Are Seed Oils Actually Healthy?

An interview with an expert on the evidence for and against.

A FEW DAYS AGO, I found myself at a beachside café in Los Angeles, asking if there might be a table available. During the brief exchange that followed—yes, there was a table—the host somehow managed to slip in a derogatory remark about canola oil.

This is what it has come to.

Is there a more controversial subject in nutrition than seed oils? According to a brigade of influencers, seed oils pose a grave danger to human health. Robert F. Kennedy Jr., the Secretary of Health and Human Services, has said Americans are being “unknowingly poisoned” by seed oils and that fast food restaurants should replace them with beef tallow.

What is wrong with seed oils? A better question is, what isn’t wrong with them?

Seed oils, it is claimed, are rancid, stale, bleached and heat damaged. They contain neurotoxins, carcinogens, heavy metals, trans fats and endocrine disruptors. And they cause diabetes, arthritis, GERD, irritable bowel syndrome, heart disease, asthma, depression, anxiety, impaired learning, autism, autoimmune disease, infertility and Alzheimer’s disease. Seed oils damage mitochondria, disturbing metabolism on the most elemental level. And they cause cancer.

Then there is mainstream science—the researchers with PhDs who work in laboratories and publish peer-reviewed papers in scientific journals. They are largely pro-seed oil. They consider seed oils a “good fat,” with the caveat that all fats are extremely calorie-dense and should not be overconsumed.

But as controversial as seed oils may be, the debate, if you can even call it that, is oddly misaligned. While the wellness gurus pump out their five-alarm warnings on Instagram, YouTube and TikTok, the scientists publish jargon-laden, impossibly dense studies in journals hardly anyone reads. It’s like two ships passing in different oceans. You can find the odd “what’s the deal with seed oils” article here or there, but they tend to be short on detail and gently dismissive.

I wanted to find a scientist who would engage with the anti-seed oil evidence head-on. I wanted a scientist willing to wrestle with the details. Is the anti-seed oil evidence completely wrong? Just a little bit wrong? Is some of it right? If so, why?

I found that scientist at the University of Toronto. His name is Richard Bazinet and he is the Canada Research Chair in Brain Lipid Metabolism. He has studied lipids (fats) since starting his master’s degree in 1998. In that time, he has published more than two hundred papers. He is the editor-in-chief of Prostaglandins, Leukotrienes & Essential Fatty Acids, an academic journal, and his lab contains more than $3 million in equipment and employs about ten scientists and graduate students, who, like Bazinet, study nothing but fats.

In 2016, Bazinet was awarded the Young Investigator Award for Outstanding Research by the Canadian Nutrition Society. Four years later, at the advanced age of forty-five, he received the Ralph Holman Lifetime Achievement Award from the American Oil Chemists’ Society. (He has since won three more achievement awards.) Right now, he’s developing a new technique to measure the rate at which the human body can turn the short-chain omega-3 fat found in plants into the long-chain omega-3 found in salmon and mackerel.

Full disclosure: I didn’t have to look very hard to find Bazinet. He is a friend, and also a colleague of sorts. I first met him in 2009 when I was researching my book Steak and trying to find out if the omega-3 content of grass-fed beef was nutritionally significant. (Bazinet thinks it probably is.) Thus began a conversation about food, science and nutrition that shows no sign of ending.

Bazinet grew up in Cornwall, Ontario, where he played hockey and wasn’t what you would call a finesse player. He has a superb wine collection, and he mixes a killer cocktail. If he delivers a lecture in France, we’ll spend more time talking about the wine and the seafood and the Armagnac than what he spoke about. He is, in short, refreshingly human. I say this because the relentless focus on data and evidence too often makes the scientists who study flesh and blood seem like they don’t have any of their own.

Warning: This is intended to be a full-throated response from a well-credentialled scientist to a popular and urgent subject. This interview is long. It is technical. It is not intended to frighten you or fill you with righteous outrage. There are no evil conspiracies revealed and no life-altering biohacks.

