WHAT THE LABEL DOESN’T TELL YOU

The American Supplement Investigation — Book One

The Promise and the Evidence

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, 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.

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

Preface

I spent sixteen years as Director of Clinical Pharmacy Services at the VA Palo Alto Health Care System, and for most of them, a patient would occasionally hand me a small paper bag before an appointment and ask me to look through it. Inside would be a prescription bottle or two — and six, eight, sometimes fifteen supplements. Fish oil. A multivitamin. Turmeric. Something with a name I didn’t recognize, bought because a friend, a podcast, or a headline had said it would help.

Almost none of those patients could tell me why they were taking a given supplement, in the sense of what evidence stood behind it. Nearly all of them could tell me who had recommended it, and how it made them feel. That gap — between the confidence with which a product is chosen and the evidence that actually exists for it — is the subject of this book.

I want to be direct about something before you read further: this is not an anti-supplement book. Some supplements work, for some people, for some purposes, and the evidence for a few of them is genuinely strong. This book is not an argument that the whole category is fraudulent, and readers looking for that argument won’t find it here. What they will find is a plain account of how a claim moves from a plausible idea to a shelf label, what kind of evidence that claim is actually required to survive along the way, and how often the answer turns out to be: less than you’d assume.

I write as an investigator, not an advocate. Forty years in medicine — sixteen of them running medication safety programs for tens of thousands of veterans — taught me that the most dangerous claims are rarely the obviously false ones. They’re the ones that are technically true, biologically plausible, and still misleading, because plausible is not the same as proven, and no one told the person holding the bottle the difference.

This book is the first of three. Its job is narrower than the other two, and it is deliberately narrow: before we can examine the regulatory gap that lets a product reach a shelf without being tested (Book Two), or apply that scrutiny to real disease claims and real case histories (Book Three), you need the tools to evaluate a health claim on your own. That’s what these seven chapters are for. By the end of them, you should be able to hear “clinically studied” or “science-backed” on a label and know exactly what those phrases do and don’t obligate anyone to have proven. You should be able to tell the difference between a mechanism and a result. And you should understand why a story about what worked for someone — even someone you trust completely — is not the same kind of evidence as a trial.

I’ve built this book around one recurring example: a vitamin that doctors recommended in good faith, for sound biological reasons, to smokers trying to protect themselves from lung cancer — and that turned out, when finally tested rigorously, to increase their risk instead. I didn’t choose that story because it’s shocking. I chose it because nobody involved was lying, cutting corners, or acting in bad faith. The scientists who proposed it, the doctors who recommended it, and the patients who took it were all doing exactly what the evidence available to them at the time seemed to support. The system worked, eventually — a trial was run, the result was clear, and the recommendation changed. What this book asks is a harder question: how much of what’s on the shelf today is still waiting for that same trial to be run at all.

Every factual claim in this book is sourced — to peer-reviewed research, federal regulatory documents, or government data — the same standard I hold the rest of this investigation to. Where the evidence is genuinely mixed, I say so, and I report the range rather than a single number that sounds more certain than the research actually is.

This is not a book that will tell you which supplements to take. It’s a book that will tell you how to ask the right question before you decide for yourself.

Robert W. Coleman, MS Pharm.

PART I — THE PROMISE

Chapter 1 — The Supplement in the Medicine Cabinet

Open the medicine cabinet in a typical American bathroom and you’ll usually find, at most, two or three prescription bottles. Open the kitchen cupboard next to it, or the shelf by the coffee maker, or the gym bag by the door, and you’ll often find a great deal more: a multivitamin, a fish oil, a bottle of vitamin D, something for joints, something for sleep, something for stress, a greens powder, a protein blend, maybe a tub of something with a name that sounds more like a mission statement than an ingredient. Fifteen products would not be unusual. None of them were prescribed. Most of them were bought on the strength of a promise rather than a diagnosis.

That cabinet, multiplied across roughly a hundred and thirty million American households, is the actual subject of this book.

The scale of the cabinet

When Congress passed the law that still governs how dietary supplements are sold in the United States, in 1994, there were roughly 4,000 such products on the market. Today there are more than 80,000. The domestic market is worth an estimated $60 billion a year; the global market, roughly $200 billion. Depending on which survey you consult, somewhere close to three in four American adults report taking a dietary supplement of some kind, and a large share take more than one.

