Are Seed Oils Bad for Blood Sugar? What the Evidence Actually Shows

This is one of the most confidently asserted claims in wellness right now, and one of the least supported by human research. I want to lay out the evidence rather than take a side by assertion.

What the research actually shows

A 2025 study in nearly 1,900 people measured linoleic acid directly in blood plasma rather than relying on food questionnaires. Higher levels were associated with lower glucose, lower insulin, and lower HOMA-IR — a marker of insulin resistance. They were also associated with lower inflammatory markers including C-reactive protein.

That's the opposite of the claim.

A meta-analysis of controlled feeding trials found that higher polyunsaturated fat intake — compared with carbohydrate, saturated fat or monounsaturated fat — improved glycemia, insulin resistance and insulin secretion capacity.

Large cohort studies consistently associate higher linoleic acid intake or tissue levels with lower risk of type 2 diabetes and cardiovascular events. Mega-cohort meta-analyses find higher polyunsaturated fat intake associated with lower all-cause mortality.

The trials that get cited against them

The anti-seed-oil case leans heavily on a few older trials — the Sydney Diet-Heart Study, the Minnesota Coronary Experiment, the LA Veterans trial — often via re-analyses conducted decades later.

A 2025 commentary in Frontiers in Nutrition examined the most prominent version of this argument and found the citations didn't support the claims made from them. The Minnesota re-analysis, cited as evidence that seed oils increase cancer and mortality, did not actually evaluate cancer outcomes.

That's a fairly direct problem with the argument.

Where the confusion comes from, and where it's fair

Here's the part I think is genuinely reasonable.

Seed oils are abundant in ultra-processed food. And ultra-processed food is robustly associated with higher type 2 diabetes risk — high-quality evidence, discussed elsewhere on this site.

So people eating a lot of seed oil are frequently eating a lot of processed food, and the association gets attributed to the oil rather than the food it came in.

Reducing ultra-processed food is well supported. Fearing the oil itself is not.

That distinction is the whole article.

What I'd actually do

Cook with what you like. Olive oil is excellent and has good evidence behind it, largely from the Mediterranean literature. Canola, sunflower and soybean oils are not harming you.

Reduce ultra-processed food, which is where the evidence actually points.

Don't replace seed oils with butter or coconut oil expecting a metabolic benefit — replacing polyunsaturated fat with saturated fat is the change with the least supporting evidence.

Be skeptical of anyone comparing a food to smoking. That comparison has been made about seed oils, and it isn't a scientific claim.

Try this week: ignore the oil in your kitchen and look at how much of your food came out of a package. That's where the evidence is.

Education, not medical advice.