Are seed oils actually bad for you?

Short answer

No. Across 15 randomized trials, higher linoleic acid intake raised not a single inflammatory marker, and both observational and Mendelian randomization data link it to lower diabetes risk, not higher. The one real concern is repeatedly reheated frying oil, which can form trans fats, and that happens in commercial fryers, not your sauté pan. If you swap to beef tallow or butter, your LDL goes up, which is the wrong direction.

Seed oils have become the one food villain that unites the internet. Canola, soybean, corn, cottonseed, sunflower, safflower, grapeseed, rice bran, the so-called hateful eight, are blamed for inflammation, obesity, diabetes, and a general sense that modern food is out to get us. The claims are specific enough to sound scientific and vague enough to never be tested.

So let's test them. But first, the framing that makes the whole question answerable: bad compared to what? You have to cook with fat. If you're not cooking with some kind of fat, you're not cooking. Every question about seed oils is really a question about substitution.

What a seed oil even is

The category is a mess. Corn is a grain. Rice bran is a husk. Avocado oil comes from the seed of the avocado and almost nobody puts it on the list. The list itself comes from Cate Shanahan, whose professional identity is built on opposing seed oils.

What the oils on the list actually share is not seediness. It's price. These are the cheap, widely available oils you buy at the supermarket and that restaurants use to fry things, which means they're also the oils in a lot of packaged and processed food. That matters enormously when you try to study their health effects, because price signaling confounds almost every health relationship in America. Rich people get better health care and better everything else.

The other thing they share is that they're high in omega-6 polyunsaturated fatty acids, mostly linoleic acid. Linoleic acid is one of two essential fatty acids. In nutrition, "essential" doesn't mean wonderful and part of a balanced diet. It means you need it to survive and your body cannot make it. If you don't get linoleic acid, you die. Deficiency is rare because the amount you need is small and almost everyone gets it without trying, but the floor is not zero.

A quick word on the chemistry, because it explains a lot of what follows. A saturated fat has only single bonds in its carbon chain, so the molecules lie straight and pack together tightly. That's why butter and beef tallow are solid at room temperature. Add one double bond and you get a monounsaturated fat like oleic acid, the main component of olive oil. The kink keeps the molecules from packing as neatly, so it's liquid, though it will congeal in your fridge. Add several double bonds and you get a polyunsaturated fat, more kinks, stays liquid.

The trans fat hangover

Much of the current panic is inherited. In 1911 the good people at Procter and Gamble figured out that you could take liquid cottonseed oil and hydrogenate it, converting double bonds back to single bonds, to make something solid and shelf stable. That's Crisco. That's margarine. Nobody wants to spread a liquid on toast.

The problem is that the old hydrogenation process created trans fats, and trans fats causally increase cardiovascular disease. They were banned. That was a real case of processing creating real harm.

But the argument has drifted. Industry took seed oils, chemically mangled them into something dangerous, and that thing is now illegal. From there people concluded that seed oils themselves are the problem, that margarine's sins are butter's vindication, and now the food pyramid gets flipped with beef tallow on top. Trans fats being bad is a different claim from canola oil being bad.

The processing scare

The first modern criticism is about how the oils are made, and there's a whole genre of Instagram video devoted to it. Roughly one thing in those videos is true.

Hexane is used as a solvent. Oils don't dissolve in water, so extracting them requires a nonpolar solvent, and hexane is the standard choice. It's then distilled off. Residual hexane is limited to less than one part per million. Yes, hexane is a neurotoxin if you're sniffing glue or you fall into a vat of it at the hexane factory. The dose makes the poison. You're getting considerably more hexane filling your gas tank than from your canola oil.

The rest of the standard script is worse. Sodium hydroxide is not a powerful carcinogen. It's a base. We add it to municipal water when the pH runs too acidic. The claim that processing makes the oil rancid or denatured is also not true; it's extracted with some heat, not the sustained extreme temperatures that actually degrade oil.

This is the influencer template. Take one true fact, extrapolate to catastrophe. Or find a tiny quantity of something and describe what would happen if it were your entire diet, or a mouse's entire diet. Those are not the same. The natural world is full of toxins and poisons at low doses. That's what your kidneys and liver are for. Processing isn't inherently good or bad. Processing can absolutely create harm, like calorie-dense food that's engineered to be easy to overeat. Processing itself is not a poison.

