Is “lowering cholesterol” the same thing as “better health”?

For decades we’ve been acting as if the answer is yes. We were wrong.

In 1982 the Australian government told you to eat less fat, and the country did as it was told. Canola spread instead of butter. Wholemeal instead of white. Skim instead of full cream. Walk any supermarket aisle today and you’ll still pass a few hundred products carrying a heart tick, a low-fat claim, or a label telling you the fat inside is the good kind.

Forty-four years of Australian dietary advice built that aisle. Most men over 40 have spent their entire adult lives steering by it, and a fair few of them are now carrying 15 or 20 kilos they can’t explain.

Here’s the part that should bother you. The advice was built on a swap: cut the saturated fat, put vegetable oil in its place, and your cholesterol will come down. The cholesterol did come down. That was never the question. The question was whether anyone lived longer, and when two randomised trials finally checked, one of them here in Sydney, the answer didn’t follow the number.

The history is how it happened and it’s worth your time. If you’d rather have the evidence than the backstory, start at the Sydney trial.

Who was Ancel Keys and how did he change the dietary guidelines?

Ancel Keys was an American physiologist who argued that saturated fat raises blood cholesterol, and that blood cholesterol causes heart disease. He didn’t write Australia’s dietary guidelines, or America’s. What he did was build the hypothesis both were eventually based on, and push it hard enough, for long enough, that it became the official position before the trials came in.

He wasn’t a quack. Keys held two doctorates, designed the K-ration that fed American soldiers through the Second World War, and ran the Minnesota Starvation Experiment, still the definitive work on what prolonged caloric restriction does to a human being. He was a serious scientist with a serious record.

Which is the point. This isn’t a story about a fraud. It’s a story about serious people committing an institution to an idea before the idea was tested, and how hard that is to walk back.

Australia adopted the same advice in 1982

Australia published its first Dietary Guidelines for Australians in 1982, two years behind the United States, carrying the same core instruction: eat less fat. They’ve been revised in 1992, 2003 and 2013, and the fat message has survived every revision.

The same year, Nutrition Australia put up the Healthy Eating Pyramid that ended up in classrooms and GP waiting rooms around the country. Cereals, breads, vegetables and fruit in the “eat most” band at the base. Butter, margarine and fats in “eat in small amounts” at the tip. A generation was told to build its diet on bread and cereal and to treat fat as something to ration.

I can’t show you a document where Australia sat down and decided to copy America, and I’m not going to pretend one exists. What I can tell you is that our guidelines landed two years after theirs, carried the same instruction, and that no Australian trial had established the benefit either.

Then the food supply moved the way the advice pointed. Australian food supply data from 1961 to 2009 shows linoleic acid, the dominant fat in seed oils, rising by roughly 177 per cent relative to where it started, almost entirely from plant oils, with canola the largest single source by 2009. Animal sources barely moved, up about 11 per cent. Availability isn’t intake, and a food supply shifting alongside a health trend doesn’t show that one caused the other. What it shows is that the food environment changed substantially, in exactly the direction the authorities were recommending.

Here’s what I see at the other end of it. Take the fat out of a meal and something fills the gap, and in the men I see it’s bread, cereal, rice and pasta, several times a day. Carbohydrate raises insulin, insulin favours storage over release, and a man eating that way rarely gets a long enough gap to burn much of what he’s put away. That’s a pattern I see in the men I work with, not an explanation for Australia’s obesity curve, and anyone offering you the second one is guessing.

What was the Sydney Diet Heart Study and what did it find?

The Sydney Diet Heart Study was an Australian randomised trial that ran from 1966 to 1973. It took 458 men aged 30 to 59 who had already had a heart event, and replaced the saturated fat in their diet with linoleic acid from safflower oil and safflower margarine. Their cholesterol fell. Their death rate rose.

It’s ours, almost nobody has heard of it, and it finished nine years before Australia adopted the advice it was testing.

