Intake: Huiles de graines

What It Is

Seed oils are vegetable oils pressed or solvent-extracted from seeds and grains and rich in linoleic acid, an omega-6 polyunsaturated fat. Soybean, sunflower, corn, canola, safflower, grapeseed, cottonseed and rice bran are the common members. Linoleic acid is an essential fat: the body cannot make it and must take it from food, and seed oils are the richest common source.

These oils are cheap to press and neutral in flavor. Over the twentieth century they became the dominant fat in home kitchens and food manufacturing, a large and recent shift in how people eat.

The suspicion has three roots:

• Linoleic acid is broken down in part into signaling molecules that can promote inflammation, so omega-6 was labeled a pro-inflammatory fat.
• The oils are industrially processed, which reads as unnatural.
• Most seed oil people eat arrives inside ultra-processed food, which tracks with worse health.

Each root has a basis, and each is narrower than the blanket claim.

What the Evidence Shows

The strongest evidence contradicts the toxicity claim. When randomized trials had people swap saturated fat for polyunsaturated fat from vegetable oils, coronary heart disease events fell by about a fifth. That is the reverse of what a toxic fat would do. When trials instead added omega-6 on its own, without the swap, it made little or no clear difference, so the benefit came from the swap itself.

Linoleic acid can be measured directly in blood and tissue. That avoids the recall errors of food questionnaires. Across 30 cohorts and about 68,000 people, more of it in the blood tracked with lower cardiovascular death. Across 20 cohorts and nearly 40,000 people, it tracked with lower risk of type 2 diabetes. These are associations, and they rank below the trials.

The claim that omega-6 drives inflammation was tested directly. Across 15 randomized controlled trials, adding linoleic acid to the diet of healthy people did not raise C-reactive protein, fibrinogen, cytokines or tumor necrosis factor-alpha.

The Research & Studies

Everything here is based on the research we have collected and checked, sorted into groups and ordered with the strongest evidence first. Click any claim to open the studies behind it.

Heart And Vascular

Remplacer les graisses saturées par des graisses polyinsaturées a réduit les événements coronariens de 19 % dans des essais randomisésStrong

In plain terms

Lorsque des essais randomisés ont fait remplacer les graisses saturées par des graisses polyinsaturées issues d'huiles végétales, les événements coronariens ont diminué d'environ un cinquième. C'est le principal argument expérimental en faveur de ces huiles.

In detail

Mozaffarian et al. (PLoS Med 2010) ont regroupé les essais ayant spécifiquement remplacé les graisses saturées par des AGPI, la substitution qui compte. La réduction concernait les événements coronariens ; les effets sur la mortalité totale à travers ces essais sont plus faibles et plus mitigés, comme le montrent des analyses ultérieures de données retrouvées. Certains des essais inclus ont été menés uniquement chez des hommes.

Who this may not transfer to:The pooled trials included both sexes in aggregate, though several older component trials enrolled men only, so the result rests more heavily on data from men.

The study · 1

Mozaffarian et al., Effects on coronary heart disease of increasing polyunsaturated fat in place of saturated fat: a systematic review and meta-analysis of randomized controlled trials · PLoS Med 2010;7(3):e1000252

Le taux sanguin d'acide linoléique le plus élevé était associé à une mortalité cardiovasculaire inférieure de 22 % chez environ 68,000 personnesModerate

In plain terms

Sur 30 études et environ 68,000 personnes, celles ayant les taux sanguins les plus élevés d'acide linoléique, la principale matière grasse des huiles de graines, présentaient une mortalité cardiovasculaire environ 22 % plus faible que celles ayant les taux les plus bas, ainsi qu'une moindre incidence globale de maladie cardiaque.

In detail

Le consortium CHARGE FORCE (Marklund et al., Circulation 2019) a réalisé des analyses de novo harmonisées sur 30 cohortes, en utilisant l'acide linoléique sanguin et tissulaire mesuré afin d'éviter l'erreur de rappel des questionnaires alimentaires. Les associations étaient cohérentes selon l'âge, le sexe, la race, le diabète, l'usage de statines et d'aspirine, les niveaux d'oméga-3 et le génotype FADS1. Il s'agit d'une étude observationnelle, elle montre donc une association, pas une cause. Deux responsables de l'étude ont déclaré un soutien à la recherche de la part d'Unilever, un fabricant de ces huiles, ce qui n'invalide pas le résultat harmonisé multi-cohortes mais constitue un intérêt déclaré à prendre en compte.

