Are seed oils bad for you?

Probably not — but both sides of this argument are overclaiming, and you deserve the honest version.

The mechanism everyone cites doesn't hold up

The entire case rests on: omega-6 (linoleic acid) → inflammation → disease.

Fifteen randomised controlled trials have raised people's linoleic acid intake and measured their inflammatory markers. Not one reported a rise in any of them — CRP, fibrinogen, TNF-α, cytokines, adhesion molecules. The foundational step of the argument has never been confirmed in a controlled human trial.

Read that with one eye open, though: the review collecting those fifteen trials was funded by an industry body, its first author consulted for a major soybean-oil processor, and it counted votes rather than pooling data. An industry-funded null that flatters its funder is exactly the result to be suspicious of. It is still the most direct test anyone has run — but it is not the slam dunk it is usually presented as.

The strongest evidence for the claim is real, and I won't strawman it

Chris Ramsden's team recovered the lost raw data from two 1960s trials and re-analysed it. The Minnesota Coronary Experiment (n=9,423, double-blind) found that swapping saturated fat for vegetable oil lowered cholesterol but did not reduce deaths — and in older participants, deaths went up. The Sydney Diet Heart re-analysis found higher all-cause mortality.

That is serious work and it is genuinely awkward for the mainstream position. It is on the evidence page as pro.

The catch: both trials used margarine containing trans fats, which we now know are harmful. That's a real confound the seed-oil camp rarely mentions.

The frying argument is weaker than it looks

Fried food really does track with heart disease — 28% more major cardiovascular events in the top intake group, across 562,445 people. But fried food is also fast food, refined carbohydrate and surplus calories, so that number cannot be pinned on the oil.

The one experiment that tries to isolate the oil fed people a meal made with oil carrying roughly three times the oxidation products of fresh oil, and measured their blood-vessel function. Comparable results — and a lower triglyceride rise. That is a real finding, but hold it loosely: 19 men, one meal, four hours. It says nothing about years of eating reheated frying oil, and the authors say so.

The honest position

• The two randomised trials that found harm are the seed-oil camp's best cards, and both are confounded by trans fat and run in men who had already had a heart attack. Set against them are larger, cleaner datasets pointing the other way. Net: no clear human evidence that seed oils harm you at normal intakes — but "no clear evidence" is not the same as "proven safe", and the trials that unsettle the picture are still on the page.
• They aren't proven health foods either. A Cochrane review found increasing omega-6 gave a mortality risk ratio of exactly 1.00 — no benefit, no harm, low-quality evidence.
• The real finding is narrower than either camp admits: seed oils look better than butter, and that's about all you can say with confidence.

If you want to avoid them, you're not harming yourself. Just don't believe you're avoiding a poison — and notice that "seed oil free" is mostly a label on ultra-processed food.