Seed Oils and Inflammation: Separating Science from Hype

The internet has decided that seed oils are the great modern poison. Vegetable oils, the argument goes, were unknown before the 20th century, are extracted with chemical solvents, contain high levels of omega-6 linoleic acid that drives systemic inflammation, and oxidize at cooking temperatures to produce toxic compounds. Like many nutrition stories, this one contains fragments of truth wrapped in a confounding conclusion. The fragment of truth is that the American diet shifted dramatically toward soybean, corn, canola, sunflower, and safflower oils in the second half of the 20th century, and that some of these oils are now consumed in quantities that bear little resemblance to traditional diets. The confounding conclusion is that the oil itself, rather than the food matrix it appears in, is the problem. The most cited serious research on this question is Ramsden and colleagues' 2010 analysis in BMJ of the recovered Minnesota Coronary Experiment data, which suggested that replacing saturated with polyunsaturated fat lowered cholesterol but did not reduce (and possibly increased) mortality. The American Heart Association has issued a formal position statement defending the substitution of polyunsaturated for saturated fat. The NIH maintains the most balanced reference on essential fatty acid requirements. The picture is more nuanced than the online debate suggests.

What seed oils actually are

"Seed oil" is a lay term for vegetable oils extracted from seeds. The dominant seed oils in the American diet are soybean (about 60 percent of vegetable oil intake), canola, corn, sunflower, safflower, cottonseed, and grapeseed. Olive, avocado, coconut, and palm oils are technically fruit oils, not seed oils, and are not usually what the wellness internet is criticizing. Sesame and peanut oils are seed oils but are consumed in smaller quantities and have different fatty acid profiles.

The fatty acid composition matters. The omega-6 linoleic acid content of common oils:

• Safflower oil: about 75 percent linoleic acid
• Sunflower oil: about 65 percent
• Corn oil: about 55 percent
• Soybean oil: about 50 percent
• Cottonseed oil: about 50 percent
• Canola oil: about 20 percent
• Olive oil: about 10 percent
• Avocado oil: about 12 percent
• Coconut oil: about 2 percent (mostly saturated)

The NIH Office of Dietary Supplements notes that linoleic acid is an essential fatty acid: the body cannot synthesize it, requires a small daily intake (about 12 to 17 g per day for adults), and uses it as a precursor to arachidonic acid and a family of signaling molecules called eicosanoids. The question is whether the modern intake, which averages 6 to 8 percent of total calories in the United States (about 14 to 18 g per day for a 2,000-calorie diet), is too high.

The omega-6 to omega-3 ratio hypothesis

The theoretical case against seed oils rests on the omega-6 to omega-3 ratio. Linoleic acid (omega-6) and alpha-linolenic acid (omega-3) compete for the same desaturase enzymes that convert them into longer-chain signaling molecules. Arachidonic acid (from linoleic acid) gives rise to series-2 prostaglandins and series-4 leukotrienes, many of which are pro-inflammatory. EPA and DHA (from alpha-linolenic acid) give rise to series-3 prostaglandins and resolvins, which are anti-inflammatory. The hypothesis is that a high omega-6 to omega-3 ratio (estimated at 15:1 to 20:1 in the modern Western diet, versus 1:1 to 4:1 in traditional diets) shifts eicosanoid balance toward inflammation.

The hypothesis was popularized by Artemis Simopoulos in a series of papers in the 1990s and 2000s. It is biochemically plausible but the human evidence is weaker than the theory. Controlled trials that increase linoleic acid intake do not reliably raise inflammatory markers. A 2017 systematic review in BMJ Open pooled 30 controlled trials and found that higher linoleic acid intake did not increase inflammatory markers (CRP, IL-6, TNF-alpha) compared to lower intake. A 2012 meta-analysis in British Journal of Nutrition reached a similar conclusion. The reason appears to be that the body tightly regulates arachidonic acid levels in cell membranes, and excess dietary linoleic acid is mostly stored in adipose tissue or oxidized for energy rather than converted to arachidonic acid.

The Ramsden 2010 Minnesota Coronary Experiment reanalysis

The most cited challenge to the polyunsaturated-fat-heart-health story is Ramsden and colleagues' 2010 reanalysis of the Minnesota Coronary Experiment, published in BMJ. The original trial, conducted from 1968 to 1973, randomized 9,423 institutionalized adults to a diet that replaced saturated fat with corn oil (high in linoleic acid) or to a control diet. The corn oil group lowered their serum cholesterol more than the control group, as expected. The surprise came decades later when Ramsden recovered the full data and found that the cholesterol reduction did not translate into fewer cardiovascular deaths. In fact, participants aged 65 and older in the corn oil group had a higher all-cause mortality, and the degree of cholesterol reduction was associated with a higher risk of death.