This is not a niche behavior. It is closer to a national default.

Why people take them

The reasons cluster into a handful of familiar categories, and it’s worth naming them plainly, because the rest of this book is going to spend a great deal of time asking what evidence stands behind each one. People take supplements for prevention — an attempt to head off disease before it arrives. They take them for energy, for longevity, for immunity, for athletic or cognitive performance. And a great many people take them as a kind of insurance — a hedge against the possibility that their diet isn’t quite giving them everything it should, just in case.

None of these motives is foolish. Wanting to stay healthy, feel better, and hedge against uncertainty is about as reasonable a set of goals as a person can have. The question this book asks isn’t whether the motive is sound. It’s whether the product in front of you is actually the right tool for it.

The appeal of “natural”

Part of what makes a supplement feel like a safe bet, compared to a prescription drug, is the word that shows up on so many of their labels: natural. It reads as a kind of built-in reassurance — nothing synthetic, nothing manufactured in a lab, nothing that could really hurt you. We’ll take that assumption apart in detail later in this book, in a chapter titled, not coincidingly, “Natural Does Not Mean Safe.” For now, it’s enough to notice how much quiet work that single word is doing on a label, and how rarely a shopper stops to ask what it’s actually claiming.

Supplement or drug? The distinction that changes everything

Here is the fact that the rest of this book depends on, and it surprises most people the first time they hear it stated plainly: under U.S. law, a dietary supplement is not classified as a drug or a medicine. It’s classified as a category of food.

That single classification decision is the hinge on which almost everything else in this investigation swings. A prescription drug cannot legally be sold in the United States until its manufacturer has proven, through years of clinical trials, that it is both safe and effective for a specific use, and the Food and Drug Administration has reviewed that evidence and approved it. A dietary supplement faces no such requirement. A manufacturer can bring a new product to market, put it on a shelf, and start selling it to the public without ever submitting evidence to anyone that it works, and in many cases without notifying the FDA that the product exists at all. We’ll walk through exactly how that came to be — and exactly what it does and doesn’t mean for your safety — in Part III of this book. For now, hold onto the plain fact of it: supplement is a legal category, not a guarantee.

The trust gap

This brings us to the actual finding this chapter exists to establish, and it isn’t about what’s inside the bottle. It’s about what’s inside the shopper’s head standing in front of the shelf.

Survey after survey finds the same thing: a majority of Americans believe, at some level, that a product sitting on a store shelf next to the vitamins and the aspirin must have cleared some kind of government safety and effectiveness check before it got there. It’s an entirely reasonable thing to assume — most of the other products on that same shelf did clear exactly that kind of check. It is also, for the supplement sitting between them, usually false.

I want to introduce you here to a story we’ll return to twice more in this book, at increasing depth, because it’s the clearest illustration I know of exactly how that trust gap can matter. In the early 1990s, doctors and researchers had good reason to believe that beta-carotene — the antioxidant compound that gives carrots their color — might protect smokers from lung cancer. The biology made sense. The early population data was encouraging. So researchers ran the trials that the best science of the time called for.

The results, when they came back, were the opposite of what everyone had expected. We’ll come back to exactly what happened, and exactly why, in Chapter 5. For now, the point isn’t the outcome. It’s the process: a plausible idea, believed in good faith by serious scientists, was tested — and the test changed the recommendation. That is the system working as it should.

The uncomfortable question underneath this entire book is how much of what’s sitting in that medicine cabinet has ever been through a test like that one at all.

Investigator’s Note — What Exactly Is a Dietary Supplement?

Legally, a dietary supplement is a product intended to supplement the diet that contains one or more of the following: a vitamin, a mineral, an herb or other botanical, an amino acid, or a dietary substance used to increase total dietary intake. It’s meant to be taken as a pill, capsule, tablet, powder, or liquid — and, critically, it must not be represented as a conventional food or as the sole item of a meal or diet. That last distinction is what keeps a protein bar marketed as “lunch” in a different legal category than a protein powder marketed as a supplement to your existing diet, even when the ingredients inside them look nearly identical.

Key Finding: The first investigative problem is not what supplements contain. It is what consumers believe the products have already been proven to do.

The next chapter follows that belief back to its source — not the ingredient list, but the advertisement that put the bottle in your hand in the first place.

