[2017] Lore of Nutrition: Challenging Conventional Dietary Beliefs

by Tim Noakes

  In 1960, an AHA position paper castigated scientists – presumably Keys – for taking ‘uncompromising stands based on evidence that does not stand up under critical examination’. Undeterred, Keys and a close colleague, Jeremiah Stamler, got themselves appointed to the AHA’s nutrition committee. Within a year, through sheer force of personality, Keys had implanted his diet-heart hypothesis into the AHA, where it still beats strongly today, as it does in heart foundations all over the world, including in South Africa.

  In 1961, the AHA began advising men not to eat saturated fat and dietary cholesterol. It was a world first, Teicholz said, telling people to avoid saturated fat and cholesterol in order to prevent heart attack. ‘This was where it was all born.’ From there, the advice spread around the globe. The AHA told people to reduce their consumption of meat, full-fat dairy and eggs; to switch to margarine instead of butter; and to use vegetable oils instead of ancient fats such as lard and tallow. It advised people to cut back on animal-based foods in general and to switch to plant-based foods, mainly grains.

  This advice was hardly logical when one considered the trends in fat consumption at the time, Teicholz said. She showed how consumption of animal fat was already dropping in the US in the 1960s when the AHA made their pronouncements. At the same time, and since the early 1900s, the consumption of polyunsaturated vegetable oils had dramatically increased. This rise ‘perfectly paralleled’ the rising heart-disease rates, she said.

  Teicholz challenged the validity of the conventional dietary recommendation that polyunsaturated vegetable oils, such as soybean, corn and sunflower oils, are beneficial for health. Strydom and the HPCSA’s expert witnesses all actively promote vegetable oils as healthy components of a ‘balanced’ diet. Noakes and the Angels’ evidence to the contrary suggested that this was another reason for the HPCSA’s case against him. The vegetable-oil industry is a powerful and influential lobby worldwide. It isn’t about to give up its lucrative cash cow without a fight.

  Teicholz showed that these oils did not even exist as foodstuffs until 1911, the year in which industry introduced the first vegetable-oil food product. Consumers in the US knew it as Crisco, a hardened form of vegetable oil that was meant to replace lard. Research going back to the 1950s documents a long list of adverse health effects from these oils, including increased rates of cancer and inflammation. Still, Keys’s beliefs about fats prevailed. In 1961, he even featured on the cover of Time magazine as ‘the most important nutrition scientist of the 20th century’.

  Teicholz spent much of her evidence describing how Keys came up with the diet-heart hypothesis and how he then needed research data to support it. As Harcombe had done, she showed how fatally flawed his Seven Countries Study was from the start, because he had cherry-picked the countries that were included. Good science, she said, requires randomisation – a way of selecting things randomly to avoid bias. In selecting countries for his study, Keys did not use randomisation. He deliberately avoided countries such as Switzerland, Germany and France, where he knew that people ate lots of saturated fat, yet had low rates of heart disease. These countries would have ruined his findings on saturated fat, Teicholz said. She once asked Keys’s right-hand man, Henry Blackburn, why they had avoided those places. Blackburn said that Keys had ‘just a personal aversion to being in those countries’. ‘I think it is fair to say that the countries in the Seven Countries Study were cherry-picked,’ Teicholz told the hearing.

  Other problems with the study included data ‘inconsistencies’ that Keys could not resolve. One error is particularly ‘emblematic’, said Teicholz. Keys studied men on the island of Crete who seemed to eat very little saturated fat and had very low rates of heart disease. Yet of the three study periods during which Keys collected data on the island, one fell during the month of Lent, when the islanders would have been religiously avoiding all meat, dairy, eggs and even fish. ‘The Greek orthodox fast is a strict one,’ Teicholz said. Keys would have therefore under-counted the amount of saturated fat the Cretans actually ate. He knew of this problem, but dismissed it without explanation. What he ‘found’ became the foundation for the Mediterranean diet as we know it today. This was just another in a litany of unscientific practices Keys used to ensure that the results of the study would demonstrate what he wanted.

