Well + Being
Are seed oils bad for you? Here's what the evidence actually tells us
By Tamar Haspel The Washington Post
Published May 16, 2025
Contact The Editor
Articles By This Author
Are seed oils bad for you, as U.S. Health Secretary Robert F. Kennedy Jr. and an army of "Make America Healthy Again" influencers claim?
The answer (tl;dr!) is that they're more likely to have some benefits than to be harmful, but it probably doesn't matter very much.
If that's the case (and I'm going to persuade you that it is), it raises another question: Why do people care so much? We'll talk about that, too.
Let's first define what we're talking about. A seed oil - duh! - is an oil that's expressed from a seed, and what opponents have dubbed the "hateful eight" are canola, corn, cottonseed, grapeseed, soy, rice bran, sunflower and safflower. But the reason for the hate varies. Sometimes it's the processing to extract and stabilize the oil, and sometimes it's simply that they're high in polyunsaturated fats (PUFAs). Or the problem could be a subset of PUFAs, omega-6 fats. Or it could be one particular kind of omega-6, linoleic acid. Different critics focus on different aspects.
In some ways, the fight over seed oils is a proxy war. Mostly, the people casting doubt are the same people who believe saturated fat isn't a health risk (and if you come here often, you know that I think it is, but a relatively small one). Case in point: Kennedy isn't pushing just for the elimination of seed oils; he's advocating rendered beef fat as a replacement.
But the health benefits of seed oils and saturated fat are separate questions, so let's tackle the seed oils on their own merits - or lack thereof.
No matter what kind of nutrition question you're trying to answer, two rules apply:
Rule No. 1: Beware of theories.
What we know about the connection between diet and health is dwarfed by what we don't know. Our bodies are complex, and Process A, over here, is affected by Hormone B, over there, and, down the road, Bacterium C. And all bets are off if Gene D comes into play. Our understanding of all those things is severely limited, and that's before we even get to the unknown unknowns.
The upshot of Rule No. 1 is that you should brace yourself for uncertainty. It also means that anyone trying to convince you something is good or bad based on, say, Process A is probably wrong. You can't really take a microscope to some tiny part of human metabolism and successfully predict health outcomes based on that.
Theories on seed oils abound. Here are a few you may have heard:
• Seed oil consumption throws off the ratio of omega-6 fats (predominant in those oils) to omega-3 fats (found in fish, nuts and other foods).
• Linoleic acid, the omega-6 fat that people tend to focus on, promotes oxidation of low-density lipoprotein (LDL) cholesterol.
• Omega-6 fats, or linoleic acid specifically, promote inflammation.
• The chemicals used in the refining of seed oils (hexane, for example) are toxic.
• Linoleic acid promotes tumor growth.
There are others. And many of them are interesting! But if a theory is correct, and matters enough to affect health outcomes, we would expect to see the results play out in actual humans. Which brings us to the second rule.
Rule No. 2: Look at all the evidence.
There's a reason doing your own research is the stuff of memes. But there are two ways to do research: to prove that something is true, or to discover whether it is. There's so much nutrition research, and most of it is varied and equivocal, so you can "prove" almost anything. I, for example, am hanging my hat on the data showing that ice cream is associated with decreased heart disease risk.
If you set out to "discover whether," you have a better chance at arriving at something true, although the human drive to confirm our beliefs can send that effort off the rails, as well. But let's see if we can't parse the body of evidence on seed oils to figure out what the evidence tells us.
I'm not going to chapter-and-verse you on evidence quality yet again. A lot of the evidence out there is terrible, so I try to focus on trials, rather than observational research. Seed oils are a bit different, though, because we have biomarkers for omega-6 fats, which means researchers don't need to rely on questionnaires to determine what people are eating. The correlations between measured fat levels and health outcomes still can't tell us whether the fat caused the outcome, in part because there are so many other confounding factors (people who eat high levels of omega-6 fats may be different, in other ways, from people who don't), but they're much more useful than questionnaire-based data.
I always start with meta-analyses. Not that they're perfect! They're only as good as the evidence they summarize, and researchers can still find ways to exclude studies that contradict their worldview. But I think they're the best jumping-off point we've got.
In the seed oil controversy, the meta-analyses speak almost with one voice. I don't know how many I read - 50, maybe? - and nearly all of them found either that the fats in seed oils were associated with decreased disease risk or that there was no association.
Over and over. And over.
Here's just a smattering:
• A 2019 meta-analysis found that higher levels of omega-6 biomarkers were associated with a decreased risk of heart disease and stroke.
• A 2023 meta-analysis found that higher linoleic acid levels correlated with reduced heart disease risk.
• A 2024 meta-analysis of trials found no association of PUFAs or omega-6s with heart disease.
• A 2018 systematic Cochrane review of trials found, essentially, no relationship between omega-6 fats and heart disease.
• A 2017 meta-analysis found that higher linoleic acid biomarkers were associated with lower risk for Type 2 diabetes.
