of devastating diseases.

So it was tried, seriously and repeatedly and with real money. Multiple sponsors developed multiple approaches: antibodies binding myostatin itself, ActRIIB-based decoy receptors that soak up myostatin and its relatives, antibodies directed at the receptor. These went into randomized, placebo-controlled human trials in dystrophies and other muscle-wasting settings over roughly a decade and a half.

And here is what happened, stated as the pattern rather than as a list of programs:

Lean mass increased. Function did not reliably follow.

Body composition endpoints moved — lean body mass by DXA, muscle volume on MRI, thigh cross-sectional area. In several programs the mass signal was clear and dose-related: the drugs did exactly what they were designed to do at the level of tissue. And the functional endpoints — timed function tests, six-minute walk distance, ambulatory assessment scales, measured strength — did not show the corresponding improvement, not consistently and not at a magnitude justifying continuation.

Multiple programs across multiple companies were discontinued for this reason. Some also met safety signals: broad ActRIIB blockade has been associated in trial settings with effects on small blood vessels, including nosebleeds and visible telangiectasias, consistent with the family's involvement in vascular biology — §16.4's specificity problem showing up in patients. But the primary reason the field emptied out was efficacy: the mass was there and the function was not.

🔬 Read the Study

```text FIGURE 16.1 — "Myostatin-pathway inhibition in muscle disease, read as a body of work" [pattern across multiple randomized trials, approx. 2010s–2020s]

THE STUDY Not one trial but a body of them: multiple randomized, placebo-controlled studies of myostatin-pathway inhibitors — anti-myostatin antibodies, ActRIIB- based decoy receptors, anti-receptor antibodies — run by several independent pharmaceutical sponsors in Duchenne and Becker muscular dystrophy and in other muscle-wasting conditions. Different molecules, different companies, different patient populations, overlapping endpoints.

THE QUESTION Does releasing the physiological brake on muscle growth slow, halt, or reverse functional decline in people whose muscle is being lost to disease?

WHAT IT SHOWS That the drugs work at the level of tissue. Across programs, measures of muscle mass and volume — DXA lean mass, MRI muscle volume, cross-sectional area — rose relative to placebo. The pharmacology is not in doubt. These compounds engage the pathway in humans and the pathway does in humans what it does in mice.

WHAT IT DOESN'T Show a corresponding, reliable improvement in what patients could do. Timed function tests, walking distance, ambulatory assessments, and measured strength did not move the way the mass changes predicted. The dissociation is the result, and it is not a single disappointing trial a better-powered study might overturn; it recurred across sponsors and mechanisms, which makes it a finding, not a fluke.

THE VERDICT The mechanism was confirmed and the therapy failed. Programs were discontinued. In dystrophy specifically there are plausible post-hoc explanations: myostatin expression is already low in dystrophic muscle, so there is less brake to release, and the underlying defect is membrane fragility, which extra bulk does not repair. Those are reasonable, and they are post hoc. Hold them as hypotheses, not as reasons to discount what was measured.

THE LESSON The most instructive negative result in Part III, and its value is entirely in the gap it exposes. Everyone involved believed adding muscle mass would add function, because that link is so intuitive it was never stated as an assumption, let alone tested. It was an assumption, these trials tested it accidentally, and it did not hold. Whenever a claim rests on a measurement that is not the thing you care about, the arrow between them is a hypothesis — and this is what it looks like when the hypothesis is wrong. ```

📊 Evidence Rating

Claim: Myostatin or follistatin pathway inhibitors improve muscle function in people with muscular dystrophies and related muscle-wasting diseases. Rating: ⚠️ Promising but preliminary (assessed 2026) Reason: Real randomized human trials have been run — which is more than almost any other compound in Part III can say — and they consistently increased lean mass while failing to reliably improve function; multiple programs were discontinued on that basis. The rating is ⚠️ rather than ❌ because there is substantial human evidence, some of it showing genuine pharmacological effect, and because the therapeutic hypothesis remains live in populations and combinations not yet adequately tested. What would change it: An adequately powered randomized trial in a muscle-wasting population reporting a clinically meaningful functional benefit — walking distance, timed function, retained ambulation, retained independence — would move this toward ✅. Continued accumulation of trials showing mass without function, or a serious class safety signal, would move it to ❌.

Read that rating carefully, because it is doing something unusual. ⚠️ here does not mean "a little data, looking hopeful." It means the human data exists and is disappointing. Compare the ❌ two sections back, awarded for the opposite situation — no human outcome data at all. A one-dimensional scale flattens two different epistemic states:

• ❌ from absence — nobody ran the trial. The claim is unsupported. It might be true.