Chapter 2 — Follow the Promise

Picture the ad. You’ve seen a version of it, whether you noticed or not. A woman in her forties, lit warmly, describes how tired she’d felt for years before “discovering” a supplement — cut to a bottle turning slowly against a clean background, a graphic of a molecule pulsing gently, a small printed badge in the corner reading “Clinically Studied.” A comment underneath from a stranger: “This changed my life.” A link. A discount code good for the next forty-eight hours only.

By the time that thirty-second video ends, it has done an enormous amount of persuasive work, almost none of which involved telling you anything specific and verifiable about what the product actually does. This chapter is about how that work gets done.

The consumer journey

Marketers who study this professionally describe the path from a stranger’s attention to a completed sale as a journey with five stages: Concern, Advertisement, Promise, Purchase, Belief. It’s worth walking through slowly, because each stage does something the next stage depends on.

Concern comes first, and it’s usually already there before the advertisement arrives — a nagging worry about energy, aging, weight, sleep, or a family history of some condition. Advertising doesn’t have to manufacture concern from nothing; mostly it has to find the concern that’s already present and speak directly to it.

Advertisement is the moment that concern meets a proposed solution. What makes supplement advertising distinct from, say, car advertising is how directly it collapses the space between problem and solution into a single frame: you’re tired, here is the reason, here is the fix.

Promise is the specific claim the advertisement makes, whether stated outright or implied strongly enough that most viewers will walk away with roughly the same idea. We’ll spend the next section of this chapter on exactly how carefully that promise is often worded.

Purchase happens fastest when the previous three stages have done their job well and a fourth pressure gets added: scarcity or urgency — a countdown, a limited stock notice, an introductory price.

Belief is the final and most important stage, because it’s the one that produces the testimonial that becomes the next advertisement’s opening scene. A single satisfied customer, genuinely convinced the product worked, is worth more to a supplement marketer than almost any other asset the company owns — which is exactly why Chapter 3 is devoted entirely to examining how reliable that belief actually is.

Two channels, examined closely

Supplement marketing runs through an enormous number of channels — retail end-caps, print ads, sponsored articles, podcast reads, direct mail. Rather than skim across all of them, it’s worth looking closely at two that have reshaped the landscape most in the last decade.

Influencer marketing works because it borrows a relationship that already exists. A viewer who follows a fitness creator, a wellness podcaster, or a parenting account for months or years has built something that functions like trust, even though the relationship is one-directional and commercial. When that creator recommends a product, the recommendation arrives wrapped in all the credibility of a friend’s advice, backed by none of a friend’s actual stake in your outcome — the creator is typically paid whether the product works for you or not, often through an affiliate link that pays them a percentage of your specific purchase.

Retail placement works through a quieter mechanism: proximity and repetition. A product placed at eye level, at the checkout counter, or next to a category leader benefits from the assumption most shoppers make without noticing they’re making it — that placement reflects merit rather than a paid arrangement between the manufacturer and the retailer. Both matter. Neither is illegal. Both are worth knowing about the next time a product seems to be “everywhere” all at once.

The vocabulary of persuasion

A handful of phrases do an enormous amount of work on supplement labels and in supplement advertising, and each one is worth defining plainly, because the phrase itself usually promises far more certainty than it’s legally required to deliver.

“Doctor recommended” can mean a single physician, paid or unpaid, was willing to say the words on camera. It does not mean a professional medical body has reviewed and endorsed the product.

“Clinically studied” can mean anything from a single small, unpublished, industry-funded study on a handful of people to a large, independent, peer-reviewed clinical trial. The phrase itself doesn’t distinguish between these, and the label is under no obligation to tell you which one it’s describing.

“Natural” describes the source of an ingredient, not its safety, its dose, or its effect on your body — a point Chapter 12 will take up in full.

“Ancient remedy” invokes tradition as if longevity of use were evidence of effectiveness. It tells you a substance has been used for a long time. It tells you nothing about whether it works, at what dose, or with what risks — traditional use predates the tools we now have to actually measure any of those things.

“Science-backed” is perhaps the vaguest phrase of all, because it makes no specific claim whatsoever about what kind of science, how much of it, or what it actually found.

Keep this short glossary close. You’ll be able to put it to use the moment you next stand in front of a supplement shelf.

The Evidence Ladder

There’s one more idea worth planting here, before we build it out completely in Chapter 4: not all evidence is the same kind of evidence. Picture a ladder with six rungs, from bottom to top: a testimonial, an anecdote, a mechanistic theory about why something should work, an observational study, a randomized clinical trial, and — at the very top — a systematic review that combines many trials together.