Read what it actually did, because the detail is the whole thing. This was not a trial of eating less fat. It was a trial of swapping one fat for another: animal fat out, polyunsaturated vegetable oil in. That swap is the central intervention the diet-heart hypothesis produced, and it’s the swap Australian households were about to be asked to make.

The controls were given no dietary instruction and no study foods. The intervention group made the swap, and it worked on the marker, exactly as the hypothesis predicted. That group also had higher all-cause mortality (hazard ratio 1.62), higher cardiovascular mortality (1.70), and higher coronary mortality (1.74). The cardiovascular and coronary results reached statistical significance. The all-cause result sat just outside it, at p equals 0.051, and I’d rather tell you that myself than have you find it in the paper.

Most of that mortality data went unpublished for four decades. It sat unexamined until a team led by Dr Christopher Ramsden recovered the original records and published the full outcomes in the British Medical Journal in 2013. I’m not going to tell you somebody buried it, because I can’t show you who or why. Incomplete publication is enough of a problem on its own: for forty years the field was reasoning from a version of this trial that left the deaths out.

Now the objection, because it’s a good one and you’d find it anyway. The safflower margarine of that era carried industrially produced trans fat, which damages arteries on its own. So the excess deaths might be a trans fat signal rather than anything to do with the oil. That’s the strongest argument against reading this trial the way I read it, and it deserves saying out loud.

Which is why the American trial matters. The Minnesota Coronary Experiment ran from 1968 to 1973 across 9,423 institutionalised adults, using liquid corn oil with no margarine confound. Keys was one of its investigators. Cholesterol fell hard in the intervention group, by 31.2 mg/dL against 5 in the controls, and the survival benefit the hypothesis predicted didn’t appear. The raw data sat in a colleague’s basement until Ramsden’s team recovered the tapes and published in the BMJ in 2016.

The two trials did not find the same thing, and it’s worth being precise about that. Sydney found more deaths among the men who made the swap. Minnesota found no benefit, which is not the same as harm. There’s a further finding in the recovered Minnesota data, that lower cholesterol tracked with higher death risk in participants over 65, but it came out of analysis after the fact rather than the randomised comparison, so it carries much less weight. And Minnesota wasn’t the population you’d choose: institutionalised adults aged 20 to 97, across a nursing home and six state hospitals, is a long way from a 52-year-old in Brisbane.

So what do the two together establish? Not that fat is harmless, and not that seed oils killed anyone. Something narrower and more useful.

Lowering cholesterol is not the same as living longer

The intervention did exactly what it was designed to do. It lowered the number. The outcome the whole hypothesis existed to improve didn’t follow. Ramsden’s team put it plainly in 2016: replacing saturated fat with linoleic acid lowers serum cholesterol effectively, and the evidence does not support the idea that this translates into fewer deaths.

That’s the sentence the guidelines needed in 1977, and nobody had it. What they had instead was a chain of reasoning that felt like proof. Saturated fat raises cholesterol, which was true. Cholesterol tracks with heart disease, which was also true. Therefore lowering cholesterol will save lives, which is where it stopped being an observation and became an assumption. The trials were the place to check the assumption, and by the time anyone read them properly the advice was forty years old.

Once you’ve seen that mistake you’ll see it everywhere, because we still make it. A number that’s easy to measure gets promoted into the thing we actually care about. Scale weight standing in for body composition. A glucose reading standing in for metabolic health. Calories burned standing in for fitness. Each one is worth knowing and none of them is the outcome.

Eisenhower’s heart attack turned a hypothesis into a national emergency

Getting from one man’s hypothesis to two governments’ policy took twenty-five years, and it started with a heart attack. In September 1955 President Dwight Eisenhower played a round of golf, thought he had indigestion, and woke with crushing chest pain. Heart disease had just become the leading cause of death in the United States, and here was the most powerful man in the world flattened by it.

