PROVE IT

The American Supplement Investigation — Book Three

The Reckoning

Joy Meier, Pharm.D.

with Robert W. Coleman, MS Pharm.

Coleman Publishing

Copyright © 2026 Joy Meier and Robert W. Coleman. All rights reserved.

No part of this publication may be reproduced, distributed, or transmitted in any form or by any means without prior written permission of the author, except for brief quotations in critical reviews or articles.

This book is a work of investigative nonfiction. Every factual claim is sourced to peer-reviewed research, federal regulatory documents, court records, or government data, as identified in the bibliography. This book does not provide personalized medical advice; readers should consult a qualified healthcare professional before beginning, changing, or discontinuing any supplement or medication.

Prove It is Book Three of The American Supplement Investigation, a three-part evidence-based series. Book One, What the Label Doesn’t Tell You, and Book Two, The Hollow Aisle, complete the investigation.

Preface

Book Two ended with a worksheet, not a conclusion — a single page meant to close the gap between what a patient takes and what a clinician actually knows about it. That gap is real, and the worksheet is worth using. But a worksheet doesn’t answer the larger question this trilogy set out to investigate, and it’s time to answer it directly: when the claims in a supplement aisle meet a real diagnosis, a real information ecosystem, and real documented cases, what does the evidence actually show — and what should change?

This book is where Books One and Two stop being separate toolkits and become a single instrument, pointed at the hardest material in this investigation. Part VII takes the evidence-literacy habits from Book One and the regulatory realism from Book Two and applies both to the questions that actually bring people to a supplement aisle in the first place: can anything here meaningfully affect a chronic disease — cardiovascular disease, cancer, diabetes, dementia, and others — and where does the vitamin, mineral, or herbal evidence genuinely stand once the marketing is set aside?

Part VIII turns outward, from the product to the information system that sells it. Why does scientific nuance lose to viral content almost every time they compete for the same attention? Is health misinformation a failure of ignorance, or a business model with its own incentives? And why does intelligence — which most people assume protects them from bad information — turn out to offer so little protection at all? Part IX asks whether the American approach is inevitable by looking at what Canada, Australia, the United Kingdom, and the European Union actually do differently, and Part X closes the investigation the way any serious one should: with documented case files, built to a consistent template, and a systematic review of everything the trilogy has found.

I want to say plainly what this book is not. It is not a list of supplements to avoid, and it is not a substitute for a conversation with your own physician or pharmacist about your own health. It is an application of a method — the same method Book One taught and Book Two tested against the regulatory system — to the diseases and decisions that matter most to the people reading it.

This book also closes with something Books One and Two deliberately withheld: a recommendation. An appendix, “What Should Change,” lays out a specific, tiered framework for supplement regulation, built from what already works in Canada and Australia rather than from an abstract ideal — the same kind of recommendation this investigation’s companion series, The American Healthcare Investigation, made in its own closing appendix. Diagnosis without a proposed fix is half an investigation. This is where we finish the other half.

Forty years in medicine taught me that the hardest conversations in a pharmacy or an exam room are rarely about medications a doctor prescribed. They’re about everything else a patient is taking on their own, believing, reasonably, that it can only help. This book is for anyone who has ever stood in that aisle believing the same thing, and wants to know, honestly, what the evidence says back.

Robert W. Coleman, MS Pharm.

PART VII — THE GREAT SUPPLEMENT QUESTIONS

Chapter 1 — Supplements and Chronic Disease

She had just been told her cholesterol was borderline high and her cardiologist wanted to talk about statins. Before that conversation happened, she stood in a supplement aisle for the first time with real urgency rather than idle curiosity, looking at fish oil, red yeast rice, and a dozen products promising to “support heart health” — the exact phrase Book Two’s Chapter 3 taught her to read skeptically. Evidence literacy is one thing in the abstract. It’s a different thing entirely when a real diagnosis is sitting behind the decision.

This chapter is where this trilogy’s two toolkits — Book One’s evidence hierarchy and Book Two’s regulatory realism — get pointed at the question that actually brings most people to that aisle: does any of this meaningfully affect a real chronic disease?

The framework, stated once, used throughout

Every disease category in this chapter, and most of the chapters that follow, gets evaluated the same way: Claim → Biological rationale → Evidence → Clinical outcome → Risk → Bottom line. State what’s promised, state why it’s biologically plausible, look at what was actually tested, look at what the trial found for outcomes that matter, weigh the risk, and arrive at an honest bottom line. This is the Evidence Ladder from Book One and the dose-and-risk thinking from Book Two, run as a single repeatable process.

Cardiovascular disease: fish oil

Claim: omega-3 fatty acid supplements support heart health and reduce cardiovascular risk. Biological rationale: genuinely strong — omega-3s have documented anti-inflammatory and triglyceride-lowering effects, and observational studies for decades linked higher dietary intake to lower cardiovascular risk. Evidence: the VITAL trial, a rigorous randomized study of nearly 26,000 adults, tested marine omega-3 supplementation directly against a placebo for the primary prevention of cardiovascular disease and cancer. Clinical outcome: the primary result was genuinely mixed rather than a clean yes or no — omega-3 supplementation did not significantly reduce the overall composite of major cardiovascular events, but a closer look at the individual outcomes inside that composite found a real 28 percent reduction in heart attack risk and a 50 percent reduction in fatal heart attack, with no benefit at all against stroke. Risk: generally low at studied doses, though Book Two’s Chapter 6 already flagged omega-3’s independent blood-thinning effect and its relevance to anyone on an anticoagulant. Bottom line: not the broad “heart health” claim on the label, but also not nothing — a real, specific, narrower benefit than the marketing implies, exactly the gap this trilogy has been naming since Book One’s Chapter 3.

Cancer: the beta-carotene lesson, applied again

Book One spent three chapters on the beta-carotene trials, so this chapter won’t repeat that story — only its lesson. The VITAL trial that tested omega-3s for cardiovascular disease also tested vitamin D and both supplements for cancer prevention in the same nearly 26,000 participants, and found no significant reduction in overall cancer incidence from either one. The pattern from Book One holds: a plausible antioxidant or anti-inflammatory rationale, a real trial, and a result that doesn’t match the hope that motivated the trial in the first place.

Diabetes: chromium and cinnamon

Claim: these supplements help maintain healthy blood sugar. Biological rationale: chromium plays a genuine, if modest, role in insulin signaling; cinnamon contains compounds with some laboratory-demonstrated effects on glucose metabolism. Evidence and clinical outcome: human trials of both have been small, short, and inconsistent — some show modest reductions in fasting blood glucose in people with diabetes, many show no meaningful effect, and none has demonstrated the kind of outcome that matters most: reduced diabetes complications over time. Risk: generally low, though Book Two’s Chapter 7 already flagged the possibility of compounding a diabetes medication’s glucose-lowering effect into dangerous hypoglycemia. Bottom line: a real but minor physiological effect, inflated by marketing into an implied treatment for a disease neither compound has been shown to meaningfully manage.

Dementia: ginkgo biloba

Claim: ginkgo biloba supports memory and cognitive function, and may help prevent dementia. Biological rationale: ginkgo has documented antioxidant properties and laboratory evidence of effects on amyloid processes implicated in Alzheimer’s disease — a genuinely reasonable hypothesis worth testing directly. Evidence: the Ginkgo Evaluation of Memory study, the largest trial ever conducted on the question, followed more than 3,000 older adults for an average of six years, randomly assigned to ginkgo or placebo. Clinical outcome: no protective effect at all — 277 participants taking ginkgo developed dementia compared with 246 taking placebo, a difference that, if anything, ran in the wrong direction, though not to a statistically meaningful degree. Risk: the same trial found no significant increase in bleeding risk, addressing one of the theoretical interaction concerns Book Two’s Chapter 7 raised for ginkgo generally. Bottom line: one of the clearest negative results in this entire book — a well-funded, well-designed, adequately large trial that simply did not find what decades of marketing had already promised.

