Chapter 8: Semaglutide: The Peptide That Changed Medicine — From GLP-1 to Ozempic to Wegovy

"We have a drug that works. That is the beginning of the interesting questions, not the end of them." — the posture this chapter tries to hold

Overview

By any reasonable measure, semaglutide is the most consequential metabolic drug of the last fifty years. It produces weight loss of a magnitude that pharmacology had failed to achieve for a century. It reduces cardiovascular events in people who do not have diabetes. It has generated more public conversation than any drug since the statins, and probably more than those.

This chapter takes it seriously in both directions, which is harder than it sounds.

Taking it seriously means saying plainly that the evidence is excellent. Multiple large randomized trials, hard endpoints, years of follow-up, regulatory review in many jurisdictions, and millions of patient-years of post-marketing experience. When this book gives semaglutide a ✅, it is not being generous — it is describing one of the best-evidenced drugs in this book.

It also means saying plainly what the trials did not establish, what the headlines dropped, what the adverse effect profile actually looks like, and what happens when people stop. That last item is the most consequential finding in the chapter, and it is the one that reframes the drug from a treatment you complete into a therapy you continue.

And it means resisting two available narratives. The first is the miracle version, in which a transformative drug has arrived and the remaining questions are details. The second is the backlash version, in which the drug is overhyped, the side effects are being concealed, and people are being sold a shortcut. Both are wrong, and Chapter 6 explained why both are available.

What this chapter gives you instead is the evidence, with its populations attached, its numbers in both relative and absolute terms, its estimands named, and its gaps stated.

In this chapter, you will learn to:

• Trace the three modifications from native GLP-1, and say what each solved
• Distinguish SUSTAIN, STEP, PIONEER, and SELECT, and what each established
• State the weight-loss figures with their populations — and explain why STEP 1 and STEP 2 differ
• Interpret SELECT in relative and absolute terms
• Describe the adverse effects honestly, including the ones that are commonly overstated
• Explain the muscle loss question and what is actually known
• Explain discontinuation and why it changes what the drug is for
• Write a complete dossier entry

Learning Paths

💊 GLP-1 — the chapter you came for. Read all of it. §8.9 is the section most likely to change a decision. 🏋️ Performance — §8.8 (muscle loss) is your section and the evidence is more nuanced than either side of that argument usually admits. 🔬 Science — full read; §8.5 and §8.7 are the technical core. 🏥 Clinical — §8.3 (population differences), §8.7 (adverse effects), and §8.9 (discontinuation) are the three sections your patients most need explained and most often have wrong.

8.1 The engineering, residue by residue

Chapter 4 gave this in full. A compressed restatement, because everything else depends on it.

Native GLP-1: 30 amino acids, half-life one to two minutes, destroyed by DPP-4 and cleared renally. Excellent hormone, useless drug.

THREE MODIFICATIONS

  ① position 8    Ala → Aib (a non-natural amino acid)
                  → abolishes the DPP-4 cut site        [defeats proteolysis]

  ② position 34   Lys → Arg
                  → leaves ONE lysine as the fatty-acid attachment site, so the
                    conjugation reaction yields a single defined product
                                                        [manufacturing control only]

  ③ position 26   C18 diacid fatty acid, attached via a spacer to that lysine
                  → binds albumin (~66,000 Da); the complex cannot be filtered
                    by the kidney and is shielded from proteases
                                                        [defeats renal clearance]

  half-life ~1–2 minutes → ~1 week.    MW ~4,100 Da.    Receptor activity: unchanged.

The receptor pharmacology was never improved. Semaglutide is not a better GLP-1 receptor agonist than GLP-1. It is the same message, delivered for a week instead of ninety seconds. Everything in this chapter follows from that difference in duration, not from a difference in what the molecule says.

8.2 Ozempic and the SUSTAIN program

Semaglutide was first approved in 2017, injectable, for type 2 diabetes, under the brand name Ozempic.

The SUSTAIN program was the development program for that indication — a series of randomized trials examining glycemic control against placebo and against active comparators including other GLP-1 receptor agonists and other diabetes drug classes.

What it established:

Substantial A1C reduction. A1C — glycated hemoglobin — reflects average blood glucose over roughly the preceding three months and is the standard endpoint in diabetes trials. Semaglutide produced reductions that compared favorably with existing agents, including other drugs in its own class.