His cardiologist, Dr Paul Dudley White of Harvard, held daily press conferences walking the public through the president’s cholesterol readings and his diet, then went looking for a fix and found Keys. Eisenhower went onto a heavily reported low-fat, low-cholesterol diet. Butter out, polyunsaturated margarine in. Good enough for the president was good enough for everyone, and millions followed.

He stayed on it for the rest of his life and died of heart disease in 1969. I won’t hand you that as evidence: one man’s outcome proves nothing, and he’d smoked heavily for decades before quitting in 1949. Eisenhower didn’t make Keys right. He made Keys urgent, which turned out to matter more.

Keys’ six-country chart left out sixteen countries with available data

Keys’ famous 1953 chart used six countries when data existed for twenty-two, and the relationship it showed didn’t survive the other sixteen.

Fat consumption on one axis, heart disease deaths on the other, the dots in a near-perfect curve. Japan at the bottom, the United States at the top, Australia up there beside it. Four years later two statisticians, Jacob Yerushalmy and Herman Hilleboe, put all twenty-two countries on the same axes and the curve turned into a shotgun blast.

I’m not interested in prosecuting him for it. He used a small sample with no stated selection criteria, which is sloppy rather than fraudulent, and plenty of good scientists have done worse. The part that matters is what came next. The Seven Countries Study followed about 12,000 men prospectively and it’s the real body of Keys’ career. It found saturated fat associated with heart disease. It did not find the same for total fat. Nearly everyone runs those together, and the difference is the whole argument. An association between saturated fat and heart disease is not the instruction “eat less fat”, and neither one is the claim that swapping the butter for vegetable oil will make you live longer. The guidelines ended up making all three.

There was industry money in it too. In 1965 the Sugar Research Foundation quietly paid Harvard researchers for a review that pinned heart disease on saturated fat and waved away the evidence on sugar, which we only know because researchers went through the industry’s own correspondence and published it in JAMA Internal Medicine in 2016. One of its authors was Mark Hegsted.

Why did the dietary guidelines change in the 1970s?

The dietary guidelines changed because a US Senate committee decided it couldn’t afford to wait for proof. That’s not just my opinion, it’s what the committee’s own scientific advisor argued at the time.

The committee was chaired by Senator George McGovern, its staff directors had no formal training in nutrition, and Hegsted, by then a decade on from the sugar review, was the scientist they leaned on hardest. His position was that public health officials couldn’t sit on their hands waiting for certainty. In February 1977 they published Dietary Goals for the United States: cut total fat to 30 per cent of energy, cut saturated fat to around 10 per cent, and lift carbohydrate to 55 or 60 per cent.

The pushback was immediate, and the cattle, egg and dairy lobbies won on wording. McGovern was blunt about why: “I did not want to disrupt the economic situation of the meat industry and engage in a battle with that industry that we could not win.” The December revision dropped “decrease consumption of meat” for advice to choose lean meats.

That revised edition also carried a sentence most people have never read. In its own front matter, the committee wrote that “the value of dietary change remains controversial, and science cannot at this time insure that an altered diet will provide improved protection from certain killer diseases such as heart disease and cancer.”

That’s the committee telling you, in its own report, that it couldn’t promise the thing it was about to recommend to two hundred million people. Three years later the USDA and the Department of Health published the first Dietary Guidelines for Americans and it stopped being advice. School lunches, hospital menus, government food procurement, the lot. Two years after that, ours said the same thing.

Did the dietary guidelines cause the obesity epidemic?

No, not on their own. The honest problem is that nobody measured Australian adults properly until 1995, thirteen years after the guidelines came in, so there’s no clean before-and-after. What we do have is two measured surveys on the same basis. In 1995, 18.9 per cent of Australian adults were obese. By 2022 to 2024 it was 32.8 per cent, and severe obesity had gone from 4.9 to 12.9 per cent. Type 2 diabetes has followed the same line.

And the charts claiming heart disease exploded after the guidelines are wrong: age-adjusted heart disease deaths peaked in the late 1960s and have fallen by more than half since.