• ⚠️ from disappointment — the trial was run, the pharmacology worked, the outcome did not follow. The claim is not merely unsupported; evidence points away from it, and any future version has to explain this result rather than ignore it.

The second is a worse position for a claim to be in despite the friendlier symbol, because the rating tracks the evidence base while the claim's plausibility has fallen. §16.8 shows what happens when you carry that across to the performance market.

16.6 Why "more muscle" is a surrogate and function is the real endpoint

A drug that reliably increases muscle mass and does not improve what a person can do is not a successful drug.

That sentence is the chapter. It sounds trivially obvious stated flatly, and yet an entire therapeutic field — staffed by excellent scientists with excellent reasons — spent fifteen years and a great deal of money finding out that it is true. The consumer market built on the same mechanism has not learned it at all.

The ladder, and the fact that every rung is a separate claim

THE LADDER FROM TISSUE TO LIFE — four different things, three fragile arrows

   MASS              how much muscle tissue exists
     │               measured by: DXA lean mass, MRI volume, ultrasound thickness
     │  ← ARROW 1: does more tissue produce more force?
     ▼
   STRENGTH          how much force can be produced
     │               measured by: dynamometry, grip strength, one-rep maximum
     │  ← ARROW 2: does more force enable more capability?
     ▼
   FUNCTION          what a person can physically do
     │               measured by: gait speed, chair rise, stair climb, 6-min walk,
     │                            timed-up-and-go, timed function tests
     │  ← ARROW 3: does more capability change a life?
     ▼
   INDEPENDENCE      living alone, not falling, not being hospitalized,
     & SURVIVAL      staying out of care, being alive

  Each arrow is a HYPOTHESIS with its own evidence requirement.
  Measuring MASS and asserting INDEPENDENCE skips three of them at once.

The four rungs are related. They are not the same, they can come apart, and this chapter is the proof that they do.

Mass is tissue quantity — and note what the standard measurements capture. DXA "lean mass" is fat-free soft tissue: muscle, but also water, glycogen, connective tissue, and in a diseased or edematous limb whatever else is there. MRI muscle volume is better but still measures volume inside a boundary, and in a dystrophic muscle that includes fatty and fibrous infiltration. A drug that increased intramuscular water, or added non-contractile tissue, would raise these numbers without adding one newton of force — not hypothetical in a field where the disease replaces muscle with fat.

Strength is force production, and not a pure function of cross-sectional area. A large fraction of strength — especially early change — is neural: how completely you recruit the motor units you have, how well they fire in sequence. Tendon stiffness matters; coordination matters. And as §16.3 flagged, specific force can fall while size rises. Bigger and proportionally weaker is documented in the animal models of exactly this pathway.

Function is what you can do, and it is limited by whatever is currently most limiting. This is the rung people underestimate most badly. Someone who cannot climb stairs may be limited by muscle, joint pain, balance, breathlessness, contractures, fear of falling, neurological control, or some combination. Adding muscle to a system whose binding constraint is elsewhere changes nothing anyone can measure, and the patient correctly reports no difference. That is one of the more compelling explanations for the dystrophy results: in a boy whose muscle membranes rupture under load, more muscle is more tissue that ruptures under load.

Independence and survival are what the whole enterprise is for, and are hardest and slowest to measure. Almost nothing in this literature reaches them.

Why the intuition is so strong, and why it is a confound

Here is the part worth thinking hardest about.

"More muscle means stronger and more capable" does not feel like an assumption; it feels like a definition. And the intuition has a real experiential basis: everyone who has trained has watched mass and strength and capability rise together, and everyone who has spent two weeks ill in bed has watched all three fall together. The correlation is real, and it is in your own memory.

But the correlation comes from a confounded source. Resistance training raises all four rungs because training is simultaneously a stimulus at every rung. It adds contractile protein, trains the nervous system to recruit, stiffens tendons, rehearses the coordination, and builds the confidence to attempt the stairs. Mass and function rise together under training because training does every job, not because one causes the other.

A drug that adds tissue without the stimulus severs precisely the link that gave the intuition its force. It performs the first job and none of the others. So the everyday experience that makes "more mass, more function" feel self-evident is, when examined, evidence about training and not about mass. The intuition was never tested; it was inherited from an intervention that happened to deliver everything at once.

That is why this failure mode is not a generic warning about surrogate endpoints. It is specific and repeated in this therapeutic area, and it recurred precisely because the assumption was so intuitive that no one wrote it down as one.