Osteoporosis: calcium and vitamin D

This is the chapter’s deliberate exception, included specifically because the framework doesn’t always end in disappointment. Claim: calcium and vitamin D support bone health and reduce fracture risk. Biological rationale: well-established — both nutrients are structurally necessary for bone formation and maintenance, not merely theoretically linked to it. Evidence and clinical outcome: in older adults with inadequate dietary intake, particularly those at elevated fracture risk, calcium and vitamin D supplementation has shown a genuine, replicated reduction in fracture incidence across multiple trials — a rare case in this book where the mechanism, the trial evidence, and the marketing claim are reasonably well aligned. Risk: not zero — several meta-analyses, including one widely discussed analysis published in The BMJ, raised a signal of modestly increased cardiovascular risk with calcium supplementation at certain doses, a genuine ongoing area of research rather than a settled dismissal. Bottom line: the supplement aisle’s rare success story, and worth naming as one, precisely because a book this skeptical of marketing claims owes its reader the honesty of saying so when the evidence actually holds up.

The pattern, named

Five conditions, five products, one recurring shape: a genuinely plausible biological rationale, a real and often rigorous trial, and a result that is either null, narrower than advertised, or — in calcium and vitamin D’s case — a legitimate but still risk-bounded benefit. None of these five stories required anyone to act in bad faith. The mechanism papers were honest. The marketing simply stopped reading at the mechanism, the same failure Book One’s Chapter 5 diagnosed in the beta-carotene story and this chapter has now shown repeats itself, condition after condition, whenever a real disease is on the line.

Claim-to-Evidence Matrix

Chronic disease claims are where the stakes are highest. But they aren’t where most supplement spending actually happens. The next chapter turns to the category everyone assumes they already understand — the humble vitamin — and asks why “necessary” so easily gets confused with “more is better.”

Chapter 2 — Vitamins: Necessary Does Not Mean More Is Better

Book Two’s Chapter 8 introduced the dose-response curve: too little of an essential nutrient causes harm, an adequate range supports normal function, and for a meaningful number of nutrients, harm reappears at the far end of the curve too. This chapter asks a sharper, more specific version of that question. If the recommended intake is enough to prevent deficiency, why do so many people take five or ten times that amount, believing more is simply more protective?

Three zones, not two

Deficiency, adequacy, and excess aren’t a straight line running from bad to good to better. They’re three distinct zones, and a vitamin’s entire value proposition depends almost entirely on which zone a person is actually in before they start supplementing. Someone genuinely deficient in a nutrient can see a real, sometimes dramatic improvement from correcting that deficiency. Someone who was already adequate before they started sees, at best, no additional benefit — and past a certain point, moves into the third zone, where the same nutrient starts working against them.

Vitamin C: hard to overdose, easy to overpromise

Vitamin C is water-soluble, and the body excretes what it doesn’t need far more efficiently than it does the fat-soluble vitamins Book Two’s Chapter 8 covered — acute toxicity from vitamin C megadosing is genuinely rare. That safety margin is precisely what makes it such an attractive megadose marketing target: a manufacturer can sell ten or twenty times the recommended intake without much fear of an obvious adverse event report. But safety isn’t the same as effectiveness. Decades of trials on high-dose vitamin C for cold prevention and cancer have produced a consistent pattern: modest or no effect on cold duration in the general population, and no demonstrated benefit against cancer specifically attributable to the vitamin at high doses. The megadose is safe. It simply isn’t doing what the label implies.

Vitamin D: the popularity surge, and what the trials actually found

Vitamin D supplementation has grown enormously popular over the past two decades, driven by genuinely legitimate concern about deficiency in populations with limited sun exposure, alongside a wave of marketing claims extending far beyond bone health into mood, immune function, cardiovascular protection, and cancer prevention. Chapter 1’s discussion of the VITAL trial already covered vitamin D’s cardiovascular and cancer findings directly: no significant reduction in either, in a nearly 26,000-person trial designed specifically to test those broader claims. Where vitamin D’s evidence remains genuinely strong is exactly where Book One’s Evidence Ladder would predict — musculoskeletal health in people who are actually deficient, which is precisely the osteoporosis case Chapter 1 named as this book’s exception.

B12 and the B-complex: treating a deficiency, not a baseline

Vitamin B12 supplementation reverses a real, well-documented deficiency — one common enough in older adults, people on certain medications, and those following a vegan diet that screening for it is a legitimate part of routine care. The marketing claim built on top of that legitimate use is broader: B-complex vitamins marketed generally for “energy,” aimed at anyone feeling tired, regardless of whether they’re actually deficient. A person who is B12-deficient and corrects that deficiency will often feel a genuine, sometimes dramatic improvement in energy — and that real experience is precisely the kind of anecdote Book One’s Chapter 3 warned against generalizing. It says nothing about what B12 will do for someone whose B12 levels were already normal to begin with.

The core reframe

A vitamin’s value is almost entirely conditional on a fact the label can’t see and doesn’t ask about: whether the person taking it is actually deficient. Correcting a real deficiency can produce a genuine, meaningful benefit. Supplementing a nutrient a person already has enough of, at the doses typically marketed, rarely replicates that benefit — it simply moves the person further along the dose-response curve toward the zone where more stops being better and starts, at the margins, being worse.

Deficiency → Adequacy → Excess

Picture the same three-zone structure across every vitamin in this chapter: a deficiency zone where supplementation genuinely helps, an adequacy zone where a person already has what they need, and an excess zone where — for the fat-soluble vitamins especially, and even for some water-soluble ones at extreme doses — more supplementation stops helping and starts, in some documented cases, causing harm. The question worth asking before buying any vitamin isn’t “could this help me?” It’s “which of these three zones am I actually in?”

Minerals follow this same three-zone logic, with one further complication vitamins mostly avoid: several essential minerals are dangerous well before they ever become beneficial. That’s where we turn next.

Chapter 3 — Minerals: Essential and Potentially Dangerous

Book Two’s Chapter 8 opened with iron: essential for preventing anemia, and simultaneously one of the leading causes of fatal accidental poisoning in young children, because a handful of adult iron tablets can deliver a toxic dose to a small child. No vitamin in the previous chapter carries quite that kind of acute risk. Several minerals do — which makes this chapter’s three-zone framework a genuinely higher-stakes application of the same idea.

Calcium: the exception’s own caveat

Chapter 1 already named calcium and vitamin D as this book’s rare success story for bone health in people at real fracture risk. It’s worth returning to briefly here for a detail that belongs with minerals specifically: calcium supplementation’s cardiovascular risk signal, flagged by several meta-analyses including the widely discussed BMJ analysis, appears tied to supplement-form calcium taken in concentrated doses — a pattern distinct from calcium consumed through diet, where no comparable signal has emerged. The same mineral, the same nutrient, a different risk profile depending on how it’s delivered — a reminder that “calcium is good for bones” is true and incomplete in exactly the way this entire book keeps finding marketing claims to be.

Iron: deficiency, overdose, and overload

Iron deficiency is real, common, and genuinely improved by supplementation — pregnant women and people with heavy menstrual bleeding are the clearest cases where correcting it matters. But iron’s risk isn’t limited to the acute pediatric poisoning Book Two’s Chapter 8 described. Chronic iron overload — hemochromatosis, whether genetic or accumulated through unnecessary long-term supplementation in someone who was never deficient — causes real, documented organ damage over time, concentrated in the liver and heart. The practical lesson is simple and easy to skip: iron supplementation should follow an actual test result, not an assumption, because this is a mineral where the excess zone carries genuine long-term consequences, not just a remote acute risk.

Magnesium: a popular claim, a modest evidence base

Magnesium has become one of the most heavily marketed minerals of the past several years, promoted for sleep, anxiety, and general “relaxation.” The biological rationale isn’t baseless — magnesium is genuinely involved in neuromuscular and nervous system function. But direct trial evidence for these specific claims, in people who aren’t magnesium-deficient, remains modest and inconsistent — a pattern that should feel familiar by this chapter, because it’s the same gap between mechanism and marketed outcome this book has now documented across nearly every category it has examined.

Zinc and selenium: real deficiency correction, real toxicity syndromes

Both minerals are genuinely essential, and correcting an actual deficiency in either produces real, measurable benefit — zinc for immune function and wound healing, selenium for thyroid and antioxidant enzyme function. Both also carry well-documented toxicity syndromes at chronic excess: selenosis, which Book Two’s Chapter 8 already introduced, produces hair and nail changes and, in severe cases, nerve damage; excess zinc can cause copper deficiency and its own neurological consequences. Essential and dangerous aren’t opposite ends of a spectrum here. They’re two facts about the same compound, true at the same time, depending entirely on dose.

