380 ‒ The seed oil debate: are they uniquely harmful relative to other dietary fats?

In this podcast episode, host Peter Attia and guest Layne Norton engage in a comprehensive discussion about seed oils and the controversy surrounding their health effects compared to other dietary fats. The conversation touches on the scientific evidence from historical and modern clinical trials, the biochemical mechanisms of LDL cholesterol and oxidation, the influence of industrial processing on seed oils, evolutionary perspectives, and finally, practical guidance for individuals trying to make sense of conflicting narratives in nutrition science and popular culture.

The Motivation and Format for the Discussion

Peter Attia begins by explaining the original concept for the episode—a formal debate with two experts holding opposing views on seed oils' health impact. Although planned for over a year, the opposing expert eventually declined to participate, citing concerns about Attia's personal leaning toward seed oils not being uniquely harmful. As a result, the format pivoted to Norton presenting the strongest arguments in favor of seed oils, while Attia played the judge and offered counterpoints, fostering a fair and evidence-based examination of the topic.

Layne Norton stresses the ubiquity of bias in science and personal beliefs. He acknowledges his own biases stemming from a background favoring animal-based, lower-carb approaches but emphasizes transparency, intellectual humility, and reliance on converging lines of evidence as crucial methods to mitigate bias and better interpret complex nutritional data.

The Historical Mortality Literature and Key Randomized Controlled Trials

The dialogue dives into major randomized controlled trials (RCTs) relevant to seed oils and cardiovascular mortality, primarily from the mid-20th century. The Minnesota Coronary Experiment (MCE) is examined as an influential study that replaced saturated fats with polyunsaturated fats (mostly linoleic acid-rich seed oils) in institutionalized patients. Although total cholesterol declined significantly in the treatment group, mortality did not decrease, an unexpected outcome that was not published for over a decade.

A critical confounding factor highlighted is the presence of industrial trans fats, which were abundant in margarine and processed seed oil products of the time (up to 25-40% trans fats). These trans fats are now well-known to be highly atherogenic, contributing negatively to cardiovascular outcomes and potentially obscuring the effects of polyunsaturated fats themselves. Additionally, changing study conditions—such as patients leaving psychiatric institutions during the trial—further complicate result interpretation.

The Sydney Heart Study, involving men recovering from myocardial infarction, similarly showed that those reducing saturated fat in favor of polyunsaturated fats (largely safflower oil) had higher mortality in some analyses, again confounded by trans fats from margarine. Due to small sample sizes and limited deaths, statistical power and confidence intervals remain a concern.

Another trial, the Rose Corn Oil trial, which lacked trans fat confounding, also suggested higher cardiovascular deaths in the corn oil group, but the small sample size limited confidence.

In contrast, longer, better-controlled studies without trans fat confounding—such as the Finnish Mental Hospital Study with a crossover design—found significant cardiovascular risk reductions by substituting saturated fat with polyunsaturated fats. Meta-analyses excluding trials confounded by trans fats generally support a cardioprotective effect of polyunsaturated fats, despite some variability from omega-3 content and study design.

The inconsistent findings are attributed largely to methodological limitations, confounders like trans and omega-3 fatty acids, and varying trial durations. Norton argues these factors explain apparent contradictions and support the conclusion that seed oils are not uniquely harmful relative to saturated fat.

Definitions and Biochemistry of Fatty Acids

To clarify terminology, Peter and Layne review the basic chemistry of fats. Saturated fats have no double bonds between carbon atoms, resulting in a "rigid," solid structure at room temperature. Monounsaturated fats have one double bond, typically cis-configured, creating a "kink" that increases fluidity. Polyunsaturated fats contain multiple cis double bonds, further enhancing fluidity.

Trans fats, a subset of unsaturated fats with trans double bonds, mimic saturated fat's linear structure but are susceptible to oxidation and have been demonstrated to be highly atherogenic. Due to overwhelming evidence, trans fats are now banned or severely restricted in many countries.

The structure of lipoproteins and their interaction with the aqueous bloodstream is explained, emphasizing how fatty acid composition influences membrane fluidity, LDL particle properties, receptor recognition, and atherogenic potential.

Mendelian Randomization and the Causality of LDL Cholesterol

The conversation turns to Mendelian randomization (MR) studies that utilize genetic variants associated with lifelong LDL cholesterol levels to investigate causal relationships with cardiovascular disease (CVD) risk. MR studies provide a form of natural lifelong randomized control, strengthening causal inference.

These studies have found a robust and linear relationship between genetically lowered LDL cholesterol and reduced cardiovascular risk, independent of other confounders. Each ~39 mg/dL reduction in LDL corresponds with approximately a 50-55% decrease in cardiovascular events, a much stronger effect than that observed in most statin trials, explained by the lifelong difference in exposure starting at birth.

The consistency of MR data with statin trials and dietary interventions that affect LDL supports the causal role of LDL particles in atherogenesis. The discussion underlines the importance of evaluating LDL particle number (apoB) alongside LDL cholesterol mass, noting that some genetic variants and therapies reduce cholesterol mass without proportionally decreasing particle count, resulting in attenuated risk reduction.

While acknowledging that LDL is not the sole driver of CVD—factors like blood pressure, insulin sensitivity, and inflammation remain important—the LDL-C/apoB axis is reinforced as a critical and independent risk factor.