What regulators do about it, and why

A surrogate endpoint is a measurement used as a stand-in for the outcome you actually care about, because it is faster, cheaper, or easier to measure. Surrogates are not illegitimate; regulators accept them routinely, under one condition — the surrogate has to be validated, meaning there is evidence that an intervention-induced change in it reliably produces the corresponding change in the outcome. Validation is a property of the surrogate-outcome-intervention triple, not of the measurement alone.

Lean mass has not been validated as a surrogate for physical function in muscle-wasting disease; the trials in §16.5 are in effect the validation study, and it came back negative. That is why agencies here require functional co-primary endpoints: a drug for a muscle-wasting condition must show patients can do something better, not merely that their thighs are larger on a scan. That reads as bureaucratic conservatism only until you know its history. It is the accumulated memory of exactly this failure, encoded as a rule so the next sponsor cannot market a mass effect as a benefit.

The Chapter 8 pairing, in the opposite direction

Now put this beside §8.8, because the pair is the most useful thing in this book for inoculating yourself against surrogate reasoning.

On GLP-1 receptor agonists, lean mass falls. A meaningful fraction of the weight lost on these drugs is fat-free mass — what happens with substantial weight loss by any route — and it generated a genuine, reasonable concern. And yet in the trials that measured it, physical function improved. People moved better and performed better on measures of capability, while the mass number went down.

On myostatin-pathway inhibitors, lean mass rises, and function does not follow.

Two chapters, two drug classes. The surrogate moves in opposite directions and in both cases fails to predict the outcome. In one, people got worse on the number and better in their lives; in the other, better on the number and no better in their lives. Same principle, cutting in opposite directions: mass is not function.

In both cases the intuitive reading was wrong, and in both the corrective was identical — someone measured the outcome instead of assuming it. Notice too that neither direction is hard to explain after the fact. On GLP-1 drugs, load on the system fell more than capacity did, so capability rose. On myostatin inhibitors, capacity was not the binding constraint, so capability did not. Both explanations are entirely sensible, and neither was reliably predicted in advance. The measurement is not a formality that confirms what the mechanism already told you. The measurement is the only part that knows.

🔍 Check Your Understanding

1. A trial reports a statistically significant increase in DXA lean body mass versus placebo over six months and no significant change in six-minute walk distance; the press release leads with the lean mass result. Name the surrogate, name the outcome, and state what the trial established.
2. Give two distinct mechanistic reasons why muscle mass could increase without strength increasing proportionally.
3. Someone argues: "For a bodybuilder, mass is the outcome, so the surrogate problem doesn't apply." State the strongest version of that argument, then give two reasons it still does not rescue the performance claim.
4. Why does ordinary resistance-training experience make the mass-equals-function assumption harder to notice rather than easier?

16.7 Sarcopenia: the indication that would matter most

If a drug in this class worked, this is where it would matter.

Sarcopenia is the age-related loss of muscle mass and function, and it is not a niche condition. It contributes to falls, fractures, hospitalization, loss of independence, entry into residential care, and mortality — among the largest sources of disability in an aging population. A safe drug that preserved older people's ability to rise from a chair, cross a road inside the light cycle, and carry groceries up a flight of stairs would be one of the most valuable medicines of the century.

There is no approved drug for it. Not in the United States, not in Europe. The reason is §16.6.

The field moved its own definition, and that tells you everything

Sarcopenia was originally defined by low muscle mass — that is what the word means, from the Greek for flesh and loss. Early diagnostic criteria were mass-based: measure appendicular lean mass, adjust for height, apply a cutoff.

It did not work well. Mass-based definitions identified a population whose outcomes were not as bad as expected while missing people who were clearly failing. Muscle mass at a point in time was a mediocre predictor of falls, disability, and death.

So the field changed the definition. Contemporary consensus criteria — the widely used European working-group definition is the clearest example — now put low muscle strength first as the primary identifying criterion, use low mass or quality as confirmatory, and use low physical performance to grade severity. Mass was demoted from the definition to a supporting role, on the evidence.

Sit with that. The research field devoted to muscle loss concluded, on its own data, that mass was not the right thing to measure. That is not a rhetorical point in an argument about peptides; it is the settled position of the specialty. Anyone selling a compound on the strength of a lean-mass number is selling against the consensus of the discipline that studies muscle loss for a living.

💊 In the Clinic — what a geriatrician actually measures

If you want to know whether a treatment for age-related muscle loss is working, here is roughly the battery a clinician or a trial would use. None of it involves a scan.