You don’t have to agree with Bazinet. But you do owe it to yourself to listen to what a well-regarded expert has to say and, if need be, to challenge your own thinking. And if you think Bazinet is ill-informed, confused or just plain wrong, leave a comment. Maybe he will respond. Let’s revive the lost art of respectful discussion.

---

Mark Schatzker: Let's start with some basics. What are seed oils?

Richard Bazinet: The term mainly refers to oil derived from the seeds of plants. A lot people exclude olive oil from the category because they consider olives a fruit. But in general, it refers to soybean oil, canola oil, corn oil, sunflower oil and a small handful of others that are in the food supply.

Mark Schatzker: What makes them distinctive or interesting from your point of view, as a scientist.

Richard Bazinet: There are two, three or four families of fats, depending on how you divide them. One we call saturated fats, and those are typically fats that are solid at room temperature. Butter has a lot of saturated fat in it, but it's not exclusively saturated fat.

If you take a fat molecule and create what’s called a double bond by removing two hydrogens, you have a monounsaturated fat. The best known example of that is called oleate, which is in olive oil. Olive oil tends to be liquid at room temperature, but if you put it in the fridge, it should become solid if your fridge is cool enough. Butter is solid at room temperature, but is a rock if you put it in the fridge.

Then we've got another class of unsaturated fats called polyunsaturated fats, which means there is more than more than one double bond. There are two types of polyunsaturated fats. One is called the omega-6s, and the number six refers to the position of that double bond—if you look at a fatty acid as a long chain, and you put a double bond six positions from the end, that becomes an omega-6. If you do the same thing three from the end, you have the other type of polyunsaturated fat, an omega-3.

Omega-6s and omega-3s are really important nutritionally, because they are essential. Your body can't make them from scratch. Doing that requires specific enzymes that plants and a few other organisms have. So you need to consume them.

Finally, when compared to butter, beef tallow and even olive oil, seed oils contain a lot of these polyunsaturated fats.

Mark Schatzker: A lot of seed oil critics are saying we should go back to using tallow instead of seed oils. Can you tell me why tallow fell into disrepute from a nutritional point of view?

Richard Bazinet: Back in the 1950s, a scientist named Ancel Keys was studying different countries and found that fat consumption sometimes correlated with coronary heart disease and sometimes it did not. For example, in Greece consuming a lot of fat didn’t appear to have that effect. The difference is they were consuming a lot of unsaturated fat—in their case, olive oil.

That was followed by research that got more specific, showing that saturated fats could essentially raise LDL cholesterol, which is a risk factor for coronary heart disease. And then there were clinical trials showing seed oils lowered LDL cholesterol whereas saturated fat raised it. One piece of evidence that stands out to me as a researcher is that when saturated fat is replaced by seed oils, the benefits are even more robust and consistent.

Is it All Ancel Keys’ Fault?

Mark Schatzker: From the lofty perch of the present, it looks like Ancel Keys got it all wrong. Today, more than one in ten people have type 2 diabetes, we have greater rates of depression, Alzheimer's, ADHD and autism. And then, of course, there's obesity. In a lot of ways, our health has never been worse. And chronologically, it looks like this is all seed oils’ fault.

Richard Bazinet: There are a couple of things to unravel here. One thing is, we have to be careful with diseases like Alzheimer's disease. Sure, there's more Alzheimer's disease today than there was in the 1950s. But the reason for that is people are living longer. So when you look at these things, you have to have age-adjusted rates. If you die young, you don’t get Alzheimer’s.

You also have to be careful with research that says, hey, the obesity epidemic started getting worse in the 1970s just as we began to consume more seed oils so it must be because of seed oils. In theory, seed oils could be contributing because they've got a lot of calories. But if you look at who’s developing obesity and who’s not and then look at what they are eating, seed oils don't jump out as being the cause.

And with heart disease, we see the opposite. If you look at European countries, as they switch from saturated fats to seed oils we see that their rates of heart disease dropped over time.