What happened is narrower, and it turns on the same distinction the trials did. Downstream, we built something genuinely world class. Stents, statins, clot-busters, fast ambulances, cardiac surgery that works: if you have a heart attack in this country today your odds are far better than your father’s were. Upstream, where men stop becoming metabolically sick in the first place, we went backwards for forty years. That’s the trade nobody agreed to. We got superb at pulling men out of the river and gave up asking who keeps pushing them in.

Plenty else changed over the same decades. Calories went up, portions grew, ultra-processed food took over the middle aisles, work got sedentary.

None of that gets the guidelines off the hook, but be precise about what they’re on the hook for. Not for causing the obesity curve, which nobody can show. For issuing a specific instruction to a whole country with a confidence the evidence never supported. Whether the whole grains men were told to build their diet on are actively harmful is a separate argument, and a longer one than this.

How much of that curve belongs to the guidelines and how much to everything else, I can’t tell you. Nobody can, and nobody will: you can’t randomise a country.

Following the guidelines made me fat while I ran 70km a week

I ate what I was told for forty-eight years. Oats for breakfast. Sandwiches at lunch. Pasta or rice at night. I was running 70 kilometres a week in a demanding corporate job, I weighed 80 kilos, and I was hungry from the moment I woke up. I paid a dietitian two hundred dollars an hour who put me on a low-fat, calorie-counted protocol and kept me where I was, only hungrier.

High carbohydrate intake, a training volume most men will never touch, and none of it mattered. My read is that it was what insulin does to a man eating that way, and I’ll call that a read rather than a proof, because one man is one man. Tim Noakes puts the practical version better than I can: if you have to exercise to regulate your weight, your diet is wrong.

In late 2014 I sat in a talk by Dr Stephen Phinney and changed everything. Six months later I was 65 kilos with completely different blood lipids. I’m 61 now, I’ve held that weight for eleven years, and my resting heart rate is 39.

I’ve since coached 895 men through the same reversal and the pattern is boringly consistent. Craig, a retired army officer at 55, had worked with a dietitian for years without losing a kilo. He went from 105 kilos to 90 in four months, body fat from 20.1 to 16.9 per cent, every marker on his blood panel improved. Rees, a general counsel in his forties, had cracked 100 kilos doing what conventional nutrition advice told him to, and dropped 12 kilos and 10 centimetres off his waist.

Neither of those men lacked discipline. They were following the model.

What should a man over 40 do if he followed the low-fat guidelines for thirty years?

Test it on yourself, and don’t take my word for it either. But start from the right place: you followed official health advice, issued by institutions that had put in writing that they couldn’t guarantee it would protect you. That’s compliance, not a character flaw.

It gets you off the blame. It doesn’t get you off the hook. Metabolic problems don’t usually resolve on their own, and you won’t know which way yours are heading until you measure them.

So run the test, and judge it the way the guidelines should have been judged: on the outcome, not the number. For four weeks, cut the bread, cereal, rice, pasta and sugar, and replace them with protein and fat you were told for forty years to be frightened of. Change nothing else: same training, same steps. Measure your waist on day one and again on day 28. If nothing moves, I’m wrong and you’ve lost a month of sandwiches. If it moves, you’ve learned something about your own body that no guideline could have told you, and you’ve learned it the right way round: outcome first.

Some men would rather not spend four weeks proving what they already suspect. Craig spent years with a dietitian and lost nothing; four months with me and he was 15 kilos down with a blood panel his GP couldn’t argue with. If you’ve got 10 kilos or more to shift, you’ve done the obvious things twice, and you want someone reading your food log every day rather than handing you a plan and disappearing, bring me three things: your waist measurement, your last blood panel, and an honest week of what you actually eat. Thirty minutes, no charge, and you’ll leave knowing what’s driving it and what I’d change first: book a session.

Frequently Asked Questions

What are the Ancel Keys dietary guidelines?

