Is saturated fat bad for your heart? Canada’s Food Guide says so. Here’s what the science says
Small, unremarkable and buried in the pages of a niche medical journal, the study appeared to show that indulging in cookies and brownies was OK — healthy, even — if they were made with vegetable fats.
Two days after it was published in December 2000, a full-page ad ran in the New York Times declaring it “groundbreaking.”
“The debate is over,” the ad said in bold print. “Margarine wins. Spread it around.”
Conducted on mainly white subjects from Texas in the late 1990s, the study showed that when 46 families ate baked goods made with margarine — rather than butter — their cholesterol was lower.
Praised at the time, no one paid much attention to its roots in a marketing campaign — it was partly funded and partly designed by the same margarine trade group that paid for the ad. As well, the study neglected to mention that the trans fats in the pale, dairy-less spread fed to these families were by then raising red flags among scientists as a potential cause of heart disease.
Despite these troubling facts, this 20-year-old study is one of many that forms what Health Canada calls its most “convincing” evidence for advising Canadians to reach past butter and grab a jug of canola oil.
The 2019 Food Guide, released to great fanfare earlier this year, said its advice was based on “the best available evidence” and that its creators tried to keep Big Food’s influence at bay.
Dr. Alfred Aziz, director general of Health Canada’s Office of Nutrition Policy and Promotion, said in an interview with the Star that the food guide’s advice to all Canadians to eat less fat from butter and more from vegetable sources, will reduce cholesterol and reduce our risk of heart disease. Health Canada said the recommendations on fat are “consistent with those of other countries — such as Australia, New Zealand, the United Kingdom and the United States.”
The Star looked at the government’s best evidence underpinning its advice on fat, a nutrient that forms the membrane of every cell in the human body, and found that some of these scientific studies were funded by the food industry and, in one case, the tobacco industry. The results of these studies raise questions about whether the sweeping advice to reduce saturated fat — found in butter, cheese, steak — and replace it with unsaturated fat is suitable for all Canadians. Design flaws and miscalculations in the studies have left cracks in their reliability.
This should come as no surprise to anyone who tries to keep up with the latest in nutrition science. What was truth last year was debunked last week. Eggs are good, then eggs are bad, and now, eggs are good again. Absolute truths about how food affects our bodies are tough to find, even in clinical trials, leaving the public wondering what to eat and what — if anything — to believe.
So, as Health Canada trumpets the food guide and its publication in 31 languages, a growing number of experts are disagreeing with the guide’s fat advice.
The most recent criticism comes from 16 nutrition scientists, including Canadian Benoît Lamarche, Chair of Nutrition at Laval University. Their paper, published in July, is a rebuke of the World Health Organization’s 2018 recommendation, which Health Canada echoes in its guide — to reduce saturated fat and replace it with unsaturated fat, saying it is “not based on evidence.” The criticismsaid the guideline fails to consider that different foods have different kinds of saturated fat, which have varying effects on the body.
The report also said the advice to reduce saturated fat had led to an explosion of processed, low-fat products, including margarine with industrial trans fats. Continuing to push this same advice could work against the guidelines and weaken their effect on chronic disease, the report said. “Scientific and policy missteps may have led to many unnecessary deaths globally, and lessons should be learned.”
The Food Guide is Canada’s trusted source of diet advice. It informs menus in public nursing homes and daycares and how nutrition is taught in elementary schools. It enables the food industry to use its recommendations to place claims on products.
Giving advice on what to eat has long been a hot-button issue. Speaking to the dairy lobby in July, Conservative party leader Andrew Scheer slammed the new guide, saying it is ideologically motivated, flawed and riddled with bias. If elected this fall he said he’ll subject it to a review. Critics, including Liberal Health Minister Ginette Petitpas Taylor, said Scheer is “spreading lies.”