Grip strength — a handheld dynamometer, squeezed. Cheap, fast, reproducible, and it predicts a startling range of outcomes (disability, hospitalization, mortality) better than most expensive measurements do. Chair-stand test — rising from a chair five times without using the arms, timed. Functional lower-body strength and power in one measure, mapping directly onto a daily task. Gait speed — walking speed over a short measured course; among the most robust single predictors of adverse outcomes in older adults that exists. Short Physical Performance Battery — gait speed, chair stands, and standing balance combined. Six-minute walk distance — how far a person covers in six minutes; a composite of muscle, cardiorespiratory capacity, and willingness.

Every one of these measures what a person can do. That is not convenience; it is a deliberate choice by the field, made after mass-based measurement disappointed. When you evaluate any claim about muscle in older adults, ask which of these was measured. If none were, the study did not examine the thing that matters, however impressive its imaging.

Why the sarcopenia trial is so hard to run

The difficulties are informative. Functional endpoints are slow: decline in an untreated older population happens over years, so trials must be long and large. Functional measures are noisier than laboratory numbers and are affected by mood, sleep, footwear, and whether the participant enjoyed the visit. Control participants often improve simply from being in a trial. And the background intervention that actually works is available to both arms — people who join a study about muscle start moving more.

That last point is a positive, not a nuisance, because it is the real answer to the sarcopenia problem as of this writing. Progressive resistance training, with adequate dietary protein, reliably improves strength and physical function in older adults, including the very old and the frail. That is a large, consistent, repeatedly replicated literature. It is the closest thing to a ✅ in this entire chapter, and it does not come in a vial. Any drug here has to demonstrate benefit on top of that, in people already doing it — the correct comparison, and a demanding one.

16.8 The performance market's evidence, examined honestly

Now the question most readers actually arrived with: does any of this do anything for a healthy adult who trains? Be specific about what evidence exists, because "no evidence" is imprecise and imprecision gets exploited.

What exists: cell-culture work showing these pathways operate as described. Rodent studies, including some showing muscle gains from myostatin blockade or IGF-1 administration. Livestock and canine genetics. The provenance of IGF-1 LR3 as a cell-culture growth reagent. Human trials in disease populations (§16.5). And a large volume of self-reported experience: forum logs, coach testimony, before-and-after photographs, and confident secondhand accounts of what unnamed competitors are supposedly using.

What does not exist: a completed, randomized, placebo-controlled trial of IGF-1 LR3, follistatin-344, or any comparable gray-market myostatin-pathway compound in healthy trained adults, reporting mass and strength and function against a control group. Not one.

Why the self-reported evidence cannot carry the claim

Worth walking through carefully, because dismissing testimony as "just anecdote" is lazy and persuades nobody who has watched a training partner change shape.

There is no control group, and muscle changes on its own. A trained adult who alters their training and eating and sleeping — which anyone starting a new compound almost always does — will change. The question is never "did he change" but "did he change more than he would have," and one person cannot answer that about himself.

There is no blinding, and the expectancy effect here is unusually large. A person who believes they are chemically enhanced trains harder. That is not a placebo in the weak sense of feeling better; it is a genuine increase in the training stimulus, mediated through effort, adherence, and willingness to add weight to the bar. It produces real hypertrophy from a real cause that is not the compound — one of the most underrated confounds in the entire performance-enhancement literature.

Attribution is impossible because these compounds are rarely used alone. In the populations generating this testimony, use of anabolic-androgenic steroids and other agents with large, well documented effects is common. When several agents are used together and one has a proven large effect while another has no human data, crediting the visible result to the second is unjustified on its face.

The reporting is selected at the source. People who feel a compound worked post about it. People who spent the money, saw nothing, and quietly stopped do not write it up. Every forum is a survivorship filter, and the filter is invisible from inside.

And you do not know what was in the vial. Per §16.4 and Chapter 34: no identity, concentration, or purity assurance. A log describing eight weeks of "follistatin-344" may be a log of eight weeks of something else entirely, including nothing.

The disanalogy argument, taken seriously

The strongest counterargument deserves its strongest form: the failed trials were in sick people. A dystrophic muscle is not a healthy muscle. Myostatin expression is already low in dystrophic tissue, so there was less brake to release. Healthy trained adults are a different substrate, and those failures do not transfer.

Parts of that are correct. Dystrophic muscle genuinely is a different substrate, and the low-baseline-myostatin explanation is plausible. But it does not do the work required of it.

First, the pattern was not confined to dystrophy. The mass-without-function result also appeared in other wasting settings and in trials involving older adults, populations that do not share dystrophy's biology. A single-disease explanation does not cover a multi-population pattern.

Second, a disanalogy is a reason for uncertainty, not for confidence. "Those results may not transfer" is an argument that we do not know what happens in healthy trained adults, not an argument that something good happens. It returns the claim to having no supporting human evidence, which is where it started.