Potassium: the mineral regulation actually restricts

Potassium closes this chapter usefully, because it’s the clearest example of Book Two’s Chapter 1 argument working exactly as intended. Unlike nearly every other mineral in this book, over-the-counter potassium supplements are capped by the FDA at a low per-dose limit, specifically because excess potassium carries a well-documented, serious cardiac arrhythmia risk. This is what premarket-style caution looks like when it’s actually applied: not a ban, not an approval process, but a dose ceiling set because the risk was clear enough to act on before harm accumulated — a rare moment in this trilogy where the regulatory system got ahead of a risk instead of documenting it after the fact.

The Essential Nutrient Paradox

The same mineral sits on both sides of this chapter’s title, essential and dangerous, separated by nothing but dose and circumstance: iron treats anemia and poisons children. Calcium builds bone and, in supplement form, carries a cardiovascular question mark. Selenium and zinc correct real deficiencies and produce real toxicity syndromes at excess. Potassium is so dangerous in concentrated form that the one mineral in this chapter with an actual regulatory dose cap is the one whose over-the-counter version you’ll never find at a meaningfully high dose in the first place.

Minerals are, at least, discrete, well-characterized compounds with known toxicity profiles. Herbal remedies introduce a problem this chapter hasn’t had to face yet: the product itself can vary meaningfully from one bottle to the next, even under the same label.

Chapter 4 — Herbal Remedies

Two bottles sit side by side, both labeled St. John’s Wort, both from established retailers. Laboratory analysis of products like these has repeatedly found meaningful differences in the actual concentration of active compounds between brands — sometimes between different batches of the same brand. This is a problem essentially unique to this category. A vitamin C tablet is a defined chemical compound at a stated dose. An herbal extract is a complex mixture of plant compounds that can vary with growing conditions, harvest timing, extraction method, and manufacturing quality control — variation a label rarely discloses in any way a shopper could use.

Standardization: the word that’s supposed to fix this

A label claiming a “standardized extract” is asserting that the product has been processed to guarantee a consistent level of one or more marker compounds from batch to batch. When standardization is genuine and verified, it meaningfully addresses the variability problem this chapter opened with. But Book Two’s Chapter 2 already showed how a proprietary blend can obscure exactly this kind of detail, and the same gap applies here: a label can say “standardized” without disclosing to what standard, verified by whom, leaving a shopper no better equipped to compare two products than before the word appeared on the package.

Traditional use as a hypothesis, not a conclusion

Book One’s Evidence Ladder placed anecdote near the bottom and a randomized trial near the top. Traditional use — centuries of a plant being used for a particular purpose across a culture — sits in an interesting position on that ladder: it’s a stronger signal than any single anecdote, because it reflects sustained, repeated observation across many people over a long period. But it’s still not a trial, and it carries its own confounders — the same regression to the mean and natural disease fluctuation Book One’s Chapter 3 named apply across a population using a remedy for generations just as they apply to one person using it for a month. Traditional use is a legitimate reason to test a compound rigorously. It is not, on its own, evidence that the compound works.

St. John’s Wort: a genuine partial success, with a serious caveat already on record

Of the herbs this chapter could examine, St. John’s Wort has some of the most genuinely supportive trial evidence in this entire book — multiple randomized trials and meta-analyses have found it comparably effective to standard antidepressant medications for mild-to-moderate depression, a real result on real, meaningful outcomes. But Book Two’s Chapter 7 already introduced this same herb from a different angle: the documented case of a heart transplant patient whose cyclosporine levels dropped low enough to trigger organ rejection, caused by St. John’s Wort’s effect on liver enzyme metabolism. Read together, the two chapters make a single point Book One’s evidence hierarchy predicts but rarely states this starkly: a supplement can have genuine efficacy and a genuinely serious risk profile at the same time. Neither fact cancels the other out.

Echinacea: popular, and mostly unconvincing

Echinacea’s cold-prevention and cold-duration claims have been tested in numerous trials over several decades, with a consistent pattern of small, inconsistent effects that most rigorous reviews describe as not clinically meaningful. This is a case where genuine, extensive testing has occurred — unlike many products in this book, Echinacea hasn’t gone untested — and the testing simply hasn’t supported the popularity the product continues to enjoy.

Saw palmetto: two large trials, one clear answer

Saw palmetto is used by more than two million American men for benign prostatic hyperplasia symptoms, commonly as an alternative to FDA-approved medications. A 2006 randomized trial published in The New England Journal of Medicine, involving 225 men with moderate-to-severe symptoms, found that saw palmetto did not improve symptoms or objective measures of the condition compared with placebo. A larger 2011 follow-up trial published in JAMA, testing escalating doses up to three times the standard amount in 369 men, reached the same conclusion at every dose tested. Two well-designed trials, run five years apart by different research teams, arrived at the identical answer — about as clear a negative result as this book has documented anywhere.

Product variability and contamination

Book Two’s Chapter 4 already established that herbal products are disproportionately represented in FDA contamination and mislabeling enforcement actions, and Book Two’s Chapter 6 showed that 40 percent of green tea extract products implicated in liver injury didn’t even list the compound on their label. Herbal remedies compound every problem this trilogy has documented — regulatory gap, marketing vocabulary, contamination risk — with one additional layer unique to the category: even an honest label may not guarantee what’s actually inside the capsule matches what’s on it.

Plant → Extract → Product → Patient

Trace an herbal remedy’s path: a plant, with genuinely variable compound concentrations depending on where and how it grew; an extraction process, which may or may not be standardized in any verifiable way; a finished product, which may or may not accurately disclose what survived that process; and a patient, who has no practical way to verify any step in that chain from the label alone. Each arrow is a place standardization can fail, and none of them is visible from the outside of the bottle.

Individual herbs, tested one at a time, are this chapter’s subject. But an entire marketing category exists that isn’t built around a specific ingredient at all — it’s built around a promise, with whatever ingredient happens to be attached to it treated as almost interchangeable. That’s where we turn next.

Chapter 5 — The Supplement for Everything

Walk down a single wellness aisle endcap and count the promises: weight loss, energy, sleep, memory, sexual performance, immunity, longevity, detoxification, anti-aging, athletic performance. Ten different products, ten different labels, and — look closely at the ingredient panels — a surprising amount of overlap in what’s actually inside them. This chapter moves faster than the ones before it, on purpose. By now, the toolkit built across four books and four chapters should let you run this analysis yourself, quickly, without needing a full chapter’s treatment of each individual claim.

Weight loss. The claimed mechanism is usually appetite suppression, fat-burning, or metabolic boost. The evidence gap: stimulant-based ingredients can produce modest, short-term effects, often accompanied by cardiovascular side effects disproportionate to the benefit — the same category of risk that ended ephedra’s run on the market, as Book Two’s Chapter 5 detailed.

Energy. Usually B-vitamins, caffeine, or adaptogens. The mechanism is often real for people who are deficient or caffeine-naive; the marketing extends the claim to anyone feeling tired, the same conditional-value problem Chapter 2 named for B12 specifically.

Sleep. Melatonin, valerian, magnesium. Some genuine evidence exists for melatonin in specific circumstances — jet lag and shift-work sleep disruption — with far thinner support for general insomnia, and Book Two’s Chapter 7 already flagged the sedative-interaction risk when these are combined with prescribed sleep aids.

Memory. Ginkgo biloba, again, alongside a rotating cast of “nootropic” ingredients — Chapter 1 already gave ginkgo’s dementia-prevention claim a full, well-sourced negative answer, and most memory-support products carry the same mechanism-outpaces-evidence pattern this trilogy keeps finding.

Sexual performance. This category carries a distinct, serious risk this book hasn’t repeated since Book Two’s Chapter 4: it’s one of the categories the FDA most frequently finds spiked with undisclosed prescription pharmaceutical ingredients, sold as “all-natural.”