The Star analyzed the best evidence that, in its technical documents, Health Canada said it relied on. Those documents show the regulator’s best evidence on fat comes from two kinds of studies: One group narrowly focused on how saturated and unsaturated fats affect certain “biomarkers,” such as blood cholesterol, that doctors look for to predict disease. The Texas study on the 46 families is one example. The others, considered the most important kind by scientists the Star spoke to, examined whether reducing saturated fat leads to fewer heart attacks, strokes and deaths.
These “clinical endpoints” are what really matter when it comes to eating for health, U.K. researcher Zoë Harcombe and other scientists say.
Nav Persaud, a doctor and scientist at St. Michael’s Hospital’s Li Ka Shing Knowledge Institute, said there is a danger in placing too much weight on the narrower “biomarker” studies as findings can be “spurious,” he said, and lead to faulty public health recommendations. A 2010 report, commissioned by the U.S. Food and Drug Administration, advised the regulator to proceed with caution when using such biomarker studiesas their findings don’t always serve as an appropriate substitute for clinical endpoints. Cholesterol levels, for example, don’t always predict cardiovascular outcomes, the report found.
But the studies of clinical endpoints, the Star has found, are outdated, have flaws and were conducted overwhelmingly on sick, disabled or institutionalized men who smoked. Women, who participated in just a handful of these trials, were either postmenopausal or inpatients in what researchers called “mental hospitals.” Some of these patients were medicated.
Most of these studies reported a lowering of cholesterol. All but one, however, showed no evidence that modifying saturated fat — either by reducing, or reducing and replacing — cuts the death rate. Some of these studies showed that modified fat diets may even be harmful: Men who ate margarine, or took daily dosesof corn oil or fish oil supplements died earlier, had more heart attacks or were at greater risk of death and disease, leading researchers of one study to speculate their diet was “toxic” and researchers of another to warn their diet could be “possibly harmful.” The one study that showed a reduction in death rate in men has been discounted by many scientists because its authors didn’t take the necessary steps to ensure the outcome wasn’t biased — skewed by anything from error to bad luck that can derail a trial and distort its results. It also failed to consider that many of its participants were on psychotropic drugs.
Aziz told the Star there is strong evidence underpinning the guide’s recommendations. Though the regulator agrees “no biomarker is a perfect substitute for a clinical endpoint,” Aziz said Health Canada’s reliance on the narrower studies that looked at cholesterol, for example, is an accepted one for creating public diet advice.His office applies this same strategy to all nutrients it studies, he said, including salt and sugar. Aziz said the regulator is acting on the best research available and that the alternative to give no advice is “not acceptable.”
The hypothesis that saturated fat in marbled meat and cheese is bad for the heart took root nearly 70 years ago with a University of Minnesota physiologist best known as a promoter of the Mediterranean Diet. Ancel Keys studied a variety of food related topics, including cholesterol and coronary heart disease.
During a visit to Europe in the 1950s, he noticed that men who ate “meagre diets” of fruit and olive oil had healthier hearts than men who ate “rich diets” full of “sugar, salt, fat, creamy stuff,” according to David Jacobs, Keys’ former colleague and Mayo professor at the University of Minnesota’s School of Public Health. Jacobs said those observations prompted Keys to make an educated guess: That saturated fats caused heart disease.
Over the next decade, he studied his “diet-heart hypothesis” among men in seven countries — the U.S., Italy, Finland, Greece, the Netherlands, Japan and Yugoslavia — and found an association between saturated fat and dying of heart disease. But Keys noted this wasn’t proof that one caused the other.
That proof could only be established by a tightly controlled experiment that examined those “clinical endpoints” — including heart attack, stroke, angina — in at least 100,000 healthy men across the U.S. for five years, according to a report from the American Heart Association committee of which Keys was a member.
That big study never happened.
Instead a much smaller “feasibility” study was conducted on 2,302 men to see if scientists could even conduct a larger trial. The results were inconclusive and the study had “several problems,” according to a report published in 1968 in the scientific journal Circulation.