Third, the burden sits with the claim. The party asserting benefit owes the evidence. Reasoning that negative evidence might not apply does not generate positive evidence, and never has.

The special asymmetry of this ❌

For most compounds in Part III, ❌ means "nobody ran the trial." Here something stronger is available. The pathway was taken into humans, by well-resourced sponsors with strong motivation, sophisticated molecules, and randomized placebo-controlled designs. The pharmacology worked — the drugs engaged the target and mass went up. And the functional benefit did not appear.

That is evidence against the performance claim, not merely absence of evidence for it. The proposition "blocking myostatin makes muscle that does more" was examined in humans with better tools than any gray-market vial, and it did not hold. A performance claim built on the same proposition inherits that result and has to answer it.

Worth noting without moralizing: both IGF-1 and agents that modify myostatin function are explicitly prohibited under the World Anti-Doping Code. That is a regulatory fact rather than an evidentiary one, but anyone competing in a tested sport should know it, and it is often omitted from the sales pitch.

📊 Evidence Rating

Claim: Follistatin-344, or myostatin inhibitors generally, increase muscle mass, strength, or performance in healthy trained adults. Rating: ❌ Hype outpaces evidence (assessed 2026) Reason: There are no randomized human outcome trials of these compounds in healthy adults; the only relevant human evidence comes from disease-population trials, which increased mass and failed on function — a result that argues against the claim rather than merely leaving it unaddressed. Product identity in the gray market is additionally unverifiable. What would change it: A randomized, placebo-controlled trial in healthy trained adults, with verified drug product, reporting strength and function alongside mass, showing benefit over training alone. A mass-only result would not be sufficient — §16.6 explains why that is now the standard this field has to meet.

16.9 Cancer risk and the growth-signaling dilemma

This section has to be careful in two directions at once, because the alarmist and the dismissive versions are both wrong and both common.

The mechanistic concern, stated accurately

IGF-1 signaling promotes cell growth and inhibits apoptosis — the two things a cancer cell benefits from most. A cell with an oncogenic mutation still has to survive, proliferate, and escape the programmed death that removes damaged cells; an environment rich in growth-and-survival signal makes each easier. The IGF-1 receptor is expressed on many tumor types, and an entire generation of IGF-1R inhibitors was developed as anticancer agents on exactly this reasoning. (That those drugs largely disappointed is itself a lesson about mechanism.) The concern is not manufactured by scaremongers; it is reasoning oncology took seriously enough to spend a decade on.

The epidemiology, stated precisely

Higher circulating IGF-1 has been associated with increased risk of several cancers in observational studies. The most consistent signals across prospective cohorts and their meta-analyses have involved colorectal, prostate, and breast cancer. The associations are generally modest and are found comparing people across the normal physiological range — not comparing people on IGF-1 injections to people who are not, because that study does not exist.

Now the part that has to be said in the same breath. Association is not causation; confounding is substantial; no causal claim is established. Circulating IGF-1 tracks with nutritional status, protein intake, body size and height, GH status, liver function, genetics, and age, any of which could drive both the level and the risk. Genetic-instrument approaches — using inherited variants that influence IGF-1 levels to reduce confounding — have offered some support for a causal contribution in certain cancers, but the field does not treat the question as settled and estimates remain imprecise.

So the honest summary has three parts: the mechanism is real, the association is real, and the causal claim is not established.

The two natural experiments

Two human populations sit at the extremes of this axis. Both are informative; neither is decisive.

At the high end: acromegaly, covered in detail in Chapter 14 and the relevant natural experiment here. Sustained pathological elevation of GH and IGF-1 in adults produces a recognizable syndrome — soft tissue and skeletal overgrowth, cardiac hypertrophy, insulin resistance, joint destruction, sleep apnea — along with increased risk of colorectal neoplasia and, historically, increased mortality that improved once treatment became effective. It is a far higher and longer exposure than anything discussed here, so it does not predict what a lower exposure does. It does demonstrate that the axis has a dose-response with a bad end.

At the low end: severe GH receptor deficiency. A cohort with an inherited defect in GH receptor signaling, and consequently very low lifetime IGF-1, has been followed for decades and reported to show strikingly low incidence of cancer and diabetes relative to unaffected relatives. Small, genetically distinctive, with very short stature and other health consequences — suggestive, not recommendation-generating. But it points the same direction as the epidemiology from the opposite end, and consilience across independent lines of weak evidence is worth something.

🩺 Safety and Risk — chronically elevating a growth axis in a healthy person

Here is the risk conversation, laid out as an accounting problem rather than a warning.

