Avoiding seed oils

• Lower intake of the omega-6 fat linoleic acid and improve the omega-6 to omega-3 ratio.
• Reduce inflammation and oxidative stress that followers attribute to seed oils.
• Lower risk of heart disease, obesity and type 2 diabetes.
• Eat fewer refined and industrially processed fats and fewer packaged and fried foods.

The phrase 'seed oils' is used mostly by critics. Physician Cate Shanahan popularized a list she calls the 'hateful eight': canola, corn, cottonseed, soy, sunflower, safflower, grapeseed and rice bran oil. Nutrition scientists usually call them vegetable oils and group them with other unsaturated fats.

What these oils share is a high content of polyunsaturated fat, mostly linoleic acid, an omega-6 fatty acid the body cannot make and needs in small amounts. Olive and avocado oil are mostly monounsaturated and are not on the list. High-oleic versions of sunflower and safflower oil are bred to be low in linoleic acid.

How much has intake changed?

A lot. Linoleic acid in the body fat of US adults rose by 136% over about 50 years, tracking the shift toward soybean oil and other seed oils [19]. This rise is a central part of the critics' case. It is a correlation over a period when almost everything about the food supply changed, including the rise of ultra-processed food.

Where seed oils show up

Most seed oil in modern diets comes from packaged and restaurant food rather than home cooking: fried foods, chips, crackers, baked goods, dressings, mayonnaise and ready meals. This is why avoiding seed oils overlaps heavily with avoiding ultra-processed food.

Core rules most followers use:

1. No seed oils at home. Remove canola, soybean, corn, sunflower, safflower, cottonseed, grapeseed and rice bran oil from the kitchen.
2. Read labels. Skip packaged foods listing those oils, which rules out most chips, crackers, dressings and mayonnaise.
3. Choose restaurants and fried foods carefully. Ask what the fryer oil is; some chains now fry in beef tallow.
4. Cook with other fats. Olive and avocado oil are accepted by both camps; animal fats and coconut oil are favored by many followers.

Variants

• Linoleic acid minimization: a target such as keeping linoleic acid to a few percent of calories, which also limits nuts, seeds and some poultry and pork.
• Refined-oil avoidance: cold-pressed or unrefined oils allowed, refined and deodorized oils avoided.
• Swap to olive oil: the mildest version, which replaces seed oils with extra virgin olive oil and keeps butter low. This is close to a Mediterranean diet.

What the protocol leaves out

The protocol is defined by what it removes, not by what replaces it. That matters: the evidence below shows that the health effect depends almost entirely on the swap. Replacing seed oils with olive oil is a very different change from replacing them with butter or tallow.

The case against seed oils rests on four ideas:

• Omega-6 and inflammation. Linoleic acid can be converted to arachidonic acid, a precursor of inflammatory mediators such as prostaglandins and leukotrienes [11].
• Omega-6 to omega-3 balance. Modern diets have far more omega-6 than omega-3. The US MAHA Commission report says seed oils contribute to an imbalanced ratio and calls this a topic of ongoing research [20].
• Oxidation. Linoleic acid has two double bonds and oxidizes easily. Its oxidized products (OXLAMs) and oxidized LDL are proposed drivers of atherosclerosis [21]. Cate Shanahan argues that oxidative stress from these oils is the root of most chronic disease.
• Refining and processing. Industrial refining reduces micronutrients such as vitamin E and phytosterols [20]. Critics also worry about solvent residues such as hexane; food scientist Eric Decker told NPR the traces are too low to be toxic.

What supports these ideas

Some evidence fits. Lowering dietary linoleic acid for 12 weeks lowered blood OXLAMs in people with chronic headaches [12]. In mice, a soybean oil diet caused more weight gain, diabetes and fatty liver than a coconut oil diet with the same calories [18]. And recovered data from two old trials found no benefit, or harm, when linoleic acid replaced saturated fat [2] [3].

Mechanisms and mouse data are a reason to test an idea, not proof of harm in people. The next sections look at what happened when people actually ate more or less of these oils.

The inflammation claim is the most popular and the most directly tested.

• Markers of inflammation. None of 15 randomized trials found that adding linoleic acid raised C-reactive protein, fibrinogen, cytokines or TNF-alpha in healthy people [1].
• Liver fat and inflammation. In 67 adults with abdominal obesity, a 10-week diet high in vegetable omega-6 fat lowered liver fat compared with a butter-based diet, with no signs of increased inflammation or oxidative stress [10].
• Overfeeding. When young adults overate muffins made with sunflower oil or palm oil for 7 weeks, palm oil raised liver and visceral fat much more [17].

Reviewers who know the arachidonic acid pathway well conclude that eating more linoleic or arachidonic acid does not raise most inflammatory markers in healthy adults, although the interplay with omega-3 is not fully understood [11].