Weight loss as a secondary finding. Participants lost weight — more than on comparator diabetes drugs. In a diabetes trial this is a welcome secondary outcome. It was also, in retrospect, the signal that redirected the entire program.

Cardiovascular safety, and more. SUSTAIN 6 was a cardiovascular outcomes trial in people with type 2 diabetes at high cardiovascular risk, run over roughly two years. Regulators require cardiovascular outcome trials for new diabetes drugs — a requirement introduced after an earlier diabetes drug raised cardiovascular concerns — and the normal expectation is that a drug demonstrates it does not increase risk.

SUSTAIN 6 showed a statistically significant reduction in major adverse cardiovascular events. Not merely safety: benefit. In a trial designed to rule out harm.

That result is why Chapter 10 exists. A drug developed to lower blood glucose reduced heart attacks and strokes, and nobody could say confidently why.

💊 In the Clinic — what A1C is and what it hides

A1C measures the fraction of hemoglobin that has glucose attached, which accumulates in proportion to average glucose exposure over roughly three months. It is the standard endpoint because it integrates over time and does not depend on when the blood was drawn.

What it does not capture: variability. Two people with identical A1C can have very different glucose patterns — one stable, one swinging between highs and lows. There is reason to think variability matters independently, and A1C is silent about it. Continuous glucose monitoring has made this visible in a way it was not before.

And it is a surrogate. Chapter 5's §5.6 applies: A1C stands in for the complications of diabetes — kidney disease, retinopathy, neuropathy, cardiovascular events. The arrow from A1C to those outcomes is better supported than most surrogates in this book, particularly for microvascular complications, but it is not perfect, and drugs have lowered A1C without delivering the expected outcome benefit.

This is why cardiovascular outcome trials became mandatory, and why SUSTAIN 6 and SELECT matter far more than any A1C result.

8.3 Wegovy and the STEP program — and the population lesson

At the higher dose of 2.4 mg weekly, semaglutide was developed specifically for weight management and approved in 2021 under the brand name Wegovy. Same molecule as Ozempic. Different dose, different label, different approved indication.

That distinction confuses nearly everyone and matters enormously for insurance coverage (Chapter 12).

The STEP program established the weight effect. Two of its trials illustrate Chapter 5's most important trap.

🔬 Read the Study — STEP 1 and STEP 2 side by side

text FIGURE 8.3 — "The same drug, two populations, two answers" [real published trials] THE STUDY STEP 1: randomized, double-blind, placebo-controlled. Adults with overweight or obesity and WITHOUT diabetes. Semaglutide 2.4 mg weekly vs placebo, 68 weeks. Both arms received lifestyle intervention. STEP 2: the same dose and duration in adults WITH type 2 diabetes. Manufacturer-sponsored. THE QUESTION How much weight does semaglutide 2.4 mg produce over 68 weeks, and does it differ by population? WHAT IT SHOWS STEP 1: mean weight change of about −15% from baseline, versus roughly −2.4% on placebo. STEP 2: roughly −10% from baseline in adults with type 2 diabetes. Both are large. They are not the same, and the difference is systematic. WHAT IT DOESN'T Neither tells you about people who are not overweight, about use without lifestyle support, about durations beyond 68 weeks, or about what happens on stopping (see 8.9). Neither was blinded perfectly in practice — gastrointestinal effects allow participants to guess assignment. THE VERDICT ✅ for the weight claim in the studied populations. THE LESSON QUOTING A TRIAL NUMBER WITHOUT ITS POPULATION IS THE MOST COMMON ERROR IN COVERAGE OF THIS DRUG. "15%" is a STEP 1 figure. A person with type 2 diabetes should expect the STEP 2 figure. They are different questions with different answers.

Why the difference? Several contributing explanations, none fully settled: people with type 2 diabetes may have altered incretin responsiveness (Chapter 7 §7.1); concurrent diabetes medications including insulin and sulfonylureas promote weight gain; and there may be differences in the metabolic state itself. The honest position is that the difference is reproducible and its causes are partly understood.

Two more things the STEP program established:

The effect continues for as long as the trials ran. Weight loss did not plateau immediately and then reverse; it continued and then largely stabilized over the trial period. Longer extension studies have supported maintenance while treatment continues.