There’s no such document, and Ancel Keys didn’t write one. What he built was the diet-heart hypothesis, that saturated fat raises cholesterol and cholesterol causes heart disease, and that’s what the 1977 US Dietary Goals, the 1980 Dietary Guidelines for Americans and the 1982 Dietary Guidelines for Australians were built on. When people say “the Ancel Keys guidelines”, they mean the low-fat, high-carbohydrate advice that came out of his work.

Does lowering cholesterol mean you’ll live longer?

Not on its own, and that assumption is the mistake at the centre of the low-fat guidelines. The diet-heart hypothesis reasoned that saturated fat raises cholesterol, cholesterol causes heart disease, so lowering cholesterol would save lives. Two randomised trials tested the intervention that was supposed to do it. The Sydney Diet Heart Study and the Minnesota Coronary Experiment both lowered cholesterol as predicted, and neither produced a survival benefit. Moving a number a hypothesis cares about is not the same as improving the outcome the hypothesis exists to improve, and it’s worth checking which one a piece of health advice has actually demonstrated.

Was Ancel Keys wrong about saturated fat?

Partly, and how he was wrong matters more than the fact of it. His 1953 six-country chart showed a strong link between fat intake and heart disease, but data existed for twenty-two countries and the correlation didn’t survive them. When statisticians showed him that in 1957, he didn’t revise, he dismissed the critique. His later Seven Countries Study did link saturated fat to heart disease, but not total fat. Those are three separate claims and they get run together constantly: an association between saturated fat and heart disease is not the instruction “eat less fat”, and neither one establishes that swapping butter for vegetable oil makes you live longer.

Why did the dietary guidelines change in the 1970s?

A US Senate committee chaired by George McGovern published Dietary Goals for the United States in February 1977, and its lead scientific advisor argued that officials couldn’t wait for proof. The report’s revised edition said in writing that the value of dietary change remained controversial and that science could not at that time ensure an altered diet would protect anyone. It became official US policy in 1980 anyway, and ours followed in 1982.

When did Australia adopt low fat dietary guidelines?

Australia published its first Dietary Guidelines for Australians in 1982, two years after the United States, with the same instruction to eat less fat. Nutrition Australia’s Healthy Eating Pyramid went up the same year, cereals and breads in the “eat most” band, butter and margarine at the tip. No Australian trial had established the benefit before the advice went out, and I can’t point you to a local inquiry that tested it either.

What was the Sydney Diet Heart Study?

An Australian randomised trial that ran from 1966 to 1973 with 458 men aged 30 to 59 who’d already had a heart event. The intervention group replaced the saturated fat in their diet with linoleic acid from safflower oil and safflower margarine. Controls got no dietary instruction and no study foods. Cholesterol fell in the intervention group, which also had higher all-cause mortality (hazard ratio 1.62, p=0.051), cardiovascular mortality (1.70) and coronary mortality (1.74). Most of the mortality data wasn’t published until researchers recovered it and put it in the British Medical Journal in 2013. The margarine of that era carried industrially produced trans fat, and that’s the main objection to the finding.

Are seed oils bad for you?

The evidence is contested, and anyone giving you a clean yes or no is overselling it. Two randomised trials tested replacing saturated fat with linoleic-acid-rich vegetable oil, one in Sydney and one in Minnesota. Both lowered cholesterol. Sydney found higher mortality in the men who made the swap. Minnesota found no survival benefit, which is not the same as finding harm. What they establish together is narrower than you’ll read online: the marker moved and the outcome didn’t follow. For the men I coach, cutting processed carbohydrate matters far more than the oil question does.

Did the dietary guidelines cause the obesity epidemic?

Not on their own. Australian obesity climbed from 18.9 per cent of adults in 1995 to 32.8 per cent in 2022 to 2024, but calories, portion sizes, ultra-processed food and sitting down all rose over the same period, and a food supply shifting alongside a health trend doesn’t show that one caused the other. What the guidelines did was issue a confident instruction that the evidence didn’t support. How much of that curve belongs to them, I can’t tell you, and neither can anyone else.