Are Seed Oils Bad for You? What the Evidence Says

Seed oils are not bad for you on the evidence we have, and the strongest studies point the other way. But the online argument mixes a few real questions in with the exaggerations, so it is worth separating them. Here is what the claims are, what the research shows, and where the honest uncertainties sit.

What are seed oils, and what is the “hateful eight”?

Seed oils are vegetable oils pressed or extracted from seeds rather than fruit: canola (rapeseed), sunflower, soybean, corn, cottonseed, safflower, rice bran and grapeseed. That list of eight is the one critics call the “hateful eight”. Olive, avocado and coconut oil come from the fruit, so they sit outside the argument.

What the eight share is a lot of polyunsaturated fat, and in particular linoleic acid, an omega-6 fatty acid that makes up 80 to 90% of the polyunsaturated fat in a typical diet. Linoleic acid is essential, meaning your body cannot make it, and the argument is about whether we now eat too much.

What do critics say is wrong with seed oils?

The claims come in four parts:

1. Inflammation. The body converts some linoleic acid into arachidonic acid, which is a building block for inflammatory signalling molecules. More omega-6 should therefore mean more inflammation.
2. Oxidation. Polyunsaturated fats oxidise more readily than saturated ones, and heating them produces aldehydes and other breakdown products.
3. Processing. Most seed oils are extracted with hexane, a petroleum solvent, then refined, bleached and deodorised.
4. Novelty and ubiquity. Industrial seed oils barely existed before the 20th century, intake has climbed steeply, and they are in almost every packaged food.

Each is a reasonable question. Here is what happens when you test them.

Does omega-6 cause inflammation?

The arachidonic acid theory makes chemical sense, and it has been tested directly. A 2012 systematic review found 15 randomised trials in healthy adults that added linoleic acid to the diet and measured inflammatory markers: C-reactive protein, fibrinogen, cytokines, TNF-alpha and others. None found a significant increase (Journal of the Academy of Nutrition and Dietetics). A 2017 meta-analysis of 30 trials and 1,377 people reached the same conclusion (Food and Function). The reason seems to be that very little dietary linoleic acid is converted to arachidonic acid; the body regulates that step tightly, a point the 2026 JACC clinician’s guide makes (JACC: Advances).

Blood arachidonic acid itself was not linked to higher heart risk in the largest pooled analysis to date, and if anything the opposite (Circulation, 2019).

The trials were mostly small and short, and a 2026 review noted that the size of the CRP change tended to track the size of the linoleic acid change, so the question is not fully closed (Critical Reviews in Food Science and Nutrition). But “seed oils cause inflammation” is a claim the controlled trials have looked for and not found.

What do heart disease studies show?

This is where the evidence is strongest, and it runs against the critics.

• Randomised trials. The American Heart Association’s 2017 advisory reviewed trials that cut saturated fat and replaced it with polyunsaturated vegetable oil, and found cardiovascular disease fell by about 30%, a reduction it compared to statin treatment (Circulation).
• Blood levels. A pooled analysis of 30 cohorts in 13 countries, 68,659 people and 15,198 cardiovascular events, found that people with higher linoleic acid in their blood or fat tissue had lower rates of total cardiovascular disease, cardiovascular death and ischaemic stroke (Circulation, 2019).
• The cautious reading. A 2018 Cochrane review of trials that increased omega-6 found little or no effect on deaths or total cardiovascular events, with a possible reduction in heart attacks on low-quality evidence (Cochrane). Neutral to mildly helpful, not harmful.
• The dissenting data. The critics’ best evidence is a 2013 reanalysis of the Sydney Diet Heart Study, a 1960s trial of 458 men who had already had a heart event. Those given safflower oil in place of saturated fat had higher death rates (BMJ). A 2016 reanalysis of the Minnesota Coronary Experiment found that corn oil lowered cholesterol without lowering deaths (ScienceDaily). Both are real results. Both are also old, small trials with wide confidence intervals, and the safflower margarine used in the 1960s may well have contained trans fats, which muddies what was being tested.

Australia’s Heart Foundation summed up the current position in a May 2026 statement: oils high in unsaturated fat, including canola, sunflower and soybean, support heart health when used in place of saturated fats such as butter, coconut oil and ghee (Heart Foundation). The 2026 JACC clinician’s guide lists seed oils among the foods with evidence of benefit for cardiovascular outcomes.

