Millions of people are obsessing over seed oils while ignoring the thing actually driving the disease curve.
The standard story you have heard on a podcast or a YouTube clip runs like this. Seed oils contain too much linoleic acid. Linoleic acid is pro-inflammatory. Pro-inflammation is upstream of heart disease, metabolic disease, and most of what kills middle-aged adults. Therefore, take the oils out and your healthspan goes up.
The story is internally tidy. The trial data does not support it. The biomarker data does not support it. And the actual food matrix that seed oils show up inside is the variable doing the damage, not the oil itself.
That does not mean every exposure to industrial oils is harmless. Repeatedly heated fryer oil, heavily oxidised oils, and ultra-processed foods built around hyperpalatable fat-salt combinations are probably a different discussion entirely. But that is a very different claim from saying seed oils themselves are independently driving modern disease.
|
|||
| Today's minimum effective dose: at your next grocery shop, ignore the oil on the ingredient list and read everything else. More than five ingredients, or anything you would not find in a working kitchen, is ultra-processed. That is the variable that matters. | |||
| Sources: Sacks et al, AHA Advisory, Circulation 2017 · Marklund et al, Circulation 2019 · Hall et al, Cell Metabolism 2019 |
What The Research Actually Says
The 2017 American Heart Association Presidential Advisory on Dietary Fats (Sacks et al, Circulation) reviewed the long-term randomised trials and concluded that replacing saturated fat with polyunsaturated vegetable oil reduces cardiovascular disease by roughly 30 percent. That is the effect size of a statin. Whatever you have been told about linoleic acid being inflammatory, the trials measuring actual cardiovascular events have gone the opposite direction.
The 2020 Cochrane review (Hooper et al, 15 randomised trials, 59,000 participants) ran the same analysis with stricter methodology and found a 21 percent reduction in combined cardiovascular events when saturated fat was lowered, with the cleanest signal coming from trials that swapped in polyunsaturated fat. The number needed to treat in secondary prevention was 32. That is a real intervention, with a real effect, in the wrong direction for the panic.
If the trial data were unreliable, the biomarker data should rescue the seed oil story. It does the opposite. The Marklund et al pooled analysis in Circulation in 2019 combined 30 prospective cohort studies, 68,659 adults, and measured linoleic acid directly in blood and tissue rather than relying on food questionnaires. Higher in vivo linoleic acid was associated with lower risk of cardiovascular events and lower cardiovascular mortality. The molecule the YouTube doctors warn you about, when actually measured in your body, tracks with better outcomes.
The food matrix is where the real signal lives. The Hall et al inpatient RCT in Cell Metabolism in 2019 housed 20 adults at the NIH and fed each of them an ultra-processed diet and an unprocessed diet for two weeks each, in random order, with both diets matched for presented calories, sugar, fat, sodium, and fibre. On the ultra-processed arm, participants ate 508 more calories per day on average and gained 0.9 kg. On the unprocessed arm, they lost 0.9 kg. Same nutrient targets. Different food matrix. A nearly two-kilo swing in a fortnight.
That is the lever. Not the oil. The processing, the texture, the engineered hyperpalatability, the things you cannot see on a food label.
The Smarter Version Of The Panic
The more sophisticated version of the argument is not about linoleic acid. It is about extraction. Industrial seed oils are pulled out with hexane, heated, deodorised, and sometimes bleached, and the processing produces oxidation products that are genuinely inflammatory in a petri dish. That chemistry is real. It is also not what the outcome trials measured. The Sacks, Hooper, and Marklund evidence is from people consuming the actual commercial seed oils the panic is about. The trials used industrially refined soybean, corn, and safflower oils. The biomarker meta tracked linoleic acid in bodies that had been consuming these processed oils for years. If processing made seed oils net-harmful, the outcome data would show it. It does not. The one place the harm signal does appear is repeatedly heated frying oil, the kind a restaurant fryer holds for a week. That is a behavioural distinction, not a molecular one.
The Bottom Line
The mechanistic story you have been sold about seed oils does not survive contact with the trial data, the biomarker data, or the food-matrix data. Linoleic acid in your bloodstream is not driving the disease curve. The category of food that linoleic acid most often arrives inside is.
Take the seed oil out of an ultra-processed product and replace it with butter and the product is still ultra-processed. Take the ultra-processed product out of your week and cook the same dinner at home in canola oil, and the seed oil stops mattering. That is the actual decision tree.
The single specific action this week is unglamorous. At your next grocery shop, ignore the oil on the ingredient list and read everything else. If a product has more than five ingredients, or any ingredient you would not find in a working kitchen, you are buying ultra-processed food. That is the variable the trials are measuring. The oil is a passenger.
Stay tuned for Friday's edition: three things we learned this week.
Longevity Daily
The Building Decades
Share this edition with someone who'd find it useful.
Instagram: @longevitydaily_ | Subscribe: join.longevitydaily.com