Peter Attia MD380 ‒ The seed oil debate: are they uniquely harmful relative to other dietary fats?
CHAPTERS
- 0:01 – 8:49
Why the debate format changed and Layne Norton's stated biases
Peter Attia explains that the planned pre-submitted-evidence debate fell through when the opposing guest withdrew, so he will steelman the anti-seed-oil case himself. Layne Norton describes his own biases, his funding sources, and his emphasis on converging lines of evidence.
- •The original debate was designed like a courtroom, with evidence submitted in advance and Attia as judge
- •The opposing guest dropped out after about a year of preparation
- •Attia will steelman the case that seed oils are uniquely harmful
- •Norton says personal beliefs can be as powerful a driver of bias as funding
- •Norton says he changed his mind on saturated fat and LDL after earlier low-carb leanings
- 8:49 – 15:44
The four arguments against seed oils and the Minnesota Coronary Experiment
Attia lays out the four arguments he will steelman: RCT mortality data, LDL oxidation, industrial processing, and the ancestral/evolutionary case. He then presents the Minnesota Coronary Experiment, where total cholesterol fell on a polyunsaturated-fat diet without a mortality benefit.
- •The four arguments are RCT mortality, oxidation, processing, and ancestral diet
- •The Minnesota study used institutionalized patients with isocaloric substitution of PUFA for saturated fat
- •Cholesterol fell but mortality did not, and publication was delayed years
- •Norton calls it the most-cited study for the anti-seed-oil case
- 15:44 – 22:02
Saturated, monounsaturated, polyunsaturated and trans fats explained
Attia and Norton pause to explain fatty-acid chemistry: double bonds, cis versus trans kinks, and how these affect membrane fluidity and lipoproteins. They explain why trans fats from margarine behave like saturated fat and are uniquely atherogenic.
- •Saturated fats have no double bonds and are more solid at room temperature
- •Cis double bonds put a kink in the tail, while trans bonds keep it straight
- •Membrane fluidity matters for LDL receptor recognition and aggregation
- •Margarine was developed as a butter substitute and later found to be harmful because of trans fats
- 22:02 – 26:53
Trans fat confounding and limitations of the Minnesota study
Norton argues that trans fats in margarine confound the Minnesota results, and that the exact trans fat intake is unknown. He adds that patients could leave the wards, and that follow-up was short relative to how slowly cardiovascular disease develops.
- •The PUFA group got much of its fat from margarine, which was 25 to 40% trans fat
- •The raw trans fat data does not appear to be available
- •Patients leaving the psychiatric wards is a confounder, but likely equal across groups
- •Short follow-up is a problem because cardiovascular disease develops over decades
- •Norton uses an investing analogy for lifetime exposure
- 26:53 – 42:55
The Sydney Diet Heart Study and the Rose Corn Oil trial
The two examine the Sydney study in men who had already had an MI, and the very small Rose Corn Oil trial. Both showed worse outcomes with PUFA, but they have trans fat confounding, wide confidence intervals and few deaths.
- •Sydney: higher mortality in the group that lowered saturated fat and used safflower oil and margarine
- •Low death counts (37 vs 28) and a confidence interval near 1 limit the finding
- •The Rose trial had about 80 participants and a confidence interval of 0.58 to 37.15
- •Norton says that when trans fat-confounded trials are lumped in, the net effect is null
- 42:55 – 51:46
Trials without trans fat confounding: VA, Oslo, Finnish Hospital and meta-analyses
Norton points to trials not confounded by trans fats that favored PUFA, including the VA study and the Finnish Hospital Study. They discuss the crossover design, omega-3 confounding, and the 2017 meta-analysis and Ramsden reanalysis.
- •The VA study showed an 18% non-significant reduction, with omega-3s included
- •The Finnish crossover study showed a 41% reduction and was not confounded by trans fats or omega-3s
- •Crossover designs increase power because each person is their own control
- •The 2017 meta-analysis excluding trans fat trials showed about a 29% reduction
- •Ramsden's analysis with trans fats left in shows a null or slightly negative result
- 51:46 – 1:00:33
Is it PUFA benefit or saturated fat harm? Substitution and net effects
Attia asks whether the results say more about saturated fat's harm than PUFA's benefit. Norton replies that nutrition studies are substitutions, so the two cannot be separated, and he uses a mutual fund analogy to argue that net effect matters more than individual mechanisms. They also clarify which trials involve omega-3s.
- •Isocaloric studies have to substitute something for saturated fat
- •Norton says excess calories from added oils are not innocuous
- •Mechanisms are like single stocks, while outcomes are like the mutual fund
- •The smoking and Parkinson's example and the aspirin example illustrate net effects
- •Only the Finnish study and STARS are free of both trans fat and omega-3 confounding
- 1:00:33 – 1:20:46
Mendelian randomization and the causal role of LDL
Norton explains that Mendelian randomization acts as a lifelong randomized trial of LDL. Attia stresses that it tests LDL causality, not nutrition, and the two discuss why genetic effects exceed statin effects and the argument that statins work through other pathways.
