By Marika Sboros
The diet-heart hypothesis is a curious creature. To some scientists and physicians, the hypothesis that saturated fat causes heart disease is a zombie. Despite all the stakes they drive deep into its heart, it just won’t die.
To others, it’s written in scientific stone. That’s even when supporters, such as Harvard nutrition and epidemiology professor Walter Willett, call it “incomplete” and “overly optimistic” in its classical form.
The hypothesis hovered in the wings at a groundbreaking conference in Zurich, Switzerland in June. That’s when it wasn’t taking centre stage. Willett and others vigorously dispute the notion of any terminal hole in the diet-heart hypothesis.
This is the second of a two-part review of an event that was a unique collaboration between global reinsurer SwissRe and The BMJ. The venue was auspicious, Swiss Re’s Centre for Global Dialogue in Rüschlikon that houses its research arm, the Swiss Re Institute.
That raised the question why the world’s second-largest reinsurer is involved in contentious nutrition science debates. Swiss Re’s chief medical officer Dr John Schoonbee said lots about that.
“If more people die, we pay out more,” Schoonbee said. “If less people die, we pay out less. So, we want to keep people living longer, healthier lives.”
He and Swiss Re’s head of life and health products Canada Emile Elefteriadis elaborated in observations in The BMJ. Their article was part a series that The BMJ commissioned and published as part of the conference. Co-authors of scientific papers had to include at least one scientist or expert with vastly divergent views from the rest.
In their observations, Schoonbee and Elefteriadis pose and answer the question: Why would a leading global reinsurer be interested in nutrition?
Questionable policy decisions
They say that understanding of nutrition is often poor because of conflicting evidence and vested interest influence. This has led to “questionable downstream policy decisions”.
These decisions have “massively influenced the outcome of world health and individual wellbeing”. And the financial health of life insurance portfolios and national health systems, they say.
The right policy decisions would have measurably different impact on individual and societal health.
Click here to read: Teicholz: How low-fat diets can cause a heart attack
To that end, Swiss Re has developed a targeted life insurance model. The aim: to assess the population effect of nutrition strategies on health. Specifically, strategies to reduce longer-term risk markers associated with type 2 diabetes, prediabetes, and metabolic syndrome.
If all those affected followed such a strategy, the model estimates an annual reduction in total death claims of 13%, Schoonbee and Elefteriadis say. That translates to substantial gains for life insurers paying billions in death claims annually.
They also say that it would reduce premature loss of life from preventable, reversible conditions.
The conference organisers clearly hoped to clear some confusion with the four-day event.
The first two days covered Food for Thought, the Science and Politics of Nutrition. Click here for a full speaker list. The last two days covered Redefining Diabetes, Putting a “Chronic, Progressive” Disease into Remission. Click here for a full speaker list.
BMJ editor-in-chief Fiona Godlee opened the first day with the statement: “There is no such thing as a miracle diet, but there is such a thing as a miracle meeting. And in my view, this is it.”
Godlee wasn’t wrong. There were “miracles” of sorts, though not in the way she may have hoped.
Speakers with vastly different opinions on optimum nutrition and on both sides of the plant- versus animal-foods divide engaged on the same stage for the first time. And they debated major scientific controversies.
Chief among these: health claims around saturated fats – in other words, the diet-heart hypothesis. And low-carbohydrate, high-fat (LCHF) diets and their polar opposite low-fat, high-carb diets that dietary guidelines enshrine.
Canadian cardiology professor Salim Yusuf said that the hypothesis is the result of “brainwashing” by a “questionable study” many years ago. Yusuf is chair in cardiovascular disease at McMaster University Medical School in Hamilton, Ontario.
He was referring, of course, to Ancel Key’s Seven Countries Study that he launched in 1958 but only published two decades later. It has become “ingrained in our DNA”, Yusuf said.
It is also the bedrock on which experts built the dietary guidelines that the US launched onto an unsuspecting public in 1977. And the rest of the English-speaking world followed thereafter.
