Practiceby Lab 10 Get started ›
Crossfire

Should you cook with seed oils, or replace them with olive oil, ghee and tallow?

What the experts say

Ben GreenfieldMS · exercise physiologist, NSCA and ISSN certified by his own account, no clinical credential

Replace them. Cook with extra-virgin olive oil, ghee or tallow, and treat fried restaurant food as the worst offender.

What it rests onHe makes two arguments and both matter. The mechanistic one: polyunsaturated oils oxidise readily, the resulting oxidised linoleic acid metabolites impair mitochondrial function, heated oils form 4-hydroxynonenal, and damaged fats are incorporated into cell membranes where they cannot be metabolised away the way excess sugar can, which is why he writes 'I would rather mow through an entire bag of sugar than consume just about any fried food on the face of the planet'. The precautionary one, stated explicitly: 'There is a huge necessity for long-term research to fully understand the health implications', which he treats as a reason to avoid now rather than a reason to wait. His main citation is Joseph Mercola's linoleic acid work in Nutrients, and Mercola sells the alternative fats at scale. He also links two tallow-fried chip brands, both of which have pages in his own product catalog.

Source: bengreenfieldlife.com ↗
Gil CarvalhoMD, PhD · nutrition and aging researcher, creator of Nutrition Made Simple!

Keep cooking with them, especially in place of butter, lard or tallow.

What it rests onHe concedes the chemistry and disputes its consequence, which is the part a headline drops. On oxidation he writes that polyunsaturated fats are chemically less stable, especially when heated, that trial results on oxidation products are inconsistent, and that seed oils also carry antioxidants such as vitamin E, and then argues that disease outcomes rather than laboratory biomarkers should drive the recommendation, noting that high consumers do not show more oxidation-related disease. On inflammation he cites dozens of randomised trials that consistently fail to detect a rise in inflammatory markers, with some detecting a fall. On hexane he notes it is used in refining many oils including coconut, is largely evaporated before sale, and that cold-pressed versions exist. He also concedes a contextual point that narrows the fight: most seed oil intake arrives inside ultra-processed and fast food, which is a different question from cooking with canola at home.

Source: gilcarvalhomdphd.substack.com ↗

Overview

Both address general adults and neither is arguing about the same eater. Greenfield is writing for someone already eating whole food and looking for the next thing to remove, and his sharpest language is about fried restaurant food, which Carvalho also treats as the problem case. Carvalho is writing for someone deciding what to put in the pan at home, and his recommendation is specifically comparative: seed oil INSTEAD OF butter, lard or tallow. Neither is defending fast food.

Read as 'are seed oils bad' this looks like a much wider fight than it is. The real question is what to cook with at home, and what to do when a plausible mechanism and a large body of outcome evidence point opposite ways.

Where they agree

Both agree the oils are chemically less stable than saturated fats and oxidise when heated. Both agree that the bulk of population-level seed oil intake arrives through ultra-processed and fried food and that this is the intake worth reducing. Both endorse extra-virgin olive oil, and both endorse whole nuts and seeds as fat sources. Neither disputes that whole foods should dominate.

What would settle it

A long-duration randomised trial feeding matched diets differing only in the cooking fat, with hard cardiovascular and metabolic endpoints rather than oxidation biomarkers, in people whose background diet is not ultra-processed.

Every large trial in the existing literature compared seed oils against a saturated-fat control inside an ordinary Western diet, which is not the comparison either man is actually arguing about. Greenfield's own stated position is that the absence of long-term data is itself decisive, so a null result would probably move him less than it would move you, which is worth knowing before you weigh him.

Practice does not pick a winner here

You do, and your call is dated, kept, and yours to change. Follow the people you already trust and what they publish turns into a short daily practice.

Get started

Free to sign up and use. We do not ask for a card.