Now we have even better studies where we follow hundreds of thousands of individuals, some of them for up to thirty years. We've watched what they eat, and we see that those consuming the highest amounts of seed oils, compared to those consuming the more saturated fats, actually have lower levels of cardiovascular disease, type 2 diabetes and cancer, and in some cases all-cause mortality. So people who consume seed oils are just less likely to get sick or die.

That said, it’s important to recognize that these types of studies don’t measure cause. So you have to be open to the possibility that seed oil consumption might just be a marker for a better overall diet. But if that’s the case, it’s hard do see how seed oils could be dangerously unhealthy given that consuming them consistently tracks with better health.

If Homo Erectus Didn’t Eat It, Should I?

Mark Schatzker: Another major criticism of seed oils is that they're alien to us from an evolutionary point of view. Hominids have been eating ruminants for more than a million years, but it’s hard to imagine a hunter-gatherer chomping enough sunflower or rapeseeds to consume a portion anywhere close to what people regularly consume today.

Richard Bazinet: We've got a really neat thing in human physiology called digestion and absorption. We do with these seed oils what we do with any fat we consume—we break them down, and then we kind of shuffle them past the intestine, and then we put them back together according to our needs. They get metabolized. And then they really look a lot like, you know, other sources of fat inside the body.

That’s one of reasons this idea that we need to eat an evolutionarily appropriate diet is a fallacy. If you walk through a grocery store, most of the foods are relatively new from that perspective.

Mark Schatzker: There were no tomatoes or carrots or bowls of chilli in the Rift Valley.

Richard Bazinet: Yes. And, like I said, we've been following the epidemiology of seed oils very closely for over thirty years, and we show that people consuming the higher amounts of seed oils have less death. When it comes to safety, less death is as good as you’re going to get.

Mark Schatzker: If you look at the body of a person who consumes a lot of seed oils, does it look physiologically or biologically vastly different from someone who doesn't?

Richard Bazinet: Not dramatically. What shows up is the omega-6 fat found in seed oils, called linoleic acid, and the omega-3 fat found in seed oils, called alpha-linolenic acid. People who consume more seed oils have higher amounts of these two fats in their blood or fat stores. And what's neat is that these correlate with less disease and total mortality.

Omega-6s and Inflammation

Mark Schatzker: But do you see both omega-6s and omega-3s? Because another worry about seed oils is that they overwhelm us with omega-6 fats but don’t have enough omega-3s. Omega-6 fats are converted by enzymes into prostaglandins which promote inflammation. Chronic inflammation is connected to a veritable who’s who of terrible diseases: atherosclerosis, heart disease, hypertension, Alzheimer's, multiple sclerosis, cancer. So, if omega-6s are a necessary piece of the inflammation puzzle, doesn't it seem reasonable that our diet could be a cause?

Richard Bazinet: Yes, and our diet possibly is a cause to some extent. We have some evidence that diet can promote or reduce markers of inflammation. But the way this works is far from clear.

With the omega-6s, you've got linoleic acid as your starting point. Through a series of steps that largely happens in your liver, it can get converted to something called arachidonic acid. And then there is a series of enzymes that can convert arachidonic acid to these pro-inflammatory prostaglandins that you just mentioned.

But everybody forgets that one of the most important anti-inflammatory molecules is called Lipoxin A 4, and it also comes from arachidonic acid. So you have to be careful just connecting these biochemical pathways because there are offshoots that are both pro-inflammation and anti-inflammation.

Mark Schatzker: To simplify ridiculously, just because you need sand to make concrete doesn't mean more sand equals more concrete.

Richard Bazinet: Exactly right. We have a lot of clinical trials in humans now, where we've tried to increase or decrease linoleic acid. And it often doesn't even change arachidonic acid, let alone these pro-inflammatory prostaglandins. And if we look at markers of inflammation, they don’t change much, if at all, with diets high or low in linoleic acid.

Now, I'm not saying you can't find a single study that doesn't show that one of these changed. But on average they're not changed. That's why I like these studies that have looked at total mortality, because if seed oils were viciously causing inflammation, we would expect people who are eating more seed oils to be more likely to die. And we see the opposite.