Mechanisms of Atherogenesis: Oxidation and Aggregation of LDL

The podcast explores the lipid hypothesis, focusing on how LDL penetrates the arterial endothelium and initiates atherosclerosis. Lipoproteins smaller than 70 nm containing apoB proteins infiltrate the subendothelial space (intima), where they may be enzymatically modified and retained via binding to proteoglycans.

The retained LDL undergoes oxidation, triggering immune cell recruitment, foam cell formation, and smooth muscle proliferation—processes integral to plaque formation.

Seed oil skeptics argue that polyunsaturated fatty acids (PUFAs), particularly linoleic acid, are more prone to oxidation inside LDL particles, increasing inflammation and atherogenicity despite lowering LDL cholesterol concentration.

Layne counters by emphasizing that oxidation in plasma is minimal (<1% of LDL oxidized), partially due to potent antioxidants (vitamin E, C, carotenoids) present in circulation. Oxidation chiefly occurs within the intima after LDL retention.

More importantly, PUFAs in LDL increase membrane fluidity, enhance LDL receptor-mediated clearance, and reduce apoB enzymatic modification, making these particles less prone to retention and aggregation compared to saturated fat-enriched LDL, which is more rigid and fosters ceramide production via sphingomyelinase activity, promoting aggregation.

Therefore, on a per particle basis, saturated fat-rich LDL are more atherogenic despite PUFA-rich LDL being more oxidation-prone. The net effect favors PUFA consumption due to overall reduced particle number and retention.

This dynamic is illustrated with an analogy of a bonfire: polyunsaturated fat lowers and "shrinks" the fire (LDL particles), which casts fewer "sparks" (particles into the arterial wall), while saturated fat creates a larger fire with more sparks more likely to kindle disease.

Industrial Processing of Seed Oils and Potential Impurities

The episode addresses concerns about industrial seed oil processing methods—involving high heat, solvent extraction with hexane, and refining steps—and whether this introduces harmful compounds or residual chemicals.

Hexane, a non-polar solvent used for oil extraction, is discussed in detail. Although small residual amounts may remain, typical concentrations in oils are exceedingly low (often below detection limits) and well under toxic thresholds. Human and animal studies suggest toxicity primarily arises from inhalation rather than ingestion, and the quantities consumed with dietary seed oils are negligible compared to toxic doses.

Refining involves removal of impurities and can reduce oxidized lipid content present in crude oils, while some minimal trans fats and polymerized triglycerides may form; however, these remain far below levels posing health risk.

Because these processed oils are stored and heated under vacuum with limited oxygen, oxidation during manufacturing is minimal. However, repeated heating in shallow pools (e.g., frying) can increase formation of oxidation products and degrade oil quality.

Overall, current evidence does not support the industrial extraction process as a significant source of harm in typical dietary seed oil consumption.

Evolutionary and First Principles Perspectives

Layne and Peter discuss the massive rise of dietary linoleic acid over the past century, increasing by as much as 75-fold from less than 3% of calories to approximately 10-15% today due to seed oil consumption.

Some argue that such rapid dietary changes are unnatural and incompatible with human evolutionary adaptation, implying harmful effects. However, the speakers caution against romanticizing ancestral diets and naturalistic fallacies. Modern humans benefit from adaptability and longevity improvements, and many "natural" exposures can be harmful.

Data from populations such as the Hadza show low LDL cholesterol levels and minimal cardiovascular disease, but their diet differs markedly from modern Western diets in numerous ways, making direct translation speculative.

The functional consequence of high linoleic acid intake is complicated. Conversion to arachidonic acid, the precursor of inflammatory prostaglandins, appears saturated at typical dietary levels, limiting concerns of inflammation arising from linoleic acid itself.

Large cohort studies show inverse relationships between linoleic acid levels and cardiovascular disease risk, further challenging the assertion of seed oil toxicity. Substituting PUFAs for saturated fats generally improves markers of inflammation, insulin sensitivity, and liver health in clinical trials.

Taken together, evolutionary and mechanistic reasoning prioritize phenotypic outcomes and converging clinical evidence over assumptions based on historical diet composition.

Practical Guidance and Context for Consumers

In closing, Norton offers actionable advice for individuals navigating conflicting information. If a person dislikes seed oils or is concerned by prevailing negative narratives, Norton suggests substituting saturated fats with monounsaturated fats like olive or avocado oil, which also appear cardioprotective, though perhaps less potent than PUFAs.

Both agree that exposure to heavily reheated oils, common in restaurant frying, especially with thin oil layers reused repeatedly, can increase oxidation products and may be best minimized. However, occasional consumption of fries or fried foods should be viewed in the broader context of overall diet quality and lifestyle.

Peter highlights that efforts to vilify seed oils often coincide with promotion of other poor dietary patterns; in many cases, reducing seed oils coincides with cutting processed foods, yielding net benefit regardless of the oil's intrinsic toxicity.

The podcast closes by urging balance: dietary factors like total caloric intake, physical activity, smoking, strength, and metabolic health exert far greater influence on chronic disease risk than specific concerns about seed oils. Communication to the public requires nuance to prevent misunderstanding, overreaction, or undue fear.

Both express frustration with the "no seed oils" marketing trend in restaurants as misleading and not evidence-based, underscoring the complexity and challenges of accurate nutritional guidance dissemination.