The history of animal fats and so-called ‘vegetable oils’ is one guided by passionate scientists, industry financing, and powerful institutions.
Testing the theory
In the 1950s, biologist Ancel Keys created the ‘diet-heart hypothesis’, the theory that saturated fat consumption causes heart disease. A prolific researcher, Keys was always careful to state that his epidemiological studies only showed an associational, not causal, connection between dietary fat and heart disease. Cause can only be established with a clinical trial which tightly controls food consumption, rather than epidemiology questionnaires. Studies ideally control for other factors, such as age, health status, smoking and exercise. The more similar the control and experimental groups are, the more accurately the differences in outcome can be attributed to the intervention.
The Anti-Coronary Club
‘The Anti-Coronary Club’ study recruited 1100 men to test the diet-heart hypothesis. The participants reduced their consumption of red meat, eggs and dairy, but not fish and poultry. They also ate at least two tablespoons of polyunsaturated seed oils per day, making the saturated to polyunsaturated fat ratio in the diet four times higher than the standard diet of the time. The initial results showed a drop in cholesterol, blood pressure and weight, which was interpreted as reducing heart disease risk. However, after a decade on the diet, 26 men on the diet had died (8 from heart attacks), compared to only 6 of the controls (none from heart attacks), and the study’s author also died of a heart attack in 1961. The final report touted the improved health markers, but failed to highlight the deaths. Even though the study was flawed, it was still used as a foundation to support the ‘diet-heart hypothesis’.
The ‘Los Angeles Veterans Trial’ found that replacing animal fat with seed oils improved risk markers, but it also caused cancer at much higher rates.
Finnish Mental Hospital Study
The 1958 ‘Finnish Mental Hospital Study’ compared traditional diets with those high in seed oils and concluded that the diet reduced serum cholesterol and halved heart disease, but that was not a correct analysis. The patients changed daily, and heart disease was only reduced in one of the two hospitals tested, and the effect was only found in men. Regardless, the study was considered “good enough” to claim that replacing animal fats with seed oils had a beneficial outcome. This study became a lynch-pin for the diet-heart hypothesis, and continues to be cited as foundational today.
The Oslo Study
Conducted in the early 1960s, ‘The Oslo Study’ recruited 412 middle-aged men who had experienced a heart attack. This group of subjects was chosen because they were more likely to stick to an unpalatable diet for health reasons, and also likely have another heart attack in the near future. As before, a traditional diet was compared to a diet that replaced animal fats with seed oils. After 5 years, the study concluded that the intervention diet reduced serum cholesterol and the incidence of heart attacks. However, the reality is that the control group was also eating margarine and hydrogenated fish oil, totaling about half a cup of trans-fats per day, whereas the intervention group was not consuming these trans-fats, something that could easily explain the difference. In addition, the intervention group reduced smoking by 45% as a part of a health campaign, which could also have greatly affected the outcome. Sadly, ‘The Oslo Study’ is simply remembered for reducing cholesterol, with these other factors forgotten.
Poor studies = bad data
These three clinical trails are the ones most commonly cited in support of the ‘diet-heart hypothesis’, yet the data generated by these studies cannot amount to convincing evidence. Instead, the trials demonstrate the enormous challenges of studying the link between nutrition and heart disease in a rigorous definitive way. Scientists agree that it is nearly impossible to control all possible variables over the amount of time required to generate statistically significant results. That is why these early studies are considered so valuable; they were performed in relatively controlled settings before ethical guidelines prevented such experiments. Yet, even in these scenarios, it was difficult to control the variables, such as control groups altering their behaviour in response to the media. In addition, neither group can be blind to what diet they are consuming, preventing the ideal double-blind placebo-controlled conditions. Also, when something is removed from the diet, such as saturated fat, the researchers must replace it with something. This results in studies that test both removal and addition, making it incredibly different to decipher cause and effect.
The National Diet Heart Study
In 1962, the NHLBI attempted to create a study that kept participants unaware of which diet they were consuming. They wanted to enroll 100,000 people in the “National Diet Heart Study”, with a follow-up period of 45 years, but first they conducted a feasibility study on 1,200 subjects which was overseen by Keys and his supporter, Jeremiah Stamler. They controlled the diets by having participants order food from a store providing special foods, such as custom-blended margarine or butter in identical containers, and burger patties made with either seed oil or tallow. Neither the participants nor the study administrators knew who was consuming which diet, an attempt at double-blinding. Stamler and Keys discussed the novelty of a diet so high in seed oil, but did not think it was an impediment. The study can be considered an effort by industry to broaden the seed oil market, as nearly every food company in the USA financially contributed to the study. However, the study was a failure because a quarter of participants dropped out due to disliking the foods and the difficulty of only eating at home. It was decided that it would be too expensive (over $1B), impractical and futile to conduct the full scale trial.