Yusuf is also principal author of the PURE study. PURE is the largest study investigating links between carbs, fats, cardiovascular disease (CVD) and death. Among conclusions: the more fat you eat, including saturated fat, the lower your risk of dying from heart disease. And the higher your carb intake, the more your risk of premature death rises.
That led to emerging consensus of sorts. Yusuf and Tufts University nutrition professor Dariush Mozaffarian were among those who agreed that there was no evidence to support current guideline caps on saturated fat.
That could have looked like the death knell for the diet-heart hypothesis. Instead, it highlighted another major conference theme: what evidence to trust – or not, as the case may be. And through them all ran the divide of plant-based diets versus animal foods.
Mozaffarian can look like he sits in the middle with his graphic on foods to prefer, eat moderately (in the middle) and avoid completely. (See right)
Among proponents of plant-based diets were Willett and Nita Ghandi Forouhi (Cambridge) in the UK who support the view that plant-based diets are healthier.
And that “much of the evidence (epidemiological) suggests that replacing saturated fat with polyunsaturated fats (including plant oils) reduces the risk of coronary heart disease”.
Jaundice view of epidemiology
On the other side were those promoting animal foods as healthier – and a jaundiced view of epidemiology. The US contingent included Prof Stephen Phinney, Dr Sarah Hallberg and science writers and authors Nina Teicholz and Gary Taubes from the US. From the UK were cardiologist Dr Aseem Malhotra, public health researcher Dr Zoë Harcombe and others.
The plant-versus-animal-foods divide was not as clear-cut as epidemiology versus randomised controlled trials (RCTs).
Harcombe said that she doesn’t automatically trust RCTs either. There should be far more reporting and promoting non-significant results that was currently the case, she said.
Harcombe cited Hooper 2011 on Reduced or modified dietary fat for preventing cardiovascular disease (CVD). The researchers found one significant result concerning CVD events – heart attack and stroke as well as a softer (subjective) measure of angina, peripheral vascular events and unplanned interventions, such as angioplasty – and 11 non-significant results.
Non-significant findings tell us “what we don’t need to worry about, which is far more than the one claim that we do (CVD events – which doesn’t hold either)”, Harcombe said.
Both epidemiology and RCTs came in for a sound thrashing via video link from Stanford University Prof John Ioannidis. Ioannidis said that 95% of nutrition research is “in bias”.
Epidemiology’s “typical recipe” of nutrition research leads to failure. He identified “large measurement error”, “cherry-picking among multiple hypotheses” and food industry influence as obstacles. Also “strong beliefs” – personal, religious and cultural – and “white hat bias”. That’s the term for “bias leading to distortion of research-based information in the service of what may be perceived as ‘righteous ends’“.
‘Hijacking’ of the evidence
RCTs also have “tremendous bias”, Ioannidis said.
Malhotra later told the conference that commercial interests have “hijacked” the best available clinical evidence. He also said that root cause of obesity and type 2 diabetes is “flawed science”.
Godlee soldiered on, looking for consensus. There was one big reversal in conventional thinking, she said: the demonization of fat. However, she expressed surprise that there did not appear to be any significant “mea culpa” from the scientific community that it had got things so wrong.
Taubes disagreed that there was any reversal. The debate has just “transformed” from saturated fat as bad to neutral, he said. But the issue is now: “neutral compared to what?”.
Saturated fat is supposed to be “neutral” compared to carbohydrates but not to polyunsaturated fats (PUFAs) and vegetable oils. This implies either that saturated fat is harmful or that PUFAs and vegetable oils are beneficial.
Therefore, the argument clearly hasn’t gone away”, he said.
Harcombe has shown that there is no robust evidence to show that replacing saturated fat with PUFAs is a good idea.
Silence a better option?
Teicholz said that where strong evidence was lacking, it would be better for experts to “remain silent” than make misguided claims. She suggested a way forward: the guidelines should contain a new low-carb advice option. That would avoid changes to other guidelines – and anyone having to admit to error.