Who this may not transfer to:Both sexes were represented across the 30 cohorts and no consistent heterogeneity by sex was found; effect sizes in any single population may still differ.

The study · 1

Marklund et al., Biomarkers of Dietary Omega-6 Fatty Acids and Incident Cardiovascular Disease and Mortality · Circulation 2019;139(21):2422-2436

Blood Sugar

Le taux sanguin d'acide linoléique le plus élevé était associé à un risque de diabète de type 2 inférieur de 35 % chez près de 40,000 personnesModerate

In plain terms

Sur 20 études et près de 40,000 personnes, celles ayant les taux sanguins d'acide linoléique les plus élevés présentaient un risque environ 35 % plus faible de développer un diabète de type 2 que celles ayant les taux les plus bas.

In detail

Wu et al. (Lancet Diabetes Endocrinol 2017) ont utilisé des biomarqueurs mesurés, et non des questionnaires alimentaires, à travers les cohortes CHARGE FORCE. L'association était forte et cohérente, mais elle est observationnelle, et le biomarqueur reflète autant le métabolisme que l'apport alimentaire. Plusieurs auteurs ont déclaré des subventions de recherche d'Unilever pour ce travail, un intérêt déclaré à mettre en balance avec la conception harmonisée multi-cohortes.

Who this may not transfer to:Both sexes were included and sex did not significantly modify the association; the effect in any single subgroup may still vary.

The study · 1

Wu et al., Omega-6 fatty acid biomarkers and incident type 2 diabetes: pooled analysis of individual-level data for 39,740 adults from 20 prospective cohort studies · Lancet Diabetes Endocrinol 2017;5(12):965-974

Cholesterol And Lipids

Adding omega-6 fat on its own left heart events and deaths unchanged across 19 trialsModerate · no effect

In plain terms

When trials looked at simply eating more omega-6 fat on its own, it made little or no clear difference to heart events or deaths, though it may lower heart attacks a little and it does lower cholesterol.

In detail

Hooper et al. (Cochrane 2018) assessed omega-6 fat in isolation, not as a swap for saturated fat, and rated most cardiovascular outcomes as low or very-low certainty. The one signal was for myocardial infarction, where about 53 people would need to increase omega-6 intake to prevent one MI. This tempers the strong 'heart-healthy' framing: the benefit is clearest as a substitution for saturated fat, less so from adding omega-6 by itself.

Who this may not transfer to:Trials included both sexes in aggregate; the review did not report a clear sex difference in effect.

The study · 1

Hooper et al., Omega-6 fats for the primary and secondary prevention of cardiovascular disease · Cochrane Database Syst Rev 2018;11:CD011094

Swapping saturated fat for corn-oil linoleic acid lowered cholesterol 14% but did not lower deathsModerate · no effect

In plain terms

In a large recovered trial, swapping saturated fat for corn-oil linoleic acid lowered cholesterol by about 14 percent but did not lower deaths, and pooling five such trials found the same. Lower cholesterol did not translate into longer life here.

In detail

Ramsden et al. (BMJ 2016) recovered and analyzed previously unpublished data from the MCE. It separates the cholesterol effect, which was present, from the mortality effect, which was absent in these trials, and the authors argue incomplete publication overstated the benefit of replacing saturated fat with linoleic-acid-rich oils. The population was older and institutionalized, and the trial predates modern cardiovascular care, which limits how far it generalizes, but it is the critics' strongest piece of randomized data.

Who this may not transfer to:The randomized cohort included both women and men; the population was institutionalized and older, so generalization to free-living adults is limited for reasons beyond sex.

