by Tim Noakes
During my HPCSA trial in November 2015, the complainant, dietitian Claire Julsing Strydom, stated under oath:
Because everyone was going on what the evidence was saying. Everybody was waiting for this publication [the Naudé review] because we could not simply go ahead and make [a] statement about Prof. Noakes’s hypothesis or diet without looking at the evidence. So everybody, all these big organisations were waiting on the publication of this information before we could make any kind of media statement …
In any science … expert opinions is [sic] your lowest level of evidence and a systematic review and meta-analysis would be the highest level. So before any media statements could be made we had to get that information and all these associations were waiting on that. It is not like you are saying it. It is not like everybody joined together to now make a statement against Prof. Noakes. We were all waiting for the evidence to be published … We would not just want to say oh, we do not agree with it [Noakes’s opinion] just because of nothing. We had to wait for that research to be complete. So the, all these associations were waiting for this information, and that is what it was [all emphasis my own].
Who these ‘big organisations’ and ‘associations’ were and why they had expressed such an interest in the Naudé review became clear in press statements they issued immediately after its publication. They were the HSFSA, ADSA and HPCSA, among others.
I first became aware of the existence of the Naudé review in the report of the Centenary Debate published in the SAMJ in February 2013.2 There, UCT professors Naomi ‘Dinky’ Levitt and Krisela Steyn had stated that the SAMRC would soon be releasing its own paper on the LCHF diet. A former student then working for a leading medical insurance company subsequently confirmed in 2014 that a meta-analysis was indeed being prepared.
I was mildly surprised, as a meta-analysis of the effects of true LCHF diets had already been published the previous year. In the British Journal of Nutrition in May 2013, Nassib Bezerra Bueno and his colleagues showed that people eating less than 50 grams of carbs per day (the diet I was promoting) lost more weight than those eating higher-carb diets. What is more, important health markers showed greater improvement in those eating the low-carb diet.3 I could not understand why we would need another meta-analysis of the same data.
Unless, of course, the study was motivated by something other than the advancement of truth and science.
On 9 July 2014 I received a phone call from journalist Wilma Stassen, who informed me that she had received a press release from the HSFSA about a new University of Stellenbosch meta-analysis that was relevant to the Banting diet. At the time I was unaware that Stassen is employed as a journalist by Stellenbosch’s Faculty of Medicine and Health Sciences, and that she is, in effect, an embedded journalist within the faculty.
In her article, published the following day, titled ‘Noakes’s low-carb diet not healthier’, Stassen referred to me as a ‘celebrity professor’ and described the Banting diet as a ‘fad’. She quoted the study’s lead author, Dr Celeste Naudé of the Centre for Evidence-based Health Care at Stellenbosch University: ‘This study shows that when the amount of energy consumed by people following the low carbohydrate and balanced diets was similar, there was no difference in weight loss.’4
This finding would not have come as a surprise to the authors, especially Dr (now Associate Professor) Naudé, who teaches that all food calories are created equal and that, for weight loss, ‘calories in’ must always be less than ‘calories out’. Based on this belief system, how could Naudé ever have expected a difference in weight loss between two diets providing the same number of daily calories? So even before they began their analysis, Naudé and her colleagues had biased the study to ensure a predictable outcome, specifically that there would be no difference in weight loss between the LFHC and LCHF diets.
Then, to make certain that the Banting/LCHF/‘Noakes’ diet had absolutely no hope of coming out on top, they did not even look at the diet that I advocate. Instead, they studied diets with an average carbohydrate content of 35 per cent, knowing full well that the upper limit of the LCHF diet that I prescribe is between 5 and 10 per cent, providing between 25 and 50 grams of carbohydrate per day. For Stassen then to describe the studied (relatively high-carbohydrate) diet as ‘Noakes’s low-carb diet’ was doubly questionable. As an embedded journalist, was she writing what she had been told to report by the Faculty of Medicine and Health Sciences at the University of Stellenbosch?
