Why We Eat (Too Much) Read online

  Let me give you an example. Say I want to sell more of a product; it could be a drug or a type of food. I go about trying to sell more of the product by telling people that it has health benefits. Let’s say I want to sell milk and I want to prove beyond scientific doubt that if you drink milk every day for five years you will be able to run faster. First, I commission a study and ask a scientist, whom I am paying, to recruit twenty young people. He splits them into two random groups of ten. He measures how fast they can run 100 metres. The scientist then tells one group to carry on as normal and come back in five years (this is the control group). He asks the other group to drink a pint of milk every day. When five years have passed, the scientist asks both groups of volunteers to come back to the track and he times how fast each group, on average, can run 100 metres. When he analyses the results, however, he gets an unfortunate surprise. My theory has not been proven! There was no improvement in the running time of the milk-drinking group. They didn’t run faster.

  Figure 8.2 Comparison of statistics from the Seven Countries study and fifteen other countries: Figure 8.2A shows the correlation between total fat consumption as a percentage of total calorie consumption, and mortality from coronary heart disease in seven countries; Figure 8.2B shows the same but for all twenty-two countries where data was available. Source: J. Yerushalmy and H. Hilleboe (1957). Fat in the diet and mortality from heart disease; a methodologic note. N Y State J Med, 57 (14), July, 2346.

  Scientific journals will not be interested in publishing this experiment because it didn’t show anything of interest. So how can I prove my theory? Well, by cunningly performing the experiment ten times instead of once – yes, ten groups, each with twenty volunteers, and ten different scientists are paid. When we analyse the results, we find that in two of the experiments there was a difference in the time the milk-drinking runners took to complete 100 metres. Unfortunately, one of them showed that milk consumption made you run slower – but the other one showed what I wanted. In this group, the subjects that drank the milk had a much faster race time after five years. I have no legal duty to publish every single experiment that I perform, especially ones that don’t show anything. However, I now have one study that has an interesting result and therefore I ask the scientist to publish it in a scientific journal. The other nine studies I discard. When the publication date is due, I give my favourite journalist the tip-off that there is an article about nutrition and health that might interest him about to be published. The newspaper headline the next day reads: ‘Drinking milk makes you run faster!’

  This hypothetical example demonstrates that if you perform enough studies, then by random chance you will come across one whose results match what you want to prove. The milk industry, which in this case I am working for, is happy because people will be drinking gallons of milk for years to come. And the real trick? Because it is a five-year study it cannot be disproved by another lab for five whole profitable years.

  Historically scientists did not have to disclose whom they were working for, or who was providing the money for their experiments. Industry, whether it be food companies or pharmaceutical companies, could control the direction of scientific research and, by perfectly legal strategies of omission or selective analysis of results, skew the outcome in their favour.

  Once the direction of research is established, scientists, guided by more industry money, will follow – often on completely the wrong track. Layer upon layer of flawed research may well underpin our understanding of many areas of medicine. The more powerful the industry, or the more money it can make available, the more influence it has over scientific direction and ‘scientific fact’.

  Nutritional research is particularly sensitive to bias and flaws, unless it is part of a controlled experiment where the subjects can be constantly observed. Most nutritional research is epidemiological in nature, meaning that associations between the way people live and the diseases they develop are looked for. Unfortunately, these associations don’t always translate into establishing genuine causes. There are often other factors involved that have not been studied. For instance, in Ancel Keys’ Seven Countries study there seemed to be an association between the fat intake of a country’s population and heart disease. However, he did not comment on the fact that the countries with the lowest fat intake and the lowest rate of heart disease, Japan and Italy, also had the lowest consumption of sugar, and the countries with the highest reported rates of fat intake and heart disease, the UK (England and Wales) and USA, were high sugar consumers. An independent analysis of Keys’ studies several years after they had been published did, however, find a strong correlation between a particular food type and heart disease. You guessed it: sugar.

  Nutritional epidemiological research, which forms the bedrock of our dietary advice on healthy eating, has many weaknesses. Eating habits are gauged by notoriously inaccurate recall questionnaires and the reporting of diseases is often based on symptoms and not actual diagnosis. If you push this ‘loose data’ through selective statistical analysis and throw out the results you don’t like, then virtually anything can become the truth. And the ‘truth’ will be determined by the biggest industry backers.

  The Diet–Heart Controversy

  Fatty deposits in the walls of blood vessels were first described by the famous German pathologist Rudolf Virchow in the nineteenth century – he linked the deposits to heart disease. About 1 in 500 people suffer with a hereditary condition that causes cholesterol levels in the blood to be very high (called familial hypercholesterolaemia). The high cholesterol in the blood eventually causes bright-yellow fatty deposits underneath the skin of the eyelids and tendons (called xanthelasma). It was first recognized by doctors in the 1930s that people with these signs commonly died very early from heart disease (at the time it was a very rare condition). Once cholesterol could be tested in the blood – in 1934 – the first definite link between high cholesterol and heart disease was established, but this was only in the people suffering from the very rare hereditary condition. A further famous study fed rabbits (who would normally eat lettuce leaves) a high-fat diet and observed that they developed atherosclerosis (the precursor to heart problems) in their arteries. Other studies suggested that cholesterol levels in the blood could be altered in some people by changing their diet. Some people who went onto a low-fat diet lowered their blood cholesterol level.