• A 2012 meta-analysis of trials concluded that "virtually no evidence is available from randomized, controlled intervention studies among healthy, non-infant human beings to show that addition of LA [linoleic acid] to the diet increases the concentration of inflammatory markers." A similar 2017 analysis found the same, with the exception that very high LA levels could increase one marker, C-reactive protein.
(The association between seed oils and cancer is hard to suss out, because there are so many kinds of cancer and much of the research is based on questionnaires rather than biomarkers. But no research that I've seen makes a strong case either way.)
Okay, so if the vast majority of the evidence points to either a small benefit, or no change at all, why the panic over seed oils?
The answer starts with the one meta-analysis that did find seed oils to be harmful, and it's cited over and over in the case against them. Dating to 2010, it finds that substituting omega-6 fats for saturated fats increased the risk of some measures of heart disease by 13 percent.
Well, that's an outlier. Where did it come from?
The answer leads us to studies done in the 1960s and '70s that are the foundation of both the anti-seed-oil and pro-saturated-fat positions. So let's hop in our DeLorean and take a look.
Between 1965 and 1992, 11 fairly large, controlled trials investigated how replacing saturated fat with PUFAs affected heart disease. We're still talking about them because we don't have many more recent studies that are as large or long-running. (Several of the studies were done on institutionalized populations, in ways that might not meet modern ethical standards.)
Conveniently, a 2023 Nature paper, "Dietary fatty acids and mortality risk from heart disease in US adults," contains a list of all 11 (in Table 1 of the Supplementary Material), including their results. If you look at that list (and, if you care about these issues, I encourage it), you see that results are mixed. The Oslo Diet-Heart Study, for example, shows a significant risk decrease, but the Minnesota Coronary Survey shows a significant increase.
As you read about the controversy over seed oil or saturated fat, you'll find that this table is ground zero. People on both sides tend to dismiss the results of the studies that don't support their position and focus on the studies that do. So, for example, the American Heart Association published an advisory in 2017: "We conclude strongly that lowering intake of saturated fat and replacing it with unsaturated fats, especially polyunsaturated fats, will lower the incidence of CVD [cardiovascular disease]." The association cites only four of the 11 studies in the table, all of which support its conclusion.
By contrast, the outlier meta-analysis from 2010, whose lead author (Christopher Ramsden, of the National Institutes of Health National Institute on Aging), was instrumental in recovering and reanalyzing data from the Sydney and Minnesota studies, cites seven of the 11. Some find benefits from PUFAs, some find harms, and Ramsden concludes: "Advice to specifically increase … omega-6 PUFA intake … is unlikely to provide the intended benefits, and may actually increase the risks" of heart disease and death.
I could write a book, or at least a pamphlet, on the strengths and weaknesses of these studies and why they should or shouldn't be included in an analysis. The Finnish one isn't properly randomized! The Minnesota one tracks most patients only for a year! But I'm guessing that would tax the attention span of even dedicated research wonks (which you must be if you've read this far).
But if we take a bird's-eye view of the body of evidence, it leads to a clear conclusion: Seed oils are more likely to be beneficial than harmful, but the effects are small enough that it probably doesn't matter very much if you use them or you don't.
We're fighting about seed oils not because they're either healthful or harmful, but because they've been associated with other controversial issues. The first is saturated fat, which has been the stuff of dietary polarization for a decade or two. If you believe saturated fat is safe, or even healthful, it's just a hop, skip and a jump to believe seed oils are dangerous; that's why sat fat is the subtext of the seed oil brouhaha.
But the second reason is simply the overall state of Americans' health and the impact highly processed foods have had. Because seed oils are a common ingredient in those foods, it's tempting to connect the dots.
I'm on record blaming obesity and disease on processed foods, but trying to find specific ingredients that are harmful is a fool's errand. It's the food in total - nutrition-challenged, calorie-dense, ubiquitous, designed to be overeaten - that's the problem. So, if you're sautéing your vegetables in canola oil every day, you're fine. If you're eating Nacho Cheese Doritos (second ingredient: seed oil) every day, you might not be fine, but the oil isn't to blame.
The seed oil problem is exacerbated by the simple fact that parsing evidence is boring (it is for me, at least, on every issue except the ones in my wheelhouse). And so the question "Are seed oils harmful?" easily morphs into a different question: "Who do I trust to tell me whether seed oils are harmful?" And the answer is likely to be someone who's aligned with you on other nutrition issues. Like, say, sat fat.
And so we get increasingly polarized factions throwing nasties over the parapet.
With seed oils, as with most nutrition issues, you can find evidence for benefits and evidence (although much less) for harms. But the evidence doesn't support zealotry, either way. So maybe stand down.
(COMMENT, BELOW)
Every weekday JewishWorldReview.com publishes what many in the media and Washington consider "must-reading". Sign up for the daily JWR update. It's free. Just click here.