Immunity. Book One’s Chapter 5 already dismantled the underlying premise directly — there is no single dial called “immunity” to turn up, and no supplement in this category has been shown to meaningfully reduce how often or how severely a healthy person gets sick.

Longevity. Perhaps the vaguest claim in the entire marketplace, promising to extend life or slow aging through mechanisms — cellular health, again, from Book Two’s Chapter 10 — too broad to test as stated, let alone prove.

Detoxification. Book Two’s Chapter 10 already gave this one its full answer: in a person with functioning kidneys and liver, no supplement has been shown to meaningfully improve on what those organs already do without help.

Anti-aging. Overlaps heavily with longevity and with the antioxidant claims Book One’s Chapter 5 examined through the beta-carotene lens — a real biological concept (cellular aging is genuine) extended into a product claim no supplement has been shown to deliver on at the level implied.

Athletic performance. The one category in this list with some of the strongest evidence behind specific compounds — creatine’s effect on short-duration high-intensity performance is genuinely well-documented — sitting on the same shelf, often in the same product line, as far weaker claims that borrow its credibility by association.

What all ten have in common

Look past the specific ingredient and the specific claim, and nearly every category on this list is selling the same underlying thing: an appeal to a universal, non-specific desire — be thinner, sharper, more energetic, more attractive, live longer — that no single compound, however genuinely useful in its narrow, well-tested application, has ever been shown to deliver comprehensively. Creatine helps with short-duration athletic output. It does not make someone generally “better.” Melatonin helps reset a disrupted sleep-wake cycle. It does not cure insomnia broadly. The gap this entire trilogy has documented, category after category, is the distance between a real, narrow, tested effect and the broad, universal promise built on top of it.

The American Supplement Promise Map

Arrange all ten categories side by side, each annotated with its actual evidentiary gap in a few words: weight loss (real short-term effect, disproportionate cardiovascular risk), energy (real only if deficient), sleep (real for circadian disruption, not general insomnia), memory (well-tested, largely negative), sexual performance (contamination risk, not just weak evidence), immunity (no dial to turn up), longevity (too vague to test), detoxification (organs already do this), anti-aging (real biology, unproven product claim), athletic performance (genuinely strong for specific compounds, weak by association for others). Ten promises. One recurring shape.

Part VII has tested these products against the evidence, category by category. Part VIII asks a different question: with this much evidence available, and this consistent a pattern, why does the marketing keep winning anyway?

PART VIII — THE INFORMATION PROBLEM

Chapter 6 — When Science Becomes Content

A research team publishes a careful finding: a compound showed a modest, statistically significant effect on one narrow outcome, in one specific population, with several caveats about generalizability clearly stated in the discussion section. Within days, a sixty-second video reduces that finding to a single, confident, wrong-shaped claim — no caveats, no population restriction, no hedging — and reaches more people in an afternoon than the original paper will reach in its entire publication life. This is not a hypothetical. It is the routine fate of nutrition and supplement research in the current information environment, and it’s worth understanding as a structural problem rather than a series of individual bad actors.

The algorithm, widened

Book Two’s Chapter 11 examined this dynamic specifically through supplement influencers — why platforms reward confident claims over hedged ones, and why that reward structure has nothing to do with accuracy. This chapter widens the lens to the information ecosystem generally: search engine results, SEO-optimized health content sites built explicitly to rank rather than to inform, and increasingly, AI-generated content produced at a volume no editorial process could realistically review. The mechanism is identical to Book Two’s Chapter 11 — engagement, not accuracy, is what gets measured and rewarded — but the scale here is the entire architecture of how people find health information at all, not one creator’s audience.

The asymmetry of effort

A rigorous study like the ones this book has cited — VITAL, GEM, the saw palmetto trials — takes years and often tens of millions of dollars to design, run, and publish. A confident video contradicting or oversimplifying that same finding takes an afternoon to produce and can be published with no review process whatsoever. This asymmetry means the volume of confident, simplified, often wrong content will structurally outpace the volume of careful, hedged, accurate content, regardless of how many rigorous trials get run — a mathematical reality of production speed that has nothing to do with which side has better evidence.

AI-generated health content: a genuinely new wrinkle

This trilogy’s companion series on American healthcare didn’t have to grapple with this problem in quite the same form, and it’s worth naming directly: AI language models can now generate plausible-sounding, confidently worded health content at a scale and speed no human content mill could previously match, and that content can appear anywhere on the evidence spectrum from accurate to entirely fabricated with no visible difference in tone or confidence between the two. A reader applying Book One’s Evidence Ladder to a piece of AI-generated content faces a genuinely harder problem than applying it to a human-written article — the confident tone that once correlated, imperfectly, with a human author’s actual conviction no longer reliably signals anything at all.

None of this requires anyone producing the content to be malicious. Confident content simply performs better than careful content, at every stage of this pipeline, for reasons that have nothing to do with which one is true. The next chapter asks a harder question about the people producing it: is this dynamic just an unfortunate byproduct of how attention works online — or, for some of the players in this ecosystem, is misinformation itself the actual business model?

Chapter 7 — The Misinformation Business

A piece of confidently wrong health content drives a wave of traffic, a spike in affiliate clicks, and a real, measurable increase in sales for whoever published it. By the metric that actually gets tracked in that publisher’s analytics dashboard, this is not a failure. It’s a success — arguably a bigger one than a careful, accurate, appropriately hedged piece of content would have produced, because Chapter 6 already established that confident claims outperform hedged ones on every engagement metric that matters to a platform’s algorithm.

A business model, not a byproduct

Book Two’s Chapter 9 traced the supplement dollar through manufacturing, subscriptions, and affiliate commissions. Book Two’s Chapter 11 showed how an affiliate marketer’s income depends on the sale, not on the product’s actual effectiveness. Put those two chapters together and widen them to the misinformation content itself, and the pipeline is direct: Attention → Trust → Traffic → Affiliate Sale. Each stage converts cleanly into the next, and at no point in that chain does accuracy factor into anyone’s revenue.

The incentive structures, named plainly

Ad revenue is tied to engagement — time on page, click-through rate, share count — not to whether the claims on the page are true. Affiliate commissions, as Book Two’s Chapter 9 documented, typically run in the range of 10 to 50 percent of a referred sale, paid regardless of whether the product delivers any benefit at all to the buyer. Neither revenue stream includes any mechanism, financial or otherwise, that rewards accuracy or penalizes error. A publisher optimizing honestly for either metric — pure engagement, or pure affiliate revenue — will, without ever intending to mislead anyone, gravitate toward exactly the confident, oversimplified content Chapter 6 described.

The consequence gap

Compare this to a regulated financial product: a false or misleading claim in an investment prospectus carries real legal and regulatory consequences for the person who made it. A false or misleading health claim that drives a supplement sale carries, in the great majority of cases, no comparable consequence for whoever published it — Book Two’s Chapter 4 already documented how thin and reactive post-market supplement enforcement actually is, and that same thinness extends to the content ecosystem selling the products, which faces even less direct regulatory attention than the products themselves.

Evidence Spotlight — Following One Claim’s Money

Picture a single viral claim, reaching even a modest fraction of a large platform’s audience: a few hundred thousand views, a small single-digit percentage click-through to an affiliate link, and a typical supplement price point with a mid-range affiliate commission. The resulting revenue to the publisher, from one piece of content, can run into thousands of dollars — generated in days, with no relationship whatsoever to whether the underlying claim survives the kind of scrutiny this book has applied, chapter after chapter, to real supplement evidence. The economics don’t require the claim to be true. They only require it to be believed long enough to click.

None of this works without an audience willing to believe the claim in the first place — and the genuinely uncomfortable finding, examined next, is that education and intelligence turn out to offer far less protection against that belief than most people assume.

Chapter 8 — Why Smart People Believe

A physician, a lawyer, an engineer — someone whose profession depends daily on evaluating evidence rigorously and whose intelligence by any conventional measure is well above average — holds a specific, confidently wrong belief about a supplement claim. Not out of laziness, and not because the information contradicting it was unavailable to them. This isn’t a rare exception this book is cherry-picking to make a point. It’s a well-documented, consistent finding, and it’s worth taking seriously precisely because it’s counterintuitive.