The inflammation claim, which has actually been tested

Here's the mechanistic argument you'll hear. Linoleic acid is a precursor to arachidonic acid, and downstream of arachidonic acid sit a bunch of inflammatory molecules. Interfering with that pathway is part of how NSAIDs reduce inflammation. Therefore linoleic acid is inflammatory.

Plausible hypothesis. Testable hypothesis. Tested hypothesis.

A systematic review in the Journal of the Academy of Nutrition and Dietetics in 2012 pulled together 15 randomized trials assigning people to higher linoleic acid intake versus other fats. Not one showed an increase in C-reactive protein. There was also no effect on IL-6, TNF-alpha, ICAM-1, L-selectin, P-selectin, fibrinogen, platelet activity, or any measured prostaglandin or thromboxane. If you have a favorite inflammatory marker, it's probably on that list.

This is how science is supposed to go. Mechanism suggests a possibility, you test it in humans, you learn something. What you don't do is get the human answer and then retreat to the mechanism.

The number one cause of obesity and diabetes, allegedly

One influencer clip claims excess linoleic acid is the number one cause of diabetes and obesity in this country. Not a contributor. Number one.

The steelman version has something behind it. There's an endocannabinoid effect, the same receptor system that gives you the munchies, and a mouse study showing that more linoleic acid exposure increases food intake. So you can get a mouse to eat a bit more.

The human data goes the other way. Observational data shows higher linoleic acid intake is associated with lower diabetes risk and lower body weight, which makes sense partly because you're eating a plant-based oil, which usually means you're eating more plants. A 2023 Mendelian randomization study found that genetically predisposed higher linoleic acid intake causally lowers diabetes risk, fasting glucose, and hemoglobin A1c.

To believe the influencer claim even directionally, you'd have to think the confounding runs hard in the opposite direction, that heavy linoleic acid consumers are so positively selected on everything else that they're masking a large harmful effect. That's a lot to carry. Also, marijuana is probably doing more to America's waistline than linoleic acid is.

The omega-6 to omega-3 ratio

The ratio argument casts omega-6 as the angry inflammatory one and omega-3 as its calming counterpart, with modern diets badly out of whack. The ancestral 1:1 comparison rests on dietary reconstruction rather than measurement, so treat it loosely, but today's higher ratio is real.

The study cited every single time is an eLife paper from 2024 using UK Biobank, about 85,000 people. It's a good design: they measured plasma levels of omega-6 and omega-3 rather than relying on dietary recall, and since these are essential fatty acids you can't synthesize, blood levels genuinely reflect intake. The headline finding is that the highest quintile of the omega-6 to omega-3 ratio had 26% higher all-cause mortality and 31% higher cardiovascular mortality.

Then you read past the abstract. Both fatty acids were inversely associated with all-cause, cancer, and cardiovascular mortality. People with the most omega-6 in their blood did better than people with the least. People with the most omega-3 did better than people with the least. Omega-3 was just more strongly protective. The scary relationship only appears when you divide one number by the other.

The most likely reading is that all of these fat measures are proxies for overall diet quality, and for income and education and exercise along with it. And if you do believe the ratio is causal, the implication isn't to cut omega-6. It's to eat more omega-3. Eat the salmon. Though note who tends to be eating salmon.

The one real concern: frying

Those double bonds that keep polyunsaturated oils liquid can break under high heat, and we're not talking about sautéing. We're talking 175 to 190 degrees Celsius, deep frying territory. When the bonds reform, they may reform in the cis position, or they may reform in the trans position, and then you have a trans fat. Trans fats causally increase cardiovascular risk. So you can create trans fats without being Procter and Gamble; you just leave the oil on the heat for hours and hours.

Rats fed sequentially reheated oil showed a dose-response effect on vascular inflammation: heat the oil once, feed rats, heat it twice, feed rats, and the damage tracks the degradation. And people who eat more fried food have higher cardiovascular risk, though separating the fryer from the fried chicken is not easy.

Where this matters is the commercial fryer. Restaurants hold oil at temperature for many hours or days, topping it up as it degrades. That treatment can generate trans fats. Home cooking is a different situation. Even if you heat oil to 375 and make excellent double-fried french fries, that's not enough time for this to become meaningful, and sautéing something in canola oil never gets close.