Inedible oil becomes food
The adoption of seed oils is one of the most dramatic changes that has occurred in our diet. Prior to 1910, only solid fats from animals were found in the kitchen, because oils were considered inedible. Oils were used to make soap, candles, cosmetics, lubricants, fuels and more. The original source of oil for this purpose was primarily whale oil, but the industry collapsed by 1860. Cotton seed oil from southern plantations became the replacement for non-food applications, but it was soon mixed with animal fat and sold as a compound spread, often secretly, due to the comparatively low cost. By the 1930s, industrial machinery that processed the seeds into oil created a product significantly cheaper than animal fat, but popularity only came in 1961 after the AHA recommended their consumption for heart health. While it is true that Keys and his associates believed in the health benefits of these oils, it is also true that the AHA received millions of dollars from seed oil manufacturers. The AHA medical director even posed with a bottle of Crisco in an AHA educational film.
Liquid seed oils become rancid quickly due to their molecular structure of multiple double bonds. It was therefore imperative to create a process that transformed it into a hardened oil. The result was hydrogenated oils, and they revolutionised the food service industry through their suitability for use in packaged foods. Although originally intended for soap, the substance has a tallow-like appearance which lent itself to food use. It took massive advertising campaigns and industry-written cookbooks that played on a housewife’s desire to be “modern” and “clean”. WW1 laws required bakers to use the new hardened oils so that animal fats could be exported to allies, and this massively boosted the industry. By the early 1940s, 1.5 billion pounds of hardened seed oils were being produced.
Cheap makes popular
In addition to Crisco, margarine appeared alongside butter in the chiller, but it was perceived as a strange chemical mixture made by ‘counterfeiting swindlers’. Sales still slowly increased due to its low price point, and the dairy industry responded by lobbying the government to restrict margarine, but they were unsuccessful. From the 1960s onward, margarine was advertised as a healthy alternative to saturated fat, changing the public perception. The AHA currently recommends consuming 5–10% of total calories from seed oils, despite strong evidence that heated seed oils are harmful. In fact, no human population had ever survived on seed oil as a major calorie source until 1976, when the Israelis were consuming the highest rates of seed oils in the world, while also presenting high rates of heart disease (which contradicts the idea that seed oils are protective).
Sick, obese, and smoking
From 1973 to 1982, Stamler directed “The Multiple Risk Factor Intervention Trial” (aka MR FIT) as a cheaper alternative to the previously proposed “National Diet Heart Study”. This time the focus was less on diet and more on other factors such as smoking, weight loss, and blood pressure. MR FIT was one of the largest and most demanding studies ever performed, involving 28 medical centers at a cost of $115 million dollars. Stamler’s teams measured the cholesterol levels of 361,000 middle-aged men, and found 12,000 whose cholesterol exceeded 290mg/dL. These men, who were mostly also obese smokers with high blood pressure, were considered to be at immediate risk of heart attack. Half of this group were given multiple interventions including medications, low-fat diet advice, and smoking cessation counseling. They used skim milk, margarine instead of animal fats, avoided meat, eggs and dessert, and aimed to consumed the low target of 8–10% of calories from saturated fat. The other 6,000 men lived as they pleased for the next seven years.
Death by healthy lifestyle
The results of the study were a massive disappointment. Although the intervention group had succeeded in quitting smoking, changing their diet, and reducing blood pressure, they died at a slightly higher rate than controls. The study administrators didn’t see this as disproving their hypothesis, instead theorising that this must have been due to other factors, such as the control group possibly changing their diet and lifestyle to more closely resemble the intervention group. Despite extensive commentary and criticism of the study following publication, the failure did not cause any re-evaluation of the dietary recommendations.