And both arms received intensive lifestyle support. Chapter 5 §5.5 covered this and it bears repeating: the placebo arm lost roughly 2.4%, which is more than an untreated population would. The trial result describes drug-plus-support versus support alone. That is not the comparison most readers have in mind, and it cuts in both directions.

8.4 Rybelsus and PIONEER: an oral peptide

Oral semaglutide was approved in 2019 as Rybelsus, and Chapter 4 covered how: a permeation enhancer (SNAC) that transiently promotes absorption across a small region of the stomach lining, plus a strict administration protocol.

Bioavailability is roughly 1%.

The PIONEER program established that this worked — that an orally administered peptide could produce clinically meaningful glycemic control.

What it demonstrates and what it costs:

Demonstrates: oral peptide delivery is possible. This is a genuine pharmaceutical achievement, and Chapter 36 treats it as the opening of a route rather than a one-off.

Costs: the tablet contains far more drug than the injection; the administration requirements are strict and easy to get wrong; and adherence to a fasting protocol in real life is a different proposition from adherence in a trial.

The general lesson: low bioavailability is a cost and variability problem, not automatically a viability problem — but the variability is where the real risk sits, and a protocol that is followed imperfectly delivers an uncertain dose.

8.5 SELECT: the result that changed what the drug is

SELECT is the most consequential trial in this chapter, and Chapter 5 used it to teach the relative-versus-absolute distinction. Here it is in full.

🔬 Read the Study — SELECT

text FIGURE 8.5 — "Twenty percent, or one and a half points" [real published trial] THE STUDY Randomized, double-blind, placebo-controlled cardiovascular outcomes trial. Semaglutide 2.4 mg weekly vs placebo in adults with established cardiovascular disease AND overweight or obesity, but WITHOUT diabetes. Roughly 17,000 participants, followed about three years. Manufacturer-sponsored. THE QUESTION Does semaglutide reduce major adverse cardiovascular events (MACE: cardiovascular death, non-fatal myocardial infarction, non-fatal stroke) in this population? WHAT IT SHOWS Yes. A 20% RELATIVE reduction — hazard ratio about 0.80. In ABSOLUTE terms, roughly 8% to roughly 6.5% over about three years: about 1.5 percentage points. A hard endpoint, a large population, long follow-up, pre-specified. WHAT IT DOESN'T It does not establish benefit in people WITHOUT established cardiovascular disease, in people with diabetes (a different trial), or beyond three years. It does not isolate the mechanism — whether benefit comes from weight loss, glucose effects, blood pressure, inflammation, or direct vascular action is NOT established by this trial. THE VERDICT ✅ for this claim, in this population. THE LESSON Both numbers, always, in the same sentence. And note what the trial changed: semaglutide stopped being a metabolic drug with a cardiovascular signal and became a cardiovascular drug that also produces weight loss. That is a different drug, clinically and commercially.

Why SELECT reorganized the field.

Before it, semaglutide's case in obesity rested on weight loss — a surrogate for the outcomes weight loss is supposed to prevent. Chapter 5 §5.6 explains why that is a real limitation: weight is a well-supported surrogate but not a perfect one, and obesity drugs have a history of improving weight and harming patients.

SELECT replaced the surrogate with a hard endpoint, in a population without diabetes. That moved the drug from "reduces a risk factor" to "reduces events," which is a categorically stronger claim and is why coverage decisions and guideline positions shifted after it.

And note what it did not do. It did not explain the mechanism, did not extend to primary prevention (people without established cardiovascular disease), and did not run beyond about three years. Chapter 10 takes up the mechanism question and does not resolve it either.

8.6 What the trials measured that the headlines dropped

Four things reported in the trials and largely absent from coverage.

Blood pressure fell. Modestly, consistently, in both arms' favor. This matters mechanistically (Chapter 10) and clinically.

Lipids improved. Modest but consistent improvements in several lipid measures.

Inflammatory markers fell. C-reactive protein and related measures declined. This is one of the threads that makes an inflammation-mediated explanation of the cardiovascular benefit plausible — and Chapter 5 §5.6 warns that inflammatory markers are among the least reliable surrogates in medicine, so this is a hypothesis-generating observation rather than an explanation.