One of those problems was self-reporting — the men kept week-long food diaries and, the article said, those subjective logs “underestimated” what and how much they ate.
Other studies from the time were also problematic.
A study published that year found that lean fish, meat, clear soup and soybean oil had no effect on men recovering from heart attack. A 1965 study, also on men with heart disease, found that those who drank three daily doses of corn oil had lowered cholesterol levels yet more of them died within two years. And a 1969 study conducted on disabled war veterans who ate a more plant-based version of the American diet — ice cream made with cottonseed oil rather than cream, for example — suffered slightly fewer cardiac events, such as heart attacks or angina, but they died at a higher rate from other causes. The authors said these deaths raised “the very important and difficult question” of whether future diet studies rich in vegetable oils should run longer than eight years.
The one study that appeared to prove Ancel Keys’ theory was published in 1968. Of roughly 400 men recovering from heart attacks, those who swapped their animal fats for vegetable oils had, after five years, a lower relapse rate than those who didn’t. But it’s unclear what the other group actually ate. As well, the information of what the men in the vegetable oils group ate was based on an analysis of the diets of only 17 selected men.
Zoë Harcombe, the U.K. researcher, recently re-analyzed the same literature on saturated fat, including these old studies that so many other scientists have pored over, finding something entirely different: no support — then or now — for Keys’ hypothesis. But what surprised her is that this hypothesis became widely viewed as fact. “It seems incomprehensible that dietary advice was introduced to hundreds of millions of people” given the “contrary” results from experiments studying just a “small number of unhealthy men,” she wrote in her 2015 review, published in Open Heart, a British Cardiovascular Society journal.
Christopher Ramsden, a U.S. National Institutes of Health researcher, likewise found no proof in the literature that replacement of saturated fat in the diet with a polyunsaturated fat lowered the risk of death.
His 2016 paper commented on the leaps made between what the science says and what ends up as guidance, saying that a better approach to givingdiet advice “might be one of humility” that highlights the “limitations of current knowledge.”
The conviction that butter,cheese and the salty, fatty edges of steak might kill us, was seeded into North America’s collective psyche at the time when a handful of American scientists and politicians decided that fat — specifically saturated fat — was to blame for an apparent epidemic of heart disease.
A report by Health and Welfare Canada around that time said that 50 per cent of deaths in Canada were caused by a hardening of the arteries. This 1976 report made the case that eating a diet that lowers cholesterol will reduce the risk of heart disease.
Its analysis went as follows: a known risk factor for atherosclerosis was elevated blood cholesterol. Studies on animals showed that diets rich in cholesterol and saturated fat caused blood cholesterol to rise, which led to this lethal hardening. Diets lower in these fats seemed to reverse that artery hardening. “Indirect” evidence in humans suggested the same thing may happen in people. Plus, studies had already demonstrated that certain types of unsaturated fats lowered blood cholesterol in humans.
Though the report noted that “studies within a country or community have failed to show a relationship between any component of the diet and the incidence of (heart disease),” it concluded there was enough information to take action.
That action was to issue diet recommendations similar to what the U.S. presented to the public in 1977.
At a public hearing that year, one influential researcher told Senator George McGovern, who was championing these new diet recommendations, that the science available was uncertain.
“I understand perfectly the position you are in, and I sympathize with it,” Pete Ahrens told the South Dakota senator. “My contention is…that this is a matter of such enormous social, economic and medical importance, that it must be evaluated with our eyes completely open…I am betting and I am hoping…for I have changed my diet to some degree, no question about it…But I have no conviction nor foreknowledge that what I am doing is prolonging my life or that of my family.”
Robert Levy, then director of the U.S. National Heart Lung and Blood Institute, said the science had yet to answer the most important question: Did changing the amount of saturated fat lead to fewer heart attacks or deaths?
McGovern was unswayed, saying that “senators don’t have the luxury that the research scientist does of waiting until every last shred of evidence is in.”
Those last shreds of evidence still might not be in.