Designing a study, the result of which is predictable before the study even begins, is not science. Rather, it is the scientific equivalent of match-fixing in sport. And, as in sport, if science is fixed, the question is, who is the ultimate beneficiary? Someone or some organisation/s is benefiting financially.
Stassen’s article continued: ‘Based on these findings the Heart and Stroke Foundation of South Africa, the Association for Dietetics in South Africa and other health groups are warning the public about the possible health risks associated with banting.’
On behalf of the HSFSA, its then chief executive, endocrinologist Dr Vash Mungal-Singh, who had recently stated that heart disease is a carbohydrate-driven disease (see Chapter 7), seemed suddenly to have changed her opinion to suit the changed circumstances: ‘Decades of research have shown the balanced diet to be safe and healthy in the long term, and along with a healthy lifestyle, is associated with a lower risk of heart disease, stroke, diabetes and certain cancers.’
Unconstrained by any real facts, she continued:
We do not have similar proof that a low-carbohydrate diet is safe and healthy in the long term, and some studies already point towards an increased risk of heart disease and death with low carbohydrate diets. Chronic diseases like heart disease, stroke, cancer and diabetes develop over many years of exposure to risk factors.
The follow-up of the trials included in the review is no longer than two years, which is too short to provide an adequate picture of the long-term risk of following a low carbohydrate diet. Based on the current evidence we cannot recommend a low carbohydrate diet to the public.
Instead, in the absence of any long-term RCT studies showing the value of a low-fat diet, and despite the findings of the WHIRCDMT, which once and for all disproved the safety and efficacy of the low-fat diet (Chapters 4 and 17), Mungal-Singh continues to burden South Africans with this false information.
ADSA spokesperson and soon-to-be-president Maryke Gallagher also gave comment for Stassen’s article, linking her and her organisation to the collective effort by individuals from the University of Stellenbosch’s Faculty of Medicine and Health Sciences, the SAMRC and the HSFSA to discredit me and the Banting diet.
The standout problem with the Naudé review, which all these expert commentators ignored, is that it was clearly designed to produce a predetermined outcome. Its design was then skilfully hidden from all but the most inquisitive readers.
The authors manipulated the analysis by including only those studies in which the energy intakes of subjects eating either the LCHF or low-fat diets were ‘isoenergetic’, i.e. the diets contained the same number of calories. For those like Naudé, who believe unequivocally in the CICO model of obesity and weight control, there can obviously be only one outcome: there should not be any difference in the amount of weight lost by people eating either LCHF or LFHC if the diets are isoenergetic.
The isoenergetic research design is devious because it negates the key advantage of the LCHF diet, which is to reduce hunger, thereby allowing one to eat fewer calories without being constantly hungry and grumpy – or ‘hangry’, as US adult and child psychiatrist Dr Ann Childers calls it. That is how I lost 20 kilograms. By not binge-eating because I was continuously hungry, I was able to cut my daily calorie intake by roughly 1 000 calories without any hunger or food cravings – indeed, without any conscious effort whatsoever. In contrast, the only way that many people are able to lose weight on the conventional LFHC diet is because they have the discipline to ignore the sens
ations of hunger and to eat fewer calories than their brains and bodies desire. Research during the Second World War showed just how difficult it is to sustain this popular ‘semi-starvation’ dietary approach to weight loss. In fact, Ancel Keys, the diet activist who introduced the world to the low-fat dogma in the 1950s, is almost as famous – or notorious – for his semi-starvation studies in the 1940s.5
In one experiment, 36 US military conscientious objectors were placed on a 1 600 calories per day diet that included only those foods that were still available to European civilians during the period of famine at the end of the war – namely, ‘whole-wheat bread, potatoes, cereals and considerable amounts of turnips and cabbage’, with only ‘token amounts’ of meat and dairy. According to modern dietetics, this prescription would be considered ideal for weight loss, as it is low in calories and very low in fat, with just 17 per cent of calories coming from fat. As US science writer and author Gary Taubes noted: ‘What happened to these men is a lesson in our ability to deal with caloric deprivation, which means, as well, a lesson in any expectation we might have about most current weight-loss advice, and perhaps particularly the kind that begins with “eat less” and “restrict fat.”’6