The Ramsden reanalysis is important because it suggests that the cholesterol-lowering effect of polyunsaturated fat does not necessarily translate into mortality benefit, at least in the form of corn oil fed in a controlled setting. The authors speculated that the oxidized cholesterol or oxidized linoleic acid metabolites may have counteracted the benefit. The study has limitations: it was a post-hoc analysis of recovered data, the population was institutionalized (mentally ill and nursing home residents), the trial duration was relatively short (about 1 to 4 years), and the diet was extreme (corn oil as the only fat). It is not direct evidence that seed oils are deadly in normal diets, but it does complicate the simple story that replacing saturated with polyunsaturated fat is always beneficial.

The Sydney Diet Heart Study reanalysis

Ramsden and colleagues published a second reanalysis in 2013 in BMJ, this time of the Sydney Diet Heart Study, which ran from 1966 to 1973 in Australia. The trial randomized 458 men with recent coronary events to a diet high in safflower oil (high linoleic acid, low saturated fat) or to a control diet. The safflower oil group lowered cholesterol more than control, but had a higher all-cause mortality (about 17 percent versus 11 percent) and higher cardiovascular mortality. The findings were consistent with the Minnesota reanalysis: lowering cholesterol by replacing saturated with linoleic acid did not reduce mortality in this population, and may have increased it.

Both reanalyses are now part of the broader literature, and the picture they paint is more nuanced than the original simple story. It is not that polyunsaturated fats are harmful in general, but that high-dose linoleic acid, in the absence of adequate omega-3 intake, in older or sicker populations, may not have the cardiovascular benefit that the cholesterol reduction suggests.

The American Heart Association position

The American Heart Association issued an advisory in 2017 in Circulation specifically defending polyunsaturated fat. The AHA reviewed the evidence from randomized trials and prospective cohorts and concluded that replacing saturated fat with polyunsaturated fat reduces cardiovascular events by about 30 percent, comparable to the effect of statins. The AHA specifically addressed the Ramsden reanalyses and argued that they were outliers in a larger literature, and that the meta-analytic evidence from controlled trials (including the Oslo Diet-Heart Study, the Finnish Mental Hospital Study, and the LA Veterans Study) supports a cardiovascular benefit of polyunsaturated fat.

The AHA position has been criticized by some researchers (notably Ramsden, Simopoulos, and others) for over-relying on old trials with methodological limitations and for under-weighting the recovered data. The honest summary is that the evidence for replacing saturated fat with polyunsaturated fat is moderate but not airtight, the effect is smaller and more context-dependent than the AHA suggests, and the form of the polyunsaturated fat matters (whole foods versus oils, fresh versus oxidized, balanced omega-3 versus not).

Oxidation and cooking

A more defensible concern about seed oils is oxidation. Polyunsaturated fats have multiple double bonds that are vulnerable to oxidation, especially at high temperatures. Heating seed oils past their smoke point produces aldehydes, lipid peroxides, and other potentially toxic compounds. A 2019 study in Preventive Medicine Reports measured aldehyde emissions from common cooking oils and found that sunflower, corn, and safflower oils produced substantially more aldehydes at frying temperatures than olive oil, avocado oil, or coconut oil. The smoke points are: safflower (510 F), sunflower (440 F), soybean (450 F), corn (450 F), canola (400 F), olive (410 F), extra-virgin olive oil (374 F), avocado (520 F).

The practical implication is that seed oils are fine for low-temperature cooking (sauteing, baking, dressing) but should be avoided for high-temperature frying, particularly deep frying. Olive oil and avocado oil are better high-temperature choices because of their higher monounsaturated content and better oxidative stability. Reusing frying oil, regardless of type, is the worst-case scenario for oxidation.

What the food matrix actually shows

The strongest evidence against seed oils does not come from the oils themselves but from the foods they appear in. The vast majority of seed oil intake in the United States is in the form of ultra-processed foods: chips, fries, commercial baked goods, fried fast food, mayonnaise-based dressings, and packaged snacks. The 2019 controlled trial by Hall and colleagues in Cell Metabolism showed that people ate about 500 more calories per day on an ultra-processed diet than on a whole-food diet matched for macronutrients, and gained weight rapidly. The seed oil was not the dominant variable, but it was a marker for the ultra-processed matrix.

