No trial shows seed oils cause disease. No trial shows they prevent it either.

There is no body of evidence showing that seed oils cause cardiovascular or metabolic disease in humans. The largest test of the question — pooled individual-level data on 68,659 people across 30 prospective studies in 13 countries — found the opposite direction, with higher circulating and tissue levels of linoleic acid, the main omega-6 fat in seed oils, associated with lower risk of cardiovascular disease and cardiovascular death [s1]. But the confident counter-claim, that seed oils are proven protective, overstates the case: the randomised trial evidence is null, sparse and graded low to very low quality by Cochrane [s2].

What the claim against seed oils is

The oils at issue — soybean, sunflower, corn, canola, cottonseed, safflower, grapeseed, rice bran and their relatives — are high in linoleic acid, an omega-6 polyunsaturated fat. The argument against them holds that linoleic acid is converted to arachidonic acid, which feeds pro-inflammatory signalling, and that oxidation products formed during processing or heating do further damage. A narrative review in Nutrition Reviews characterises the resulting public campaign as a shift in opinion away from unsaturated fat-rich foods and toward saturated fats, and sets out to weigh it against the human research record [s3].

The mechanistic story is not fabricated. Linoleic acid is a precursor to arachidonic acid, and arachidonic acid does sit upstream of inflammatory eicosanoids. The question is whether that biochemistry produces measurable disease in people who eat these oils.

The biomarker evidence

The strongest data on that question avoid dietary questionnaires entirely. A global consortium ran harmonised, individual-level analyses of 30 prospective observational studies from 13 countries, measuring linoleic acid and arachidonic acid directly in blood and adipose tissue as a percentage of total fatty acids [s1]. Across follow-up periods ranging from 2.5 to 31.9 years, 15,198 cardiovascular events occurred among 68,659 participants [s1].

Higher linoleic acid levels were associated with lower risk of total cardiovascular disease (hazard ratio 0.93, 95% CI 0.88 to 0.99 per interquintile range), cardiovascular mortality (0.78, 0.70 to 0.85) and ischaemic stroke (0.88, 0.79 to 0.98), and non-significantly with lower coronary heart disease risk (0.94, 0.88 to 1.00) [s1]. Arachidonic acid — the metabolite that the inflammation argument turns on — was not associated with higher risk of any cardiovascular outcome; comparing extreme quintiles, higher levels went with lower total cardiovascular disease (0.92, 0.86 to 0.99) [s1]. No consistent heterogeneity appeared across age, sex, race, diabetes status, statin or aspirin use, omega-3 levels, or fatty acid desaturase 1 genotype [s1].

This is observational. Tissue linoleic acid is a reasonable proxy for intake, but people with high levels differ from people with low levels in ways a statistical model can only partly absorb.

The randomised evidence, which is thin

Cochrane's review asked the narrower question that trials can answer: what happens when you randomise people to eat more omega-6 fat. It found 19 randomised controlled trials in 6,461 participants followed for one to eight years — seven testing gamma-linolenic acid, 12 testing linoleic acid, none testing dihomo-gamma-linolenic or arachidonic acid, with the omega-6 fat usually displacing saturated or monounsaturated fat [s2]. Only three of the 19 were judged at low summary risk of bias [s2].

The results were null. Low-quality evidence suggested increasing omega-6 intake makes little or no difference to all-cause mortality (risk ratio 1.00, 0.88 to 1.12; 740 deaths across 4,506 randomised in 10 trials) or cardiovascular events (0.97, 0.81 to 1.15; 1,404 people with events of 4,962 randomised in seven trials) [s2]. For cardiovascular mortality (1.09, 0.76 to 1.55), coronary heart disease events (0.88, 0.66 to 1.17), major adverse cardiac and cerebrovascular events (0.84, 0.59 to 1.20) and stroke (1.36, 0.45 to 4.11), the review rated the evidence very low quality and stated it was uncertain whether omega-6 fats have any effect [s2]. No dose-response or duration effect was found for any primary outcome, though the reviewers noted a suggestion of greater protection in participants with lower baseline omega-6 intake [s2].

What follows from a null with wide intervals

The Cochrane result is frequently cited by both camps, and neither reading is right. It does not show seed oils are harmful — the point estimate for mortality is exactly 1.00 [s2]. It does not show they are protective either. It shows that the trials that exist are too few, too small, too short and too poorly conducted to detect a modest effect in either direction, which is what "low quality" and "very low quality" mean in the GRADE framework the review uses [s2].

The practical consequence is that the biomarker consortium is doing most of the work in this literature, and it is observational [s1]. The Nutrition Reviews narrative — which is a narrative review, not a systematic one, and argues an explicit position — makes the displacement point that the trials structurally cannot: what replaces seed oil matters, and the comparator most likely to be used in practice is saturated fat, whose relationship with serum cholesterol and atherosclerotic risk is far better established [s3].

What is genuinely unresolved

Almost nothing in this literature isolates industrially refined seed oil as consumed in ultra-processed foods from linoleic acid as a nutrient. The biomarker studies measure the fatty acid in tissue, not the oil in the bottle or the fryer [s1]. The Cochrane trials mostly substituted one fat for another under research conditions, not one food environment for another [s2]. Anyone claiming a settled answer about the health effect of the oils as they are actually eaten — reheated in commercial frying, embedded in packaged foods — is claiming more than these three papers deliver.

This article is informational and is not medical advice.

Sources

• Biomarkers of Dietary Omega-6 Fatty Acids and Incident Cardiovascular Disease and Mortality — Circulation, 2019
• Omega-6 fats for the primary and secondary prevention of cardiovascular disease — Cochrane Database of Systematic Reviews, 2018-11-29
• Are Seed Oils the Culprit in Cardiometabolic and Chronic Diseases? A Narrative Review — Nutrition Reviews, 2025-01-16