Over the six months of the experiment, in addition to a multitude of adverse physiological effects, these men developed what Keys labelled ‘semi-starvation neurosis’, characterised by ‘weakness and lack of energy (which are somewhat but not adequately covered by the item “tiredness”), general slowing down, sensitivity to cold, concerns with thoughts about food, and decrease in sociability’.7 After six months of semi-starvation, almost all subjects reported that they tired quickly, felt unsteady when walking and had sensations of being ‘weak all over’. In addition, a majority found it hard ‘to keep their mind on the job at hand’, felt ‘down-hearted frequently’ and were ‘frequently bored with people’ so that 50 per cent ‘preferred to be left alone’.8
I do not believe that it was the low-calorie content alone of the semi-starvation diet that caused these effects. I know of many people eating low-calorie LCHF diets who have not experienced these changes. Quite the opposite, they have been energised. Instead, I suggest that it was the extremely low fat content of the diet that caused these psychological effects. My point is that it is possible to eat fewer calories and not be hungry, provided the calorie-restricted diet is also high in fat.
A result of this unique, hunger-satisfying effect of the low-calorie LCHF diet – as described by Anne Stock and John Yudkin,9 Atkins10 and Westman, Phinney and Volek11 – is that when LFHC and LCHF diets are used in isoenergetic weight-loss studies, those eating the LCHF diet must ingest more calories than are required to satisfy their hunger. In other words, the design of the experiment forces them to eat more calories than they need. As a result, they will lose less weight than they would have had they ingested the fewer calories that their satiated brains now found acceptable. Alternatively, if those on LCHF eat only the number of calories their brains direct them to eat, then to make the diets isoenergetic, those on LFHC must eat fewer calories than their brains desire, with the result that they would be perpetually hungry, experiencing a measure of the semi-starvation neurosis described by Keys.
Typically, the authors of isoenergetic studies do not appreciate this critical difference between the LFHC and LCHF diets. And why should they? They ‘know’ that hunger is driven by calories. Therefore, if both groups are eating the same number of calories, then both will be equally hungry or equally satiated. Unfortunately for them, science and experimentation with LFHC and LCHF diets show their theory to be incorrect. Furthermore, LFHC proponents never consider that their balanced isoenergetic diets, containing at least 45 per cent carbohydrate, might actually be driving hunger. Hence, they needlessly drive calorie consumption beyond what is necessary, thereby preventing weight loss and promoting weight gain and obesity.
Stassen did at least have the decency to quote my opinion on this in her article:
But Noakes argues that a high fat and protein diet has been known to reduce hunger, leading to less food intake and thus less energy intake. He argues this diet is the easiest to follow. ‘Unless the diet takes away hunger it will not produce the change in lifestyle necessary to sustain weight loss,’ Noakes told Health-e News. ‘Low-carb (diets) take away hunger in a way that no other diet does, that’s why it is the easiest diet to follow and the most effective.’
Noakes says that particularly favourable results are achieved on the low-carbohydrate diet in people with insulin resistance (pre-diabetes) …
Another way in which the Naudé review was designed to produce a predictable outcome unfavourable to those of us promoting the LCHF/Banting diet was that it conveniently failed to investigate true low-carbohydrate diets. Instead, the range of carbohydrate intakes in the ‘low-carbohydrate’ interventions was 90 to 200 grams per day, comprising an average of 35 per cent of ingested calories. According to the LCHF/Banting definition, these constitute high-carbohydrate diets, because the carbohydrate content is way in excess of the range of 20 to 50 grams carbohydrate per day that we prescribe for people with significant weight problems and higher levels of IR.