Mark Schatzker: But could we lower our omega-6 intake to a point where we could reduce inflammation because there wouldn't be enough to make those prostaglandins?

Richard Bazinet: We have been working on this in animals and if we go pretty extreme we can change arachidonic acid levels, but the effects on prostaglandins and inflammation are very subtle.

There's a little bit of evidence that if people who get migraines lower their omega-6s and raise their intake of fish oil, a long-chain omega-3, it can improve some of their scores on migraines. That said, in normal people omega-6s clearly are not killing them or causing disease.

The Oxidation Issue

Mark Schatzker: As you mentioned earlier, seed oils are rich in polyunsaturated fat, which have more of these double bonds. This makes them less stable and chemically prone to oxidation, which is also called rancidity. Critics claim that the seed oils are rancid because they've been heated to extremely high temperatures during manufacturing, and that when we consume these rancid oils they get in our body, and, for example, get converted to oxidized LDL cholesterol.

Richard Bazinet: All lipids with a double bond can become oxidized to some extent. In theory, the more double bonds you have, the more oxidized you can get. So it's kind of funny that the omega-3s from fish, which a lot of people think are good for you, are actually the most susceptible to oxidation.

If you breathe on your hand right now, you're going to increase the oxidized lipids on your skin. We can measure oxidized lipids to a ridiculous degree. If you divide a gram by a thousand, you’ve got a microgram. If you divide a microgram by a thousand, you've got a nanogram. If you divide a nanogram by a thousand, you have a picogram. If you divide a picogram by a thousand, you have a femtogram. Can we measure oxidized lipids in seed oils? Yes, we can. But if you look at the guidelines for oxidation levels in all these oils, they are well below all the established safe guidelines. One of the ways we know that is if you open up a bottle of seed oils it doesn’t stink. If something really goes off and is rancid, it's kind of disgusting. Butter can get oxidized, even though it's predominantly a saturated fat. I would say as a chemist there's not a single food with fat in it that's not oxidized to some degree.

Mark Schatzker: What about seed oils that have been used over and over for frying?

Richard Bazinet: This is a legitimate concern. Heating seed oils repeatedly will cause them to oxidize. That’s why restaurants should change their oil regularly.

That being said, the research isn’t clear to what degree oxidized lipids accumulate in the body. There are some small markers we see once in a while, especially in animal studies. But again, if oxidized seed oils were causing disease, we'd see the inflammation scores shoot up and we'd see the disease risk scores shoot up with people who consume a lot of seed oils, and we don't see that. Disease is lower and death is lower.

Fried foods show up more as a risk for obesity but particular diseases don’t jump out.

Toxic Industrial Manufacturing

Mark Schatzker: There are several other alarming aspects of seed oil manufacturing. For example, processors use hexane or other toxic solvents to extract the maximum amount of oil. They use other chemicals to deodorize the oil. They have to filter out waxes. Then they add other chemicals to keep seed oils shelf-stable, most of which are difficult to pronounce. Critics claim these compounds are carcinogenic.

Richard Bazinet: There's a couple of ways to look at this. Like I’ve said already, people who consume these seed oils have less total mortality, less cardiovascular mortality and less cancer. If somebody's telling me that if you eat this food you're going to get cancer, but then I look at people who eat that food and they're getting less cancer, that's not a trivial piece of evidence against the argument.

Nevertheless, there's some validity to these worries. They do use hexane to extract some of the fats. We use it in my lab for the same reason: to extract fats. Hexane is really volatile, though, so it evaporates quite easily. The amounts detected in these oils are well below safety guidelines. And then if you heat it in your frying pan, it won’t even be detectable.

The claim about bleaching is one of my favorites. I see people online talking about literally adding Javex bleach to the seed oils. But it's more like a charcoal filtration system.

Food Cravings and Flooded Mitochondria

Mark Schatzker: There is a claim that seed oils increase the production of endocannabinoids, specifically 2 AG and Anandamide, and that these compounds hijack the satiety receptors in our brain in the same way that cannabis does, leading to increased food cravings.