This is the core confounding problem in the seed oil debate. Observational studies linking higher seed oil intake to worse health outcomes are heavily confounded by the fact that most seed oil intake comes from ultra-processed food. People who eat more whole-food sources of polyunsaturated fat (nuts, seeds, fish) have better health outcomes, and people who eat more ultra-processed sources of polyunsaturated fat (chips, fries, commercial baked goods) have worse health outcomes. The oil itself is difficult to isolate from the matrix.

What to actually do

A reasonable, evidence-based approach to seed oils:

• Use olive oil as the default cooking and dressing fat. The PREDIMED trial showed that 4 or more tablespoons per day of extra-virgin olive oil reduced cardiovascular events by 30 percent. Olive oil is mostly monounsaturated, low in linoleic acid, and high in polyphenols.
• Use avocado oil for high-temperature cooking. It has the highest smoke point (520 F) and an excellent fatty acid profile.
• Avoid deep frying regardless of the oil, and never reuse frying oil.
• Limit ultra-processed foods regardless of the oil they contain. This is the dominant variable.
• Do not worry about modest seed oil intake in salad dressings, mayonnaise, or low-temperature cooking. The evidence does not support the claim that 1 to 2 tablespoons of canola or soybean oil per day is harmful.
• Ensure adequate omega-3 intake to balance the omega-6 load. Aim for 2 to 3 servings of oily fish per week, or 1 to 2 g of EPA plus DHA per day from supplements. Use the Omega-3 Calculator to check your intake.
• Replace omega-6 with monounsaturated where possible. Olive, avocado, and macadamia nut oils are better choices than corn, sunflower, or safflower oils.

Score your overall diet pattern with the Mediterranean Diet Score Calculator, which captures most of the relevant variables. The Mediterranean pattern, which uses olive oil as the default fat and limits ultra-processed foods, has the strongest cardiovascular evidence of any diet pattern.

Common pitfalls

Pitfall 1: Treating all seed oils as equivalent

They are not. Canola oil is about 20 percent linoleic acid, while safflower oil is 75 percent. Canola also contains omega-3 alpha-linolenic acid (about 10 percent), making it less omega-6-heavy than the others. The lumping of all seed oils into one category obscures real differences.

Pitfall 2: Conflating the oil with the food matrix

Most seed oil intake in the United States comes from ultra-processed foods. The seed oil is a marker for ultra-processing, not necessarily the causal agent. Studies that substitute whole-food sources of polyunsaturated fat (nuts, seeds, fish) for saturated fat consistently show benefit. Studies that substitute ultra-processed seed oil products do not. The matrix matters as much as the oil.

Pitfall 3: Throwing out all polyunsaturated fat

Polyunsaturated fats are essential. The body cannot synthesize linoleic acid or alpha-linolenic acid and requires both. The goal is balance, not elimination. Aim for a ratio of omega-6 to omega-3 closer to 4:1 or 5:1 rather than 15:1, by increasing omega-3 intake and reducing ultra-processed seed oil intake.

Pitfall 4: Using seed oils for high-temperature frying

This is the one use case where the oxidation concern is well-founded. Sunflower, corn, and safflower oils produce substantially more aldehydes at frying temperatures than olive or avocado oil. Use olive or avocado oil for frying, and never reuse frying oil.

Pitfall 5: Believing the most extreme claims

The internet is full of claims that seed oils are uniquely toxic, that they cause every chronic disease, and that eliminating them will transform your health. The evidence does not support these claims. The evidence supports a more measured position: use olive oil as the default, limit ultra-processed foods, ensure adequate omega-3, and avoid high-temperature frying with high-linoleic oils.

The bottom line

Seed oils are not the great modern poison, but the modern pattern of consuming them in ultra-processed foods at high doses, in the absence of adequate omega-3 intake, is genuinely problematic. The Ramsden 2010 reanalysis of the Minnesota Coronary Experiment and the 2013 Sydney Diet Heart Study reanalysis suggest that high-dose linoleic acid, in older or sicker populations, may not have the cardiovascular benefit that the cholesterol reduction suggests. The American Heart Association maintains that replacing saturated with polyunsaturated fat reduces cardiovascular events, but the evidence is more nuanced than their position allows. The omega-6 to omega-3 ratio hypothesis is biochemically plausible but not strongly supported by controlled trials, which generally show that increasing linoleic acid does not raise inflammatory markers. The more defensible concerns are oxidation at high cooking temperatures and the confounding of seed oil intake with ultra-processed food intake. Use olive oil as the default, avocado oil for high-temperature cooking, limit ultra-processed foods, ensure adequate omega-3 intake, and do not worry about modest seed oil intake in low-temperature cooking. Score your pattern with the Mediterranean Diet Score Calculator and Omega-3 Calculator.