Chapter 6 — Further Reading

Tier 1 — Verified canonical

On compounding and the semaglutide episode. The U.S. Food and Drug Administration publishes material on compounding, including the distinction between 503A compounding pharmacies and 503B outsourcing facilities, the drug shortage list and its role in compounding permissions, and specific communications regarding compounded semaglutide — including warnings about semaglutide salt forms (semaglutide sodium and semaglutide acetate) and about dosing errors arising from vial-and-syringe administration. All free on fda.gov. This is the primary source for Case Study 2 and it is worth reading directly; the language regulators use is narrower and more precise than most reporting on it.

The America's Poison Centers network publishes aggregate data on exposure calls, including semaglutide-related calls during the relevant period.

On the underlying science of the three cases. For NAD+ precursors: the biochemistry of NAD+ as a coenzyme is standard textbook material (Lehninger, Berg). Human trials of nicotinamide riboside and nicotinamide mononucleotide are indexed on PubMed and registered on ClinicalTrials.gov — search both and note the trial sizes and durations for yourself. This is a case where doing the check is more instructive than reading about it.

For GHK-Cu: the underlying wound-healing and matrix-biology literature is genuine and dates back decades; it is indexed on PubMed. Chapter 30 covers the cosmetic-claims question in full.

For BPC-157: the preclinical literature is likewise on PubMed. Search it, then filter for "Randomized Controlled Trial" and note what returns. Chapter 17 works this in full.

On why this is hard. Daniel Kahneman, Thinking, Fast and Slow covers the availability heuristic, regression to the mean, and the general architecture of the biases §6.3 and §6.9 describe. The chapters on regression to the mean are directly relevant and unusually clear.

Robert Cialdini, Influence is the standard treatment of persuasion mechanics — social proof and commitment-consistency in particular, both of which operate strongly in the testimonial and forum environment.

Ben Goldacre, Bad Science (again) has an excellent chapter on how media coverage of health science degrades, and on why the degradation is structural rather than a matter of individual bad faith.

Tier 2 — Attributed, specifics unverified

On the scale of the compounded semaglutide market. Estimates of how many patients received compounded semaglutide during the shortage period vary widely across sources and none is quoted here. That the market was large, that it spanned a wide range of legitimacy, and that it contracted substantially when the shortage resolved are all well reported.

On dosing errors. Reports of dosing errors involving compounded semaglutide, including errors arising from confusion between units of measurement when drawing from vials, were described by regulators and poison control centers. Specific case counts vary by source and reporting period.

On the timing of shortage-list changes and subsequent litigation. The FDA's removal of semaglutide from the shortage list and the transition periods and litigation that followed are matters of public record; specific dates and the current status of any legal proceeding should be checked directly, as this book's information is current only as of 2026.

On survivorship in testimonial records. The 100-person diagram in §6.3 is a constructed illustration. That online testimonial records are systematically filtered toward positive outcomes is well established in principle; the specific proportions are not measured quantities.

On the demographics of peptide interest. The characterization in §6.9 — that engagement skews toward educated, physically active, institutionally skeptical readers — reflects a widely observed pattern rather than a measured survey result, and is presented as characterization.

On "over a hundred studies" claims. The number of preclinical publications on any given popular compound is checkable on PubMed; the point in this chapter is about what such a count does and does not establish, not about any particular figure.

Tier 3 — Illustrative and constructed

• The four vials (Case Study 1) is a constructed teaching scenario, built from documented real failure modes. No specific product, vendor, or incident is described. The failure modes listed in its Safety and Risk callout — wrong identity, wrong quantity, endotoxin, non-sterility, aggregation, residual process chemicals — are the genuine, documented categories of manufacturing failure, and Chapters 19 and 34 support them.
• Figure 6.2 ("the limitations paragraph nobody quotes") is a constructed teaching example. The quoted limitations text is written for this book to be representative of the genre; it is not a quotation from any specific paper.
• The 100-person testimonial diagram in §6.3, the five-stage pipeline diagram, and the incentive diagram in §6.7 are this book's own teaching devices.
• The sixty-second first response protocol in §6.10 is this book's own construction.

If you only do one thing

Take one compound and look at it twice.

First, search for it the way anyone would — a general web search, and whatever video platform you use. Note the first ten results. Note what they are selling.

Then search PubMed for the same compound, filter for "Randomized Controlled Trial," and note what returns. Then open one preclinical paper and read its limitations paragraph.

Hold the two pictures side by side. That gap — not any argument in this chapter — is the thing worth seeing, and you cannot un-see it once you have.

A note on this book's own position

§6.7's incentive diagram applies to this book. It is free, it sells nothing, and it has no affiliate relationships — which removes one class of bias and leaves others.

The specific bias to watch for: an academic preference for rigor that shades into excessive conservatism, and an author who finds the ❌ more satisfying to write than the ⚠️. Chapter 5's rule 4 — never downgrade a rating with distaste — applies to this book as much as to anyone, and readers are invited to hold it to that standard. Where you think a rating in these pages is too harsh, the contribution guide explains how to say so.

Looking ahead

Chapter 7 opens Part II with incretin biology. No preparation needed — but if you want a useful contrast, note that everything you are about to read about GLP-1 came out of roughly forty years of publicly funded gut-hormone physiology that had no commercial application when it began. It is a different kind of story from this chapter's, and reading them consecutively is the point.