Quality of life and physical function improved on validated instruments in several trials. Widely underreported, and arguably more meaningful to patients than the weight figure.

And one thing that is reported and consistently misread: the trials report discontinuation rates, including discontinuation for adverse events. A meaningful minority of participants stopped. Chapter 5 §5.9's estimand discussion is exactly this — the people who stopped are counted in the treatment-regimen analysis and not in the efficacy analysis, and that is why the two figures differ.

8.7 Adverse effects, honestly

This section is written to be neither reassuring nor alarming. Both would be a failure.

The common ones

Gastrointestinal effects are the dominant adverse effect: nausea, vomiting, diarrhea, constipation, abdominal pain. They occur in a substantial minority of patients — commonly reported in the range of a third to a half depending on the effect and the trial — are usually mild to moderate, and typically attenuate over weeks.

They are the mechanism, overshooting. Chapter 7 action ③: slowed gastric emptying, plus hindbrain receptor activation in a region that is also the vomiting trigger zone. This is why titration exists — starting low and increasing stepwise allows accommodation. It is also why the effects fade while the metabolic effects persist: tolerance is a property of drug-tissue pairs (Chapter 2 §2.8).

The less common ones that matter

Gallbladder disease. Increased rates of gallstones and cholecystitis. Rapid weight loss increases gallstone risk from any cause, and it is not fully settled how much of the signal is the drug and how much is the weight loss. Either way, it is real and is on the label.

Pancreatitis. A concern from early in the class's history. The signal has been examined extensively and is less clear-cut than the early alarm suggested — large trials and observational studies have not established a substantial increased risk, though it remains labeled and monitored. This is an instructive case of a safety concern that was appropriately raised and then not confirmed at the magnitude feared.

Gastroparesis. Severely delayed gastric emptying. Given that slowing gastric emptying is the mechanism, this is the mechanism at its extreme rather than an unrelated event. Rare but genuinely disabling when it occurs.

Aspiration under anesthesia. A practical issue that emerged from post-marketing experience: a stomach that empties slowly may contain food when it is assumed to be empty, which matters for anesthesia. Professional societies have issued guidance on managing this. This is a good example of something Phase III could not have detected — it requires the drug to be in wide use among people undergoing procedures.

The rodent thyroid signal

This one requires care because it is both real and routinely misstated.

What was observed: in rodent studies, GLP-1 receptor agonists produced C-cell tumors of the thyroid, including medullary thyroid carcinoma. This is a class effect in rodents and it is why the labels carry a boxed warning and a contraindication in people with a personal or family history of medullary thyroid carcinoma or MEN2 (multiple endocrine neoplasia type 2).

What is not established: that this translates to humans. Rodent thyroid C-cells express GLP-1 receptors far more abundantly than human ones do — a species difference that provides a plausible reason the finding may not transfer. Human data has not demonstrated the increase that the rodent finding would predict, though the event is rare enough that detecting a modest increase is genuinely difficult.

⚠️ Hype Check — the thyroid warning, in both directions

The alarmist version: "There's a black box warning for thyroid cancer." True, and it describes a rodent finding with a species-difference explanation and no demonstrated human increase. Stated without that context, it is misleading.

The dismissive version: "That's just a rat thing, it doesn't apply to humans." Also overreaching. "Not demonstrated in humans" is not "shown not to occur in humans," and Chapter 5's distinction applies with full force. Medullary thyroid carcinoma is rare enough that a modest relative increase would be hard to detect even in very large datasets.

The honest version: a real rodent finding, a plausible species-difference reason it may not transfer, no demonstrated human signal, a contraindication in the specific high-risk groups where the theoretical concern is concentrated, and continued monitoring. The contraindication is not theatre — for someone with MEN2 or a family history of medullary thyroid carcinoma, it is the operative fact.

This is the book's best example of a safety signal that both sides of a public argument have incentive to misstate, and it is worth practicing the accurate version.

8.8 Muscle loss and "Ozempic face"

Two related concerns that circulate widely and deserve a careful answer.

The muscle question. Weight loss of any kind — by any method, drug or diet or surgery — comprises both fat and lean mass. This is not specific to GLP-1 agonists; it is a property of losing weight. Studies using body composition measurement have found that a meaningful proportion of weight lost on these drugs is lean mass, in a range broadly comparable to what other weight-loss interventions produce.

