According to the World Health Organisation, nearly two billion adults and more than 40m children under the age of five were overweight and obese in 2014.
Over the past four decades, the prevalence of obesity has doubled, bringing with it significant increases in chronic “lifestyle” diseases, including type 2 diabetes and heart disease. Once considered an ailment of the rich, the obesity epidemic recognises no national borders and now afflicts most populations throughout the world.
While the United States leads the global community in obesity prevalence (about 35% of the population), the rising rates in most nations suggest that within one generation most of Europe and Australia will rival the US. Emerging economic powers such as Brazil and India that currently suffer from the double burden of malnutrition and accelerating rates of obesity will experience exponential increases in the burden of “lifestyle” diseases.
Funding of nutrition and obesity research
In response to the global pandemic, nations around the world have increased their investment in nutrition and obesity research. For example, the US spends more than US$2 billion annually, and the UK, with the highest rates of obesity in Europe, has dramatically increased spending. Yet the massive increases in funding have not been apportioned equally across scientific domains and the oft-criticised field of epidemiology has received the lion’s share.
In Europe, the most expensive undertaking to examine the effects of diet on health has been the WHO’s European Prospective Investigation into Cancer and Nutrition (EPIC). This study, started in the 1990s, was funded by various governments and non-governmental agencies and has examined more than half a million people across Europe. This immense undertaking has produced more than 400 research publications on the effects of the European diet on health and obesity.
In the US, the most influential government-funded epidemiologic studies are the National Health and Nutrition Examination Survey (NHANES) conducted by the Centers for Disease Control and Prevention (CDC) and The Nurses' Health Study, conducted by the Harvard School of Public Health and associated medical schools. These studies have produced thousands of research publications that have informed US dietary guidelines and global public health policy over the past four decades.
Despite the diversity of the populations being studied, there are two fundamental commonalities to these massive and expensive research efforts: each collects dietary data via memory-based methods (interviews and questionnaires), and this is consistently not “physiologically plausible”. This is a scientific way of saying that many of the study’s participants could not survive on the amount of foods and beverages they recall consuming. An editorial in the British Medical Journal stated dietary data can often be “incompatible with life”.
The dietary data collection methods used by these expensive studies are surprisingly simplistic and naïve. The researchers, via interviews, surveys or questionnaires, merely ask participants to recall whatever they think they ate or drank over the past day, week, or year. It defies both scientific and common sense to think anyone can accurately remember (and honestly report) the exact amount and specific types of foods and beverages they consumed in the past year.
Not surprisingly, decades of evidence demonstrate the lack of validity of memory-based methods. As such, these methods produce little more than uncorroborated anecdotes and meaningless numbers, and when this is presented as scientific evidence, the conclusions are both pseudoscientific and specious.
Science and the politics of food
In a recent paper in Mayo Clinic Proceedings, my colleagues and I argued that the essence of science is the ability to discern fact from fiction and presented evidence from multiple scientific fields to support the position that the data generated by dietary surveys and questionnaires are not independently observable, measurable, or falsifiable – which is why we called them pseudo-scientific.
Without objective corroboration, it is impossible to discern or quantify what percentage of the recalled foods and beverages are completely false memories, grossly inaccurate estimations, or somewhat congruent with actual consumption. Stated simply, no one, not even the participant, knows the amount of “fact or fiction” in memory-based data.
Nevertheless, this constitutes a majority of the evidence base for dietary guidelines around the world. As such, we think the greatest problem in nutrition and obesity research is not ignorance; it is the illusion of knowledge created by these memory-based methods.
The greatest tragedy is that these methods create fear of foods and nutrients that have been a part of a human diet since the dawn of human history – meat, milk, eggs, sugar and salt. The shifting sands of recommendations against fat, salt, and cholesterol are exemplars of pseudoscientific research leading to failed public health messages. While fears sell better than facts, the fears generated by nutrition researchers direct attention and resources away from the actual cause of “lifestyle” diseases.
Over the past 50 years, scientific evidence from humans and other animals suggests that metabolic dysfunction such as obesity and type 2 diabetes are not driven by diet but are predisposed at birth by “maternal effects” – where non-genetic characteristics in a mother can be inherited by her offspring. A large body of research is demonstrating how a mother’s characteristics and behaviour before, during, and after pregnancy can determine the health of her children and grandchildren.
My theory, something I call the maternal resources hypothesis, brings together this evidence to explain why the amount of body fat a pregnant woman has, as well as the amount of exercise she does are the determining factors for the health of future generations.
For example, when a pregnant women exercises the increased energy demands redirect calories to her muscles and away from her developing foetus' fat cells. This competition between the mother’s muscle cells and her foetus' fat cells produces leaner, healthier children. When this competition does not occur because of maternal inactivity, the foetus' fat cells increase in size and number, which can predispose obesity and type 2 diabetes later in life. This metabolic dysfunction will be progressively passed on and each generation will be born fatter and increasingly inactive and unhealthy as adults. I argue that this explains why preconception and prenatal exercise are essential components of a healthy pregnancy.
In this evolutionary context, diet is merely a necessary but trivial component to obesity and disease, therefore we must prioritise resources to improving the lives and health behaviours of girls and young women rather than spending as much money as we do trying, and failing, to find out what people had for lunch.