What is known: IGF-1 promotes growth and inhibits apoptosis. Higher levels across the normal range are associated with modestly increased incidence of several cancers. Very high sustained levels (acromegaly) carry clear multi-system harm. Very low lifetime levels are associated with unusually low cancer incidence. Acutely, exogenous IGF-1 carries a real hypoglycemia risk.

What is not known: whether raising IGF-1 exogenously, above one's own normal, for months or years, in an otherwise healthy adult, increases cancer risk — and if so by how much. Nobody has measured it, and nobody is going to; the trial would require thousands of healthy volunteers followed for decades, and no ethics committee would approve it for a physique indication.

What that gap means. Not that the risk is small. Not that it is large. That the number does not exist, and a person elevating this axis is deciding under irreducible uncertainty — with consequences, if any, that arrive years later and are individually unattributable, because you cannot tell whether a cancer at fifty-five was caused by anything.

The asymmetry that decides it. Risk is only ever evaluated against benefit. For a child with severe primary IGF-1 deficiency, an unquantified long-term risk is accepted for a definite, life-altering benefit, under supervision, with monitoring. For a healthy adult seeking a physique change, the same unquantified risk is accepted for a benefit that has never been demonstrated in a human trial. The uncertainty is identical; the ledger is not remotely the same.

Add the nearer-term items, which get less attention than cancer and are more likely: hypoglycemia; the acromegalic changes of Chapter 14 if the axis is elevated enough for long enough; cardiac tissue that also responds to growth signaling; joint and tendon complaints; and everything in §16.4 about what is in an unregulated vial. As always, this belongs in a conversation with a clinician who knows your history and can order labs — not with a book, and not with a vendor.

The dilemma, named

There is a genuine dilemma here and it should not be smoothed over. Growth signaling is not a bug — it is how tissue is built and maintained, and it declines with age in ways plausibly connected to frailty. Longevity research points toward lower growth signaling extending lifespan in model organisms; clinical experience points toward higher anabolic capacity preserving function in aging humans. Reconciling those is an open scientific problem.

So: an axis you probably do not want maximally high across a lifetime and probably do not want minimally low either, with no established way to locate the right place on that curve for an individual, and no drug that reliably improves function by moving anyone along it. Whoever tells you the answer is simply "more" has skipped the whole literature.

📋 Your Evidence Dossier

Field 6 — What the Evidence Actually Measured

Field 6 is the endpoint field, and this chapter is the reason it exists. Previous fields established what a molecule is, what it is claimed to do, and what studies have been run. Field 6 asks a narrower, more discriminating question: did the supporting evidence measure the thing you care about, or something correlated with it? It is the field that separates a claim you can rely on from a claim that merely has studies behind it.

FIELD 6 — WHAT THE EVIDENCE ACTUALLY MEASURED

  Endpoint(s) reported      what the study's headline result is a measurement OF
  Surrogate or outcome?     is it a stand-in, or the thing that matters to a person?
  If surrogate: validated?  is there evidence that changing THIS reliably changes
                            the outcome, for THIS intervention, in THIS population?
  Was the outcome measured  yes / no / measured but not reported prominently
    too?
  Direction check           did surrogate and outcome move together, or apart?
  Rating consequence        which of the three cases below applies

The three cases, and what each one does to a rating

Case A — surrogate improved, outcome never measured. The commonest case in this book. You know the intervention does something; you do not know whether that something matters. The outcome claim cannot rise above ⚠️, and asserted confidently it is usually ❌, because the confident version claims more than the study can support. Most of Chapter 15's secretagogue material sits here.

Case B — surrogate improved, outcome measured, outcome improved. The arrow held. This is what a successful development program looks like, and it is what licenses more confidence in the surrogate next time. Ratings can rise to ✅ here, given adequate size and replication.

Case C — surrogate improved, outcome measured, outcome did NOT improve. The arrow broke. This is worse than Case A, and the friendlier-looking symbol is the trap. The surrogate is now positively discredited for that outcome in that setting, and any future claim relying on it inherits a debt it must pay before it is taken seriously. That case is not hypothetical. It is §16.5.

Worked demonstration — two entries, deliberately contrasted

FIELD 6 — MYOSTATIN-PATHWAY INHIBITORS, muscle disease    [worked demonstration]

  Endpoint(s) reported      Lean body mass (DXA); muscle volume / cross-sectional
                            area (MRI). Functional endpoints also collected: timed
                            function tests, six-minute walk distance, ambulatory
                            assessment scales, measured strength.