  Despite all its flaws, the Seven Countries Study became ‘extremely influential’ and is still widely cited today, Teicholz said. But while large and seemingly persuasive, it is still only observational. ‘It’s a basic principle of science that observational studies can only show association but not causation,’ she said. Causation requires clinical trials, preferably RCTs. Teicholz described how governments around the world undertook large, multicentre RCTs on the diet-heart hypothesis throughout the 1960s and 1970s. They replaced animal foods with inferior food items, such as faux meat, margarine and soy-filled cheese. The US National Institutes of Health alone spent billions of dollars on these studies, trying – and ultimately failing – to prove Keys’s hypothesis. The NIH trials were ‘remarkably special’ in that they were conducted in highly controlled, in-patient environments where researchers served all meals to participants.

  Additionally, almost all of the NIH studies had ‘hard endpoints’, Teicholz said. Hard endpoints refer to ‘indisputable outcomes’, such as death, which cannot be contested. The diagnosis of a heart attack is another endpoint, but is ‘a little more disputable’, she explained, given the range in diagnoses among doctors. Many studies today use far less reliable, softer, intermediary endpoints, such as lipid markers like LDL and HDL cholesterol. There is disagreement about which of these best predicts heart-attack risk, said Teicholz. The early large, well-controlled RCTs looking at hard endpoints are therefore a valuable source of data.

  Altogether, the NIH-funded trials on the diet-heart hypothesis included more than 75 000 subjects, mostly men. They showed that while restricting saturated fats did successfully reduce total blood cholesterol, there was no impact on the ultimate outcome – whether people died of a heart attack – and mortality rates overall. For decades, those in positions of power and influence suppressed and ignored this data, said Teicholz.

  In recent years, researchers around the world have gone back to examine the data and there are now more than a dozen published meta-analyses and systematic reviews of the evidence, she said. Nearly all have concluded that saturated fat and dietary cholesterol do not cause death from heart disease. Yet if you read 10 000 nutrition papers, as Teicholz has done, you will find that they all telescope back to the Seven Countries Study. Keys’s research was thus the ‘Big Bang of modern nutrition studies’.

  Teicholz introduced – and demolished – another key element in the case against Noakes: the role of sugar and other carbohydrates in a healthy diet. Alessandro Menotti, one of Keys’s fellow project leaders, headed up the Italian part of the Seven Countries Study. Years later, in 1999, Menotti went back to re-analyse the dietary data. He found that the food that best correlated with heart disease was not saturated fat, but ‘sweets’, Teicholz said. She interviewed Menotti to ask how that finding had escaped Keys. He told her that the Seven Countries Study leaders ‘did not know how to treat [sugar]. We reported the facts and had some difficulty explaining our findings.’

  Keys clearly understood that any hypothesis associating heart disease with sugar would compete with his own, Teicholz said. He also knew that only one was likely to be right. He therefore did what he always had when faced with inconvenient data: he ignored it and went on the attack, suggesting that all those who promoted the idea that sugar caused heart disease were financially motivated. Or ‘just plain wrong’.

  Teicholz described how British physiologist and nutritionist Dr John Yudkin, one of the most prominent proponents of the sugar hypothesis in the 1970s, fell afoul of Keys. Yudkin was a professor at Imperial College, London, and author of Pure, White and Deadly: How Sugar Is Killing Us and What We Can Do to Stop It. Keys dismissed Yudkin’s theory as ‘a mountain
of nonsense’ and accused him of being motivated by unnamed financial backers. Yudkin eventually paid a heavy price for going head to head with Keys. As Ian Leslie noted in The Guardian in 2016: ‘Prominent nutritionists combined with the food industry to destroy his reputation, and his career never recovered. He died, in 1995, a disappointed, largely forgotten man.’10 (Big Sugar loomed large in the wings of the case against Noakes, as Russ Greene’s 2017 exposé showed.11)

  Teicholz noted that Yudkin’s experience was not uncommon. Others who challenged establishment dogma faced similar vilification. The whole scientific community, not just Keys, tried to diminish critics by calling them names, dismissing the quality of their work, exaggerating errors, and denigrating their opinions as strange or bizarre.