The study · 1

Ramsden et al., Re-evaluation of the traditional diet-heart hypothesis: analysis of recovered data from Minnesota Coronary Experiment (1968-73) · BMJ 2016;353:i1246

How it works

Linoleic acid did not raise inflammatory markers across 15 controlled feeding trialsModerate · no effect

In plain terms

The common worry that omega-6 fat drives inflammation was tested directly: across 15 trials, adding linoleic acid to healthy people did not raise inflammation markers like C-reactive protein.

In detail

Johnson and Fritsche (J Acad Nutr Diet 2012) addressed the mechanistic core of the anti-seed-oil argument, that omega-6 feeds pro-inflammatory eicosanoid production. Across randomized trials in healthy noninfant adults, the predicted rise in inflammatory markers did not appear. This is about markers in healthy people, not clinical outcomes in disease, and it does not speak to oxidized or repeatedly heated oil, which is a separate question.

Who this may not transfer to:Trials were in healthy adults of both sexes; the review did not analyze a sex difference.

The study · 1

Johnson and Fritsche, Effect of dietary linoleic acid on markers of inflammation in healthy persons: a systematic review of randomized controlled trials · J Acad Nutr Diet 2012;112(7):1029-1041

The omega-6 to omega-3 ratio hypothesis is not settled by the evidencePreliminary · mixed

In plain terms

One influential idea holds that the modern diet has too much omega-6 relative to omega-3, and that this imbalance itself promotes disease. It is a serious hypothesis, and direct measurements of linoleic acid so far do not support its alarming version.

In detail

Simopoulos (Biomed Pharmacother 2002) is the widely cited statement of the ratio hypothesis. It is a serious and unsettled line of argument, and it is the strongest conceptual point the critics have. It sits in tension with the biomarker findings, where higher measured linoleic acid tracks with lower disease, and with the trial finding that adding linoleic acid did not raise inflammation. The practical implication most consistent with the data is to raise omega-3 intake, not to eliminate omega-6.

The study · 1

Simopoulos, The importance of the ratio of omega-6/omega-3 essential fatty acids · Biomed Pharmacother 2002;56(8):365-379

Where Caution Is Warranted

Oil reused through many fry cycles builds up reactive aldehydes; a single home fry with fresh oil does not.

Cautions For This Practice

Extra restraint

Everything to be aware of is here, in one place. This practice suits most healthy people; a few situations call for real care.

Repeatedly reheated frying oil raised blood pressure and cholesterol and damaged blood vessels in animals

Ng et al. (Vascul Pharmacol 2014) reviews the harm signal from repeatedly heated oil, which is about how the oil is treated, not the fresh oil itself. The evidence is mostly from animal feeding studies using oil reheated to a degree typical of commercial deep-frying, not a spoon of fresh oil used once at home, so it should be read as a caution about reheating practice, not about seed oils in general.Ng et al., Heated vegetable oils and cardiovascular disease risk factors

The toxicity claim goes beyond the evidence

The strong version says these oils are uniquely toxic and a main driver of chronic disease. Two specific concerns are narrower and better supported: reheated frying oil, and a diet heavy in omega-6 and light in omega-3.

Repeatedly reheated frying oil is the clearest harm signal

Oil reused through many fry cycles builds up reactive aldehydes; a single home fry with fresh oil does not. In animals fed reused oil over time, blood pressure and cholesterol rose and blood vessels were damaged. The data is mostly from animals and from extreme reheating, so the practical point is narrow: fry with fresh oil, and retire it once it darkens.

The population benefit is association, and some is industry-funded

The blood-marker findings come from observational studies. Those people also tend to eat more whole plant food, smoke less and move more, and the statistics cannot fully strip that out. Several of the largest analyses had research support from companies that sell these oils. Both are reasons to weight the trials more heavily, and to read the reassurance with the funding in view.

Clearer for heart attacks than for length of life

Swapping saturated fat for polyunsaturated fat cut coronary events. But a recovered corn-oil trial lowered cholesterol by about 14 percent without lowering deaths, and pooling five such trials found the same. That gap is why a blanket heart-healthy claim overstates the trials.