What about oxidation?

Lowering linoleic acid does lower oxidized linoleic acid metabolites in blood [12]. No trial has shown that this change prevents heart disease or other illness. Oxidation during frying is real, which is one reason to limit deep-fried food whatever the oil.

If you test hs-CRP, it is a reasonable way to see whether your own diet change moves inflammation; see the hs-CRP guide.

These two trials are the strongest evidence for the seed oil critics, and they deserve a careful reading.

Minnesota Coronary Experiment (1968 to 1973). 9,423 residents of a nursing home and six state mental hospitals were randomized to a diet where corn oil and corn oil margarine replaced saturated fat, or to a control diet. The original 1989 report found no difference in cardiovascular events or deaths; the mean time on the diet was only 384 days [22]. Christopher Ramsden's team recovered the raw data in 2016. Serum cholesterol fell 13.8% versus 1.0%, but there was no mortality benefit, and each 30 mg/dL (0.78 mmol/L)0.78 mmol/L (30 mg/dL) drop in cholesterol was linked to 22% higher risk of death [2].

Sydney Diet Heart Study (1966 to 1973). 458 men with a recent coronary event were told to replace saturated fat with safflower oil. All-cause death was 17.6% versus 11.8% in controls (hazard ratio 1.62), and coronary death was also higher [3].

How they fit with the other trials

The split comes from which trials count. Pooled analyses that include trials where omega-3 also rose, or where trans fat fell, show benefit. Trials that raised mainly linoleic acid show none [23] [9]. The old trials also had problems: in Minnesota the average time on the diet was only 384 days [22], and in all eight datasets of one analysis the swap replaced a mix of saturated and trans fat, not saturated fat alone [23]. No modern trial has been run.

Observational studies cannot prove cause, but the biomarker studies avoid one common weakness: they measure linoleic acid in blood or fat tissue instead of relying on food questionnaires.

• Heart and stroke. In 68,659 people from 30 cohorts in 13 countries, higher linoleic acid levels were linked to 7% lower total cardiovascular disease, 22% lower cardiovascular death and 12% lower ischemic stroke risk [5].
• Diabetes. In 39,740 adults from 20 cohorts, higher linoleic acid levels were linked to 35% lower type 2 diabetes risk [16].
• Mortality. Across 44 cohorts, the highest linoleic acid intake was linked to 13% lower total and cardiovascular mortality and 11% lower cancer mortality [13].
• Swaps. Replacing 5% of energy from saturated fat with polyunsaturated fat was linked to 27% lower total mortality [14]. In 221,054 US adults followed for up to 33 years, the highest plant oil intake was linked to 16% lower mortality and the highest butter intake to 15% higher; soybean and canola oil were each linked to lower mortality [7].

Why critics doubt the cohorts

People who choose vegetable oils over butter may differ in many other ways, and early guideline advice may have nudged health-conscious people toward them. That is a fair limitation. The biomarker studies are harder to explain away, because blood linoleic acid is not a lifestyle choice people report, but they still cannot fully separate oils from the diets they come in.

If you decide to avoid seed oils, what you eat instead decides the effect.

The American Heart Association concluded that trials replacing saturated fat with polyunsaturated vegetable oil cut cardiovascular disease by about 30% [8]. Christopher Gardner of Stanford has said he thinks junk food, not seed oils, explains the health problems critics attribute to seed oils, and that seed oil in a salad dressing that gets people eating vegetables is a win.

What to watch

If you switch to animal fats, check your lipids. A rise in LDL cholesterol or ApoB after the change is the main known risk. Many people who avoid seed oils also cut packaged and fried food, which may improve triglycerides and weight for reasons unrelated to the oil itself.

• Replacing seed oils with large amounts of butter, tallow or coconut oil and not checking LDL cholesterol or ApoB.
• Assuming 'seed-oil-free' packaged snacks or fried foods are healthy; they are often still ultra-processed.
• Cutting nuts and seeds, which are whole foods rich in fiber and unsaturated fat, in an effort to minimize linoleic acid.
• Crediting seed oil removal for improvements that come from eating less packaged and fried food.
• Treating omega-6 as toxic; linoleic acid is an essential fat.

• People with high LDL cholesterol, high ApoB, familial hypercholesterolemia or heart disease should not replace seed oils with saturated animal fats without talking to their clinician.
• People with a current or past eating disorder should be careful with restrictive food rules.

• Higher LDL cholesterol and ApoB if seed oils are replaced with butter, tallow, lard or coconut oil.
• Higher grocery costs and fewer options when eating out.
• Possible drop in vitamin E intake if nuts, seeds and their oils are all removed.