  This was the historical background to the diet–heart hypothesis debate of the 1960s and onwards. It was suggested that because high cholesterol caused heart disease (in a rare genetic condition) and that cholesterol could be lowered in some people by a low-fat diet, then by association a low-fat diet could lower a population’s risk of heart disease. This seemed to be confirmed when Ancel Keys published his Seven Countries study, which seemed to strongly link dietary cholesterol and heart disease.6

  Unfortunately, for the proponents of the hypothesis, things were not that simple. They assumed that high saturated-fat intake was to blame for the rise in heart disease in the 1950s, though in fact red-meat consumption had already been decreasing for some time.7 In addition, their understanding that heart disease was due to a slow furring up of the arteries of the heart did not fit in with the sudden decrease in heart disease seen during the Second World War (when food, including sugar, was rationed). If the condition was chronic, then how could it be reversed so quickly? Finally, epidemiologists had failed initially to link smoking, which was at a post-war peak in the 1960s, to heart disease.

  In later years, it was shown that cholesterol had many different forms in the blood depending on how it was being carried (being insoluble it needed a vehicle, called lipoprotein, to travel in the blood). One of the vehicles, high density lipoprotein (HDL) cholesterol, was very good for you and protected you from heart disease. The other, low density lipoprotein (LDL) cholesterol, was thought to be harmful. However, recently LDL has been discovered to be made up of two further subtypes: LDL type A, which is small and dense, and LDL type B, which is large
and buoyant. LDL type B is not linked to atherosclerosis because it is too large to get into the blood vessel lining and cause inflammation. Dietary saturated fat does increase LDL cholesterol, but it’s the harmless type B subtype. The small and dense LDL type A cholesterol is the cause of atherosclerosis and ultimately heart disease. The latest research suggests it is increased not by fat, or by cholesterol either, but by carbohydrate and sugar,8 exactly what Dr John Yudkin had been saying in the 1950s before being discredited by the sugar scientists.

  (For more information on the cholesterol controversy and its effect on our eating habits, see Appendix 1: Cholesterol.)

  The New Science of Nutritionism

  The debate on cholesterol in the 1950s and beyond still resonates in our way of life today. Just as the researchers in infectious diseases had traditionally used epidemiological studies of the environment to understand and treat infection, so nutrition scientists now looked at the food supply to the population as a potential cause of disease. This was a new way of thinking about food and disease: identifying which individual components of foods contributed to which disease.

  This time, there were constellations of interested parties: politicians and their lobbyists, food manufacturers (paying the lobbyists) and their profits, the scientists and their funding (paid by the food industry), and lastly the confused consumers (who generated the profits for the food companies). The man or woman on the street now had to gauge what was best to eat in an era when food choice had gone from simple traditional seasonal fare to a confusing range of preserved, processed and imported foods. The outcome of the ‘fat versus sugar’ debates of this time, as we will see, has had a profound effect on the type of foods that we eat today – and an adverse effect on our health and waistline.

  The Diet–Heart Hypothesis Becomes Official Policy

  Despite Ancel Keys’ study, the New England Journal of Medicine paper and several other epidemiological studies showing an association between saturated fat and heart disease, there was still concern in the scientific community that there was not enough evidence to make the diet–heart hypothesis public policy. Many scientists based in Britain did not share the view that the evidence was overwhelming. They looked at the evidence from more robust trials, called controlled trials. These trials compared heart disease in two groups after several years of observation. One group was put on a diet with reduced saturated fat, while the other group continued with their regular diet. Many of these experiments were carried out on thousands of people over many years. The size of the trials meant that they were quite accurate and less prone to error or bias. The outcome showed no decrease in the rate of heart disease in people placed on low-fat diets. The only consistent result seemed to be a trend in the development of cancer among those subjects eating low fat. The Lancet, a respected British medical journal, commenting on the diet–heart controversy at the time, stated that ‘The cure should not be worse than the disease’, reminding doctors and scientists of their Hippocratic oath – ‘First, do no harm’.9

  In the late 1960s, a Senate Select Committee in the US had been formed to issue guidance on nutrition. Its chairman was Senator George McGovern. The committee’s original remit had been to advise government on malnutrition and ways to prevent it, but by the 1970s its sights had switched to the role of diet in disease, particularly heart disease. In 1977, after much debate with leading scientists of the time, including Ancel Keys and John Yudkin, they issued the very first national guidance on nutrition. Dietary Goals for the United States was a validation of the diet–heart hypothesis, despite it not being proven, and although many scientists were sceptical about the evidence it became government and national policy. The McGovern Report, as it became known, suggested cutting back on fat, particularly saturated fats containing cholesterol.