Intelligence and bias are mostly separate systems

Psychological research on what researchers call dysrationalia — the gap between raw cognitive ability and the tendency to reason well in practice — has found consistently that measures of general intelligence correlate only weakly with resistance to cognitive bias. The biases most relevant to this book — confirmation bias and availability bias, both introduced in Book One’s Chapter 3 — operate largely on motivation and attention, not on raw reasoning capacity. A highly intelligent person is just as capable of noticing only the evidence that confirms what they already believe, and just as susceptible to overweighting a vivid, available anecdote over a duller, more representative statistic, as anyone else. Intelligence helps you construct a more sophisticated justification for a belief. It does not reliably protect you from forming the belief incorrectly in the first place.

Six distinct mechanisms, not one

Authority persuades independently of evidence quality — Book Two’s Chapters 11 and 12 already showed how a following or a fame can substitute for expertise entirely. Fear and hope both short-circuit careful evaluation, for opposite reasons: fear pushes toward immediate action to escape a perceived threat, hope pushes toward immediate belief in a longed-for solution, and neither leaves much room for Book One’s Evidence Ladder to operate before a decision gets made. Personal experience, examined in full in Book One’s Chapter 3, remains the most persuasive category of all, precisely because it’s the hardest to argue against — it happened to you. Identity and community reinforcement are the two mechanisms this chapter treats in the most depth, because they explain why the other four don’t fade once the initial claim is examined.

Identity: the mechanism that resists correction

Once a health practice becomes part of how someone understands themselves — “a person who takes control of their own health,” “someone who doesn’t just trust doctors blindly,” “someone who does their own research” — contrary evidence stops functioning as a correction of fact and starts functioning as an attack on identity. This reframing changes how the brain processes the new information entirely: instead of updating a belief, the person defends it, often more strongly than before the contrary evidence appeared. Community reinforcement compounds this — a person embedded in a community organized around a particular health belief faces a real social cost for updating that belief that has nothing to do with the evidence itself, and everything to do with maintaining standing within a group that matters to them.

The Belief Cycle

Picture a loop rather than a line: a claim is encountered → it’s evaluated through the lens of existing identity and community rather than neutrally → confirming evidence is noticed and remembered, disconfirming evidence is discounted or forgotten → the belief strengthens rather than updates → the strengthened belief becomes more central to identity → the next piece of contrary evidence faces an even higher bar for consideration than the last one did. Each pass through the loop makes the next correction harder, not easier — which is precisely why simply presenting better evidence, on its own, so often fails to change a firmly held belief.

Part VIII has examined why the information ecosystem favors confident claims and why belief in them resists correction once formed. Neither of these dynamics is uniquely American — which raises a natural question this trilogy hasn’t yet answered: do other countries’ regulatory systems account for any of this, or does the United States face a genuinely unique version of the problem?

PART IX — THE INTERNATIONAL PERSPECTIVE

Chapter 9 — What Other Countries Do Differently

Book Two’s Chapter 1 established the American starting point plainly: a dietary supplement can reach a U.S. shelf with no premarket proof of safety or effectiveness required. It’s worth asking a question this trilogy hasn’t yet answered directly — is that the only way a modern country can regulate this category, or is it a choice, made once in 1994, that most of the rest of the world simply didn’t make the same way?

Canada: a license before a sale

Health Canada regulates supplements under its Natural Health Products Regulations, and the difference from the American approach starts at the most basic level: a product cannot legally be sold in Canada without first obtaining a Natural Product Number, issued only after Health Canada reviews the product against a monograph — a pre-established evidence standard for that ingredient and claim, or a company-submitted evidence package if no monograph exists yet. This is genuine premarket review, scaled to the product: a well-established ingredient with a modest, already-vetted claim can move through review quickly against its monograph, while a novel ingredient or a more assertive claim requires the company to submit its own supporting evidence before a license is granted. Nothing in this system requires the multi-year clinical trial process a new pharmaceutical drug faces. But nothing reaches a Canadian shelf without somebody outside the manufacturer looking at the evidence first — the single requirement DSHEA never included at all.

Australia: sorting by risk, not treating everything alike

The Therapeutic Goods Administration runs what is probably the clearest working example of the risk-tiered model this trilogy’s appendix ultimately recommends. Low-risk products with well-established ingredients and modest, pre-approved claims can be self-certified as “Listed” (AUST L) — fast to market, lightly reviewed, roughly comparable to the current American baseline. Products making a more specific health claim beyond basic nutrient support move into an “Assessed Listed” tier, requiring the company to submit the evidence behind that specific claim for review before the product can carry it. Products with the highest risk profile — genuinely novel ingredients, or claims that approach disease treatment — require full “Registered” status, comparable to the deeper review the highest-risk end of Canada’s system requires. Nothing in Australia’s approach treats a standard multivitamin the same way it treats a novel compound making an aggressive disease-adjacent claim, which is precisely the distinction Book Two’s Chapter 3 showed American structure/function claim law fails to draw at all.

The European Union and United Kingdom: a positive list, not a presumption

The EU takes a different approach again, and in some ways a stricter one on paper: under Directive 2002/46/EC, only the specific vitamins, minerals, and forms listed in the directive’s own annexes may be used in a food supplement at all — trading in an unlisted vitamin or mineral source has been prohibited across the EU since August 2005. Any genuinely novel ingredient, one without a documented history of use in the EU before 1997, requires a formal safety evaluation by the European Food Safety Authority before it can be added to the list at all. This is the inverse of DSHEA’s default: the American system presumes an ingredient may be sold unless proven unsafe after the fact, while the EU’s system presumes an ingredient may not be sold until it’s been added to an approved list in the first place. The United Kingdom, following its departure from the EU, largely preserved this same positive-list structure in its own domestic regulations rather than adopting anything resembling the American model.

International Supplement Regulation

Set side by side, the contrast is stark: the United States requires no premarket safety or efficacy review for the overwhelming majority of products. Canada requires a license, evidence-scaled to the product. Australia sorts every product into one of three risk tiers before it reaches a shelf. The EU and UK restrict the ingredients that may be used at all to a pre-approved list, full stop. None of these systems eliminated their supplement markets. None of them requires a drug-trial-length process for an ordinary vitamin. All of them require something — anything — before a product reaches a consumer, which is precisely the one requirement DSHEA has never included in three decades.

An accident of politics, not a reflection of unique values

It would be convenient to conclude that America’s approach reflects some genuinely distinct national value — a stronger commitment to consumer choice, a deeper skepticism of regulatory overreach. But DSHEA’s own legislative history, briefly touched in Book Two’s Chapter 1, tells a narrower story: a stricter FDA posture toward supplements in the early 1990s provoked a consumer and industry backlash large enough to produce a law removing premarket authority almost entirely — a specific political outcome, in a specific political moment, that most other developed countries simply never replicated in the years since. The American approach isn’t a reflection of a value no other democracy shares. It’s the surviving result of one legislative fight, thirty years old, that the rest of the developed world resolved differently.

With the comparative evidence established, it’s time to apply everything this trilogy has built — the evidence hierarchy, the regulatory realism, the risk data, the information-ecosystem analysis — to real, individually documented cases.

PART X — THE INVESTIGATION

Chapter 10 — Case Files

Every case in this chapter is built from court records, FDA enforcement documents, congressional testimony, and contemporaneous reporting — never from a single dramatic anecdote presented as though it proved something about an entire category. That standard, first stated in Book Two’s Chapter 6, matters most here, in the chapter where it would be easiest to abandon.

Case One: Ephedra and the Death of Steve Bechler

The Claim: Ephedra-based weight-loss products, marketed under names including Xenadrine, promised rapid fat loss and enhanced athletic performance through a natural herbal stimulant.

The Product: Ephedra, the botanical source of ephedrine alkaloids, sold as an over-the-counter dietary supplement with no premarket safety review required under DSHEA — Book Two’s Chapter 1 opening example, now shown at its most consequential.

The Evidence: Well before 2003, ephedra had already been linked to a documented pattern of adverse cardiovascular events serious enough that the FDA had begun collecting adverse event reports and commissioned an independent review by the RAND Corporation.

The Marketing: Sold as a natural, over-the-counter aid to weight loss and athletic performance, available without any of the warnings or monitoring that would accompany a stimulant of comparable potency sold as a prescription drug.