That’s one reason why the fat recommendations today rest, in part, on those old studies. Another reason is that doing a new study is complicated — especially one that examines the important clinical endpoints in a population that’s big and diverse enough, and does so over a long enough stretch of time, to draw broad conclusions. Such a study could cost an estimated $300 million, scientists told the Star.
To fill in the incomplete picture scientists and policy makers, including Health Canada, rely on the narrower biomarker studies.
Single studies don’t have the heft to underpin public diet advice but when combined with others and processed through a “systematic review,” they become “more reliable evidence” and “reduce bias” according to Health Canada. The regulator calls these reviews the “gold standard” approach to evaluating all evidence on a complex topic.
Systematic reviews, also called “expert summaries,” aggregate data from many studies into one clean, coherent message.
Researchers set their criteria, sifting through sometimes tens of thousands of studies. The goal is a streamlined collection that seeks to answer a question, such as: Is saturated fat unhealthy?
After weighting the data from each individual study, according to size, duration and the number of important outcomes it reveals, they toss the fruits of their labours into something like a giant scientific juicer.
Health Canada looked at 15 such systematic reviews.
Some did not find improved health outcomes. Others focused on issues that didn’t make it into the guidelines. Two of these reviews examined those important clinical outcomes, finding that modifying dietary fat lowered the risk of heart disease by 14 and 19 per cent respectively.
When you look at what went into these two scientific juicers, you might think twice about drinking what comes out.
The author of one of these two reviews said the studies were “not ideally valid” though plentiful. After her review, she graded their collective evidence that showed a reduction in cardiovascular events, such as heart attack and stroke, as “low quality.” The author of the other said that design faults, possibly inaccurate results and mixed evidence from the studies he analyzed made it impossible to distinguish if the risk reduction he found was a result of eating less saturated fat or eating more of a certain type of unsaturated fat.
All of the studies that went into these two systematic reviews had flaws.
A 1979 study on patients in two Finnish mental institutions — the only study to report its modified fat diet reduced deaths — failed to account for the considerable psychotropic drug use among some of its participants. A 1968 study on disabled war veterans living in Los Angeles institutions had a high dropout rate. Some subjects found the food “objectionable” and its authors mistakenly placed more and heavier smokers into the non-dieting group, creating a plausible reason, other than eating butter, for why some of the men died.
Some of the studies didn’t keep proper records on participants’ smoking habits. And, a majority of participants smoked. They were also ill. Most of the studies were conducted on around or fewer than 1,000 men who were either recovering from a heart attack or had been institutionalized for disability or mental illness. And most also relied on some degree of self-reporting.
Two of the studies have since been re-evaluated, producing different results.
When it was originally published in 1989, The Minnesota Coronary Survey found that a modified fat diet lowered cholesterol but did not reduce deaths or cardiac events in around 9,000 men and women living in mental institutions. But, after recovering a trove of unpublished, original data from the trial, U.S. researcher Ramsden co-authored a 2016 re-analysis that showed an increased risk of death among older men whose cholesterol was lowering. Ramsden concluded his findings “add to growing evidence” that outdated research not only overestimates the benefits of replacing saturated fat with vegetable oils that contain linoleic acid but underestimates its “potential risks.”
David Jacobs, Keys’ former colleague, said he remembers when the Minnesota’s authors tabulated their results. “I don’t think anyone was hiding anything,” he told the Star. Jacobs said that among many problems, the study was too short and there were too few participants that stuck around long enough to detect a change in anyone’s health. “It was a flawed study that didn’t work…I would ignore it.”
The second revisited study was the Diet and Reinfarction Trial (DART). Published in 1989, it found that men who ate fatty fish a few times a week or took fish oil supplements, which are rich in unsaturated fats, had the lowest mortality even though they suffered the most heart attacks. A follow-up a decade later found that benefit had disappeared after two years. In 2003, a co-author published results from a similar trial that resulted in conflicting findings. DART-2 also found that the fish diet advice was “associated with a significant increase in sudden cardiac death.” The co-author also concluded that “it should not be assumed that the same nutritional interventions are equally appropriate” for all people at high risk of heart disease.