Marketing claims cluster overwhelmingly at the bottom two rungs of that ladder. That’s not an accident, and it’s not necessarily dishonest — the bottom rungs are simply cheaper, faster, and easier to produce than the top ones. But a claim that never climbs higher than “someone said it worked” or “here’s a story about someone it worked for” is a claim that has not yet been tested in any way that could actually rule out the explanations Chapter 3 is about to walk through in detail.

Evidence Spotlight — What Does “Clinically Studied” Actually Mean?

Nothing specific, and that’s the honest answer. The phrase carries no legal definition and no minimum evidentiary bar. It could describe an eight-week, industry-funded pilot study on twelve people that was never published anywhere a scientist could scrutinize it. It could also describe a large, independent, peer-reviewed randomized trial. The label will not tell you which. The only way to find out is to look for the actual study — and Chapter 4 will show you exactly what to look for once you find it.

The advertisement, the testimonial, and the label vocabulary all point toward the same underlying resource: somebody’s personal story about what happened when they took the product. That story is the single most persuasive ingredient in the entire marketing apparatus this chapter just described — and it’s also, on its own, worth far less than it feels like in the moment. That’s where we turn next.

Chapter 3 — The Power of Anecdote

She’d had trouble sleeping for years. Nothing dangerous, nothing diagnosed — just the slow accumulation of tired mornings that comes from lying awake too long too often. A coworker mentioned a magnesium supplement. She bought a bottle on the way home. Within a week and a half, she was sleeping better, and she has taken it every night since. Ask her whether it works and she won’t hesitate: absolutely, without question, and she’ll tell anyone who asks the same thing.

I want to be clear about something before this chapter goes any further: I have no reason to doubt that she’s sleeping better. The improvement is almost certainly real. What this chapter is about is a much narrower and more uncomfortable question — not whether something changed, but whether the magnesium is what changed it.

Why one story beats a thousand pages

There’s a reason a single vivid account can outweigh an entire body of published research in most people’s minds, and it isn’t stupidity or gullibility. Human beings are built to learn from narrative. A specific person, with a specific problem, telling a specific story with a beginning and an end, engages attention and memory in a way that an abstract statistic simply doesn’t. “Forty-one percent of participants in a twelve-week trial reported improved sleep quality on a validated questionnaire” is accurate and almost instantly forgettable. “It worked for me” is neither of those things, and it works on nearly everyone, including people who know better.

That’s what makes anecdote worth taking seriously as a subject, rather than dismissing as simple gullibility. The pull is real, and it works on careful thinkers as readily as anyone else. The question this chapter exists to answer is what else, besides the supplement itself, might explain the exact same story.

Four explanations, before you reach for the supplement

Confirmation bias. Once someone believes a product is helping, they tend to notice the days that confirm it and explain away the days that don’t. A good night’s sleep becomes evidence the magnesium is working; a bad one becomes evidence of an unrelated stressful day. The belief filters the evidence rather than the evidence shaping the belief.

The placebo effect. This is not “it’s all in your head” in the dismissive sense that phrase usually carries — placebo effects are real, measurable, and often substantial, particularly for exactly the kind of subjective symptoms supplements are most often marketed against: fatigue, mood, pain, and sleep quality. Expecting to feel better is itself a mechanism that can make a person feel better, entirely apart from anything in the capsule.

Regression to the mean. People overwhelmingly try a new supplement at their worst moment — the worst stretch of insomnia, the lowest energy, the most stressed. Statistically, a bad stretch is often followed by some natural improvement even if nothing changes at all, simply because extreme periods tend to move back toward a person’s normal baseline over time. Starting a supplement right before that natural improvement arrives makes the supplement look responsible for something that was already on its way.

Natural fluctuation of disease and symptoms. Chronic conditions wax and wane on their own schedule. Joint pain, digestive symptoms, mood, and energy all have good weeks and bad weeks independent of anything a person does. A supplement started during a bad stretch, followed by an ordinary good stretch, earns credit for a recovery that had very little to do with it.

None of these four explanations require anyone to be lying, careless, or foolish. They require only that the human brain works the way it actually works — which is to say, they apply to everyone, including the person writing this book.