The Risk: Cardiovascular events — heart attack, stroke, and dangerous elevations in body temperature during exertion, a risk multiplied by combining the stimulant with intense physical activity and dietary restriction.

The Outcome: On February 16, 2003, Baltimore Orioles pitching prospect Steve Bechler, 23 years old, collapsed during a spring training workout after taking an ephedra-containing product while trying to lose weight quickly; his body temperature reached 108 degrees, and he died the following day. His death became the specific catalyst — following 155 previously documented deaths already tied to the ingredient — for the FDA’s December 2003 decision banning ephedra from dietary supplements, effective in 2004. A subsequent legal challenge from a supplement manufacturer succeeded in striking down the ban in a federal district court in 2005, on procedural rather than safety grounds; the ban was substantially restored on appeal shortly after, and ephedra has not returned to the U.S. supplement market.

The Lesson: The regulatory system Book Two’s Chapter 1 described — no premarket proof of safety required — allowed a documented pattern of serious harm to accumulate for years before enforcement followed, and even then, only after a case specific and public enough to generate sustained political pressure. The FDA’s ephedra ban remains, to this day, the only time the agency has used its DSHEA authority to remove an entire ingredient category from the market — a measure of how high the bar for that kind of action actually is.

Case Two: Red Yeast Rice and the Pharmanex Litigation

The Claim: Cholestin, a red yeast rice supplement manufactured by Pharmanex, promised to help maintain healthy cholesterol levels.

The Product: Red yeast rice fermented and processed to concentrate mevinolin — a compound chemically identical to lovastatin, the active ingredient in the prescription drug Mevacor, approved by the FDA in 1987.

The Evidence: The FDA determined in 1998 that Pharmanex’s manufacturing process yielded concentrations of the drug compound far beyond what traditional red yeast rice naturally contains, and concluded the product was, in substance, an unapproved version of an already-approved drug.

The Marketing: Sold as a natural supplement for cholesterol support — the same “supports heart health” register Book Two’s Chapter 3 examined — despite containing a standardized, concentrated dose of a compound with genuine pharmaceutical potency.

The Risk: The same risks associated with the prescription drug lovastatin — including, notably, the interaction and dosing concerns any statin carries — sold without the monitoring, dosing precision, or physician oversight that accompanies the prescription version.

The Outcome: Pharmanex sued the FDA; a federal district court initially ruled in the company’s favor in 1999, finding Cholestin qualified as a dietary supplement. The FDA appealed, and the Tenth Circuit Court of Appeals reversed that decision in 2000, ultimately siding with the FDA on remand — but on narrow grounds specific to Pharmanex’s particular concentrated manufacturing process, not a blanket ruling against red yeast rice as an ingredient. That narrow scope is precisely why red yeast rice supplements, containing the same compound at lower, less standardized concentrations, remain legally sold today — the exact product Book Two’s Chapter 1 used as its opening illustration of the blurred line between drug and supplement.

The Lesson: Even when the FDA wins an enforcement case, the resulting precedent can be narrow enough that the underlying regulatory ambiguity survives largely intact. A court can rule against one company’s specific product without resolving the broader question of whether an entire ingredient category deserves the same scrutiny.

Case Three: The Tainted Product Pattern

The Claim: Products marketed for sexual enhancement, weight loss, and bodybuilding, sold as “all-natural” alternatives to prescription medication.

The Product: This case, unlike the first two, isn’t built from a single company or product — it’s built from a documented pattern across hundreds of FDA enforcement actions, compiled in the agency’s own public warning database of products found to contain hidden pharmaceutical ingredients.

The Evidence: Laboratory testing conducted as part of routine FDA surveillance has repeatedly found sexual-enhancement products secretly spiked with sildenafil or tadalafil — the active ingredients in prescription erectile dysfunction medications — and bodybuilding and weight-loss products spiked with undisclosed steroids, stimulants, or prescription weight-loss drug compounds, sometimes at doses exceeding what a physician would ever prescribe.

The Marketing: Sold explicitly as natural and herbal, frequently with claims implying the product is a safer alternative to the very prescription medication secretly hidden inside it.

The Risk: Undisclosed prescription-strength drug interactions — a consumer taking a nitrate medication for heart disease, for instance, who unknowingly takes a product secretly spiked with sildenafil, faces a documented, serious cardiovascular interaction risk they had no way to anticipate because the label told them they were taking an herbal product.

The Outcome: The FDA’s warning list for these products grows continuously, with new entries added on an ongoing basis, and recalls — mostly voluntary, as Book Two’s Chapter 4 described — following each discovery. The pattern has proven persistent rather than solved; new brands and formulations replace warned-against ones with some regularity.

The Lesson: This is the starkest possible illustration of Book One’s central finding — that consumers assume premarket testing occurred where none did — because these products don’t just lack evidence for their claims. They actively conceal the one ingredient that might make them work, specifically because disclosing it would reveal them as unapproved drugs rather than supplements.

Case Four: NMN and the Cost of Regulatory Ambiguity

The Claim: Nicotinamide mononucleotide, marketed as a leading longevity and anti-aging compound, promised to boost cellular NAD+ levels and slow aspects of cellular aging — precisely the vague, appealing promise Chapter 5 named as the “longevity” category’s core pitch.

The Product: NMN, a form of vitamin B3 sold as a dietary supplement, widely available and popular through the early 2020s.

The Evidence: This case is unusual in this chapter because it isn’t primarily about safety evidence at all — it’s about a legal technicality inside DSHEA itself. In November 2022, the FDA determined that NMN was excluded from the legal definition of a dietary supplement under DSHEA’s “race to market” clause, which bars an ingredient from being sold as a supplement if it was first authorized for investigation as a new drug before being marketed as a supplement.

The Marketing: Sold as one of the most heavily promoted anti-aging supplements of the early 2020s, with claims drawing directly on genuine, published cellular-aging research — the same mechanism-to-marketing gap Chapter 1 documented across every chronic disease category in this book, here applied to aging itself.

The Risk: Not a safety risk in the traditional sense of this chapter’s other cases — the risk here was entirely regulatory and economic. Following the FDA’s determination, major retailers including Amazon, along with payment processors including PayPal and Shopify, stopped supporting NMN sales, causing significant disruption to an entire product category with no new safety finding behind the action at all.

The Outcome: The Natural Products Association filed a citizen petition in 2023 and then a federal lawsuit in 2024 challenging the FDA’s determination. A federal court halted FDA enforcement against NMN in November 2024 while the agency reconsidered. In September 2025, the FDA reversed its own 2022 position entirely, concluding that NMN had, in fact, been marketed as a supplement in the U.S. as early as 2017 — before the drug investigation the original exclusion had been based on — restoring its lawful status nearly three years after the disruption began.

The Lesson: Not every case in this book is a story of underregulation causing harm. This one is a story of regulatory uncertainty itself imposing a real cost — years of market disruption, litigation expense, and consumer confusion — over a legal technicality that had nothing to do with whether the product was actually safe or effective. A regulatory system can fail in more than one direction, and Book Three’s closing appendix, “What Should Change,” has to account for both failure modes, not just the one this book has spent the most time documenting.

Four cases, four different failure patterns: a documented death from an entire ingredient category taking years to face a ban; a legal victory narrow enough that the underlying product remains on shelves; a persistent enforcement pattern that regenerates faster than any single case can resolve it; and a case where the regulatory system itself, rather than the marketplace, produced years of costly ambiguity over a technicality. One final chapter draws these cases, and every chapter before them, into a single closing synthesis.

Chapter 11 — Following the Evidence

Return, one last time, to the medicine cabinet that opened this trilogy — fifteen bottles, none prescribed, most bought on a promise rather than a diagnosis. Three books ago, the question sitting behind that cabinet was simple to state and hard to answer: how much of what’s inside has actually been tested, and how much has a shopper simply assumed was tested because the packaging looked the way regulated medicine looks? This chapter closes the investigation by answering it as plainly as the evidence allows.

