''
In 1967, the New England Journal of Medicine published a literature review that reshaped how Americans ate for the next 50 years.
It was titled "Dietary Fats, Carbohydrates and Atherosclerotic Disease," and it concluded that saturated fat was the dietary culprit in heart disease.
Not sugar. Saturated fat.
Hence, the diet-heart hypothesis that governs much of modern nutritional thinking:
Saturated fat raises blood cholesterol, and blood cholesterol drives heart disease. Therefore, reducing saturated fat reduces heart disease.
The journal didn't disclose (and wasn't required to at the time) that the review had been funded by the Sugar Research Foundation (SRF), the trade association for the US sugar industry.
The SRF paid the equivalent of ~$48,000 in 2026 dollars to three Harvard nutrition scientists:
Frederick Stare (chair of Harvard Nutrition), D. Mark Hegsted (who’d later help write the first US Dietary Guidelines), and Robert McGandy.
According to internal industry documents uncovered by Cristin Kearns and her UCSF colleagues in 2016, the SRF hand-selected which studies the reviewers analyzed. Internal correspondence showed the industry communicating what they wanted the outcome to be. The Harvard scientists delivered it.
For five decades, doctors, dietitians, and government agencies built the case against animal fats — and the case for the low-fat, ultra-processed , seed-oil-enriched American diet — on top of that paper and the others that followed in its wake.
Now, the diet-heart hypothesis was accelerated by the 1967 review, but is itself older than that…
The strongest support came from Ancel Keys' Seven Countries Study, begun in 1958.
Keys tracked dietary patterns and cardiovascular outcomes across populations in Greece, Italy, the former Yugoslavia, the Netherlands, Finland, the United States, and Japan.
He found a correlation between saturated fat intake and heart disease, and that correlation became the empirical backbone of half a century of dietary policy.
However…
The Seven Countries Study was observational.
It compared whole populations, so it couldn't isolate saturated fat from everything else those populations ate (like sugar, refined grains, processed food, and alcohol) or from how they lived, smoked, worked, and slept.
When you compare countries to each other, you compare ways of life. Critics raised this at the time.
In 1957, before the Seven Countries Study had even begun, Jacob Yerushalmy and Herman Hilleboe published a methodological critique of Keys' earlier cross-country work. They pointed out that broader data showed weaker and less consistent associations than Keys' initial analyses suggested.
The critique was largely ignored. The hypothesis kept moving.
By 1977, the US Senate's McGovern Committee report recommended reducing saturated fat consumption.
By 1980, the first official US Dietary Guidelines made that recommendation national policy.
The UK followed in 1983. The American Heart Association followed shortly after.
The policy was built on observational data, industry-influenced reviews, and a methodological framework that couldn't isolate cause from correlation. That was the foundation.
The randomized controlled trials, which are the kind of evidence required to actually establish causation, came later…
And they didn't validate the hypothesis.
A 2025 systematic review and meta-analysis published in JMA Journal pooled 9 randomized controlled trials with 13,532 participants.
The conclusion:
Saturated fat restriction did not produce significant reductions in cardiovascular mortality (RR 0.94), all-cause mortality (RR 1.01), myocardial infarction (RR 0.85), or coronary artery events (RR 0.85).
Earlier meta-analyses found similar things.
A 2022 review by Valk and colleagues concluded there was no conclusive link between saturated fat intake and cardiovascular risk.
A meta-analysis by Harcombe and colleagues across seven cohort studies and roughly 90,000 participants found no significant difference in coronary heart disease mortality based on saturated fat intake.
Across more than two dozen meta-analyses and systematic reviews now in the literature, the consistent finding is that the diet-heart hypothesis — saturated fat causes cardiovascular disease — isn't supported by the level of evidence required to enforce it as universal policy.
The science required to back the dietary guidelines, it turns out, was never assembled.
The guidelines remain in place anyway.
I’ve trained inside this system. My father was a doctor, too. The advice he passed to his patients came directly from these guidelines.
He wasn't lying to anyone, though.
He was repeating what his training told him was settled science.
The system trains brilliant, well-meaning people to repeat a consensus built on a paper, paid for by an industry, using a methodology that couldn't support the conclusion in the first place.
By the time the original premise was seriously questioned, three generations of doctors had already built their practices on top of it.
This is what I mean when I say the system isn't the people. The people inside it are mostly trying to help.
So if saturated fat isn't what we should have been focusing on, what is?
The newer markers tell a much more accurate story…
ApoB is the most important. It measures the number of atherogenic particles (the ones that drive arterial plaque buildup) in your blood. The 2019 European Society of Cardiology guidelines and the 2024 National Lipid Association consensus both concluded that ApoB is a more accurate cardiovascular risk marker than LDL cholesterol.
A 2025 review found 9 of 9 discordance studies showed ApoB outperforms LDL-C. When ApoB and LDL-C disagree on risk, the actual events track with ApoB.
Triglyceride/HDL ratio is one of the simplest and most informative insulin resistance markers.
A ratio above 2 (in American units) suggests metabolic dysfunction. Below 1.5 is good. This is on almost every standard panel and almost nobody uses it.
Fasting insulin is rarely measured in standard panels but tells you more about metabolic health than fasting glucose.
By the time fasting glucose is elevated, insulin resistance has usually been compensating for years.
Lp(a) is genetically determined and an independent cardiovascular risk factor. Many cardiologists don't test for it.
These are the markers that matter. The cholesterol panel your doctor runs yearly mostly measures the wrong thing.
Here's what to do with this information:
Get the right tests. ApoB, triglyceride/HDL ratio, fasting insulin, Lp(a), and hs-CRP (an inflammation marker) tell far more about your cardiovascular trajectory than total cholesterol. Most labs run them on request. Insurance won't often cover them, but they're inexpensive out of pocket.
Eat real food. Single-ingredient, animal-based foundation. Ruminant meat, organs, eggs, dairy if tolerated, fruit, raw honey, seasonal vegetables you tolerate well. This is what humans evolved to eat. The body recognizes it. The cardiovascular biomarkers respond to it.
Cut the actual drivers. Seed oils (canola, soybean, corn, sunflower, safflower) and ultra-processed foods are the dietary inputs most strongly associated with metabolic dysfunction. They drive insulin resistance, inflammation, and the apoB particle elevations that actually predict events.
Move your insulin sensitivity. Resistance training, sleep, time-restricted eating, walking. The mechanism that's broken in most cardiovascular disease is metabolic, not lipid-specific. Address the metabolism and the lipids follow.
The 1967 paper that started this is the foundation of dietary guidelines still in place today.
The science required to defend it was never built. The science that did get built tells a different story.
You don't have to fear animal foods.''
Paul Saladino