What anecdote is actually good for

None of this means personal stories are worthless. They serve a real and legitimate scientific purpose: generating hypotheses worth testing. If enough people independently report the same specific effect from the same specific substance, that pattern is worth a researcher’s attention — it’s exactly how many productive lines of medical research have started. The failure isn’t in noticing the pattern. It’s in treating the noticed pattern as though it were already the tested conclusion, skipping every step between “worth investigating” and “proven to work” that Chapters 4 through 6 are about to walk through.

Questions to Consider

If a person improves after taking a supplement, what else might explain the improvement? Before turning the page, try building your own list. Chances are it will look a great deal like the four explanations above — which is itself worth noticing. The tools for skepticism aren’t exotic or specialized. Most people already have them. The trick is remembering to use them at the exact moment a story feels most convincing.

Key Finding: Anecdotes can generate hypotheses. They cannot establish effectiveness.

If a personal story isn’t enough, the next four chapters build, piece by piece, what actually is — starting with the most basic question of all: what does the word “evidence” mean in the first place?

PART II — THE EVIDENCE

Chapter 4 — What Does “Evidence” Really Mean?

Go back to the phrase from Chapter 1’s teaser: “scientifically proven.” It’s one of the most common claims in health marketing, and one of the least examined. Proven how? By whom? To what standard? This chapter builds the framework you need to actually answer those questions the next time you see the phrase — starting with the idea that not all evidence occupies the same rank.

The Evidence Pyramid

Picture research evidence arranged as a pyramid, with the least reliable forms at the wide base and the most reliable at the narrow top.

At the base sit laboratory and test-tube studies — research on cells or isolated compounds, useful for understanding a possible mechanism, but many steps removed from what happens inside an actual human body. Just above them, animal studies add a layer of biological realism but still can’t tell you how a substance behaves in people, whose metabolism, dosing tolerance, and disease processes often differ in ways that matter enormously.

Next come case reports — a single patient’s documented experience, useful for raising a question, incapable of answering one. Above those sit observational studies, which follow real groups of people over time and look for associations — people who take more of X tend to have less of Y — without controlling who takes what. Observational studies can be large and genuinely valuable, but they can’t rule out the possibility that something else entirely explains the pattern they find.

Near the top sits the randomized controlled trial: a study that assigns people to receive either the substance being tested or a placebo, by chance, and compares what happens to each group. Randomization is what lets a trial rule out the confounders from Chapter 3 — if the two groups are alike in every respect except what they were given, a difference in outcome is far more likely to be caused by what they were given.

At the very top sits the systematic review and meta-analysis — a study that combines the results of many trials together, in an effort to see what the whole body of evidence shows rather than relying on any single trial that might have been unusually lucky or unlucky.

Statistical significance is not the same as mattering

One distinction gets collapsed constantly in consumer health coverage, and it’s worth stating plainly: a result can be statistically significant — meaning the difference measured is unlikely to be due to chance alone — while being clinically insignificant, meaning the actual size of the difference is too small to matter to how a person feels, functions, or lives. A supplement might “significantly” lower a cholesterol number by an amount so small that no doctor would consider it clinically meaningful. Both things can be true about the exact same study, and marketing will reliably lead with the first one and omit the second.

A preview of publication bias

One more distortion is worth flagging here, because it applies to how research becomes visible in the first place, not just to how any single study is designed. Positive results — the ones that find an effect — are simply more likely to be published, publicized, and cited than negative or null results, particularly when the research is funded by a company with a product to sell. We’ll return to this directly in Chapter 6, once you’ve seen what a well-designed trial actually looks like and have a clearer sense of what might be missing from the published record around it.

The ladder, completed

Return now to the Evidence Ladder from Chapter 2 — testimonial, anecdote, mechanistic theory, observational study, clinical trial, systematic review — and walk it top to bottom using the beta-carotene story first introduced in Chapter 1.

It began, decades ago, with observational data: researchers noticed that people whose diets were naturally rich in beta-carotene — from vegetables, not pills — tended to have lower rates of lung cancer. That’s a real association, honestly reported, sitting at the observational rung of the ladder. It was not, on its own, proof that beta-carotene itself was the protective ingredient, or that taking it as an isolated supplement would produce the same benefit. The only way to find out was to climb to the top of the ladder and run the trial — randomizing smokers to receive either beta-carotene or a placebo and watching what happened. We’ll see exactly what that trial found, and why, in the next chapter.

