Chapter 19 — Key Takeaways

One page. 🪞 Learning Check-In chapter.

The Category Is Wrong Before the Argument Starts

"Seed oils" is defined by provenance — seed origin, industrial extraction, refining — not by composition.

⚠️ A sevenfold spread inside a category treated as homogeneous. Chapter 18 Case Study 1's failure exactly: a provenance claim answering a dose question.

The Four Claims

✅ §19.3 — What Actually Changed (take this seriously)

1. LA-rich oil consumption rose enormously across the twentieth century — soybean oil in particular. 2. ⚠️ Human adipose tissue linoleic acid content rose substantially over the past half-century — measured in fat biopsies, not estimated from food surveys. ⚠️ This bypasses Chapter 2 §2.5's biggest problem and is one of the better exposure measurements in the entire field. 3. It is genuinely without historical precedent.

⚠️ "Unprecedented" ≠ "harmful" (so are refrigeration, vaccination and iodized salt) — but it is an excellent reason to investigate, and the people who noticed it were right to notice.

⚠️ §19.5 — The Broken Link (the chapter's decisive finding)

The chain: LA → arachidonic acid → inflammatory eicosanoids → disease. The shared enzyme: delta-6 desaturase, competed for by the omega-3 pathway. All real biochemistry.

Two predictions, both tested:

⚠️ Adipose LA rises with intake (§19.3 measured it). Membrane AA largely doesn't follow. The second arrow is where the chain breaks — and it's the load-bearing one.

Chapter 6's lesson again: a pathway existing is not the same as a pathway being rate-limiting. A diagram shows what CAN happen, not what does, how much, or what regulates it. ⚠️ This is Chapter 13's antioxidant supplements, in a different system.

✅ The Fact the "Toxic" Framing Has to Explain

Linoleic acid is ESSENTIAL. Removing it causes a named clinical syndrome — scaly dermatitis, impaired wound healing, infection susceptibility, growth failure in infants — documented in fat-free parenteral nutrition and inadequate infant formulas, and it resolves when LA is provided.

⚠️ So the coherent sceptical argument is about DOSE — requirement 1–2% of energy, intake 5–8% — not toxicity. Chapter 18's threshold, demonstrated as cleanly as it can be.

🟢 Fix the Ratio at the Other End

⚠️ A ratio can be changed from either side and the two are not equivalent.

ALA→EPA/DHA conversion is poor — worse for DHA, lower in men. ⚠️ Cutting omega-6 does not reliably put EPA/DHA in your tissues. Eating fish does. (And note: the supplement trials are far more mixed than the food evidence — Ch 13, Ch 16, again.)

🟡 §19.6 — The Recovered Trials

Sydney Diet Heart (BMJ 2013) and Minnesota Coronary Experiment (BMJ 2016), Ramsden et al. — original data recovered and reanalyzed. Replacing saturated fat with LA-rich oil lowered cholesterol and did not reduce mortality.

⚠️ Real, published in a major journal, uncomfortable for the mainstream — and anyone presenting this debate without them is not giving you the argument.

Contested because: ⚠️ the intervention oils included trans-rich margarine of the era · high dropout, institutionalized populations, short exposure · outliers against a larger body · ⚠️ and their decades of non-reporting is a genuine publication-bias scandal that applies to ALL the literature, including the parts sceptics like (Ch 2 §2.8).

Against them: ✅ SFA→PUFA lowers LDL-C; 🟢 and probably reduces CVD events — moderate certainty, which means further research could change the estimate.

§19.8 — The Real Concern, Precisely Stated

The honest chain: ✅ chemistry certain · 🟢 plausibility strong · 🟡 human dose-response unquantified. ⚠️ Three confidence levels; collapsing them in either direction misreports it.

⚠️ "Seed oils oxidize and produce toxic aldehydes" is TRUE. "Therefore your cupboard is delivering them" does not follow from those studies.

§19.9 — Smoke Point Is Half Backwards

Refined canola/sunflower ~220–230°C · refined avocado ~260°C · ghee ~250°C · beef tallow ~200–205°C · ⚠️ EVOO ~190–200°C · ⚠️ butter ~150°C.

⚠️ Refining REMOVES the free fatty acids and particulates that smoke — so the process the argument objects to is the one that raises the smoke point.

But smoke point ≠ oxidative stability. Stability depends on degree of unsaturation and antioxidant content (which refining strips). ⚠️ EVOO smokes early and oxidizes slowly; refined high-PUFA oils smoke late and oxidize readily; saturated fats are most stable — the sceptics' best point on this question. For home cooking at ordinary temperatures with oil used once, the difference is unlikely to matter much.

§19.10 — Processing

❌ Hexane residue — ppm range, far below limits. Not a meaningful risk in the finished product. 🟡 Bleaching/deodorizing strips antioxidants and phytochemicals — a real nutritional cost. Chapter 7's intact-vs-refined continuum, applied to oils. 🟡 ⚠️ The honest concession: deodorizing generates small amounts of trans fat — well under a few percent, not zero. Direction right, magnitude unestablished.

⚠️ §19.11 — The Confounding That Swallows Everything

Seed oils are in: fried takeaway · crisps · packaged baked goods · dressings · mayonnaise · fast food · restaurant food · essentially all UPF.

Seed oil intake and ultra-processed food intake are not separable exposures. No adjustment fixes it. ⚠️ The confounding is SYMMETRIC — it prevents indictment and exoneration equally.

And it explains why elimination works: ⚠️ you cannot remove seed oils without removing six other things (Ch 17 shape 8 — count the changes). The intervention works and the explanation is wrong — the most common pattern in this book. Chapter 22 has the real mechanism.

§19.11's harder lesson: ⚠️ the Chapter 17 tells identify how a claim is SOLD, not whether it's RIGHT. Seed-oil rhetoric fires shapes 1, 4 and 5 on a claim whose factual premise is true. Run the tells to decide how much scrutiny is warranted. Then do the scrutiny.

§19.12 — Four of Five Slots Empty

Dose (5–8% of energy vs a 1–2% requirement — is the concern 8%? 4%? any excess?) · ⚠️ Form (cupboard vs commercial fryer — the most important omission) · Period · Population (desaturase variation is real and never mentioned) · ⚠️ Comparator (butter? olive oil? tallow? or not eating the fried food at all — four different interventions).

What To Do

1. ⚠️ Form over oil — deep-fried food eaten out is the exposure. 2. Olive oil as default, on a positive case (Ch 10 §10.5). 3. A high-smoke-point oil for high heat. 4. ⚠️ Don't reuse frying oil, and don't buy from places that do. 5. Store cool, dark, sealed; smell for rancidity — crayons, old putty. 5b. 🟢 Oily fish twice a week — the best-evidenced action here. 6. ⚠️ Don't bin oil you already own — change when the bottle empties. 7. Check the denominator: if you removed seed oils and removed fried takeaway, know which one you did.

🧾 Economics

Eighth consecutive chapter where the best-evidenced change is the cheapest.

One Thing to Remember

⚠️ Theo binned $34 of oil and spent $48 replacing it, for a claim whose central mechanism failed direct testing.

Meanwhile the finding that IS solid is about a restaurant fryer he'll eat from on Friday — and it costs nothing to act on.