The more unsaturated the fat, the bigger this issue, which is why monounsaturated oleic acid is more robust. Olive oil is mostly oleic acid but smokes too readily for deep frying. High-oleic sunflower and safflower oils have high smoke points and hold up well. They are, technically, seed oils. You could also fry in beef tallow, which being saturated won't form trans fats with extended heating.

If you swap out seed oils, what do you swap in?

Four candidates.

Beef tallow is having a moment. Whole Foods named tallow foods its top food trend of 2026. It's about 50% saturated fat, 42% monounsaturated, and only 4% PUFA. Its most abundant fatty acid is oleic acid, same as olive oil. But 26% is palmitic acid, which causally raises LDL. Substitution studies show that swapping tallow in for seed oils raises LDL. There are no outcome trials, so I can tell you your cholesterol will be higher and I cannot tell you how many more heart attacks that buys.

Butter is even higher in saturated fat than tallow, and raises cholesterol for the same reason. It is also delicious, and all foods can be sometimes food.

Olive oil is predominantly oleic acid, which is omega-9, not omega-3. This is the one with a real randomized trial behind it. PREDIMED randomized roughly 7,500 Spanish adults to a low-fat diet, a Mediterranean diet with olive oil, or a Mediterranean diet with mixed nuts. Both Mediterranean arms had substantially lower rates of major cardiac events. Olive oil is a component of the best-evidenced diet we have for people at high cardiovascular risk. Just don't deep fry with it.

Avocado oil looks ideal on paper: 70% oleic acid, high smoke point, neutral flavor. The problem is the shelf. A UC Davis study found that 93% of products labeled as containing avocado oil contained other oils, sometimes unreported. Even bottles sold as pure avocado oil were widely adulterated with canola, sunflower, and safflower. The same group found very little adulteration in olive oil, where the industry has more monitoring and standardization. Avocados cost much more than olives, so the incentive to cheat is much larger. You may be paying a premium for secret seed oil.

The substitution trials themselves

A few trials have actually replaced animal fat with seed oils and followed outcomes. The famous one is a 1969 trial in about 850 men where roughly two thirds of animal fat was swapped for corn, soybean, safflower, and cottonseed oils. Cholesterol fell substantially and stayed down, and the group had lower rates of myocardial infarction. That's a point for seed oils.

Others point the other way. The Sydney Diet Heart Study, published in the BMJ in 2013, randomized 458 men to a diet high in safflower oil and safflower oil margarine replacing animal fats, and found higher overall mortality in the intervention group, barely statistically significant. It's small, it's men only, it's Sydney, and the control group received no specific dietary instruction and no study foods, which is a fairly uncontrolled control.

When trials point in both directions with effects this size, the honest read is usually that the thing doesn't matter much either way. Nobody is going to save their life by adding margarine or avocado oil to their diet. Meaningful changes in outcomes come from larger lifestyle shifts, not from switching a single fat.

Bottom line

Seed oils are fine. The processing scares fall apart on contact with the dose, the inflammation hypothesis has been tested in 15 randomized trials and failed, and the human evidence on linoleic acid and diabetes runs protective rather than harmful. The omega-6 to omega-3 ratio finding is an artifact of division; both fatty acids look good on their own. The genuine concern, repeatedly reheated frying oil generating trans fats, belongs to commercial fryers, not your kitchen.

Plant-based fats are probably better for you than animal-based fats, and seed oils are members of that group. If you switch to tallow or butter, your LDL goes up. If you switch to avocado oil, there's a good chance you're buying seed oil anyway. The most useful advice I heard on this came secondhand from a listener's friend who keeps several fats around and matches the oil to the job. Sometimes that's a seed oil, sometimes butter, sometimes, apparently, duck fat.

It's never one thing.

Here's our discussion from the episode:

I covered this in depth on Wellness, Actually. Listen below.

Wellness, Actually Podcast

"What's the deal with seed oils?" — Listen to the full episode, including the week's health news and listener Q&A.

Related episodes

• Is red meat actually bad for you? What the evidence really shows
• What can your poop actually tell you about your health?
• What ticks actually do to your body, from Lyme to alpha-gal

Topics

Metabolic & diet  ·  Environmental exposures & health scares