What is genuinely uncertain and matters:

Lean mass is not the same as muscle function. Chapter 5 §5.6 and Chapter 16 both turn on this: lean mass is a surrogate, and strength and function are the outcomes. Whether the lean mass lost translates into meaningful functional decline is not well established, and the trials that measured physical function generally found it improved rather than worsened — which is what you would expect if the functional benefit of losing fat mass outweighs the cost of losing some lean mass.

Populations differ. In an older adult with low baseline muscle mass, lean mass loss is a substantially greater concern than in a younger person with ample reserve. "Sarcopenic obesity" is a real clinical entity and is exactly the population where this question is most pressing and least studied.

And the interventions that would address it are not mysterious. Adequate protein intake and resistance training are the standard recommendations for preserving lean mass during weight loss, from any cause. Whether they fully offset the effect on these drugs specifically is an active research question. What an individual should do about it belongs with their clinician, and this is one of the clearest cases in the book where that is not a hedge — the answer genuinely depends on the person.

"Ozempic face." Facial volume loss with substantial weight reduction. It is real, it is not specific to the drug — the same appearance follows comparable weight loss by any means — and it reflects loss of facial fat pads. That it acquired a drug-specific name is a cultural artifact rather than a pharmacological finding.

8.9 Discontinuation: the finding that reframes the drug

This is the most important section in the chapter.

Trials that followed participants after stopping semaglutide found that weight returns. Not entirely, not instantly, but substantially — with a meaningful proportion of lost weight regained over the following year, and cardiometabolic improvements moving back toward baseline alongside it.

This surprised the public and did not surprise anyone who had read Chapters 2 and 3.

WHY WEIGHT RETURNS — the override-versus-replace rule, applied

  INSULIN in type 1 diabetes          SEMAGLUTIDE for obesity
  ─────────────────────────           ────────────────────────
  REPLACES an absent signal            OVERRIDES an intact system
  (beta cells destroyed)               (appetite regulation is working —
                                        it is defending a set point)
        │                                    │
        ▼                                    ▼
  no counter-regulation provoked       the system was never removed;
  effect persists indefinitely         it was outweighed
        │                                    │
        ▼                                    ▼
  stopping = the deficiency returns    stopping = the override is removed
                                       and the intact system resumes

  Chapter 2 §2.8 predicted this. Chapter 13 shows a long history of
  appetite drugs defeated by exactly this mechanism.

Three consequences follow, and they are the practical content of this chapter.

First, this is chronic therapy. Semaglutide for obesity is not a course of treatment that ends. It is comparable to an antihypertensive or a statin — a drug that manages a condition for as long as it is taken. That reframing has consequences for cost (Chapter 12), for the decision to start, and for how the drug should be described.

Second, it is not a failure of the drug. A blood pressure medication whose effect ends on discontinuation is not thereby a bad drug. Nobody says statins "don't work" because cholesterol rises when you stop. The expectation that a weight drug should produce permanent change after stopping is a expectation applied to no other chronic therapy, and it comes from the moral framing of weight rather than from pharmacology.

Third, it complicates the access question enormously. A drug that must be continued indefinitely, at high cost, with inconsistent insurance coverage, in a population where discontinuation is common for reasons of cost rather than tolerance, produces a specific and predictable pattern of outcomes. Chapter 12.

📊 Evidence Rating — semaglutide for weight loss in adults with obesity

Claim: Semaglutide 2.4 mg weekly produces substantial, sustained weight loss in adults with obesity, or overweight with a weight-related comorbidity, alongside lifestyle support, for as long as treatment continues.

Rating: ✅ Strong clinical evidence.

Why: Multiple large randomized, double-blind, placebo-controlled trials (STEP). About −15% from baseline at 68 weeks versus roughly −2.4% on placebo in adults with overweight or obesity without diabetes; roughly −10% in adults with type 2 diabetes. Consistent across trials, approved in multiple jurisdictions, with a characterized adverse effect profile.

What would change it: long-term data showing the effect does not persist during continued treatment, or a serious harm emerging with multi-year use at a rate outweighing benefit.

What does NOT change it: weight regain after discontinuation. That is expected physiology and it changes what the drug is for, not whether it works.

📊 Evidence Rating — semaglutide for glycemic control in type 2 diabetes

Claim: Semaglutide improves glycemic control in adults with type 2 diabetes.