As well, the original DART had problematic funders.
One of them was the Flora Project, a 1970s campaign to sell margarine. The brainchild of a London, England, public relations firm and executives of Unilever, then makers of Flora margarine, it aimed to give margarine a health halo by using science to show it had cholesterol-lowering powers.
The DART’s second problematic funder was the Health Promotion Research Trust. Founded in 1983 by the British government, it had 11 million pounds of Big Tobacco money earmarked for medical research, provided the research didn’t study the use and effects of tobacco, according to a 1987 commentary written by Simon Chapman, an Australian academic. He also wrote that accepting this “blood money” was tantamount to “collusion” with Big Tobacco’s desire to appear to be concerned about health when all it wanted was good PR.
Lee Hooper, an author of one of the two systematic reviews, and a researcher with The Cochrane Collaboration, a British charity that parses health science for policy makers, said she noticed the “funding issues” while conducting her research. But, she said, the Cochrane said, “that you shouldn’t downgrade evidence for funding issues. So we didn’t.”
The DART wasn’t the only study to get money from industry.
Health Canada has said that in developing the 2019 guide, it refused meetings with industry, didn’t consult with experts who had been paid by it in the previous two years and excluded industry-commissioned reports from its review.
Two of the systematic reviews that Health Canada’s advice on fat relies on had a close relationship with industry.
Commissioned by industry and submitted to Health Canada, the regulator consulted the evidence from these reviews when putting together the Food Guide. Health Canada refused to reveal to the Star the identities of these petitioners or the data they provided, saying that its refusal was meant to ensure that in the future “industry is not discouraged from submitting information.”
The Star has learned that one of these reviews was commissioned and submitted by the now defunct Vegetable Oil Producers of Canada (VOIC). The trade group had hoped to claim its low-saturated fat vegetable oils reduced cholesterol, said David Ma, a University of Guelph nutrition professor who conducted the review on VOIC’s behalf before the group submitted it to the regulator in 2009.
Ma said it’s important for researchers and scientists to work with a variety of stakeholders to advance understanding of an important public issue. He sifted through the literature and found that certain types of unsaturated fat lowered blood cholesterol when they replaced saturated fat in the diet.
He said that while there’s growing recognition that food is complex and it’s not ideal to base public diet advice on studies that narrowly examine a single interaction, “it is still the best science out there.” We have a responsibility,” he said, to use it to guide the public.
Laval University’s Benoît Lamarche, a supporter of the current food guide, says that despite drawing their opinions from the same bank of evidence, scientists take either black or white positions on saturated fat’s potential harms and benefits. His personal view is more a “shade of grey,” he said, because at the moment there isn’t enough research to swing the pendulum in one direction. “We just don’t know yet.” He suspects the source of the saturated fat makes a difference to health, but more studies are needed to prove that.
He said that focusing public diet advice on a single nutrient, such as saturated fat, is meaningless to the eating public. It might be a more prudent strategy to simply steer them away from “ultra-processed foods,” like hot dogs, candy and cheesies.
Health Canada’s Aziz recommends Canadians avoid those foods but reiterated that it also delivers additional recommendations based on the best available science.
“We’re not going rogue here in Canada,” he said. “At the end of the day, we, government, cannot wait for the perfect, perfect answer. We have to be willing to work with the best available evidence at our disposal.”
Meanwhile, the evidence continues its evolution. Fresh off the nutrition-research presses earlier this month came a study published in the British Medical Journal. It found that vegans and vegetarians had a lower risk of heart disease but a higher risk of stroke compared to those who eat meat. The authors speculated it could be because of lowered cholesterol.
Data analysis by Andrew Bailey
Michele Henry is a Toronto-based investigative reporter. Reach her at email@example.com. Follow her on Twitter: @michelehenry
This content was originally published here.