  The report was a defining moment in public health. For the first time a government was advising its people what they should eat. The US dietary goals were: to increase carbohydrate consumption to 55–60 per cent of energy intake; to decrease fat consumption from 40 per cent to 30 per cent of energy intake; to decrease saturated fat to 10 per cent; decrease cholesterol consumption to 300mg/day; and to decrease sugar and salt consumption.

  What happened to heart disease? Well, from 1980 to 2000 the rates of heart disease dropped from approximately 250 per 100,000 of the population to 160 per 100,000. Rates are still dropping to this day. Supporters of the low-cholesterol diet (i.e. most people who have not analysed the research) would point to the drop in heart disease as proof that it does indeed work and that it is a good public health measure. But, as in Ancel Keys’ original epidemiological study, there may have been other factors in play which caused the improvement in heart health. One overlooked fact was that by the time the McGovern Report was published heart disease had already begun to drop. In 1960, it was 400 cases (per 100,000 of population); by 1970 it was down to 300 cases.

  In 1964, a seminal public health event occurred. Another famous report, by the Surgeon General in the US, warned people of the real health dangers of smoking. Let’s see what happened to smoking rates compared to rates of heart disease.

  Smoking Rate

  Cigarettes/year Heart Disease Rate

  Events per 100,000

  1960 4,400 400

  1964 Surgeon General’s Report on Smoking

  1970 4,000 300

  1977 McGovern Report – Dietary Goals

  1980 3,000 250

  2000 2,000 160

  Table 8.1 The effects of smoking on rates of heart disease Sources: For smoking rates: CDC (2012). National Health and Nutrition Examination Survey, 2011–2012. CDC/NCHS. For heart disease rates: C. S. Fox et al. (2004). Temporal trends in coronary heart disease mortality and sudden cardiac death from 1950 to 1999: the Framingham Heart Study. Circulation, 110 (5), August, 522–7.

  It may well be that the statistics for smoking cigarettes also explain why our dietary changes coincided with improved heart health, though this was happening even before Dietary Goals was published.

  But why has so much research been conducted on cholesterol? Why is the diet–heart hypothesis still important enough for many scientists to defend it? The recent funding for research in this area is not from the sugar industry but from pharmaceutical companies. Currently the biggest-grossing drugs class in the world is statins. These drugs decrease cholesterol levels in order to reduce the risk of heart attacks and grossed $35 billion worldwide in 2010. The research direction for profit is still undoubtedly cholesterol. The pharmaceutical industry would see a big drop in profit if the diet–heart hypothesis were to be disproved, and this is why they employ many of the top scientists and top labs in the world – to try and support the crumbling edifice of this fragile theory.

  What Happened to Our Food after the Low-Cholesterol Advice?

  The government advice spurred the food industry into action. Aware that people’s food choices would change after the McGovern Report, they were quick to move to adjust their products. In fact, the dietary guidelines had given them an opportunity to market foods as officially healthy according to government guidelines. The slight problem was that fats, particularly saturated fats, accounted for a large proportion of the ingredients in most processed foods. Once you reduced the fat content of the food, it had an adverse effect on palatability – basically it tasted like cardboard. But they quickly came up with a solution: they started to replace the fat with sugar. Yes, the energy food that had been exonerated as a cause of heart disease!

  Newly designed and ‘healthy’ processed foods started to line the food aisles for the discerning consumer who had recently been ‘educated’ by the scientists and journalists. ‘Low cholesterol’ and ‘low fat’ labels jostled for prime position. The products tasted OK as well, if suspiciously sweet. All seemed to be well in nutritional public health and it looked like scientists and politicians had made a brave and wise decision to intervene in the eating behaviour of the population.

  In 1980, as the changes to the co
nstituents of our processed foods reached the consumer, sugar consumption, which had been steady for thirty years, started to rise again. For the next twenty-five years, more and more, then even more, sugar would be consumed: to rise from 80lb per person per year to 100lb. The McGovern Report had taken effect.

  Food companies not only had to reduce the total fat content of their foods (from 40 per cent to 30 per cent of total calories) but also to decrease the proportion of the fat that was of the cholesterol-containing saturated variety. The report had recommended replacing these unhealthy saturated fats with what they identified as healthy polyunsaturated oils. Fortunately, vegetable oils, containing polyunsaturates, were cheap and available. Advances in genetic engineering of rapeseed in Canada meant canola oil (Can-ola is short for Canada- oil) became a new staple alongside soya oil.

  Most nutritionists will develop a fanatical gleam in their eyes when explaining the health benefits of polyunsaturated vegetable oils: low in saturated fats (the diet–heart hypothesis lives on), high in ‘good fats’. We are now consuming vast amounts of these oils when only 100 years ago they were used as lantern fuel and to make candles. What exactly is this new type of food that has been integrated into our diet thanks to the McGovern Report?

  If you think that vegetable oils are simply made from compressing plant seeds (rapeseed, soya, sunflower), think again. Olive oil (which is a natural and healthy monounsaturated fat) is produced in this way, the simple extraction technique going back to the days of the ancient Greek civilization. Vegetable oil production is a little more ‘industrial’. You may need a degree in chemistry, or a background in petroleum engineering, to understand its production.