In a conference review, Harcombe noted that “in any business strategy book, negotiation advice includes the principle “always give your opponent an out”. Teicholz’s suggestion does “exactly that”.
She has also said that there’s little chance of the “mea culpa” that Godlee suggests might be in order ever materialising.
In an email to me, Willett said that the classical diet-heart hypothesis is proven. And there are “layers of evidence” to support benefits of a largely plant-based diet (such as a traditional Mediterranean diet) versus a primarily animal-based diet, he said.
The evidence includes superior effects on blood lipids, long-term epidemiologic studies and RCTs. These show “significant benefit of replacing saturated fat with PUFA that includes both N-6 and N-3 fatty acids”.
“All the evidence is in this direction, none in the opposite direction,” he said.
For proof, he referred to his Nutritional Epidemiology textbook, now in its third edition.
“Replacing saturated fat with refined starch and sugar will not reduce risk of heart disease and could increase risk,” Willett said. Replacement, therefore, “is key”.
Meat and dairy fat not optimal?
At the core of much of discussion, high dietary glycemic load is a “serious problem”, Willett said. “However, it doesn’t mean that eating a lot of red meat and dairy fat is an optimal solution. There is a huge body of data of many types showing that it is not.”
He also described Ioannidis as “completely ignorant” in his presentation to the conference.
Ioannidis misrepresents how epidemiology is actually conducted and the results of studies, Willett said. “He makes his living by trashing other people’s work and has never done any serious research himself – the greatest COI (conflict of interest) of all.”
US specialist Dr David Diamond watched the conference on live-streaming and shared his impressions with me via email.
Diamond is professor of psychology and molecular pharmacology and physiology at the University of South Florida. He has a personal and professional interest in diet and cardiovascular disease. He lost weight and improved his triglycerides by cutting carbs and eating butter, eggs and red meat.
Diamond shares his knowledge in online lectures, hospital “grand rounds” and invited lectures at international medical conferences.
He delivered the keynote lecture on statins and heart disease at the World Congress on Diabetes & Obesity in Riga, Latvia in 2015. His lecture so impressed clinicians that they honoured him with their award for “outstanding contribution to science”.
Diamond has his share of detractors who support the diet-heart hypothesis. Here’s his rebuttal to a cardiologist who said his views on statins are “dangerous”.
Flaws in diet-heart hypothesis
Diamond described the diet-heart hypothesis as “flawed and antiquated”. Keys based the simplicity of the hypothesis solely on his personal beliefs, rather than strong empirical science, he said.
In one study, Diamond and 16 co-authors (all MDs and/or PhDs, including cardiologists) rigorously assessed the hypothesis that people over the age of 60 with high levels of LDL-C ( so-called “bad cholesterol”) would have a higher rate of death from heart disease and all-causes.
They found the opposite. Elderly individuals with the highest levels of LDL-C had an equivalent or, in most cases, a lower rate of death, than those with the lowest levels. That was another stake in the diet-heart hypothesis.
Diamond was strongly critical of University of Sydney nutrition professor Jenni Brand-Miller’s presentation.
Brand-Miller said that processed food had great value because it made women’s lives easier by “allowing them to go back to work”.
(Diamond wasn’t the only who found Brand-Miller’s opinions astonishing and even primitive from a scientific perspective. She was lucky to be speaking to a well-behaved audience.)
Brand-Miller further claimed that a pregnant woman could harm her growing foetus if she reduced her carbohydrate intake.
New Zealand-based dietitian and academic Dr Caryn Zinn found that claim extraordinary, as Diamond and others found it.
Diamond called it “perhaps the (conference’s) worst scientific moment”. Brand-Miller was “perhaps paid well by the food industry”, he said – a sentiment that others at the conference echoed. She probably didn’t help her crediblity much by also saying that we can’t blame the food industry for doing what we told it to do: produce low-fat, high-carb foods.