  Teicholz drew parallels with the HPCSA’s prosecution of Noakes. ‘You can’t walk down a half block with Professor Noakes without someone calling out “20 kilograms, Professor”, “15 kilograms, Professor”,’ she said. Many people have lost weight on the Noakes diet, she pointed out, and there is a ‘whole body of research’ that supports his advice.

  ‘What, then, is the response by experts?’ she asked rhetorically. ‘To deny his work, to make fun of him, to pretend his work is full of errors, instead of reckoning with him and saying: “Okay, here are a number of observations that our hypothesis does not explain. We need to explain it.”’ The latter is what one expects of good science, said Teicholz, but it does not happen in nutrition science. That’s due mostly to the hostile and aggressive way in which contesting views are treated. It’s an attitude that goes back to the formative days of the field, to Keys and his colleagues. The bullying and denials have prevailed for decades, Teicholz said, with the result that few know that Keys’s diet-heart hypothesis is ‘the most tested hypothesis in the history of nutrition and disease’, and that ‘the results were all null’.

  The consequences for public health have been nothing short of tragic, Teicholz said. It’s quite likely, she argued, that by shifting towards a greater consumption of grains and other carbohydrates, the US guidelines have been a major contributor to the pandemics of obesity, diabetes and heart disease.

  Teicholz also presented evidence on the safety and efficacy of low-carb diets. There have been more than 74 RCTs, virtually all on Western populations, including at least 32 that lasted six months or longer, and three that lasted two years. (RCTs are considered the gold standard when it comes to judging whether there are any adverse side effects.) These trials have all established the efficacy of the low-carb diet for fighting obesity, diabetes and heart disease. Some official bodies are now in fact taking notice of the risks of low-fat diets, Teicholz said. In 2015, for example, the Heart and Stroke Foundation of Canada lifted the cap (as a percentage of calories) on saturated fats.

  Teicholz also addressed common criticisms of low-carb diets. One is that certain ancient civilisations ate high-carb diets, yet did not have high rates of obesity and diabetes. These included the Japanese, who ate rice, and the Egyptians, who ate lots of bread. The criticism is reasonable enough, Teicholz observed. Experts don’t know exactly why carbohydrates are now driving disease more than they did in the past. ‘We do not know if it is some combination of total carbohydrates plus sugar that has an especially negative metabolic effect,’ she said. ‘We do not know if it is that we have changed the way we produce wheat. We do not know if it is something about food processing, or if it is vegetable oils plus carbohydrates … We really do not know. All we know is that if you restrict carbohydrates, you see benefit.’

  Bhoopchand seemed almost overwhelmed by the sheer volume of Teicholz’s evidence. His cross-examination was uncharacteristically brief. Apart from a couple of questions, he left the bulk of her evidence unchallenged. He made a valiant attempt to undermine Teicholz on South Africa’s paediatric guidelines by suggesting that her expertise related only to the US dietary guidelines and not to those of the World Health Organization or to paediatric advice in South Africa. Teicholz dismissed that out of hand. The US dietary guidelines are considered the global gold standard, she said. They have been ‘exported, imitated and copied all over the world’ by nearly all Western countries and also by the WHO. ‘You will find that most guidelines internationally mirror the US dietary guidelines,’ she told Bhoopchand.

  Bhoopchand then tried to interpret her evidence and personal views on nutrition as contradicting Noakes’s tweeted advice. ‘What I will not stay away from asking,’ he said, ‘is that [in] your evidence you do not personally recommend a diet high in meat and fat, do you recall that?’ In reply, Teicholz said that she is not a nutritionist and does not make recommendations, but rather talks about what the evidence supports. When Bhoopchand then suggested that she avoids meat and high fat in her own diet, Teicholz told him that her diet is ‘generally low in carbohydrates, and higher in fat’. He pushed her to concede that while her diet is ‘higher in fat’, it is not high fat. Teicholz responded that ‘high fat’ could mean anywhere from 40 to 90 per cent fat. ‘I am in that range,’ she said.