Notice what the ladder does here: it doesn’t say the early observation was wrong to report, or that the scientists who noticed it were careless. It says that an association sitting at one rung of the ladder cannot be treated as though it already occupied a much higher one — and that the distance between those two rungs is exactly where a great deal of supplement marketing chooses to stop climbing.

Research Update

[Editorial note: this box is a template. At final drafting, insert the most current relevant finding on evidence quality or a recent high-profile supplement trial result, sourced and dated, to keep the volume’s examples current through publication.]

Key Finding: “Science-backed” is not a scientific category. The quality of evidence matters.

Evidence quality explains method. But long before most claims ever reach a trial, they rest on something upstream of any of it: a biological explanation that sounds, on its face, like it has to be true. The next chapter takes that apart.

Chapter 5 — When Biology Sounds Like Proof

Here is the reasoning, laid out exactly as researchers made it in the 1980s, because it deserves to be taken seriously rather than mocked in hindsight: beta-carotene is an antioxidant. Oxidative damage to cells is understood to contribute to cancer development. Smokers are exposed to enormous oxidative stress from cigarette smoke. Therefore, giving smokers beta-carotene should help protect them against lung cancer.

Every link in that chain was reasonable. The biochemistry was real. The epidemiological signal — people who ate more beta-carotene-rich vegetables tended to have lower lung cancer rates — was real too, and had been observed independently by multiple research groups. By the standards of the evidence available at the time, this was about as promising as a preventive hypothesis gets. It was plausible enough that public health messaging began encouraging antioxidant supplementation well before the definitive trials were run.

What the trials found

Two major randomized trials set out to test the hypothesis directly. The Alpha-Tocopherol, Beta-Carotene Cancer Prevention Study — ATBC — enrolled more than 29,000 male smokers in Finland and randomly assigned them to receive beta-carotene, vitamin E, both, or a placebo, then followed them for years. The Beta-Carotene and Retinol Efficacy Trial — CARET — enrolled a comparable population of smokers and people with significant asbestos exposure in the United States.

Both trials found the opposite of what they were designed to confirm. In the ATBC trial, participants who received beta-carotene had 18 percent more lung cancer diagnoses and 8 percent more deaths than those who received a placebo. CARET found a similarly elevated risk, strong enough that both trials stopped their beta-carotene arms early rather than continue exposing participants to what now looked like harm rather than protection. A later review by the U.S. Preventive Services Task Force, examining beta-carotene trials together, found an overall 20 percent increased risk of lung cancer associated with supplementation, concentrated most strongly among people already at elevated risk — smokers, above all.

Why the mechanism was wrong in practice, without being wrong in theory

This is the part worth sitting with. The antioxidant theory wasn’t nonsense — beta-carotene genuinely does have antioxidant properties, and oxidative stress genuinely is implicated in cancer biology. What the theory missed is that a compound behaves differently in the specific chemical environment of a smoker’s already-stressed lung tissue than it does in a laboratory dish, and differently again as an isolated, concentrated supplement than as one component of a whole vegetable eaten alongside hundreds of other compounds. Researchers have since proposed that breakdown products formed when isolated beta-carotene is metabolized may behave in ways the original antioxidant theory never accounted for — a mechanism discovered only after the trial result forced scientists to go looking for one.

Notice the order of operations here, because it’s the entire point of this chapter: the mechanism didn’t predict the outcome. The trial produced a surprising outcome, and the mechanism had to be revised to explain it. A biological explanation that sounds airtight beforehand is not the same thing as a result that has actually been observed.

The same pattern, briefly, elsewhere

The beta-carotene story is not an isolated case of biology overpromising. “Antioxidants fight free radicals” is true as far as it goes, but it says nothing about whether swallowing a concentrated dose of any particular antioxidant, at any particular time, produces a measurable health benefit for a given person — a claim that has to be tested separately for every compound and every proposed use. “Boosting immunity” invokes an image of a single dial that can be turned up, when immune function is in reality a vast, balanced system in which “more active” is not straightforwardly “better” — an overactive immune response is the underlying mechanism of autoimmune disease. Claims about gut health and brain health frequently follow the identical shape: a real, interesting mechanism, described accurately, extended into a specific product claim that mechanism alone cannot support.

