Start · Catalogue · Profile · Table
Food BODY HANDBOOK
Food · §347
Seed Oils
Half a century ago, the fat in your kitchen came from animals or olives. Now most of it comes from seeds — soybean, sunflower, canola — extracted with a petroleum solvent and bleached clear at 250°C. They make up roughly 10% of the calories the average American eats. Whether that swap is a cardiovascular win or a slow-motion industrial accident is one of the most genuinely contested fights in nutrition right now: cardiologists publish against each other in the same medical journals. The honest read of the literature: home cooking with olive oil and butter is a defensible default, deep-fried restaurant food is where the case against seed oils is strongest, and a lot of the rest sits in real uncertainty.
Avoid · Daily Evidence Mixed Chapter Food

What's at stake isn't the canola in your salad dressing — it's the repeatedly reheated fryer oil at the fast-food window, and the toxic aldehydes that come off it onto your fries. Skip the worst exposures and the rest is honest middle ground. The home pantry swap takes a minute. Restaurant choices take longer. Expect no consensus this decade.

Linoleic acid is the molecule the fight is about. It's an essential omega-6 fat — your body can't make it, you have to eat some — and it makes up the bulk of every seed oil on a grocery shelf. Safflower and grapeseed run 70–75% linoleic acid. Soybean and corn, about half. Sunflower, roughly two-thirds (the high-oleic varieties are the exception). Canola is the outlier — only 20% linoleic acid, chemically closer to olive oil than to soybean.

The thing that matters about its chemistry: linoleic acid has two weak spots in its structure that make it easy to break apart. Heat, air, light, and your own body's metabolism all attack those weak spots and set off chain reactions. The end products are a family of small, sticky, protein-damaging molecules called aldehydes — and they show up at higher concentrations inside diseased artery walls than in healthy tissue.

That's the contrarian mechanism in plain terms — eating heated linoleic acid is a way of eating its breakdown products. The mainstream mechanism is different: replacing animal fat with linoleic acid lowers LDL-C by 10–15% in tightly controlled feeding studies Mensink 2016 (WHO), and lower LDL is supposed to mean less heart disease. Both mechanisms can be true at once. The fight is over which one wins at the population scale — and whether the LDL drop actually buys you years.

Two camps, same journals

The evidence is split, and it's split deeper than most "nutrition is confusing" debates. Both sides publish in the same medical journals, cite many of the same trials, and reach opposite conclusions about whether the substitution that's been standard public-health advice since the 1960s actually works.

The mainstream case rests on two pillars. First, large prospective studies that measured linoleic acid directly in people's blood and body fat — they consistently find that people with higher tissue levels have lower cardiovascular mortality. Second, randomised trials of swapping animal fat for vegetable oil reduce cardiovascular events (heart attacks, strokes, procedures) by about 17%, per the most recent Cochrane review Hooper et al. 2020. The American Heart Association's 2017 advisory landed on this evidence base: seed oils are heart-healthy, eat less butter Sacks et al. 2017.

The contrarian case rests on two trials that were buried for decades. The Sydney Diet Heart Study ran from 1966 to 1973 on 458 men who'd just had a heart attack. The intervention swapped their animal fat for near-pure linoleic acid from safflower oil. The safflower group's risk of dying was about 60% higher than the controls'. The full mortality data wasn't published until 2013, when an NIH researcher tracked down the original records and ran the analysis Ramsden et al. 2013.

The reconcilable middle: the event trials and the mortality trials have been quietly diverging for fifteen years, and nobody has run the modern trial that would settle it. The Cochrane review — the most authoritative summary going — confirms the event-reduction signal but finds no detectable mortality effect either way (RR 0.96, 95% CI 0.90–1.03, moderate-quality evidence) Hooper et al. 2020. That's a strange outcome under the original logic: if cholesterol drops and heart attacks fall, people should live longer. The fact that they don't is what gives the contrarian camp a real opening — and what the mainstream camp has not yet answered convincingly.

Three lines that don't survive contact with the literature

"Smoke point is the safety number." No. Oxidation starts well below the smoke point. The smoke point tells you when the oil visibly degrades into something you wouldn't want to breathe — it's not when the toxic aldehydes start forming. Those start at normal frying temperatures, hours before any visible smoke Grootveld et al. 2020. A high refined smoke point is a marketing number, not a safety number.

"Seed oils are heart-healthy because they lower cholesterol." They reliably lower LDL-C by 10–15% when you swap them for animal fat — that part isn't contested. What's contested is whether that translates into a longer life. The Minnesota trial got the full cholesterol drop and no mortality benefit Ramsden et al. 2016. Cholesterol response and mortality response are not the same outcome, and the original advice treated them as if they were.

