Saturated fat: why the number on the label says almost nothing

The saturated fat figure tells you one thing for certain: whether a fat is solid or liquid at room temperature. Above 40%, it is solid. Below 20%, it is liquid. Among common fats, there is nothing in between.

This is no coincidence. A saturated chain is straight and stacks with its neighbors, an unsaturated chain is kinked and does not. The saturated fat figure is, quite literally, the melting point.

That is useful information in the kitchen. It is not information about health, and that is where the misunderstanding starts.

"Saturated fat" is not the name of a molecule, it is the name of a family. The single figure printed on the label adds up at least six different acids, which neither follow the same path through the body nor have the same effect on cholesterol.

The sharpest contrast is between two fats nobody would think of comparing. Cocoa butter is 59.7% saturated, coconut oil 84%. Both count as "highly saturated".

Cocoa butter is mostly stearic acid, 33.2 g. That acid is neutral on LDL cholesterol: it is poorly absorbed, and the liver turns what does get through into oleic acid, the one in olive oil. Coconut oil is lauric acid, 45.8 g, the acid that raises LDL more than any other.

Three readings follow straight from the table.

Palm and coconut are not interchangeable. They reach a similar figure through different acids, palmitic for one, lauric for the other. Lumping them together is like confusing two foods because they weigh the same.

Olive oil is not unsaturated. It holds 15.2 g of saturated fat, 11.8 of it palmitic, exactly the same acid as palm oil. It simply has three times less.

Butter is the only one containing short-chain fatty acids, 8.9 g. Those skip the normal route for fats and go straight to the liver. It is a sixth category, entirely invisible in the total.

A double bond is a place where oxygen can attack. A saturated fat has none, a monounsaturated fat has one, a polyunsaturated fat has two or three. The polyunsaturated figure therefore measures the surface offered to oxidation.

That figure, not the saturated one, is what predicts what an oil becomes in a hot pan.

The saturated figure, by contrast, does not rank oils by quality. It even ranks them backwards: coconut, the most saturated of all, is the one that best withstands heat.

Both families use the same enzymes. An excess of omega-6 takes up the room and blocks the conversion of omega-3, which is why the ratio matters and not just the amount of omega-6.

Rapeseed is thus the best balanced of the common oils, with a ratio of 3. In absolute terms it holds three times more omega-6 than olive oil, but it also brings 7.5 g of omega-3 where olive brings 0.7 and sunflower almost none.

Its weakness lies elsewhere: those omega-3 are fragile and heat destroys them. That is why rapeseed is an excellent unheated oil and a poor one in the pan, the exact opposite of what its smoke point would suggest.

Its stability makes it a good cooking oil, with nothing to oxidize. Its lauric acid, on the other hand, raises LDL cholesterol, and that is one of the best measured effects in the field.

A meta-analysis of sixteen controlled trials gives, against non-tropical vegetable oils: LDL +10.5 mg/dL, total cholesterol +14.7. No effect, however, on triglycerides, weight, blood sugar or inflammation.

Two caveats. HDL rises too, by 4.0 mg/dL, so the total-to-HDL ratio moves less than LDL alone. And the link between higher LDL and an actual cardiovascular event remains the contested step in this whole story, the one our article on Ancel Keys tells.

What is solid is that coconut oil raises LDL. What is less solid is by how much it changes the risk.

The saturated figure tells you for certain whether a fat will be solid or liquid, and almost nothing about what it does to the body. For that you need the breakdown, and that is exactly why the debate has run for sixty years.

In practice, three rules of thumb are enough.

Olive stays the default oil. It is the only one combining decent stability under heat with a small effect on LDL.

Rapeseed for unheated use. Its omega-6 to omega-3 ratio is the best of the common oils, provided you do not heat it.

Coconut for high heat, in modest amounts. Its stability is real, and so is its effect on LDL.

Sunflower is the real loser in these tables: it has neither the stability of coconut, nor the profile of olive, nor the omega-3 of rapeseed. To choose in the moment at the stove, the Choosing your cooking fats page compares the common cooking fats by use.