  Surrogate or outcome?     Mass endpoints = SURROGATE. Function endpoints = OUTCOME.
  If surrogate: validated?  NO. Lean mass is not a validated surrogate for physical
                            function in muscle-wasting disease. These trials are
                            effectively the validation attempt.
  Was the outcome measured  YES — and this is the crucial entry. Function was
    too?                    measured, across multiple programs and sponsors.
  Direction check           APART. Mass rose; function did not reliably follow.
  Rating consequence        CASE C. Disease claim: ⚠️ — real human data, engaged the
                            target, disappointed on outcome, programs discontinued.
                            Performance claim in healthy adults: ❌ — no human data
                            AND an adjacent human result pointing away from it.

FIELD 6 — A GH SECRETAGOGUE "BODY COMPOSITION" CLAIM      [worked demonstration]
                                                          [see Chapter 15]
  Endpoint(s) reported      Serum GH pulse amplitude; serum IGF-1 concentration.
                            Sometimes lean body mass. Rarely anything else.
  Surrogate or outcome?     ALL SURROGATE. GH and IGF-1 levels sit two rungs below
                            even lean mass, which is itself a surrogate (§16.6). The
                            claim being sold is about strength, appearance, recovery,
                            and vitality — none of which appear here.
  If surrogate: validated?  NO, doubly. "Raises GH/IGF-1" has not been shown to
                            reliably produce functional benefit in healthy adults
                            (Ch 14, 15), and lean mass has not been shown to
                            reliably produce function (this chapter).
  Was the outcome measured  Largely NO. Where measured, effects on strength and
    too?                    function in healthy adults have been unimpressive.
  Direction check           Cannot be performed — the outcome is mostly absent.
  Rating consequence        CASE A. Mechanism confirmed, claim unsupported: ❌.

  THE CONTRAST: both entries end near ❌ for the performance claim, by different
  routes. The first failed a test; the second was never given one. Say WHICH — only
  one of those can be fixed by running a study.

Do this now, for your own list

For every peptide in your dossier, go back to the studies you recorded and answer one question in writing: what was actually measured? Not what the abstract concluded, not what the press release said — what number the researchers wrote down. Then ask whether that number is the thing you care about or a stand-in for it, and whether anyone ever checked that the stand-in stands in.

For a substantial fraction of your entries you will find the outcome you care about was never measured by anyone. That is not a failure of your research. It is the finding.

Conclusion

Muscle grows when protein accretion outpaces breakdown, driven principally by mechanical loading, with satellite cells supplying nuclei. Resistance training is the dominant intervention and no peptide substitutes for it; everything here is studied as an adjunct or as a therapy for disease.

IGF-1 mediates much of GH's anabolic effect, travels under tight control by a family of binding proteins, and exists as an approved drug — mecasermin — for one narrow rare pediatric indication. That approval proves the mechanism and licenses nothing beyond itself. IGF-1 LR3, engineered specifically to escape those binding proteins, is unapproved, known by a code rather than a generic name, and has no completed randomized human outcome trials for muscle or performance.

Myostatin is a genuine brake, and the natural knockouts — mice, cattle breeds, a racing dog breed, a documented human case — are real and striking. But a developmental knockout is not adult blockade, and the animals came with costs the photographs do not show, including reduced force per unit of muscle.

And then the trials happened. Multiple sponsors, multiple molecules, real randomized human studies in muscular dystrophy and related conditions. Lean mass rose. Function did not reliably follow. Programs were discontinued. The pharmacology worked and the therapy failed, at the one link nobody had thought to examine. That link is the whole point: mass, strength, function, and independence are four different things, and the arrows between them are hypotheses requiring separate evidence. Chapter 8 gave you the same principle in reverse. Mass is not function. Neither direction of that dissociation is intuitive, and both were discovered only by measuring.

Sarcopenia is where a working drug would matter most, there is not one, and the field moved its own diagnostic definition away from mass toward strength and performance because mass did not predict what mattered. In the performance market the honest position is a ❌ of unusual character: not merely absent evidence, but a closely related body of human evidence pointing away from the claim. And the growth-signaling question stays open in both directions — the risk of chronically elevating this axis in a healthy person is unquantified, which is not the same as small.

Chapters 17 and 18 turn to the repair peptides — BPC-157 and TB-500 — where the evidence base is thinner and the enthusiasm greater, and where you will need everything Field 6 just taught you.

Key Terms

Insulin-like growth factor 1 (IGF-1) — a 70-amino-acid polypeptide produced largely in the liver in response to growth hormone, mediating many of GH's anabolic effects; structurally related to proinsulin.