Claim	What the evidence shows
Omega-6 drives inflammation	Not found in dozens of randomised trials measuring inflammatory markers
Seed oils cause heart disease	Trials and blood-level studies link them to lower risk when replacing saturated fat
Hexane residues are dangerous	Residues in oil are far below the EU limit of 1 mg/kg; EFSA is re-checking the old toxicity data
Heated seed oil is toxic	Repeated high-heat frying does form harmful aldehydes; fresh oil at normal cooking heat does not
They are in all junk food	True, and that is the processed food’s problem rather than the oil’s

Is the hexane in seed oils dangerous?

Hexane is used to pull the last of the oil out of crushed seeds, then boiled off during refining. What remains is a trace. In the EU the legal limit is 1 mg/kg in oil, and surveys of canola, soybean and even olive oil have found residues well under it, with a maximum of 0.58 mg/kg in one review (EUFIC, OCL).

What is true is that the toxicity study behind the EU limit is decades old. EFSA said in 2024 that it needs re-evaluation, held a stakeholder meeting with industry in March 2026, and is expected to give an opinion in 2027 (EFSA). That is a regulator updating old data, not a finding of harm. In the US, hexane is permitted as a processing solvent and we did not find a specific residue limit in oil, which is a fair thing to want. If it bothers you, cold-pressed and expeller-pressed oils are made without any solvent and are widely sold.

What about frying, and reheating oil?

This is the concern with the most science behind it, and it applies to every oil to some degree. Heating any fat for long enough breaks it down. Polyunsaturated oils do this faster, and the breakdown products include aldehydes such as 4-HNE that are genuinely toxic in the lab. In one often-cited University of Minnesota experiment, three hours at 190C produced the highest 4-HNE levels in corn oil and soybean oil, and in butter (Critical Reviews in Food Science and Nutrition).

Three hours at deep-fry temperature is a chip shop, not a stir-fry. The practical advice, which even the pro-seed-oil reviews give, is not to reheat the same oil again and again at high heat. A fresh splash of canola in a pan for ten minutes is not the same chemistry. If you deep-fry often, high-oleic sunflower or canola, which are bred to have less polyunsaturated fat, and peanut or olive oil, hold up better.

Are seed oils the problem, or the food they are in?

Here the critics have half a point. Seed oils are cheap and stable, so they are in chips, biscuits, fried fast food, sauces and most ultra-processed products. People who eat a lot of those foods do have worse heart outcomes, and a 2026 review found that association holds even after accounting for the salt, sugar and saturated fat (News-Medical).

But the oil is a passenger in that data. When researchers separate the oil from the food it rides in, the oil comes out fine or better than fine, which is why the British Heart Foundation and US dietitians both describe the ultra-processed food link as a problem with the whole product rather than its fat (British Heart Foundation, Healio).

The same logic applies to newer products. Plant-based meats, including the 3D printed steaks now on US shelves, use added oils and binders to mimic fat and texture, and the honest question is about the whole ingredient list rather than one line of it. We go through that list in Perfecta 3D printed steak: what’s in it?

Common questions

Which oils are the “hateful eight” seed oils?

Canola, sunflower, soybean, corn, cottonseed, safflower, rice bran and grapeseed. The label was coined by critics online; it is not a scientific category.

Is canola oil bad for you?

The evidence says no. Canola is low in saturated fat and has some omega-3 as well as omega-6. The Heart Foundation names it among the oils that support heart health when used instead of butter or coconut oil.

Does omega-6 cause inflammation?

Randomised trials that added linoleic acid to people’s diets and measured inflammatory markers have not found an increase. The body converts very little of it to arachidonic acid, the compound the theory relies on.

Is it safe to cook with seed oils?

Yes for normal cooking. The real concern is oil heated to deep-fry temperatures for hours and reused, which produces harmful aldehydes. Use fresh oil, do not let it smoke, and do not keep reusing frying oil.

Are seed oils worse than butter or beef tallow?

Butter and tallow are mostly saturated fat, and trials that replaced saturated fat with polyunsaturated oil cut heart disease by about 30%. Heart groups in Australia, the UK and the US all recommend the swap in that direction, not the reverse.

The bottom line

Seed oils are not the villain the posts describe. Trials and blood-level studies tie linoleic acid to lower heart risk, the inflammation claim has been tested and not found, and hexane residues are trace amounts under review rather than a discovered hazard. Keep your frying oil fresh, eat less ultra-processed food, and the oil in your pantry is the least of your worries.

Keep reading: Food Science · Perfecta 3D printed steak: what’s in it? · Do olives and olive oil give you gas? · Is sesame oil always gluten free?