Walk into a supermarket in 2026 and the shelf has started talking back. Bottles stamped “seed oil free.” Restaurant menus that advertise tallow instead. A cookbook aisle in which “industrial oils” is the villain of the season, blamed in turn for inflammation, gut trouble and a list of chronic diseases.
The claim is not subtle: that sunflower, canola, soybean and corn oils — the cooking fats in most of the packaged food supply — are a modern poison. What the published literature has done this year is quieter. Three studies, spread across the past eight months, keep arriving at the same place from different directions.
The chain the argument rests on
The popular case leans on one piece of biochemistry. Linoleic acid, the dominant omega-6 fat in seed oils, is a precursor to arachidonic acid, and arachidonic acid is a precursor to signalling molecules involved in inflammation. From that chain, the leap is made to chronic inflammation, and from chronic inflammation to heart disease.
The chain is real. The leap is where the evidence thins — and it is a leap that has now been tested several ways.
What 144,900 adults showed
The largest new contribution is a Mexican prospective cohort published in Clinical Nutrition ESPEN on 8 October. Researchers measured circulating fatty acids by nuclear magnetic resonance in 144,900 adults aged 35 and older, using blood drawn between 1998 and 2004, and followed deaths through October 2022.
Higher omega-3 levels went with lower all-cause mortality. Per interquintile range the hazard ratios were 0.94 for DHA, 0.83 for the non-DHA omega-3 fraction and 0.88 for total omega-3; people in the top fifth for non-DHA omega-3 had a 20% lower hazard than those in the bottom fifth.
Linoleic acid itself showed no association with mortality. The other omega-6 measures ran the other way from the “toxic” framing, with a hazard ratio of 0.90 (95% CI 0.87–0.92) for non-linoleic omega-6 per interquintile range.
What a cohort like this cannot do
One plasma measurement is a snapshot of metabolism, not a food diary, and people with higher linoleic acid could differ from the rest in ways no model captures. Observational data of this kind cannot establish cause. What it can do is fail to find harm in a very large, very long dataset — which is exactly what happened here, in a population that is under-studied compared with North American and European cohorts.
The inflammation question, tested rather than argued
Trials that shift the fat in someone’s diet are a different kind of evidence, and on this question the experimental record is thin and unglamorous.
A meta-analysis in Frontiers in Nutrition pooled nine randomized controlled trials with 504 participants that measured inflammatory markers after omega-3 or omega-6 supplementation. Across the pooled group, neither fatty acid moved interleukin-6, C-reactive protein or TNF-alpha to a statistically significant degree. Interleukin-1 beta fell by a small margin — a mean difference of −0.04 (95% CI −0.07 to −0.01) — and the effect was slightly larger in the omega-6 subgroup. The authors describe the overall picture as limited effects on most inflammatory markers.
Nine trials and 504 people is not a verdict on a food supply. It is, though, the opposite of a demonstration of harm.
A cross-sectional analysis in The Journal of Nutrition took a different route. It compared serum omega-6 and linoleic acid against five inflammation markers in 2,133 adults, and found higher fatty acid concentrations tracking with lower high-sensitivity C-reactive protein, glycoprotein acetyls and sICAM-1, with no significant trend for serum amyloid A or sVCAM-1. The authors concluded that their findings “do not support the hypothesis” that dietary omega-6 is pro-inflammatory.
That study should be read with the disclosure page open. Several of its authors consult for food and pharmaceutical companies, and the listed relationships include ACH Food Companies and Seed Inc. — commercial interests in the very category under discussion. The design also cannot establish which direction the association runs. Treat it as corroboration with a conflict of interest attached, not as proof.
The other side of the ledger is no sturdier. A double-masked crossover pilot in the same journal put ten adults on 30 g a day of soybean or palm oil, delivered through study foods, for four weeks at a time. Linoleic acid biomarkers moved as expected and body weight did not change. Ten participants, funded by the United Soybean Board. That is roughly what the experimental literature on this question currently looks like.
The guidance never changed
None of this is news to the agencies. The World Health Organization’s healthy-diet fact sheet treats linoleic acid as an essential fatty acid that “can only be obtained from the diet,” sets a floor of 15% of daily energy from fat, caps saturated fat at 10%, and names soybean, canola, corn, safflower and sunflower oils among the preferred replacements for butter, lard and ghee.
MedlinePlus, the National Library of Medicine’s consumer encyclopedia, is blunter still: avoid saturated fats such as butter and lard, and “try to replace them with oils such as canola, olive, safflower, sesame, or sunflower.”
Harvard’s Nutrition Source adds numbers. The American Heart Association suggests 8–10% of daily calories from polyunsaturated fat, with evidence that up to 15%, substituted for saturated fat, can lower heart disease risk. Its table of cooking fats shows why seed oils sit at the centre of this argument in the first place: sunflower oil is about 66% polyunsaturated fat, corn oil 60%, soybean oil 37%, against 8% for olive oil and 5% for butter.
Where the criticism still lands
Two parts of the sceptics’ case survive contact with the evidence, and neither one is about inflammation.
The first is a labelling problem. “Highly processed” is a category, not a chemical property, and the classification systems do not agree on where refined oils belong — the same argument that complicates ultra-processed food research. It matters most for what oils are used to build: the crisps, the fried snacks, the bakery case.
The second is quantity. Oil is the most calorie-dense thing in the kitchen, so a drizzle and a pour are not the same act, and any fried food carries the oil it was cooked in. If the goal is trimming calories without turning dinner into a spreadsheet, the bottle is a place to look — for reasons that have nothing to do with omega-6.
What remains unaddressed here is what happens to oil at high heat, and to oil used over and over in a fryer. Those are questions about cooking practice rather than about the seed the oil came from, and they are not settled by any of the studies above.
What this settles
Nothing in this year’s record says seed oils are a health food, and none of it is a reason to reorganise anyone’s diet on its own authority. The narrower conclusion is that the specific proposition driving the trend — that linoleic acid inflames the body and damages the heart — has not held up in the studies designed to test it, including a cohort of nearly 145,000 people followed for up to two decades.
What the evidence does not provide is a personal prescription. Anyone weighing a change to fat intake because of cardiovascular disease, a lipid disorder or a family history should have that conversation with a clinician or a registered dietitian who knows the rest of their history. This is news coverage of published research, not medical advice — the boundaries are set out in our medical disclaimer.