Case Study 1 — The Green Folder, Reopened: Theo's Four Headline Pairs, Answered

⚠️ Theo Vasquez is the book's anchor case and an illustrative composite, established in Chapter 1. ⚠️ His green folder — four pairs of contradictory headlines collected over eighteen months — was the question this book was written to answer. This case study answers it, pair by pair.**

Setup

⚠️ The folder has been in a drawer since Chapter 1. He brought it back.

⚠️ His instruction to himself, written on the inside cover in Chapter 1 and unchanged:

"Work out which one is right."

⚠️ That instruction turns out to be wrong, and how it is wrong is this case study's finding.

⚠️ In three of the four pairs, neither headline was right. ⚠️ In the fourth, both were, about different questions.

Pair 1 — "Eggs cause heart disease" / "Eggs are a superfood"

⚠️ Six questions from §37.7, applied.

⚠️ The answer, from Chapter 9: ⚠️ dietary cholesterol has a much smaller effect on blood cholesterol than was assumed for decades, the guidance limiting it was withdrawn, and eggs are an inexpensive complete-protein food with a modest and unremarkable place in a diet.

⚠️ Neither headline. And Theo's own reaction:

"So the actual answer is 'eggs are fine and not important.' That's not a headline."

⚠️ No. §37.6 explains why it never was one — and this is the pattern for three of his four pairs.

Pair 2 — "Red meat causes cancer" / "Red meat is essential nutrition"

⚠️ The most instructive of the four, because here the honest answer is a distinction rather than a verdict.

⚠️ Where each headline went wrong:

⚠️ Headline A conflated processed and unprocessed, and reported a relative risk without a baseline — ⚠️ which is the single most common misreporting in nutrition and Chapter 2 §2.6's subject.

⚠️ Headline B was true of nutrients and false as a dietary conclusion. ⚠️ "Contains essential nutrients" does not establish "you need this food," because Chapter 8 and Chapter 14 list the alternatives.

⚠️ The honest answer: eat less processed meat specifically, unprocessed red meat is a moderate question, and neither headline was about the same thing.

Pair 3 — "Fat makes you fat" / "Carbs make you fat"

⚠️ The pair Theo found most frustrating, because both had books behind them.

⚠️ The answer, from Chapters 4, 7, 9 and 10:

⚠️ Neither macronutrient has been shown to have the special causal role each camp claims. ⚠️ Controlled feeding trials at matched energy find small differences. Free-living trials find that adherence dominates (Ch 10's threshold).

⚠️ And the thing both books were partly right about, which is why both had converts:

⚠️ Each diet works for the people who can stick to it, and each removes a large category of energy-dense food that its adherents had been eating a lot of.

💡 Aha moment. ⚠️ "They're both descriptions of what happens when you stop eating one of the two things most of your junk food was made of."

⚠️ That is close to right, and it is Chapter 10's finding in the reader's own words.

⚠️ §37.2b's row for this pair: ⚠️ the insulin-carbohydrate mechanism LARGELY FAILED; the practical advice about refined carbohydrate was ABSORBED.

⚠️ And DIETFITS (Ch 35 §35.3b) is the direct test: a year, both arms eating well, no meaningful average difference, enormous individual variation, and no predictor of who did better on which.

Pair 4 — "Coffee is good for you" / "Coffee is bad for you"

⚠️ The pair where both headlines were defensible, and the reason is worth understanding.

⚠️ Both headlines cited real findings about different outcomes, and neither said which outcome.

⚠️ The honest answer: coffee is fine, it is not a health intervention, the timing matters more than the amount for most people, and it is a preference.

⚠️ Which is §37.4's optional list, and is the least exciting possible resolution.

⚠️ What the folder actually contained

⚠️ Sorted after the fact, and this is the case study's real finding.

⚠️ Chapter 17's finding, in one man's drawer: the problem is over-extension, not fabrication.

⚠️ And the structural observation:

⚠️ Every one of the eight was exciting. ⚠️ Not one of them said "this adjusts the existing advice slightly," and §37.2b says that is what the true ones almost always do.

