Case Study 2 — Reading a Diet Book: Anatomy of a Bestseller

A media and rhetoric case. The book described is a composite assembled from the structural features that recur across the genre; the mechanics are not composites.

Setup

I have read a great many diet books, mostly because patients bring them to me, and after about the twentieth you stop reading them as arguments and start reading them as architecture.

They are built to a plan. Not because their authors are cynical — most are sincere, many are clinicians, and several have genuinely helped people — but because this is the structure that works, and the market selects for it.

Here is the plan, in nine parts. Once you can see it, you can read any book in the genre in about twenty minutes and know exactly what you're holding.

The nine-part structure

1. The origin story (Chapter 1)

The author was sick, or fat, or exhausted, or watched a parent die. Conventional medicine failed them. They went looking themselves.

What it does: establishes credibility that isn't credentials, and — crucially — establishes that the author is on your side against an establishment. It is genuinely moving and it is doing rhetorical work.

What to notice: an origin story is rung 3 evidence (Chapter 2) about one person, positioned where the strongest evidence would normally go.

2. The villain (Chapters 2–3)

One substance. Sugar. Carbohydrate. Seed oils. Lectins. Grains. Dairy. Insulin.

What it does: converts a multi-causal, probabilistic, boring problem into a single-cause story with an enemy. This is Chapter 1 §1.6's single cause shape, and it is the load-bearing element of the whole book.

What to notice: whether the villain is ever described in terms of dose or comparator. It almost never is, because both would dissolve it.

3. The suppressed truth (Chapter 4)

Why haven't you heard this? Industry funding. Captured guidelines. Doctors who get no nutrition training. Sometimes the Sugar Research Foundation episode, which is real and is asked to carry the whole edifice.

What it does: immunizes the book against contrary evidence in advance. If the mainstream disagrees, that's confirmation.

What to notice: this is the only unfalsifiable element, and it's the one that makes the rest durable. A book with this chapter cannot be refuted by any subsequent finding.

4. The mechanism (Chapters 5–8)

Now the science. Insulin. mTOR. Inflammation. Leaky gut. Mitochondria. Usually accurate for the first several pages, with real citations.

What it does: supplies the intellectual satisfaction. This is the part readers quote.

What to notice: where the accurate mechanism stops and the extrapolation begins. It is almost always at the point where a dial becomes a switch (Chapter 6, Case Study 2). Count the words switch, mode, stuck, reset, activated, unlocked.

5. The evidence (scattered)

Real studies, cited accurately. Mechanistic studies presented as if they were outcome studies. Animal work presented without species. Small trials presented without sample size. Occasionally one large trial that genuinely supports the position, and it will be mentioned repeatedly.

What to notice: run the Chapter 2 §2.9 abstract checklist on three citations at random. Not all of them — three. The result generalizes.

6. The protocol (the middle third)

Phases. Elimination, then reintroduction. Lists of yes-foods and no-foods. A 21-day or 30-day structure.

What it does: the elimination phase removes ultra-processed food, alcohol, added sugar, and most snacking simultaneously. It works, and Chapter 3's Case Study 1 explains why — it's a bundled intervention, and the bundling means nobody ever learns which element mattered.

What to notice: how many things change at once. Count them. It's usually six to ten.

7. The testimonials (throughout)

Boxed. First names. Sometimes photographs.

What to notice: these are rung 3, and there is no denominator. You are seeing the people it worked for. The people who stopped at week seven — Dominic, in Case Study 1 — did not write in.

8. The objection-handling (near the end)

"But what about…?" Cholesterol. Fiber. Bone density. Kidneys. Each addressed briefly and confidently.

What to notice: whether any objection is ever conceded. In the genre, almost none are — and a book that never concedes anything is not evaluating, it's advocating. Chapter 2's test applies: does the author ever describe a finding that complicates their own position?

9. The recipes and the product ecosystem (the back third, and the URL)

Recipes, meal plans, and a website. Often a supplement line, a testing service, a coaching programme, or a certification.

What to notice: what happens on day thirty-one. The genre is structured around a defined protocol with an end date, and the exit is where the outcome is determined (§10.12's reversibility test) — and it is almost never addressed, because the answer is buy the next thing.

Running the structure on a real book

Take whatever diet book is nearest to you. Twenty minutes:

Score it out of nine. In my experience most bestsellers in the genre score seven or higher, and the two most commonly missing are the product ecosystem (in books by academics) and the suppressed truth (in books by people who still work in institutions).

The honest half

Here is what makes this case study worth including rather than a hatchet job.

These books work. Not for the reasons they give — but people who follow them lose weight, feel better, and sometimes resolve genuine symptoms. That's real, and dismissing it makes you wrong as well as unhelpful.

Why they work, in descending order:

1. They remove ultra-processed food. Every one of them, whatever the villain. This is probably most of the effect (§10.5, Chapter 22).
2. They provide structure. Decisions are exhausting; rules are restful. This is a genuine benefit and the mainstream's "everything in moderation" provides none of it.
3. They provide an identity and a community (§10.10), which sustains behaviour far better than information does.
4. They provide a start date. "Begin Monday, thirty days" is a commitment device, and commitment devices work.
5. They increase attention to food, which is worth something on its own.

Not one of those five is the mechanism in Chapter 5 of the book. And a book that said "I'm going to remove ultra-processed food, give you rules, a community, and a start date, and the biochemistry is mostly decorative" would be more honest, would work identically, and would sell nothing.

💡 Aha moment. The genre's central bargain: it delivers real benefits through mechanisms it doesn't name, using a story it can't support, and the story is what makes people do it.

Which produces a genuinely uncomfortable question. If the false story is what generates the adherence, and adherence is what produces the outcome (§10.3) — is the false story doing good?

My answer, for what it's worth: it does good once, and harm afterwards. It works until it stops, and then the person has learned nothing transferable, has no floor, has no plan for day thirty-one, and has acquired one more piece of evidence that they can't do this. Theo had five of those. The transferable understanding is the thing worth having, and it's the thing the genre cannot sell.

Discussion Questions

1. Score a real diet book against the nine elements. Which did it score on, and which did it skip? What does the pattern of omissions tell you about its author?

2. The chapter argues the false story generates real adherence. If so, is a book that tells a false story but improves health doing net good? Argue both, and consider what happens at year two.

3. Element 3 — the suppressed truth — is described as "the only unfalsifiable element" and the one that makes the rest durable. Is there a legitimate version of this chapter? What would it look like?

4. The genre provides structure, identity, community, and a start date — four things mainstream advice provides none of. Design a way to deliver those four without a false mechanism. Then explain why it doesn't exist commercially.

5. Testimonials have no denominator. Estimate, for a book claiming a 30-day protocol: what fraction of purchasers finish it, and what fraction are still doing anything a year later? What would you need to find out, and who has that data?

Your Turn

Find a diet book — yours, a library's, a friend's — and give it twenty minutes with the table above.

Then answer the three questions that matter most:

1. At which sentence does the accurate mechanism become an extrapolation? Mark the page. It's usually in the first third and it's usually one paragraph.
2. How many things does the protocol change at once? Count them.
3. What does it say about day thirty-one? If the answer is nothing, you have found the book's most important omission.

And then, if you've ever followed one: go back to your Chapter 10 autopsy table and add a column — what did I learn from this that transferred? For most people, across most attempts, the honest answer is nothing, and that's not a criticism of them. It's a description of what the genre is built to deliver.