"All seed oils are equivalent." They aren't. Safflower and grapeseed are nearly all linoleic acid. Soybean and corn are about half. Canola has only 20% linoleic acid and 10% omega-3 — chemically, canola sits closer to olive oil than to soybean. The "seed oils" shorthand collapses meaningful differences. The strongest case against any of them is against industrially heated, repeatedly reused soybean and corn fryer oil; the weakest case is against canola in moderate home use.

"Your body needs linoleic acid, so eating more is better." Your body needs about 1–2% of your calories as linoleic acid; the average American now eats around 7–8% Blasbalg et al. 2011. Essential doesn't mean dose-independent. Iron is essential too; you can still get too much.

What to actually do

In descending order of how confident the evidence lets you be.

Cost honestly: extra-virgin olive oil runs 5–10× the per-litre cost of soybean or canola. For a typical household swapping its cooking-oil pantry, that's $50–200 a year extra. Butter and ghee are cheap. The bigger cost line, if you take this seriously, is eating out less.

Effort honestly: the home swap takes one minute and one grocery trip. The sustained effort lives in restaurants — asking what oil the kitchen uses, accepting that the answer is almost always soybean, choosing the grilled item over the fried one. Choice by choice, not hour by hour.

What's already happening

The unusual thing about this exposure is that it's already been running on you, mostly silently, for whatever fraction of your life you've been eating standard American food. Body-fat measurements of US adults show the linoleic acid stored in their subcutaneous tissue roughly doubled between 1959 and 2008 Guyenet & Carlson 2015. That tissue has a slow half-life — roughly two years — so what's in you now is a multi-year average of what you've been eating, not what you ate last week.

If the mainstream case is right, this is approximately fine, and the conversation is about how much extra benefit you'd squeeze out of optimising further. If the contrarian case is right, the next decade in the version of you who keeps eating restaurant fries twice a week is a slow drift the standard lipid panel will miss — fatty-liver readings the doctor calls "borderline," insulin sensitivity quietly compressing, the kind of low-grade inflammation that registers as "just getting older." Friends in their fifties have lab numbers that gradually look like they belong to someone older. Nobody knows for sure which case is closer to right.

The asymmetry is what to act on. If mainstream is right, cutting the deep-fryer channel costs you almost nothing. If contrarian is right, it's a quiet rescue you were not going to feel landing.

What changes when you swap

The first thing most people who make the switch report is feeling lighter and less bloated within a few weeks — and feeling like restaurant fried food is harder to come back from when they do eat it. The community signal is consistent, and the mechanism (less aldehyde and oxidised-fat load on the gut) is plausible. The randomised trial confirming this specific felt change hasn't been run.

What has been run, and is closer to a settled win than anything else in this debate, is the Mediterranean-pattern trial that's effectively a seed-oil-displacement trial in disguise. Five years in, the olive-oil arm took about a third off their cardiovascular-event rate compared to the low-fat control. The "cooking with olive oil" effect and the "not cooking with seed oils" effect are not separable inside the data, but you don't have to separate them to take the benefit.

The honest payoff sentence is conjunctive: you swap the oil, you eat more fish and nuts, you cook at home more, you eat less restaurant food, and over five to ten years you take roughly the same edge on your cardiovascular risk that the olive-oil arm took on theirs. The long-tail aesthetic benefit (Mediterranean eaters look measurably less aged at the decade mark) rides along with the same package. None of the components are isolated, and you don't have to isolate them to get the win.

Adjacent topics worth knowing about: omega-3 supplementation (the other half of the ratio question — whether adding fish oil on top of high seed-oil intake helps the way replacing seed oils does); industrial trans fats (a separate substance — partially hydrogenated oils are not modern refined seed oils, but they co-occur in the food supply and have their own settled-bad evidence base); ultraprocessed food generally (the largest delivery channel for seed oils, and an independent cardiovascular risk on its own); and ApoB and the lipid panel (what your standard cholesterol numbers actually tell you about heart risk and where they miss).

Related in the handbook
Related
  • — ApoB and the lipid panel are what actually tell you whether your fat choices are landing where it counts.
  • — The other half of the fat-ratio question is whether adding omega-3 on top helps the way cutting seed oils does.
  • — Most seed-oil exposure comes through ultra-processed food; the deep fryer and the packaging are the real concern.
  • — Cooking vegetables with fat boosts their nutrients — but which oil you use is its own contested debate.
·
347