IGF binding protein (IGFBP) — one of six proteins that bind circulating IGF-1, extending its half-life, restricting its distribution, and keeping the great majority of it inactive until released.

Mecasermin — recombinant human IGF-1, approved for severe primary IGF-1 deficiency in children. Its label carries a prominent hypoglycemia warning.

Severe primary IGF-1 deficiency — a rare condition in which IGF-1 production fails despite normal or elevated GH, usually from a defect in GH receptor signaling; mecasermin's approved indication.

IGF-1 LR3 — a modified IGF-1 variant with an N-terminal extension and a position-3 substitution, giving reduced binding-protein affinity and extended activity. Originally a cell-culture reagent; sold on the gray market; not approved anywhere; known by a code rather than a generic name.

Myostatin (GDF-8) — a TGF-β superfamily member produced by skeletal muscle that negatively regulates muscle growth. Loss of function produces marked hypertrophy in mice, cattle, dogs, and rare human cases.

Activin receptor type IIB (ActRIIB) — the type II receptor through which myostatin signals, shared with activin A and GDF-11, which is why broad blockade there is potent and unselective.

Follistatin — an endogenous glycoprotein, roughly 35–40 kDa, that binds and inhibits myostatin, activin A, GDF-11, and some bone morphogenetic proteins.

Follistatin-344 — a gray-market compound named for the 344-residue follistatin precursor isoform. Not approved anywhere. Its size and glycosylation make product identity a serious open question.

Double muscling — the extreme muscularity of certain cattle breeds, notably Belgian Blue and Piedmontese, from naturally occurring loss-of-function mutations in the myostatin gene.

Hypertrophy / hyperplasia — growth of existing muscle fibers versus an increase in their number. Hypertrophy is the dominant mode in adult humans; hyperplasia is largely developmental and a substantial contributor to double-muscled animal phenotypes.

Satellite cell — a resident muscle stem cell that activates in response to loading and damage and fuses into existing fibers, supplying nuclei. The cytoplasmic volume each nucleus governs is its myonuclear domain.

Net protein balance — the running difference between muscle protein synthesis and breakdown.

Specific force — force per unit of muscle cross-sectional area; muscle quality rather than quantity. Reported to fall in myostatin-null animals.

Lean mass — fat-free soft tissue, commonly measured by DXA. Includes muscle but also water, glycogen, and connective tissue, which is why it is an imperfect proxy for contractile tissue.

Surrogate endpoint — a measurement used as a stand-in for the outcome that matters, assuming changing it changes the outcome. Legitimate only when that assumption has been validated for the specific intervention and population.

Functional endpoint — a measurement of what a person can actually do: gait speed, chair rise, stair climb, six-minute walk distance, timed function tests.

Sarcopenia — age-related loss of muscle mass and function. Consensus definitions now lead with low strength, use low mass as confirmatory, and grade severity by physical performance.

Laron syndrome (GH receptor deficiency) — an inherited defect in GH receptor signaling producing very low lifetime IGF-1 and severe short stature; the low end of the axis.

Acromegaly — sustained pathological GH and IGF-1 excess in adults; the natural experiment at the high end of the axis, covered in Chapter 14.

Spaced Review

1. (Ch 14, 15, 16) Chapter 14 established that "raises GH" is a surrogate claim; Chapter 15 applied it to the secretagogues; this chapter establishes that "raises lean mass" is also one. Arrange GH level, IGF-1 level, lean mass, strength, and independent living into a chain, and state how many separate hypotheses a marketer skips when advertising a secretagogue with an IGF-1 number and a photograph.

2. (Ch 16) Myostatin inhibitors for muscle disease got ⚠️ while IGF-1 LR3 for performance got ❌, even though the myostatin trials failed and the IGF-1 trials were never run. Explain why the ratings are correct as written, then why the ⚠️ claim is nonetheless in a worse position.

3. (Ch 5, 16) Rule 3 says never upgrade a rating with mechanism. Apply it to the Belgian Blue: one sentence on what the cattle genuinely establish, one on the upgrade someone would be tempted to make. Then name which of Chapter 5's six rules the tempting version violates.

4. (Ch 8, 16) In one paragraph, explain to someone who has read neither chapter why lean mass falling on a GLP-1 drug while function improves, and lean mass rising on a myostatin inhibitor while function does not, are the same finding rather than two contradictory ones.

5. (Ch 14, 16) A clinic advertises "IGF-1 optimization" for adults over fifty, citing mecasermin's approval as proof that IGF-1 therapy is established medicine. Identify three distinct errors, drawing on Chapter 14's replacement-versus-enhancement distinction, §16.2, and §16.9.