Indeed, a key point that few people, other than those of us who have lost significant weight on the LCHF diet, understand is the exquisite sensitivity of weight loss to the level of carbohydrate intake in people with varying degrees of IR or appestat (the region of the brain that is believed to control appetite) malfunction. Figure 6.1 shows that those who are insulin sensitive (bottom line) can probably ingest carbohydrates over a wide range (0 to 500 grams per day) without gaining weight. But as they become progressively more insulin resistant, so increasingly fewer carbohydrates ingested daily will cause them to become overweight (second line from the bottom), obese (third line) or morbidly obese (top line).
Figure 6.1
A theoretical model to project changes in body mass index (kg/m2) with changes in daily carbohydrate intake (grams/day) in people with different degrees of IR
The notable feature of this graph is the response to changes in dietary carbohydrate intake for a person with severe IR and appestat malfunction (top line). Should that individual decide to reduce her daily carbohydrate intake from over 400 to about 200 grams per day (horizontal axis), her body weight (vertical axis) will stubbornly refuse to change. She will be as morbidly obese eating 200 grams carbohydrate per day as she was when she ate 400 grams per day. But should she reduce her carbohydrate intake to less than 50 grams per day, the miracle occurs and dramatic weight loss begins.
Our current explanation is that a dramatic reduction in carbohydrate intake reduces blood insulin concentrations, reversing the persistent hyperinsulinaemia that is the defining characteristic of IR. But there are probably other factors as well. For example, the LCHF diet removes hunger, so the desire to eat excess calories is reduced, and this effect might be independent of the insulin effect. Removal of sugar and grains may also reduce the addictive drive to eat. Furthermore, the dramatic change in the nature of the foods eaten on LCHF may beneficially alter the gut microbiome in such a way that it increases weight loss, for we now know that the bacteria in our gut ‘talk’ to our brains!
Recall how Billy Tosh lost 83 kilograms in 28 weeks the moment he reduced his carbohydrate intake to about 25 grams per day and removed all sugar from his diet (see Chapter 5, Figure 5.1). He ascribed at least some of the effect to the removal of addictive foods from his diet.
From this we can infer that daily carbohydrate intakes well above 50 grams per day – the studies included by Naudé and her colleagues in their meta-analysis – are unlikely to have any real effect, particularly in those who are seriously overweight as a result of significant IR or appestat malfunction or, more usually, both.
Design flaws aside, it was the sweeping statements made by the representatives of the various health associations in Stassen’s article that really made me despair of the authors’ intentions. The implication that the Naudé review proved t
hat the LCHF diet was dangerous and could cause increased rates of heart disease and other dread diseases was misleading. As Stassen should have pointed out, the Naudé meta-analysis did not ever set out to measure whether the LCHF diet was ‘dangerous’. Instead, it measured changes in surrogate biological markers, such as blood LDL cholesterol concentrations, none of which can quantify danger. Also, the study found no difference in any of the surrogate markers that might have been interpreted as indicators of danger in people eating either ‘low-carbohydrate’ (according to their definition) or LFHC diets. Most interestingly, there were no differences in LDL cholesterol concentrations between those eating ‘low-carbohydrate’ and those eating LFHC diets.
The scientific conclusion should have been that, because there was no increase in LDL cholesterol concentrations (their marker of danger) in those on the ‘low-carbohydrate’ diets, according to their diet-heart hypothesis, the ‘low-carbohydrate’ diet could not possibly be considered dangerous. This is not how real science is meant to work. If there is no difference, there is no difference, however much you may personally wish that there was one.
On the same day, 10 July 2014, what appeared to be a second version of Stassen’s article was published on SA Breaking News under the sensational headline ‘New research shows Noakes diet no more than dangerous fad’.12 According to this report, 3 200 participants in 19 international trials were placed on either the Banting diet or a balanced weight-loss diet for a period of between three months and two years. Because the review found that weight loss was the same on both diets, the writer concluded: ‘This means Noakes’ diet, which drastically reduces the intake of carbohydrates in favour of fats, resulted in weight loss by limiting the amount of calories (kilojoules) the body took in. In effect, a person lost weight because they reduced their caloric intake rather than restricting the amount of carbohydrates.’