A systematic review of the investigation

Book One built the toolkit: an Evidence Ladder running from testimonial to systematic review, an Evidence Pyramid ranking study designs by reliability, and the beta-carotene trials as a single worked example of a plausible mechanism failing a rigorous test. Book Two applied that toolkit to the system itself, and found a regulatory structure that requires no premarket proof of safety or effectiveness, a label that discloses less than it appears to, documented liver injury and drug interactions running well ahead of what most patients or physicians track, and a business built on subscription revenue, affiliate commissions, and multi-level marketing that profits identically whether or not the product helps anyone. Book Three pointed both toolkits at the hardest material: real chronic disease claims, most of which turned out narrower or weaker than marketed, with calcium and vitamin D standing as the deliberate exception that proves the rest of the pattern; an information ecosystem that structurally rewards confident claims over accurate ones; a documented psychological reason intelligence doesn’t protect against this; a comparative regulatory picture showing the American approach is a political accident rather than a reflection of any distinctly American value; and four real case files showing the same failure patterns playing out with real products and real consequences.

Evidence Matrix: the trilogy, compressed

What changed, book by book

Book One gave you evidence literacy — the ability to hear “clinically studied” and know to ask which kind of study, to recognize a mechanism dressed as a result, and to understand why your own experience, however real, isn’t proof. Book Two gave you regulatory and economic realism — the understanding that a supplement’s presence on a shelf next to regulated medicine implies nothing about equivalent scrutiny, and that the industry’s dominant business models profit independent of whether any given product works. Book Three gave you applied judgment — the ability to run the full framework, quickly and independently, against a real diagnosis, a viral claim, or a specific product, the way Chapter 5’s rapid survey of ten categories and Chapter 10’s four case files were both built to demonstrate you now can.

What this investigation is not

It is not an argument to empty the medicine cabinet this trilogy opened with. Calcium and vitamin D, for people at real fracture risk, work. Omega-3s show a real, narrower cardiovascular benefit than marketed. Creatine does what the evidence says it does for athletic performance. St. John’s Wort has genuine antidepressant efficacy alongside a genuinely serious interaction risk. The point was never that nothing in this aisle works. The point is that the confidence with which a product gets chosen and the evidence that actually exists for it are, across case after case in this trilogy, two very different things — and that gap is close-able, by anyone willing to do what this trilogy’s seven-step framework asks: check the claim against the evidence before checking it against the price.

What comes next

Diagnosis without a recommendation is half an investigation, and this trilogy has one more page. The appendix that follows, “What Should Change,” takes Chapter 9’s international comparison and proposes a specific, tiered evidence framework for U.S. supplement regulation — modeled on what’s already working in Canada and Australia, not on an abstract ideal, and built to survive the same political reality that produced DSHEA in the first place rather than simply demanding its reversal. That appendix is where this investigation’s diagnosis becomes a proposal.

— End of Book Three, and of The American Supplement Investigation —

Appendix: What Should Change

Toward a Recommended Framework — After Following the Evidence on What Has Actually Worked

This appendix could end with the easy answer that runs under most of the preceding chapters: require the same kind of approval for supplements that we require for drugs. That intuition is largely right — a system where a manufacturer can sell a product to millions of people on the strength of a label they wrote themselves cannot be trusted to police itself. But looking at what other countries actually do — not what advocates on either side say should happen, but what has been running, successfully, for years — changes the shape of that answer. Full drug-style premarket approval for every product in an 80,000-product marketplace has no working precedent anywhere in the world. A more targeted model does. This appendix proposes that model.

Robert W. Coleman, MS Pharm.

What the evidence actually shows

I won’t repeat the case made in Book Two’s Chapters 1 through 4 and this book’s Chapter 10 here — it lives in the regulatory history and case files already documented. But four findings drive everything that follows.

First: the regulatory gap isn’t a design flaw that crept in over time — it’s the original design, and the market it was built for no longer exists. The Dietary Supplement Health and Education Act of 1994 (DSHEA) gave the FDA no authority to review a supplement’s safety or effectiveness before it reaches a shelf. That was written for a marketplace of about 4,000 products.

4,000 → 80,000+ The growth in the number of dietary supplement products on the U.S. market since DSHEA passed in 1994 — a twenty-fold expansion the law’s oversight authority never expanded to match.

$60B / $200B The estimated size of the U.S. domestic and global dietary supplement markets today.

Second: post-market enforcement, by design, can only respond after harm is already documented — and the documentation itself is thin relative to the market’s size. Between 2004 and 2021, the FDA’s adverse event system received over 79,000 reports tied to dietary supplements. A widely cited analysis of the 2004–2013 window found more than 15,000 reports that included 339 deaths and nearly 4,000 hospitalizations — and researchers who have studied the system describe it as a likely undercount, not an overcount, given how few adverse events ever get reported or connected back to a supplement at all.

Third: a workable middle path already exists, and it isn’t hypothetical. Two of the United States’ closest peer countries already run functioning premarket systems for exactly this category of product, without collapsing their markets or requiring years-long drug trials for a bottle of fish oil. Canada requires every natural health product to obtain a licence — a Natural Product Number — before sale, with the depth of evidence required scaled to the product and the claim through a monograph system. Australia runs a three-tier model: low-risk products can self-certify against a pre-approved list of ingredients and claims; products making a more specific health claim must have that claim’s evidence assessed before the product can carry it; and only products with the highest-risk profile face full registration. Neither country requires what DSHEA-era critics feared and what current maximalist reform proposals imply: that every vitamin be run through the same approval pathway as a new chemotherapy drug.

Fourth: Congress itself is already moving toward the first, most modest piece of this — a mandatory listing requirement, not premarket approval. The Dietary Supplement Listing Act, introduced in 2026, would require manufacturers to file their products with the FDA or have them treated as misbranded. It would not require anyone to prove a product works before selling it. That it is the piece currently gaining traction is itself informative: it suggests the politically viable sequence starts with knowing what is on the market, not with re-litigating DSHEA’s core bargain all at once.

None of this changes the conclusion that runs through this book: American consumers are, right now, trusting a system to have already checked something it was never authorized to check. What it does is point toward a specific, sequenced fix rather than an all-or-nothing rewrite of the 1994 law that has failed twice before in Congress.

The design principles this model is built on

• Classify by risk, not treat every claim identically. A single-ingredient vitamin at a standard dose and a compound marketed for a specific disease-adjacent outcome do not carry the same risk, and shouldn’t carry the same review burden. Australia’s tiered system is the clearest working example of this in practice.
• Require evidence in proportion to the claim, not the ingredient’s novelty. Canada’s monograph model shows that a well-established ingredient with a modest, pre-vetted claim can move through review quickly, while a new or more assertive claim draws a heavier evidence requirement — without shutting down the low-risk end of the market.
• Start with knowing what’s on the market before regulating what’s on it. Mandatory product listing is the precondition for everything else on this list, and it’s the piece already closest to becoming law.
• Fund post-market surveillance to match the market’s actual size, not its 1994 size. A twenty-fold increase in products with no corresponding increase in the FDA’s capacity to monitor them is not a sustainable trade.
• Treat the political failure mode as seriously as the safety failure mode. DSHEA itself was the direct result of a stricter FDA posture in the early 1990s provoking a consumer and industry backlash that produced weaker oversight than existed before. Any recommended model has to survive contact with that same politics, not just survive contact with the evidence.