Rating: ✅ Strong clinical evidence.

Why: The SUSTAIN program, including active-comparator trials; substantial A1C reduction; approved for this indication since 2017; low hypoglycemia risk when used without insulin or a sulfonylurea, consistent with glucose-dependence (Chapter 7 §7.4).

What would change it: essentially nothing at this point for the A1C claim. Note that A1C is a surrogate — but the outcome question is answered separately and favorably by SUSTAIN 6.

📊 Evidence Rating — semaglutide for cardiovascular risk reduction

Claim: Semaglutide 2.4 mg reduces major adverse cardiovascular events in adults with established cardiovascular disease and overweight or obesity, without diabetes.

Rating: ✅ Strong clinical evidence — for this population.

Why: SELECT: roughly 17,000 participants, about three years, pre-specified hard endpoint, a 20% relative reduction (hazard ratio about 0.80), roughly 8% to 6.5% absolute — about 1.5 percentage points. SUSTAIN 6 provides supporting evidence in type 2 diabetes.

What would change it: a failure to replicate, or long-term follow-up reversing the finding.

What this rating does NOT cover: primary prevention — people without established cardiovascular disease. That population was not studied and the rating does not extend to it.

Three ratings, one molecule. This is Chapter 5's rule 6 in action, and Chapter 10 adds several more — including two that are 🔬 rather than ✅.

8.10 What is still unknown

A chapter this favorable owes the reader an explicit list of what has not been established. These are not quibbles; several would change clinical practice if they were answered.

Long-term safety. SELECT ran about three years. STEP ran 68 weeks with extensions. These are drugs intended for decades of continuous use in very large populations. Three years is genuinely reassuring about the harms that appear in three years, and Chapter 3's cadaver growth hormone case is the standing reminder that some harms have incubation periods measured in decades. Nothing here predicts a problem. It states the limit of what has been observed.

Effects in populations the trials under-enrolled. Older adults with low baseline muscle mass, people with significant renal impairment, people with eating disorder histories, adolescents, and people who are not overweight are all under-represented or excluded. Off-label use has run well ahead of the evidence in several of these directions, and Chapter 12 takes up why.

Whether the cardiovascular benefit extends beyond the studied population. Not rated here, for the arithmetic reason given above. A primary-prevention outcomes trial would settle it and would have to be substantially larger and longer than SELECT.

The mechanism of the non-glycemic benefits. Chapter 8's Case Study 1 lists six candidates and settles none. This limits extrapolation more than it appears to: if the benefit is mediated by weight loss, it should be reproducible by other means of losing weight; if by a direct receptor effect, it would be specific to the class. Nobody can currently say which.

Bone. Rapid weight loss affects bone density, and the long-term skeletal consequences of sustained pharmacological weight reduction — particularly in older adults and postmenopausal women — are not well characterized.

Pregnancy. These drugs are not recommended in pregnancy, and there is limited human data. The practical issue is that improved metabolic health and reduced weight can restore fertility in people who were not expecting it, which is a real clinical scenario and a genuine counseling requirement.

Psychological and behavioral effects. The "food noise" reports (Chapter 7 §7.7) describe a substantial change in the relationship with eating. That change is welcomed by many people and is under-studied in both directions — including in people for whom appetite suppression interacts badly with a disordered eating history (Chapter 12).

And what happens at population scale. A drug taken by a very large fraction of a population produces effects that no trial measures: on food systems, on how obesity is understood and discussed, on what insurance covers, and on how people who do not take it are regarded. Those are not pharmacological questions and they are not trivial, and Chapter 12 is where this book attempts them.

🔍 Check Your Understanding

1. Which of the unknowns above would most change your view if it resolved unfavorably?
2. Why does not knowing the mechanism limit extrapolation more than it appears to?
3. Name one unknown on this list that a trial could resolve and one that no trial could.

📋 Your Evidence Dossier

This chapter is the model entry: all twelve fields, on one peptide, completely.

Every other entry in your dossier should be measured against this one. Where yours is thinner, that is information about the compound, not about your effort.