Evidence Spotlight — Plausible Does Not Mean Proven

A biological mechanism tells you an effect is possible. It cannot tell you whether the effect actually occurs in real people, at the dose sold, with the balance of benefit and risk a consumer deserves to know about before taking it. Only a trial can tell you that — and until one has been run, “here’s why it should work” and “here’s proof that it does” remain two entirely different sentences, however similar they sound on a label.

The trial is where a plausible idea like this one finally meets reality. The next chapter goes inside one — using ATBC itself as the model — to show exactly how that test is built, and what it takes for a result to be trusted.

Chapter 6 — The Clinical Trial Test

Return one more time to the ATBC study, not for its result now, but for its architecture. More than 29,000 male smokers, enrolled in Finland, randomly assigned to one of four groups — beta-carotene, vitamin E, both, or a placebo — and followed for years, with lung cancer diagnosis as the outcome the researchers were actually watching for. Almost every feature of a trustworthy clinical trial is visible in that one sentence, and it’s worth pulling each one out individually.

Randomization and blinding

Assigning participants to a group by chance, rather than letting them choose or letting a researcher decide, is what makes a trial capable of ruling out the confounders from Chapter 3. If the two groups are alike in every respect — age, smoking history, diet, health status — except which pill they received, then a difference in outcome between them is very unlikely to be explained by anything other than the pill. Blinding — keeping participants, and ideally the researchers assessing outcomes, unaware of who received what — closes off the placebo effect and any unconscious bias in how results get measured or reported. A great many studies cited in supplement marketing lack one or both of these features, which is precisely why their results carry far less weight than a headline claiming “clinically proven” would suggest.

Sample size and endpoints

ATBC studied tens of thousands of people and measured a hard, unambiguous endpoint: a diagnosis of cancer, confirmed by medical records. Compare that to the shape of a typical study behind a supplement marketing claim — often several dozen participants, followed for a few weeks, measuring a change in a lab value or a self-reported symptom score rather than an outcome like disease diagnosis or death. Both kinds of study can be legitimate science. They are not, however, equally capable of answering the question a consumer actually cares about, and a label that cites “a clinical study” without specifying its size or endpoint is inviting you to assume the ATBC kind when the reality is very often the other kind.

Absolute risk versus relative risk

This distinction deserves a worked example, because it’s one of the most common ways a technically true number misleads. Imagine a supplement is said to “cut your risk of a condition in half.” If the condition affects 2 people in 1,000 without the supplement, cutting that risk in half means it affects roughly 1 person in 1,000 with it — a real effect, but a difference of one additional person out of a thousand, not the dramatic transformation “cut in half” implies to most readers. The relative risk reduction is 50 percent. The absolute risk reduction is 0.1 percent. Both numbers describe the same result honestly. Only one of them is likely to appear on the label.

Publication bias, in full

Chapter 4 flagged this; here is the complete picture. Studies that find a positive effect are substantially more likely to be published, and published quickly, than studies that find no effect or a harmful one — a pattern documented extensively across pharmaceutical research and, if anything, harder to track in the supplement industry, which faces far less regulatory pressure to register a trial’s existence in advance or disclose its results regardless of outcome. A company that funds five small studies and publicizes only the one with a favorable result is not lying about that one study. It is allowing the published record to tell a story the full set of evidence would not support.

What Consumers Hear vs. What the Trial Actually Found

Every method in this chapter assumes that evidence, once gathered, stays settled — that a trial result is a fixed point you can build a permanent recommendation on. It isn’t always. The next chapter turns to what happens when the evidence itself changes.

Chapter 7 — When the Evidence Changes

Close the loop now on the story this book has followed since Chapter 1. For years, based on sound biological reasoning and encouraging early data, doctors and public health voices encouraged beta-carotene supplementation, including specifically for smokers hoping to protect themselves from lung cancer. Then two large, well-designed trials found the opposite of what everyone expected, and the recommendation reversed — not gradually, and not reluctantly, but as directly as the evidence demanded.

I want to name something before going further, because it’s easy to miss if you’re reading this story primarily for outrage: this is what the scientific process is supposed to look like. Nobody involved was dishonest. The hypothesis was reasonable. The trial was run at real cost and real effort specifically to find out whether the hypothesis held up, and when it didn’t, the recommendation changed to match what the trial found rather than what everyone had hoped for. A system incapable of that kind of reversal would be a far more dangerous system than one that occasionally has to correct itself in public.