Richard Bazinet: This is something I have studied. We talked a little bit about how linoleic acid can get converted to arachidonic acid and make these molecules called prostaglandins. There's another series of enzymes that can convert arachidonic acid to things like Anandamide and 2-AG. The biochemical pathway does exist. We did a study in which we increased levels of anandamide in mice, but they did not eat more or gain weight. In another study, we found that these anandamide-like molecules derived from linoleic and alpha-linolenic acid actually decrease food intake in rats.

Mark Schatzker: Some of the online criticism over seed oils can get very technical. One theory I found claims that seed oils cause something called “reductive stress” because the excessive polyunsaturated fats stimulate a compound called PPAR-alpha, which drives too much fuel into the mitochondria.

Richard Bazinet: There are a lot of people who have a rich imagination when it comes to biochemical pathways. It’s true that linoleic and alpha-linolenic acid can bind to PPAR and stimulate it. And if you take PPAR and insert it into a mouse transgenically so that you massively increase PPAR levels, then you will find that these mice develop a diabetes type of condition. So in a theoretical back-of-the-napkin kind of way, you can connect it all back to seed oils.

But I don't think that's nutritionally relevant. These studies generate levels of PPAR stimulation we could never hope to do with nutrition. When we just feed mice diets high in linoleic or alpha-linolenic acid, they don't develop diabetes compared to other fats. In fact, sometimes they do a little better compared to those other fats. And again, not to sound like a broken record, but higher intakes and blood levels of linoleic acid are associated with lower rates of diabetes.

Mark Schatzker: Someone I know sent me an email that was being sent around by some very prominent people in the wellness community. It cites evidence—a peer-reviewed clinical trial published in a scientific journal, no less—that found that linoleic acid, the omega-6 fat, “shows a 700% higher rate of strokes, heart attacks, and cardiovascular deaths when comparing diets high in vegetable oils vs omega-3-rich fish oils.”

Richard Bazinet: You said fish oils, but the omega-3s in that study were actually from seed oils.

Mark Schatzker: I was just quoting the email.

Richard Bazinet: So this study looked at a margarine that was rich in omega-6s and a margarine that was rich in omega-3s. Essentially, it was comparing one seed oil to another, which, from a technical point of view, is not a good way to demonstrate that seed oils are harmful. But let’s look things in more detail since whoever sent that email seems to be interested in omega-6s. The study looked at biomarkers, things like cholesterol and blood pressure. And it showed that the biomarkers of people consuming the omega-3 margarine looked a little better than the biomarkers of the omega-6 group in some cases. But the authors conclude the effect of the omega-3 intake on heart disease risk is the same, not better, as the omega-6 margarine.

Mark Schatzker: What about the 700 percent higher rate of strokes, heart attacks, and cardiovascular deaths?

Richard Bazinet: Good question. I searched for the number 700 in the paper and could not find it anywhere. I worry that the people who circulate these things never actually read the studies.

But there is something else important to say, which is that this study was done in older people, it only lasted a couple of years, it wasn’t very large, and it was designed to look at biomarkers. If you do a study in older people for a few years, a couple people will probably die. That's just how it works. And that’s what happened in this study. One person died of a heart attack in the omega-6 group, but one person also died from a heart attack in the omega-3 group. That’s not a 700-percent difference, that’s a 0 percent difference.

How Much Saturated Fat is Too Much?

Mark Schatzker: If saturated fat is so bad, why does dairy consumption show up as being healthy in a lot of the observational studies?

Richard Bazinet: This is a great question, and the answer is not entirely clear. Firstly, some of the findings are specific to low-fat dairy, which is low in saturated fat. But we also have data showing that people who eat foods like yogurt tend to have healthier lifestyles and diets overall. I also wonder if milk is displacing soda in some cases. Dairy also contains a lot of other important nutrients or what we call the matrix might be important. Clearly more research needs to be done.

Mark Schatzker: Were you ever concerned about seed oils or omega-6 fats?