Diamond said that Brand-Miller’s opinions represented a “piffle quotient”. It reached its highest level when an MD in the audience stated that it was “well-known that the Atkins diet (with its reckless sanctioning of consumption of saturated fat) caused every form of disease known to Western civilization, including cancer, obesity and diabetes”.
Click here to read: PHC lays low-fat diets to rest
Diamond paid tribute to Hallberg for responding to “this twaddle” by stating the obvious: Carefully conducted research has shown that the Atkins diet has not been linked to these diseases.
The coup de grâce for Diamond was criticism from a panel of Hallberg’s recent Virta study showing an improvement of all primary biomarkers of health for diabetics and pre-diabetics with a very low-carb diet compared to usual care.
Hallberg showed that a personalised (vegetarian or omnivorous) low-carb diet could be of great value. Yet panelists “obsessed over the fact that subjects could choose their own treatment option, rather than the study being sterilized with a random design”, Diamond said.
He described Willett as “a force in the nutrition field by virtue of the fact that he’s a Harvard professor and has published a vast amount of (largely epidemiological) research to support his plant-based diet agenda”.
‘Weapons of mass cardiovascular destruction’
When it comes to saturated fat, Willett considers beef and butter to be “weapons of mass cardiovascular destruction”, Diamond said.
The conference gathering reminded him of his primary area of research: one of the most famous, intense neuroscience debates. These involved two great neuroanatomists of the 19th century who had opposing views of the brain’s structure.
Spanish anatomist Ramon y Cajal provided strong evidence that brain cells were unique and separate elements. Despite the overwhelming evidence in favour of Cajal’s position, Italian anatomist Camillo Golgi believed that the brain was composed of a continuous network of neural fibres.
Click here to read: Saturated fat causes heart disease? Bollocks – Kendrick
When they shared a Nobel Prize for their research in 1906, the debate continued in their acceptance speeches. In it, each one said he was right and the other was wrong.
Decades later, after both had died, the invention of the electron microscope ended the debate, Diamond said. Photographs of individual neurons unequivocally demonstrated that each brain cell was a separate element physically isolated from other cells. Therefore, the evidence proved Cajal right, he says.
Golgi refused to accept overwhelming evidence that his reticular theory of the brain was wrong. Similarly, Willett’s assertion that a plant-based diet is healthier than a diet with animal fat is based on “a myopic approach that ignores or dismisses contrary findings”, Diamond said.
The ‘Golgi’ of nutrition?
Willett will “prove to be the Camillo Golgi of the nutrition field”, he said.
Nutrition science needs the equivalent of an electron microscope photograph to end the debate on issues of saturated fat and the value of the low-carbohydrate diet as a diabetes and obesity treatment.
“Ultimately, the equivalent of the decisive photograph in the nutrition field would be recognition that food that raises blood sugar, not consumption of meat, cheese, butter and eggs in large part cause the diseases of Western civilization,” Diamond said.
From my vantage point, it was instructive to hear an “eclectic” group (as Godlee described it ) of global experts supporting plant-based diets, giving evidence and debating alongside animal-foods “enemies”. Harcombe noted in her review that the conference was essentially an opportunity that allowed “some foes to become friends or at least acquaintances”.
Many, if not most, divisions disappeared during the last two conference days. Hallberg, medical director at Virta Health, took centre stage to present her team’s research. It is the first clinically-proven treatment to show that type 2 diabetes really is reversible safely and sustainably without drugs or surgery.
There were no dissenting voices and rightly so as Hallberg did what good scientists do. She let the evidence speak for itself.
She also dismissed the diet-heart hypothesis. The saturated fat debate is “over”, Hallberg said. “We are pretty much done with it and we now need to get that message across.”
Hallberg also said that low-carb is just one of three effective options to treat and reverse type 2 diabetes that include bariatric (stomach) surgery and caloric restriction.
No diet is one-size-fits-all and patient choice “is key”, she said.