  With that, Bhoopchand said that he was done. It surprised everyone, especially Adams, who thought that she had misheard him. ‘I was just stunned for a moment,’ Adams remarked once he had taken his seat. Teicholz was equally surprised when, after having dismissed The Big Fat Surprise as the work of ‘a mere journalist’ just a few days earlier, Bhoopchand asked her to autograph his copy.

  Caryn Zinn gave her evidence on the final day of the hearing, 26 October. As a practising dietitian and a nutrition academic, Zinn was able to provide a two-pronged perspective. Like Harcombe and Teicholz before her, she aimed her evidence at the pillars of the charge against Noakes, and effectively undermined them without making a single major concession under cross-examination.

  Zinn told the hearing that three things embarrass her these days: as a university student, she never questioned what any of her lecturers told her about the benefits of diet and nutrition; then, as a university lecturer, she told her students that low-carb diets were dangerous; and later, in her private practice, she prescribed low-fat diets to adults and children for 15 years.

  For the past five years, she has used LCHF in her practice for adults and children with no adverse side effects. Zinn said that she prefers to talk about low-carb, healthy-fat – rather than high-fat – foods. LCHF is about eating ‘real’ food, she said. Like Noakes, Harcombe and Teicholz, she believes that when people eat real food, they tend naturally to eat fewer carbohydrates and more fats.

  Zinn said that she had been loyal to the conventional LFHC dietary guidelines until she experienced her own ‘Damascene moment’ in 2011. That was when an AUT colleague, Professor Grant Schofield, asked her opinion on LCHF research, some of which she had not come across before. Zinn thought she would easily ‘set [Schofield] straight’ with research that refuted any benefits of LCHF. She expected to be able to tell him to ‘take your low-carb diet and go somewhere else. I am the dietitian and I know because my lecturers taught in my dietetics degree that this [LFHC] is how it is.’

  At the time, Zinn believed that dietitians were best placed to give dietary advice. She was therefore ‘flabbergasted’ to discover that most of what she had learnt and thought she knew about diet and nutrition was wrong. ‘The evidence that led to mainstream dietary guidelines was largely observational, correlation-based research,’ Zinn said. ‘From a quality viewpoint, it did not compare to the solid, randomised controlled trials that were available to support low-carb, high-fat diets.’

  Zinn began to realise that there was good reason why patients in her own practice were battling to lose weight and keep it off: they were constantly hungry because they were eating too many carbohydrates and too little fat. In the face of this realisation, Zinn made what she described as ‘a logical, biological and scientific leap to LCHF’.

  Her research over the last five years has focused almost exclusively on the LCHF paradigm in public health for all ages. She and her postgraduate students have led studies
on hyperinsulinaemia that demonstrate how a substantial subset of the population with normal glucose curves (supposedly healthy) have elevated insulin levels. This could be the start of chronic disease, she said.

  In her review of the literature around LCHF research, Zinn said that she uncovered some ‘staggering outcomes’. One RCT in particular compared a low-carb, high-fat diet with a low-fat diet, and found that LCHF outperformed low fat in all metabolic parameters, including weight loss, HDL cholesterol and triglycerides (blood fats). Especially relevant was the fact that the LCHF diet contained triple the amount of saturated fat compared with the low-fat diet. ‘Even more staggering’ was that LCHF showed a substantial reduction in all inflammatory markers, while the low-fat diet showed an increase in some of them. ‘When you are in practice and you read this kind of research, you start thinking about your own ethics and morals,’ Zinn said. She now believes that for dietitians to be ethical, they must include all evidence-based dietary options in their advice to patients, not just the ones they believe to be true based on what they have been taught.

  The HPCSA’s expert witnesses had made it clear that one of the reasons for charging Noakes with giving ‘unconventional advice’ was because he did not promote cereals as good first foods for infant weaning. Zinn argued that there is no evidence whatsoever that cereals are good first foods for infants. She also dismissed the HPCSA experts’ contention that cereals fortified with iron are beneficial for infant weaning. She argued that the advice to make starchy foods the basis of a child’s main meals could, in fact, cause malnutrition. That’s because phytates (antioxidant compounds found in wholegrains, legumes, nuts and seeds, and that bind to certain dietary minerals) reduce the bioavailability of important nutrients, such as iron, zinc and the B vitamins.