⚠️ Theo's instruction to himself — work out which one is right — assumed the answer was in the folder. ⚠️ It was not in the folder. The folder was the phenomenon.

⚠️ What he does now

⚠️ He kept the folder and changed what goes in it.

⚠️ He no longer collects contradictory headlines. ⚠️ He runs §37.7's six questions on anything that makes him want to change what he eats, and he writes down which question it failed on.

⚠️ Eighteen months of the new folder:

⚠️ The one that survived: increasing pulses, on the fibre evidence (Ch 11). ⚠️ Which was already in the boring answer, and which he had read in Chapter 11 and not acted on until a claim survived his own filter.

⚠️ That is the chapter's argument as a personal outcome. ⚠️ Thirty of thirty-one exciting claims failed a ten-minute check, and the one that survived was something the boring answer already contained.

Analysis

1. ⚠️ In three of four pairs, NEITHER headline was right. ⚠️ The instruction "work out which one is right" presupposed a false structure.

2. ⚠️ None of the eight was fabricated. ⚠️ Four omitted a comparison or baseline, two reported nutrient content as an outcome, one reported a mechanism as a result (Ch 17's finding, replicated in a drawer).

3. ⚠️ Pair 2 needed a DISTINCTION, not a verdict — processed versus unprocessed — and both headlines had collapsed it.

4. ⚠️ Pair 3's two camps were each describing the removal of a category of energy-dense food, ⚠️ and DIETFITS tested the mechanism directly and refuted it (Ch 35 §35.3b).

5. ⚠️ Pair 4's headlines were both defensible about different outcomes and neither named its outcome — ⚠️ which is §37.7's question 3 in a different form.

6. ⚠️ All eight were exciting, and none said "this adjusts the existing advice slightly." ⚠️ §37.2b says the true ones almost always do.

7. ⚠️ Thirty-one claims over eighteen months; one survived; it was already in the boring answer.

8. And he acted on Chapter 11's advice only after it survived his own filter. ⚠️ Which suggests what the book was actually for, and Chapter 38 has to make that argument explicitly.

Discussion Questions

1. ⚠️ "Work out which one is right" presupposed a false structure. What is the better question? ⚠️ Write the instruction Theo should have put on the folder in Chapter 1.

2. ⚠️ None of the eight headlines was fabricated. ⚠️ Does that make the situation better or worse? Which is easier to defend against?

3. ⚠️ Pair 2 required a distinction rather than a verdict. ⚠️ Find two other places in this book where the honest answer was a distinction. (Chapters 18 and 22.)

4. ⚠️ Pair 3's camps both had converts and both were partly right. ⚠️ Is "adherence explains it" a satisfying answer? Why is it so rarely the headline?

5. ⚠️ Thirty of thirty-one claims failed a ten-minute check. ⚠️ Is that ratio plausible, or is Theo now over-filtering? How would he tell?

6. ⚠️ He acted on Chapter 11 only after it survived his own filter. ⚠️ What does that say about how advice becomes behaviour? (Ch 33, and Ch 35's Theo said something similar about conviction.)

Your Turn

Step 1 — ⚠️ Find your own folder.

⚠️ Two nutrition claims you have encountered that contradict each other.

A: _ B: _

Step 2 — ⚠️ Run §37.7's six questions on both.

Step 3 — ⚠️ Ask whether the question was even the right one.

⚠️ Are A and B about the same outcome, the same population, and the same dose?

⚠️ In three of Theo's four pairs they were not.

Step 4 — ⚠️ Classify, don't adjudicate.

⚠️ Accurate · missing comparison or baseline · nutrient content as outcome · mechanism as result · fabricated.

⚠️ Count how many land in the last category. For Theo it was zero.

Step 5 — ⚠️ Start the new folder.

⚠️ For the next six months, log anything that makes you want to change what you eat, and write down which of the six questions it failed on.

⚠️ Expect most to fail at question 1.

Step 6 — ⚠️ And notice what survives.

⚠️ If something survives all six, check it against §37.3's plate before doing anything.

⚠️ Theo's survivor was already in it, and the useful realization was not the claim — it was that he had read it, agreed with it, and not acted on it for eighteen months.