- •Genetic variants are randomly assigned at birth, giving lifelong LDL exposure
- •MR requires that the genes not act through other pathways (pleiotropy)
- •About a 50-55% risk reduction per mmol/L lower LDL in MR versus about 22% for statins
- •Norton attributes the gap to the compounding effect of decades of earlier exposure
- •Consistent dose response across mechanisms argues against alternative explanations
- 1:20:46 – 1:30:45
Does linoleic acid drive inflammation via arachidonic acid?
Attia raises the concern that linoleic acid could convert to inflammatory arachidonic acid and increase oxidized LDL. Norton takes the high-level view that cohort and tissue-biomarker studies show higher linoleic acid is associated with lower cardiovascular disease.
- •Attia frames inflammation and oxidation as central to atherosclerosis
- •Cohort studies show higher linoleic acid intake associated with lower cardiovascular disease
- •Tissue biomarkers of linoleic acid show the same protective association
- •Norton says linoleic acid intake does not appear to raise arachidonic acid
- •He argues for consistent standards: the same logic would be applied to saturated fat
- 1:30:45 – 1:38:23
Oxidized LDL, aggregation and the bonfire analogy
Norton lays out the lipid hypothesis and argues that most LDL oxidation happens inside the intima, not in plasma. He says PUFA lowers particle number, retention and aggregation, then offers a bonfire-and-sparks analogy.
- •ApoB particles under 70 nm can enter the intima, with entry dependent on concentration
- •Retention follows enzymatic modification of ApoB, then oxidation, aggregation and foam cells
- •Saturated-fat-enriched particles are stiffer and form ceramides that promote clumping
- •Bonfire analogy: PUFA-enriched particles may be slightly more flammable, but there are far fewer sparks
- •Antioxidant trials have not reduced cardiovascular disease
- 1:38:23 – 1:45:36
Attia presses on peripheral oxidized LDL
Attia challenges Norton on whether a small fraction of oxidized LDL in the periphery could still be disproportionately harmful. They discuss hypothetical comparisons, the soybean oil trial, antioxidant studies and where measured oxLDL originates.
- •Attia asks whether PUFA diets raise the oxidized fraction even at lower LDL
- •Norton cites one trial in which soybean oil raised peripheral oxLDL slightly
- •He says intimal oxidation rates are far higher than plasma rates
- •Whether measured oxLDL escaped from the intima is unknown without tracer studies
- •Norton says on balance PUFA still wins on entry, retention and aggregation
- 1:45:36 – 1:53:55
Industrial processing and hexane in seed oils
Attia argues that industrial refining, heat and solvent extraction could cause chronic harm. Norton explains hexane extraction, residual levels and the amounts needed for toxicity, and says refining lowers oxidation products.
- •Hexane is used because it is a nonpolar solvent with a low boiling point
- •Residual hexane is typically well under 1 ppm, and the danger is from inhalation
- •Norton calculates that mild effects would need an implausibly large amount of oil
- •Hexane does not bioaccumulate, unlike LDL, which is always present
- •Refining reduces peroxides and aldehydes
- 1:53:55 – 2:05:17
The ancestral diet and evolutionary argument
Attia presents the first-principles argument that linoleic acid intake has risen enormously and was absent in ancestral diets. Norton rebuts with the naturalistic fallacy, longer lifespans, and the low LDL of the Hadza, and ties it back to the converging evidence.
- •Linoleic acid intake has risen dramatically, as both speakers note
- •Norton says modern foods and animals are far from ancestral anyway
- •Cardiovascular disease needs time to develop, and people now live longer
- •Hadza LDL levels are around 50-70, which seed oil critics' ancestral diet would not predict
- •Norton reviews minor processing byproducts and concludes the net evidence favors PUFA over saturated fat
- 2:05:17 – 2:15:41
Practical takeaways: alternatives, frying and food marketing
Attia asks how to land the discussion for listeners. Norton suggests displacing saturated fat with leaner proteins or monounsaturated oils if people avoid seed oils, and the two discuss frying, lard versus seed oils, and marketing claims.
- •Norton says that if someone avoids seed oils, they should still displace saturated fat, for example with olive or avocado oil
- •Oxidation depends on frying conditions, with thin layers and repeated reheating producing more oxidized products
- •Norton thinks the lard versus seed oil frying question is hard to answer without trials
- •Norton warns that marketing 'no seed oils' can make people think fries are healthy
- •Social media hot takes omit context that reading the actual studies provides
- 2:15:41 – 2:25:58
Bigger levers for health and closing thoughts
Norton argues that calorie balance, activity, strength and fitness matter far more than seed oils, while LDL and ApoB remain worth controlling. Attia agrees and adds that seed oils travel with junk food, so avoiding them can help, but one need not be maniacal.
- •Norton cites much larger mortality hazard ratios for obesity and fitness than for seed oils
- •He suggests limiting saturated fat, eating enough fiber and getting ApoB measured
- •LDL is an independent risk factor but not the only one, and risk is probabilistic
- •Attia says seed oils tend to show up in low-quality foods, so restricting them can help via substitution
- •Attia asks restaurants to drop 'no seed oils' menu claims
