Seed Oils: What the Science Actually Says

What Are Seed Oils and How Are They Made? Seed oils — also called vegetable oils — are oils extracted from the seeds of plants: soybeans, sunflowers, corn, rapeseeds (canola), cottonseed, safflower, and grapeseed are the most common. They are distinguished from fruit oils (olive oil, avocado oil, coconut oil), which are extracted from the flesh of the fruit rather than the seed. The extraction process matters enormously for the quality of the final product. Traditional cold-pressing or expeller-pressing uses mechanical pressure to extract oil without heat or chemicals. This method is used for olive oil, avocado oil, and some specialty seed oils. Industrial extraction — used for the vast majority of commercial seed oils — involves: solvent extraction with hexane (a petroleum-derived solvent) to maximise yield; degumming (removing phospholipids); refining with sodium hydroxide (caustic soda) to remove free fatty acids; bleaching with activated clay to remove colour and impurities; deodorisation at temperatures of 200–270°C (392–518°F) to remove volatile compounds that cause off-flavours. This process produces a neutral-tasting, shelf-stable oil — but also removes most of the natural antioxidants (vitamin E, polyphenols) that protect the oil from oxidation, and generates trans fats and oxidation products during the high-temperature deodorisation step. The resulting oil is chemically different from cold-pressed oils, and this distinction is central to understanding the health debate.

The Linoleic Acid Question: Is Omega-6 Inflammatory? The primary concern raised about seed oils is their high content of linoleic acid (LA) — an omega-6 polyunsaturated fatty acid. Soybean oil is approximately 51% LA; sunflower oil 65%; corn oil 58%; safflower oil 75%. The concern: linoleic acid is a precursor to arachidonic acid (AA), which is a precursor to pro-inflammatory eicosanoids (prostaglandins, leukotrienes, thromboxanes). The argument follows that high LA intake → high AA → high pro-inflammatory eicosanoids → chronic inflammation. This is the theoretical basis for the claim that seed oils cause inflammation. The evidence complicates this picture significantly. A 2012 systematic review found that increasing dietary linoleic acid did not increase plasma or tissue arachidonic acid in most studies — the conversion of LA to AA is tightly regulated and does not increase linearly with LA intake. A 2019 meta-analysis of 30 RCTs found that replacing saturated fat with polyunsaturated fat (primarily LA) reduced cardiovascular events by 29%. The large Women's Health Initiative and PREDIMED trials found no association between LA intake and inflammatory markers. However: these trials used whole dietary patterns, not isolated seed oils; they did not distinguish between cold-pressed and industrially refined oils; and they did not account for the omega-6:omega-3 ratio, which may matter more than absolute LA intake. The current evidence does not support the claim that dietary linoleic acid per se is pro-inflammatory at typical intake levels. The more legitimate concern is the omega-6:omega-3 ratio and the oxidation products generated during cooking.

Oxidation: The More Legitimate Concern. Polyunsaturated fatty acids (PUFAs) — the primary fat in seed oils — are chemically unstable. Each double bond in the fatty acid chain is a site of vulnerability to oxidation. Linoleic acid has two double bonds; alpha-linolenic acid (ALA, the omega-3 in flaxseed oil) has three. Compare this to oleic acid (the primary fat in olive oil and avocado oil), which has one double bond and is far more stable. When PUFAs are exposed to heat, light, or oxygen, they undergo lipid peroxidation — a chain reaction that generates: aldehydes (including 4-hydroxynonenal/4-HNE and malondialdehyde/MDA), which are cytotoxic and genotoxic; acrolein, a respiratory irritant; trans fatty acids; and other oxidation products. 4-HNE is particularly concerning: it forms protein adducts, damages DNA, and has been found in elevated concentrations in the brains of Alzheimer's patients and in atherosclerotic plaques. A 2015 study found that sunflower oil heated to frying temperature (180°C) generated significantly higher levels of aldehydes than olive oil or coconut oil at the same temperature. The key variable is the smoke point relative to cooking temperature, but more importantly, the degree of polyunsaturation — highly polyunsaturated oils (sunflower, safflower, flaxseed) generate far more oxidation products at cooking temperatures than monounsaturated oils (olive, avocado) or saturated fats (coconut, butter, ghee). This is the most scientifically grounded concern about seed oils: not that LA is inherently inflammatory, but that cooking with highly polyunsaturated oils generates toxic oxidation products.

What the Clinical Trials Actually Show. The clinical evidence on seed oils is genuinely mixed, and the interpretation depends heavily on which trials you include and how you weight them. Evidence supporting seed oil consumption: The Seven Countries Study (Ancel Keys, 1958–1970) — found that saturated fat intake correlated with cardiovascular mortality; replacing saturated fat with polyunsaturated fat (seed oils) was associated with lower CVD risk; this study has been heavily criticised for cherry-picking data and not accounting for sugar intake. The PREDIMED trial (2013, 7,447 participants) — Mediterranean diet with EVOO or nuts reduced CVD events by 30%; this trial used olive oil and nuts, not seed oils. Multiple meta-analyses of RCTs — replacing saturated fat with polyunsaturated fat reduces LDL cholesterol and CVD events. Evidence raising concerns about seed oils: The Sydney Diet Heart Study (1966–1973) — replaced saturated fat with safflower oil (high LA); the intervention group had significantly higher all-cause mortality and cardiovascular mortality despite lower LDL cholesterol. The Minnesota Coronary Experiment (1968–1973, recovered data published 2016) — replaced saturated fat with corn oil; LDL fell significantly; but for every 30 mg/dL reduction in LDL, the risk of death increased by 22%; the intervention group had higher all-cause mortality. The Rose Corn Oil Trial (1965) — corn oil group had higher mortality than control. These trials are older and have methodological limitations, but they raise a genuine question: does replacing saturated fat with omega-6 seed oils reduce cardiovascular mortality, or does it reduce LDL while increasing other risks? The honest answer is that the evidence is not as settled as mainstream nutrition guidelines suggest.

A Practical Framework: Which Oils to Use and When. Given the evidence, a practical framework for oil selection: For high-heat cooking (frying, searing, roasting above 200°C): use oils with high saturated fat content (coconut oil, ghee, butter, lard) or high monounsaturated fat content (avocado oil, refined olive oil); these are the most stable at high temperatures and generate the fewest oxidation products. For medium-heat cooking (sautéing, roasting at 160–200°C): extra virgin olive oil is the best choice — its high oleic acid content and natural antioxidants (polyphenols, vitamin E) make it more stable than seed oils at moderate cooking temperatures; avocado oil is also excellent. For cold use (dressings, finishing, dips): extra virgin olive oil, avocado oil, walnut oil, and flaxseed oil (for omega-3); these are best consumed without heating to preserve their polyphenols and omega-3 content. Oils to minimise or avoid for cooking: refined sunflower, safflower, corn, soybean, and cottonseed oils — highly polyunsaturated, prone to oxidation at cooking temperatures, and stripped of natural antioxidants during industrial refining. The nuanced position: seed oils are not the singular cause of modern chronic disease that some online commentators claim. But the evidence that highly refined, highly polyunsaturated seed oils are the optimal cooking fat is also weaker than mainstream nutrition guidelines suggest. The precautionary principle supports preferring olive oil, avocado oil, and stable animal fats for cooking, while reserving seed oils for cold applications if used at all.