The Scientific Correction Cycle

Map beta-carotene onto the general pattern, stage by stage, because you’ll see this same shape recur throughout the rest of this trilogy. First, an early observational signal — people with beta-carotene-rich diets had lower lung cancer rates. Second, public enthusiasm and commercial adoption — the biological story was compelling enough that supplementation was encouraged before the definitive trials existed to support it. Third, a definitive trial — ATBC and CARET, large, randomized, and designed specifically to test the hypothesis directly. Fourth, reassessment — researchers had to explain a result nobody had predicted. Fifth, a changed recommendation — beta-carotene supplementation for cancer prevention, particularly among smokers, is no longer advised.

A second example, briefly

The pattern isn’t unique to one vitamin. Vitamin E followed a similar arc with cardiovascular disease: encouraging early observational data suggested it might protect the heart, enthusiasm and supplementation followed, and large randomized trials run afterward found no cardiovascular benefit — with some evidence of increased risk in certain populations. The details differ from beta-carotene’s story, but the shape doesn’t: a plausible early signal, followed by a rigorous test, followed by an honest revision.

Why this should increase your trust in the process, not decrease it

Here is the argument I want to leave you with at the close of this book. The alternative to a system capable of reversing a recommendation is not a system that’s always right the first time. It’s a system that never admits being wrong — which is a far worse outcome for everyone depending on it. Beta-carotene’s story is, in the end, a story about science doing exactly what it’s supposed to do.

But notice what that same story does not tell you: it doesn’t tell you how many other claims currently sitting on supplement shelves have ever been tested with anything resembling ATBC’s rigor at all. The correction cycle worked, eventually, for beta-carotene — because someone funded a trial large enough and rigorous enough to force an answer. The uncomfortable question this evidence-literacy volume leaves you with is how much of what’s for sale today is still waiting for that same trial to be run, and whether the law actually requires anyone to run it.

Research Update

[Editorial note: template, as in Chapter 4 — insert the most current relevant evidence-reversal example at final drafting to keep the volume current through publication.]

How to Use the Rest of This Investigation

You now have the tools this book set out to build. You can read a label and recognize which rung of the Evidence Ladder a claim is actually standing on. You can hear “clinically studied” and know to ask which kind of study. You can spot a mechanism dressed up as a result, and you understand why your own experience — or anyone else’s — however real, isn’t the same kind of evidence as a trial.

Book Two turns to a harder and, in some ways, more uncomfortable question. Everything in this volume has assumed a world where evidence eventually gets tested — where enough scrutiny, enough funding, and enough time produces an answer, the way it did for beta-carotene. What Book Two investigates is what the law actually requires anyone to prove before a product reaches your shelf in the first place, and what happens in the enormous space where that requirement turns out to be far smaller than most Americans assume.

— End of Book One —

How This Book Sources Its Claims

Every factual claim in this book is sourced — to peer-reviewed research, federal regulatory documents, 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

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

U.S. Food and Drug Administration. “Information for Consumers on Using Dietary Supplements.” FDA.gov.

U.S. Food and Drug Administration. “Dietary Supplements.” FDA.gov.

The Alpha-Tocopherol, Beta Carotene Cancer Prevention Study Group. “The Effect of Vitamin E and Beta Carotene on the Incidence of Lung Cancer and Other Cancers in Male Smokers.” New England Journal of Medicine 330, no. 15 (1994): 1029–1035.

Omenn, Gilbert S., et al. “Effects of a Combination of Beta Carotene and Vitamin A on Lung Cancer and Cardiovascular Disease.” New England Journal of Medicine 334, no. 18 (1996): 1150–1155.

Tanvetyanon, Tawee, and Gerold Bepler. “Beta-Carotene in Multivitamins and the Possible Risk of Lung Cancer Among Smokers Versus Former Smokers: A Meta-Analysis and Evaluation of National Brands.” Cancer 113, no. 1 (2008): 150–157.

U.S. Preventive Services Task Force. Recommendation Statement: Vitamin, Mineral, and Multivitamin Supplementation to Prevent Cancer and Cardiovascular Disease.

Yusuf, Salim, et al. “Vitamin E Supplementation and Cardiovascular Events in High-Risk Patients” (The Heart Outcomes Prevention Evaluation Study Investigators). New England Journal of Medicine 342, no. 3 (2000): 154–160.

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. What the Label Doesn’t Tell You is the first 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.