Practiceby Lab 10 Get started ›
Crossfire

If I eat more butter and cheese, is that bad for my heart?

What the experts say

Nina TeicholzPhD · investigative science journalist and author

No. The trials designed to prove it never did, and LDL alone cannot settle it.

What it rests onGovernments ran large randomised trials in the 1960s and 1970s replacing saturated fat with polyunsaturated vegetable oils, together testing about 67,000 people with hard endpoints. Compliance was demonstrated by an average 29 mg/dl fall in cholesterol, yet the expected mortality reductions did not appear in most trials. More than twenty subsequent review papers have on the whole found no effect on cardiovascular events, cardiovascular mortality or total mortality. Reviews finding harm mainly do so by including the non-randomised Finnish Mental Hospital Study, and Cochrane 2020's effect on cardiovascular events disappeared in its own sensitivity analysis. She adds that diet-induced LDL lowering has not been shown to reduce outcomes the way drug-induced lowering has, that the LDL saturated fat raises is the large buoyant kind, and that saturated fat raises HDL.

Source: pmc.ncbi.nlm.nih.gov ↗
Gil CarvalhoMD, PhD · nutrition and aging researcher, creator of Nutrition Made Simple!

Yes, enough to swap it. The lipoproteins are the causal agent and saturated fat raises their number.

What it rests onObservational data showing less heart disease and death when saturated fat is replaced with unsaturated fat or whole carbohydrate; a meta-analysis of the fourteen trials meeting minimum quality standards finding a modest reduction in cardiovascular events; and the mechanism that saturated fat raises the number of the lipoproteins he calls the main causal agent of heart disease. He discounts the same 1960s trials she relies on, because Minnesota lost most of its participants and was interrupted before outcomes could be recorded, and because saturated fat was sometimes replaced with trans-fat margarine. He is careful in the same breath: he calls flat condemnation of saturated fat a gross oversimplification and grades the randomised evidence as modest.

Overview

Both are addressing the general adult population choosing everyday cooking fats, which is rare and makes this an unusually clean disagreement: there is no population mismatch to hide behind. The one real scope difference is that her stronger claims are aimed at people who are already metabolically ill, where she argues the whole calculation changes, and his are aimed at everyone.

Where they agree

Real and specific. Both reject reasoning from a nutrient to a plate without the food matrix (tei-satfat-food-matrix and car-satfat-nuance make structurally identical arguments from opposite sides). Both demand hard outcomes rather than markers, and both say so as a principle.

Both treat the 1960s trials as central rather than peripheral. Both reject the two loudest camps: she says do not gorge, he says one side calls meat poison and the other says there is no risk whatsoever and both are exaggerations.

What would settle it

For her: a randomised trial with hard endpoints, adequate power, and a design that separates saturated-fat reduction from everything else it usually travels with. For him: the same. Both sides are asking for a trial that will not be run, which is why this fight is stable, and why setting both positions out rather than declaring a winner is the honest answer.

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.