Chapter 17 — Further Reading

Real sources only. Where I'm unsure of exact details, I describe the source rather than inventing a citation.

Start here — the two that do the most work

NHS "Behind the Headlines" — the NHS website's service that takes a health story from the press and compares it to the study it was based on. ⚠️ This is Chapter 1's five-step pipeline, documented case by case, and it is the fastest way to internalize where over-extension happens. Free, archived, and covers years of stories.

Sense About Science (senseaboutscience.org) — a UK charity producing plain-language guides on evaluating evidence, including specific material on detox claims. Their "Making Sense of…" series is short and unusually well-written for the genre.

On specific claims in this chapter

The alkaline diet. Search for reviews on "dietary acid load and bone health" — the genuine PRAL literature, which is real and does not support the diet's claims. ⚠️ Then search for the Robert Young conviction for practising medicine without a licence, which is the documented harm.

ORAC and antioxidant capacity. Search "USDA ORAC database withdrawn 2012" — the USDA's own statement is short and explains its reasoning clearly. Read it alongside Chapter 13's supplement trials; they are the same lesson from two directions.

The blood type diet. The 2014 PLOS ONE paper — search "ABO genotype blood-type diet cardiometabolic risk". Open access. ⚠️ Read the design before the result; the reason it's a clean refutation is entirely in how the hypothesis was specified.

Cooking and nutrient retention. Search "cooking method vitamin C retention broccoli" and "lycopene bioavailability cooked tomato". The findings run in opposite directions for different nutrients, which is the point.

Arsenic in rice. FDA guidance on inorganic arsenic in rice and rice cereals (fda.gov) and EFSA's opinions on inorganic arsenic (efsa.europa.eu). ⚠️ Both address infant exposure specifically; read that section if you feed a child rice cereal.

Ginger and nausea. Search Cochrane "interventions for nausea and vomiting in early pregnancy" and your national obstetric guideline. The plain-language summaries are free.

On the contamination cluster

PFAS — this is the section where I'd most want you to read past the chapter, because the real literature is larger and more concerning than the cookware claim it's usually attached to.

• EPA PFAS pages (epa.gov/pfas) — drinking water limits and their basis
• EFSA's opinion on perfluoroalkyl substances in food — search "EFSA PFAS tolerable weekly intake"
• ATSDR toxicological profile for perfluoroalkyls — long, thorough, free

BPA — read both, deliberately. ⚠️ This is the most useful exercise on this page.

• EFSA's 2023 re-evaluation of BPA — search "EFSA bisphenol A tolerable daily intake 2023"
• The FDA's BPA position (fda.gov)

Then read them against each other. Two competent regulators, largely the same evidence, materially different conclusions. This is what genuine scientific disagreement looks like from the outside, and learning to recognize it is worth more than either verdict — because Part IV is full of it.

On non-stick specifically: search "polymer fume fever PTFE" for the real hazard, which is overheating rather than ingestion.

On why people believe things — and whether correction works

Denise Minger, Death by Food Pyramid. ⚠️ I recommend this with a caveat and I mean both halves. Minger is a sharp, funny, genuinely skilled critic of nutritional epidemiology, and her dissection of how weak studies became policy is worth reading. She also lands on conclusions this book doesn't share. Read it as a demonstration of method — including where you think the method is being applied selectively. Chapter 9 §9.6's treatment of the Keys cherry-picking charge is the worked example of doing that.

On the backfire effect: search for work by Brendan Nyhan and Jason Reifler, and importantly for the subsequent replication literature, which has substantially complicated the original finding. ⚠️ This is itself a good case study: a striking result that spread widely, was then found to be much weaker and more conditional than reported, and is still cited in its original strong form. Including by people who should know better, and probably including me somewhere.

Hugo Mercier and Dan Sperber, The Enigma of Reason — argues that reasoning evolved for argumentation rather than for individual truth-seeking. It reframes §17.11 substantially, and it is the most interesting thing I've read on why intelligent people hold badly-supported beliefs.

On motivational interviewing — the clinical technique behind most of §17.13. Search "motivational interviewing brief introduction" or read Miller and Rollnick. ⚠️ This is a trained skill with an evidence base, not folk wisdom, and it is genuinely more effective than argument. If any part of §17.13 was useful to you, this is where it comes from.

Checking a claim yourself

PubMed (pubmed.ncbi.nlm.nih.gov) — free abstracts. Filter by publication type: Systematic Review, Meta-Analysis. Chapter 2 §2.9 covers reading an abstract.

Cochrane (cochranelibrary.com) — plain-language summaries free.

Retraction Watch (retractionwatch.com) — ⚠️ worth a browse specifically to calibrate. Papers do get retracted, and papers get cited for years afterwards.

Your national dietetic association and Health Canada / NHS / EFSA / NIH consumer pages — unglamorous, cautious, usually right, and free.

A note on what to be careful with

⚠️ The risk this chapter creates is not credulity. It's the opposite.

A reader who finishes here armed with eight tells can dismiss almost anything, and the tells work faster than evidence does. That's a genuine hazard, and §17.9 exists specifically because of it — PFAS are real, BPA is unresolved, and industry influence on nutrition science is documented (Chapter 18 has the case).

Three checks worth keeping:

1. Can you name the narrower true version? ⚠️ If your response to a claim is "that's nonsense" and you cannot state what real finding it over-extends, you haven't finished — you've pattern-matched.

2. Are you using ⚗️ and 🟡? A verdict distribution that is all ❌ is Chapter 12's warning: a different flavour of overconfidence. The honest answer is "unclear" far more often than either side of any argument finds tolerable.

3. Would you accept this evidence if it pointed the other way? ⚠️ The single hardest question in the book, and the only one that reliably catches motivated scepticism.

Part IV is where this gets tested, because Chapters 18–22 contain genuine disagreements among serious people. The tells will not settle them, and reaching for a verdict where the honest answer is 🟡 is the same error this chapter is about — with the sign flipped.