DOSSIER ENTRY — SEMAGLUTIDE                            [the model entry]

 1. IDENTITY        Semaglutide. Brands: Ozempic (T2D, injectable), Wegovy (weight
                    management, injectable), Rybelsus (T2D, oral). GLP-1 receptor
                    agonist. 31 residues incl. Aib. ~4,100 Da. Sequence public.

 2. ORIGIN          Engineered analog of human GLP-1 (endogenous, 30 aa, from
                    intestinal L cells). Not endogenous itself. No classical axis
                    suppression — GLP-1 release is nutrient-triggered, not
                    set-point regulated (Ch 3 §3.9).

 3. MECHANISM       GLP-1 receptor (class B GPCR) agonist. Four actions:
                    glucose-dependent insulin secretion; glucagon suppression;
                    slowed gastric emptying; reduced appetite. Receptors also in
                    heart, kidney, vasculature.
                    LOAD-BEARING for weight: gastric emptying + appetite.
                    LOAD-BEARING for cardiovascular benefit: NOT ESTABLISHED.
                    ROUTES: sustained systemic exposure incl. leaky brain regions —
                    may NOT be the same signal as endogenous GLP-1 (Ch 7 §7.6).

 4. PHARMACOLOGY    Subcutaneous weekly (~1 week half-life) or oral daily (~1%
                    bioavailability, strict fasting protocol). Three modifications:
                    Aib at 8 (blocks DPP-4); Lys→Arg at 34 (single attachment site,
                    manufacturing); C18 diacid at 26 (albumin binding). Steady state
                    ~4–5 weeks. Refrigerated before use.

 5. EVIDENCE        SUSTAIN (T2D glycemic + CV outcomes), STEP (weight), PIONEER
                    (oral), SELECT (CV outcomes without diabetes, ~17,000, ~3 years).
                    Multiple large RCTs, hard endpoints, active comparators, years of
                    follow-up, millions of patient-years post-marketing.
                    WHAT DOES NOT EXIST: primary-prevention CV data; data beyond ~3–4
                    years for most endpoints; adequately powered data in older adults
                    with low muscle mass; head-to-head against lifestyle intervention
                    of equivalent intensity without a drug.
                    FUNDING: manufacturer-sponsored throughout.

 6. RATING          Weight loss in obesity: ✅
                    Glycemic control in T2D: ✅
                    CV risk reduction (established CVD + overweight, no diabetes): ✅
                    CV primary prevention: NOT RATED — not studied
                    (further indications: Ch 10)

 7. APPROVED USE    T2D (2017, injectable; 2019, oral); weight management at 2.4 mg
                    (2021); cardiovascular risk reduction in the SELECT population.
                    Jurisdictions and exact wording vary — check locally.

 8. CLAIMED USE     Widely used off-label for weight loss below approved BMI
                    thresholds and for cosmetic weight reduction. Not studied in those
                    populations. Claims about anti-aging, longevity, and general
                    "metabolic optimization" are unsupported.

 9. RISKS           Common: GI effects (nausea, vomiting, diarrhea, constipation) in a
                    substantial minority; usually attenuate. Less common: gallbladder
                    disease; gastroparesis. Rare/monitored: pancreatitis (signal not
                    confirmed at the feared magnitude). Boxed warning: rodent thyroid
                    C-cell tumors; contraindicated with personal/family history of
                    medullary thyroid carcinoma or MEN2 — human translation not
                    established. Practical: aspiration risk under anesthesia. Lean
                    mass loss occurs, as with any weight loss; functional consequence
                    not well established, and most concerning in sarcopenic obesity.
                    Hypoglycemia is uncommon alone; real in combination with insulin
                    or a sulfonylurea.

10. STATUS          Approved and marketed widely. Was in shortage; compounded versions
                    proliferated under the shortage provision and contracted when it
                    resolved (Ch 6 CS2, Ch 12, Ch 38). WADA: check current Prohibited
                    List — status of GLP-1 agonists in sport is a live question (Ch 38).
                    Cost varies enormously by jurisdiction and coverage (Ch 12).

11. VERDICT         [the reader's own two sentences]

12. WHAT WOULD      Would change ✅ → lower: long-term data showing loss of effect
    CHANGE MY MIND  during continued treatment, or a serious harm at multi-year use
                    outweighing benefit.
                    Would NOT change it: weight regain after stopping (expected
                    physiology); anecdotal side effect reports (already in the trials);
                    a newspaper article.