Low cholesterol means cancer
The 1997 follow-up study brought even more bad news; the intervention group had higher rates of lung cancer than controls. Stamler considered this a ‘troublesome chance finding, unexpected, not explained, not rationalised’, but the results were not unusual. By 1981, nearly a dozen studies had found that reducing cholesterol increased rates of cancer. ‘The Framingham Study’ showed that men with cholesterol under 190mg/dL were 3 times more likely to get colon cancer than men with cholesterol over 220mg/dL. Cholesterol-lowering interventions also seem to lead to gall stones and strokes. An NHLBI committee, which included Keys and Stamler, discussed the cancer issues, but decided that it was less concerning than the ‘common sense’ advice to eat low-fat to prevent heart disease. Further NHBLI meetings were held in 1990 to discuss the significantly increasing in deaths from cancer and other non-cardiac diseases in low cholesterol populations, but there were no follow-ups or changes to recommendations that resulted.
Largest diet study ever, fails
There has been much selective reporting and suppression of data over the years, and “The Minnesota Coronary Survey” is one of the greatest examples. This 1968 NIH study was the largest ever clinical trial of the ‘diet-heart hypothesis’, yet it is rarely referenced. Over four years, 9,000 mental hospital patients were fed either a traditional diet or a diet high in seed oils, but researchers could not find any benefit from the intervention. Due to the disappointing results, the study was not published until 16 years when the lead researcher had retired, and then only in a niche journal.
Was there an effect or not!?
Another epidemiological study from Stamler followed 2000 men working at a Chicago electric company for 20 years. While the study’s abstract states that the outcome supports the use of diet to lower cholesterol, the study’s results showed that the effect was tiny, and that the “amount of saturated fatty acids in the diet was not significantly associated with risk of death from coronary heart disease”. In the discussion section of the study, Stamler dismissed his own data, and instead referred to other studies that found the ‘correct’ outcome. When questioned about it directly, Stamler stated “saturated fat had no independent effect”. When asked to clarify whether this means that saturated fat in the diet didn’t matter, Stamler yelled back “it had no independent effect”.
Japanese not such a good example
The ‘Ni-Hon-San’ study compared Japanese living in Japan and overseas, but used poor methodology to reach unsupported conclusions. In fact, the Japanese population has been increasing their consumption of meat, eggs and dairy over recent decades, but their rates of heart disease have been declining. This contradicts Keys’ original theory that the Japanese had low rates of heart disease because they consumed low amounts of saturated fat.
The Lancet responds
In response to the growing volume of studies, the Lancet published an analysis of the evidence for the ‘diet-heart hypothesis’ in 1974 which read, “so far, despite all the effort and money that has been spent, the evidence that eliminating risk factors will eliminate heart disease adds up to little more than zero. One thing is clear, statistical association must not be immediately equated with cause and effect”. This British publication was consistently outspoken about the validity of the ‘diet-heart hypothesis’, with lively debate taking place over the years, a stark contrast to the attitude in America. Many British scientists considered the approach of Keys and Stamler to be ‘emotional, irrational and unscientific’, with the Lancet stating that “the cure should not be worse than the disease”. These British scientists were concerned that the proposed restrictions would lead to a lack of essential fatty acids in the diet, nutrients required for proper functioning of the body. Another concern was the inevitable increase in carbohydrate consumption, which would be necessary to make up the caloric difference.
An overwhelming amount of studies had been generated by the 1970s, and most could be interpreted to support whatever bias the researchers had. “There were simply believers, and non-believers”, according to prominent cholesterol expert Daniel Steinberg. A number of interpretations were possible and equally compelling, but there was only one for believers, while disbelievers became heretics outside the establishment. Thus the normal defences of modern science had been flattened by a perfect storm of forces in the post-war USA. The infancy of nutrition research allowed it to be led by a few strong characters with compelling ideas that promised to solve the rising tide of heart disease. This powerful idea was quickly supported by industries backed by millions of dollars, which left little room for debate. The result was a nationwide dietary experiment to cut out meat, eggs, dairy and animal fats, and replace them with seed oils, grains, sugar, fruits and vegetables. Sufficient evidence emerged to show that the new diet was not so healthy after all, it was too late by then, because the ‘diet-heart hypothesis’ had now been enshrined in heavily-funded institutions for decades.
This article was based on the incredible book ‘The Big Fat Surprise’ by Nina Teicholz. Her deeply researched book is filled with data and stories detailing the evolution of the ‘diet-heart hypothesis’, which is the basis for modern food guidelines and medical treatments. Please consider purchasing a copy from her website, a local book store, or Audible.