If you have heart or metabolic disease, one small step

These oils are ordinary food for most people, and cooking with them needs no supervision. If you are managing heart or metabolic disease and want to act on the balance, the step you control is small. Eat more oily fish for omega-3, and cut back on deep-fried and packaged food. Direct-to-consumer testing can measure your own omega-3 index if you want to track it. Bring in a professional for tailored guidance if you have a condition or a question.

Start slow, be smart, read the research, and consult a professional if you have any concerns. This is here to inform your choice, not make it for you.

How It Works

A seed oil acts on the body in three ways: as a fat built into cell membranes, as a replacement for saturated fat, and as something that oxidizes when overheated.

Anatomy of the Practice

1What the fat does in the body

Linoleic acid is built into cell membranes and converted in small amounts into signaling molecules. Some of those molecules promote inflammation and some resolve it. Because both directions are produced, adding linoleic acid does not simply raise inflammation.

2What replacing saturated fat does

Swap butter, lard or palm fat for a polyunsaturated oil and blood LDL cholesterol falls, one of the most reliable findings in nutrition. Whether that lower cholesterol becomes fewer heart attacks depends on the study. Coronary events fell in the pooled trials, while the effect on total deaths stayed weaker and mixed.

3What heat and processing do

The double bonds that make these fats good to eat cold also make them oxidize when they are heated hard and repeatedly. Oil used fresh and once carries a light load of oxidation products; repeated high heat drives that load up. Nearly all seed oil is eaten inside packaged and fried food, so the population data cannot separate the oil from the food carrying it.

Ways to Do It

The useful moves here are free or cheap, and both sides of the argument can share them. Start at the first rung.

1

Cut deep-fried and ultra-processed food firstFreeEasy

This is the highest-value move. Shift your plate toward whole, minimally processed food, and you handle most of your seed-oil exposure at once, along with a lot else. It costs nothing and singles out no ingredient.

2

Do not reuse frying oil over and overFreeEasy

Frying at home now and then with fresh oil is fine. Trouble builds when a fryer recycles the same oil batch after batch. Use frying oil once or a few times, then discard it before it turns acrid.

3

Use olive oil as your everyday fat$Easy

Extra-virgin olive oil carries the strongest whole-diet evidence of any fat, and it is mostly monounsaturated. It is not a seed oil, so this debate does not touch it. It is the main fat of the [Mediterranean](/go/integrative/practice/mediterranean-diet) diet, the best-tested eating pattern we have.

4

Match the fat to the heat$Easy

For high-heat searing and repeated frying, a more heat-stable fat holds up longer before it oxidizes: olive oil, avocado oil, or modest amounts of butter, ghee or tallow. For dressings, low-heat cooking and baking, a fresh seed oil is ordinary.

5

Raise omega-3: oily fish twice a week$Easy

If the omega-6 to omega-3 balance is what concerns you, the direct move is to lift the omega-3 side: oily fish twice a week, or the plant sources. Raising [omega-3](/go/integrative/practice/omega-3-fish-oil) shifts the ratio and brings benefits of its own.

Go Deeper

Five related pages go further:

• The truth about dietary fat, how saturated, monounsaturated and polyunsaturated fats compare.
• Whole foods and ultra-processed food, the food most seed oil rides inside.
• The Mediterranean diet, the best-tested pattern, built on olive oil.
• Omega-3 and fish oil, the other end of the ratio.
• Inflammation, what the markers in these trials measure.

Common Questions

Are seed oils toxic or bad for you?

Toxic overstates it. For most people these are ordinary cooking fats, and no controlled trial has linked the fresh oil to disease. Two real concerns remain: oil that has been reheated hard and often, and a diet that runs heavy on omega-6 with little omega-3.

Is canola oil bad, or different from the others?

Canola, pressed from rapeseed, is a seed oil, but it sits a little apart from the group. It carries less linoleic acid than sunflower or corn oil, and a small amount of omega-3. Beyond that it follows the same picture as the rest of the list.

What is the omega-6 to omega-3 ratio argument?

One line of thinking holds that shifting this balance feeds chronic inflammatory disease, a serious but unsettled idea. The blood measurements and feeding trials argue against it. The traditional human diet carried the two fat families near one-to-one, and no wider than four-to-one; the modern diet has pushed that far past it.