Richard Bazinet: Yes, and in some ways I’m still concerned. About 15 years ago, Health Canada said vegetable oil makers could use a label that says they lower cholesterol. Then a scientist by the name of Chris Ramsden, who is a colleague of mine, dug up two old studies, two incredible old studies that are amazing for a few reasons, including the fact they weren't published.

Mark Schatzker: Why weren’t they published?

Richard Bazinet: Some people think it is because the scientists conducting them didn’t like the results, but we don’t really know.

These studies were trying to test the heart health hypothesis, which was that that replacing saturated fats with polyunsaturated fats—specifically corn and safflower oil very rich in omega-6s—was going to be heart healthy. And they didn't show that! In fact, if you look at the studies, they kind of show the opposite. There was a little more heart disease and death with seed oil consumption. But it gets tricky because one study also tracked smoking, and it found that it was only the people who smoked and consumed seed oils very rich in omega-6s who had more heart attacks. If you look at the people who weren't smokers, they didn't have more heart attacks.

So eleven years ago a cardiac surgeon and I wrote a letter to the Canadian Medical Association Journal saying we need to think about this a little bit more, we need more data.

A couple of things have happened since then. One is that seed oils have changed. The safflower and corn oil used in those studies, which were extremely rich in omega-6s, are very hard to find in North America today. They have almost all been replaced by what we call mid or high oleic-acid varieties in which the omega-6s have been replaced by the monounsaturated fat found in olive oil.

Also, the research has improved. Instead of asking people what they eat, we are measuring biomarkers in their blood or in their fat. And they all tend to show the same thing, which is that seed oils are associated with a lower mortality, a lower cardiovascular mortality and a lower cancer mortality.

Mark Schatzker: A lot of people are saying people should consume oil that comes from fruits—olive oil or avocado oil. Would you say those are a good move or a bad move?

Richard Bazinet: I think they're generally a good move. We don't have much data on avocado oil, at least not in human populations. But olive oil shows up in the epidemiology as being generally good. It shows up being very comparable to seed oils, and it's part of some randomized controlled trials that show some good outcomes. So I think olive oil is a totally fine choice.

But there is a practical aspect to consider. For certain recipes, some oils work better than others. As a grad student, I got excited about olive oil, and I tried to make pancakes with extra-virgin olive oil. They were disgusting. It started to smoke. It started to oxidize. The pancakes were inedible.

Mark Schatzker: How worried should people be about saturated fat?

Richard Bazinet: Since the 1960s, saturated fat intake in North America has dropped by several percentage points. A lot of people are now in the Goldilocks zone or sweet spot. That said, some people are still consuming high levels, and others have gone lower than they need to go.

Mark Schatzker: Could you argue, then, that consuming seed oils in a way gives us permission to eat butter and steak?

Richard Bazinet: If seed oils or olive oil are your everyday go-to oil, then a certain amount of saturated fats will fit into a diet just fine.

Mark Schatzker: Do you eat butter?

Richard Bazinet: Yes, probably a couple of times a week, especially in recipes that require butter. I like to make pancakes. I like to make waffles. I like to make crepes. And they usually require butter. Every once in a while I treat myself and buy a grass-fed butter. They have some neat flavours. Spreading that on fresh, warm bread is a treat.

Mark Schatzker: If I could grant you a single wish, and you could be teleported somewhere for dinner, where would it be and what would you order?

Richard Bazinet: It's a tough call. If the weather's nice I'd like to be on a dock somewhere in the Mediterranean where someone presents you with a bucket of fish and you point at one and they grill it for you. I'd pair that with a glass of white wine.

If it was a little cooler, it would be a bistro, and it would probably be some sort of a piece of duck with potatoes cooked in duck fat, which is also a saturated fat. This time the wine would be red wine, maybe from Bordeaux, but anything local will do.

***

Mark Schatzker is the Writer in Residence and Associate Director of Communications at the Modern Diet and Physiology Research Centre (MDPRC), McGill University.

Follow me on X and Instagram: @MarkSchatzker

---