Your task

Write one complete twelve-field entry, on any peptide in your dossier, to this standard.

Then compare it to this one and note where yours is thinner. Some thinness is your effort and some is the compound's evidence base, and telling the difference is the point. If Field 5's "what does not exist" line is longer than the rest of the field, that is a finding.

Conclusion

Semaglutide is native GLP-1 with three modifications: a non-natural residue at position 8 that abolishes the DPP-4 cut site, a substitution at position 34 that leaves a single attachment site for manufacturing control, and a fatty acid at position 26 that binds albumin and defeats renal clearance. Half-life goes from about two minutes to about a week. The receptor pharmacology is unchanged.

Four trial programs established what it does. SUSTAIN: glycemic control in type 2 diabetes, plus a cardiovascular benefit in a trial designed to rule out harm. STEP: about −15% body weight at 68 weeks in adults with overweight or obesity without diabetes, and about −10% in adults with diabetes — the same drug, two populations, two answers, and quoting one for the other is the most common error in coverage of this drug. PIONEER: an oral peptide, at roughly 1% bioavailability. SELECT: a 20% relative and about 1.5 percentage point absolute reduction in cardiovascular events over about three years, in people without diabetes — which moved semaglutide from a drug that improves a risk factor to one that reduces events.

The adverse effects are dominated by gastrointestinal effects that are the mechanism overshooting and that typically attenuate. Gallbladder disease and gastroparesis are real. The pancreatitis signal was appropriately raised and not confirmed at the magnitude feared. The thyroid warning describes a rodent finding with a species-difference explanation and no demonstrated human signal — and the contraindication in MEN2 and medullary thyroid carcinoma is not theatre.

Lean mass is lost, as in any weight loss; whether it translates into functional decline is not well established, and the population where it matters most — older adults with low baseline muscle — is the least studied.

And weight returns when the drug stops, which Chapters 2 and 3 predicted, which makes this chronic therapy rather than a course of treatment, and which is not a failure of the drug any more than blood pressure rising after stopping an antihypertensive is.

Three ✅ ratings, one molecule, three claims. Chapter 9 adds the drugs that came next — and the statistical trap that one of their trials sets.

Key Terms

SUSTAIN — the clinical development program for semaglutide in type 2 diabetes.

STEP — the clinical development program for semaglutide 2.4 mg in weight management.

PIONEER — the clinical development program for oral semaglutide.

SELECT — the cardiovascular outcomes trial of semaglutide 2.4 mg in adults with established cardiovascular disease and overweight or obesity without diabetes.

A1C (glycated hemoglobin) — the fraction of hemoglobin with glucose attached, reflecting average glucose over roughly three months; the standard endpoint in diabetes trials, and a surrogate.

Titration — stepwise dose increase, used here to allow gastrointestinal tolerance to develop.

Gastroparesis — severely delayed gastric emptying; the class's mechanism at its extreme.

Medullary thyroid carcinoma — a rare thyroid cancer arising from C-cells; the subject of the class's boxed warning, based on rodent findings.

MEN2 (multiple endocrine neoplasia type 2) — an inherited syndrome carrying high medullary thyroid carcinoma risk; a contraindication for this drug class.

Lean mass — non-fat body mass, including muscle; a surrogate for strength and function, not a substitute for measuring them.

Sarcopenic obesity — the combination of excess adiposity and low muscle mass, typically in older adults; the population where lean mass loss is of greatest concern.

Chronic therapy — a treatment whose effect persists only while it is taken.

Indication — the specific approved use of a drug, which may be narrower than how it is prescribed.

Spaced Review

1. (Ch 4) State semaglutide's three modifications and what each solves. Which one provides no benefit to the patient, and why does it exist?

2. (Ch 7) Explain why the nausea is not a random side effect, and why it typically fades while the metabolic effects persist.

3. (Ch 5) SELECT reported a 20% relative reduction and about 1.5 percentage points absolute over about three years. Estimate the NNT and write the sentence you would use to explain the result to a patient.

4. (Ch 2, Ch 3) Explain why weight returns on discontinuation, using the override-versus-replace rule. Then explain why this is not evidence that the drug does not work.

5. Someone tells you "semaglutide produces 15% weight loss." State the two questions you would ask before accepting the figure, and explain what each protects against.