The Recommended Model: A Tiered Evidence Framework

1. Pass mandatory product listing first. Every dietary supplement sold in the U.S. must be filed with the FDA — ingredient, dosage, manufacturer, and label claim — before sale. No listing, no legal sale. This is the single reform closest to enactment and the foundation everything else depends on.
2. Sort products into three evidence tiers, modeled directly on Australia’s Listed / Assessed Listed / Registered structure.
   • Tier 1 (Listed): single, well-established ingredients at standard doses making only pre-approved, low-level structure/function claims. Self-certified, fast to market, lightly reviewed — the current system, but only for products that actually qualify for this tier.
   • Tier 2 (Assessed): any product making a specific claim beyond basic nutrient support (cognitive function, joint health, immune “support” claims that go beyond general wellness) must submit the evidence behind that claim for FDA review before using it on a label.
   • Tier 3 (Registered): new ingredients, compounds with a documented history of adverse events or drug interactions, and any claim that approaches a disease-treatment claim require the fuller premarket safety and efficacy review Canada already applies to comparable products.
3. Give new ingredients a real premarket safety filing, modeled on Canada’s NPN evidence requirement — not a drug trial, but a documented safety dossier before a genuinely new compound reaches consumers, rather than the current system’s near-total reliance on the ingredient having existed before October 1994.
4. Fund CAERS and FDA post-market surveillance at a level proportional to an 80,000-product, $60 billion market, and extend mandatory reporting beyond “serious” adverse events to close the gap researchers have identified between reported harm and actual harm.
5. Require standardized interaction warnings on labels for ingredient categories with documented interaction risk — anticoagulants, chemotherapy agents, immunosuppressants, sedatives — the same way OTC drug labels already must disclose comparable risks.
6. Sunset the 1994 grandfather clause. Ingredients marketed before October 1994 currently face no obligation to ever produce the safety data a new ingredient would need. Require a basic safety re-review within a defined transition window, not full re-approval, so “grandfathered” doesn’t mean “permanently exempt.”

What changes for you

For consumers, the clearest visible change is a label marker — something like Australia’s AUST L / AUST L(A) distinction — that tells you at a glance whether a claim has been reviewed or simply self-certified. That single piece of information does more to close the gap this book documents than almost anything else in this framework, because it turns an invisible distinction into a visible one at the point of purchase.

For responsible manufacturers, the added burden is real but bounded: a listing filing, and — for products making a specific claim — an evidence submission most legitimate companies already possess internally, since they’re the basis for the marketing claim in the first place. Companies selling standard, well-established, modestly-claimed products see the least change. Companies whose entire model depends on an untested claim see the most friction, which is the intended effect, not a side effect.

For the industry as a whole, this model does not ban a category of product, does not require multi-year clinical trials for a multivitamin, and does not import the FDA’s drug-approval timeline wholesale into a food-regulated category. It imports the pieces of two working foreign models that are narrowest and best-evidenced, and it starts with the piece of reform already closest to passing in Congress.

Why not go straight to full drug-style premarket approval

The most direct answer to everything documented in this book — require every supplement to clear the same bar as a prescription drug — is also the version of reform that has failed twice before and has no successful precedent anywhere. No country in Part IX’s international comparison runs its entire supplement category through a full drug-approval pathway; even the stricter models in Canada and the EU reserve genuinely rigorous premarket review for the highest-risk tier, not the whole market.

There is also a political history worth taking seriously here rather than repeating: DSHEA exists because a stricter FDA posture toward supplements in the early 1990s produced a consumer and industry backlash large enough to pass a law that removed premarket authority almost entirely. A reform proposal that asks Congress to reverse that outcome in one step, rather than through the sequenced, risk-tiered model above, is the version most likely to produce the same result it produced in 1994 — not stronger oversight, but a codified retreat from it.

What this model does not solve

No framework built from an honest reading of the evidence should claim to solve everything. This model does not resolve the enforcement problem posed by imported products sold through online marketplaces outside FDA’s practical reach — a jurisdictional problem no relabeling scheme fixes by itself. It does not stop influencer-driven misinformation, which is a platform and cultural problem, not a labeling problem, and requires the kind of information-ecosystem response discussed in Part VIII, not a regulatory one. It does not guarantee its own funding — a tiered evidence framework is only as good as the appropriations behind the FDA office that has to run it, and Congress has underfunded this exact function before. And it does not, by itself, close the gap in physician and pharmacist training on supplement-drug interactions documented in Book Two’s Chapter 13 — that requires a professional-education fix this framework can support but not deliver on its own.

What it offers instead is a sequence: know what’s on the market, sort it by actual risk, require evidence in proportion to the claim, and fund the surveillance that catches what evidence review misses. It is a narrower promise than “treat supplements like drugs,” and it is the more achievable one.

This appendix draws on the regulatory history in Book Two’s Part III, the risk evidence in Book Two’s Part IV, and the international comparison in this book’s Chapter 9. It is intended, like Beyond Reform’s own updated appendix, to close the trilogy with a recommended model rather than a diagnosis alone.

How This Book Sources Its Claims

Every factual claim in this book is sourced — to peer-reviewed research, federal regulatory documents, court records, or government data. Where a figure represents an estimate or a range rather than a single precisely observed count, this book says so, and reports the range rather than a single headline number that would sound more certain than the underlying research actually is. The full bibliography follows this note.

If you believe a citation in this book has been misrepresented, I want to know — reach out through Coleman Publishing or brokenpromiseshealthcare.org, and it will be reviewed.

Bibliography

Manson, JoAnn E., et al. “Vitamin D Supplements and Prevention of Cancer and Cardiovascular Disease.” New England Journal of Medicine 380, no. 1 (2019): 33–44.

Manson, JoAnn E., et al. “Marine n-3 Fatty Acids and Prevention of Cardiovascular Disease and Cancer.” New England Journal of Medicine 380, no. 1 (2019): 23–32.

DeKosky, Steven T., et al. “Ginkgo biloba for Prevention of Dementia: A Randomized Controlled Trial.” JAMA 300, no. 19 (2008): 2253–2262.

Bent, Stephen, et al. “Saw Palmetto for Benign Prostatic Hyperplasia.” New England Journal of Medicine 354, no. 6 (2006): 557–566.

Barry, Michael J., et al. “Effect of Increasing Doses of Saw Palmetto Extract on Lower Urinary Tract Symptoms: A Randomized Trial.” JAMA 306, no. 12 (2011): 1344–1351.

Bolland, Mark J., et al. “Calcium Supplements With or Without Vitamin D and Risk of Cardiovascular Events.” BMJ 341 (2010): c3691.

Pharmanex, Inc. v. Shalala, 35 F. Supp. 2d 1341 (D. Utah 1999), rev’d, 221 F.3d 1151 (10th Cir. 2000).

U.S. Food and Drug Administration. Final Decision on Cholestin (Docket No. 97P-0441), 1998.

U.S. Food and Drug Administration. Final Rule Declaring Dietary Supplements Containing Ephedrine Alkaloids Adulterated, 2004.

RAND Corporation. Ephedra and Ephedrine for Weight Loss and Athletic Performance Enhancement: Clinical Efficacy and Side Effects, 2003.

U.S. Food and Drug Administration. Tainted Products Marketed as Dietary Supplements (public warning database).

Natural Products Association. Correspondence and citizen petition records regarding NMN’s status under the Dietary Supplement Health and Education Act, 2022–2025.

U.S. Food and Drug Administration. Letters to New Dietary Ingredient Notification holders regarding NMN, November 2022 and December 2025.

Health Canada. Natural Health Products Regulations and Natural Product Number licensing framework.

Australian Therapeutic Goods Administration. Listed, Assessed Listed, and Registered medicines framework.

European Parliament and Council. Directive 2002/46/EC on the approximation of the laws of the Member States relating to food supplements.

European Food Safety Authority. Novel Foods Regulation (EC) No. 2015/2283, safety assessment procedures.

Stanovich, Keith E. What Intelligence Tests Miss: The Psychology of Rational Thought. Yale University Press, 2009.

Dietary Supplement Health and Education Act of 1994, Public Law 103-417, 103rd Congress.

About the Authors

Joy Meier, Pharm.D.

[Placeholder — bio to be supplied. Suggested elements to include: clinical background and credentials, current or most recent practice setting, relevant specialization (e.g., ambulatory care, nutrition, clinical pharmacology), any supplement- or dietary-science-specific expertise or publications, and how she came to co-author this investigation.]

Robert W. Coleman, MS Pharm.

Robert W. Coleman spent forty years in medicine, sixteen of them as Director of Clinical Pharmacy Services at the VA Palo Alto Health Care System, where he ran medication safety programs for tens of thousands of veterans. His peer-reviewed research spans infectious disease, pharmacokinetics, anticoagulation therapy, and medical informatics.

He writes as an investigator, not an advocate. Prove It is the third and closing volume of The American Supplement Investigation, a companion series to his healthcare-financing trilogy, The American Healthcare Investigation. His work is published under the Coleman Publishing imprint and can be found, alongside his full research library and sourcing archive, at brokenpromiseshealthcare.org.