Big sugar, big data

The oldest of cravings

Sugar is a cautionary tale of what can happen when an addictive energy source that used to be very scarce becomes cheap and abundant. Sugar is also a cautionary tale about the fragility of data, of how quickly it can go out of date, of how it can be corrupted, manipulated and misused. Sugar is the ultimate story of modern “branding,” of how something sweet on the surface can have such a sour and dark underbelly.

There is nothing we crave more than energy because energy is life. Sugar is one of the most ancient and deepest of our cravings for energy. How we have responded to the explosive growth of cheap sugar has many similarities to how we have responded to the equally explosive growth of cheap digital.

Our ape ancestors adored the sugar they found in fruits and honey because sugar has two uniquely powerful properties. Firstly, it gives you an immediate energy boost. You take it and bang! you feel it acting. Like digital, sugar acts as speed. Secondly, what sugar energy you don’t immediately consume is very efficiently converted into fat. In prehistoric times, there were many periods of severe hunger and famine, so our ape cousins learned to love and crave sugar because getting enough of it often meant the difference between life and death.

Over millions of years, a sugar craving was hardwired into our DNA. The pleasure centers of our brains were taught to light up at its taste. Sugar used to be as rare and precious as computing power. Our ancestors learned that whenever they got an opportunity they should gobble up as much sugar as they could get. This was the case for most of human existence.

In the thirteenth century, it would have cost about 360 eggs to buy a pound of sugar, according to Gary Taubes. In the last 200 years, we’ve become fabulously good at producing sugar, the price plummeted, and demand soared. In 2012, over 50% of US citizens consumed 180 pounds a year.

Sugar has a cruel and bitter history. Of the roughly 13 million slaves brought across the Atlantic, two-thirds of them worked on sugar cane plantations. There are some parallels with digital here too. Many of the raw materials that make our digital devices are mined by poor people in poor countries, by child laborers in war zones, working in toxic and dangerous conditions, so that we can have vibrant colors in our smartphone screens, so that our batteries will last longer.

Like digital, sugar is everywhere. It is hard to buy a processed product today that is not literally saturated with sugar and smothered in single-use plastic. Sugar makes everything more addictive and the profits are clear. Wholesome foods such as vegetables tend to yield profits of 3–6%, whereas highly processed, sugary foods can yield as much as 15%. As with Facebook, the business model of sugar is addiction, stoking our deepest and most ancient cravings.

Big Sugar advertising has been used with great efficiency to trap children. There are four dominant categories for TV ads aimed at children, three out of four of which are sugar-saturated: toys, cereals, candy and fast foods. In recent years, the targeting of children—particularly poor children—has exploded. By 2020, marketing budgets in the US targeted at children were 150 times bigger than in the 1980s, while portion sizes in US restaurants have increased fourfold since the 1950s. Making poor children fat and sick is a Big Sugar advertiser’s dream.

Not surprisingly, by 2016 almost 40% of the US population was obese. In 2019, the World Health Organization estimated that almost one-third of the world’s population were either overweight or obese. It’s not just obesity that sugar is linked to. Even at much, much lower quantities than we consume today, sugar triggers a whole range of disorders, from diabetes to cancer, according to Gary Taubes, author of The Case Against Sugar.

Dodgy data

Before there was Big Data there was Big Sugar. Big Sugar (the large organizations that produce and use sugar in their products) have been manipulating and spinning data on sugar’s health impacts for decades. Combine that with the fact that we are notoriously bad at maintaining the quality of digital data over time, and that digital data has grown explosively, and you have a challenging data environment in front of you when it comes to getting accurate information about sugar.

If we look at the Web as a living system, we could say that it is obese. While the tools, systems and processes for publishing and storing are abundant and cheap, the tools and processes for review and removal of what is out of date are anemic and weak. That’s because we are addicted to publishing and we hate cleaning up after ourselves. If the Web was a digestive system, it would have no capacity to poop.

“The web made lots of content available for ‘free’,” medical journal editor, Susanna Guzman explains. “Whether it was good or reliable or not was of secondary concern, it seemed. As an editor working on the staff of a medical journal, I did everything I could to ensure that the journal’s brand for being evidence-based and transparent was upheld. That publication, and others that stayed true to their brands and didn’t sell out, are now benefiting from that decision. If they’re still in business, that is. Many are not, having not been able to compete with free.”

Digital bloat is everywhere. Going on the Web to find accurate information is an increasing challenge. Quality control is often very poor when it comes to Web content. Once it’s up there, it’s hardly ever reviewed. If new information emerges that makes what is already published out of date, misleading or wrong, there are rarely proper procedures in place to update, and where necessary remove, the out-of-date content.

That’s surely not the case with health information, you say. Think again. I’ve worked with numerous health organizations over the years and their management of web content was patchy at best. At one stage, the US Department of Health had 200,000 pages on its website. It finally got around to reviewing what they had and deleted 150,000 of them. Nobody noticed. Not a single enquiry for one of those 150,000 pages. Why were they there? What purpose were these pages serving?

I have known other health organizations that hadn’t reviewed their Web-based health information in five years or more. Here’s the scary thing. In 25 years of working with organizations in 40 countries, I have found that in the majority of cases nobody is responsible for reviewing and removing out-of-date content. I have often pointed to out-of-date content. I would check again in another six months and the very same out-of-date content would still be there. When I stressed the need to keep the content up to date, the digital team would explain that management simply didn’t care and would not provide enough resources for review. It was all publish, publish, publish.

It was thus with some trepidation that I decided to examine sugar-related health content. To create more focus, I decided to narrow my research to the potential relationship between sugar and cancer.

Sugar and cancer data

“There’s a lot of confusing and misleading information on the Internet about the relationship between sugar and cancer,” an article in Memorial Sloan Kettering stated in December 2016. “The notion that refined sugar causes cancer or that cutting sugar from the diet is a good way to treat cancer are two common—and incorrect—claims that turn up in a Google search.” So, according to Sloan Kettering, it is incorrect to state that there is a connection between sugar and cancer. I decided to get in touch with Sloan Kettering about this blog post because I was finding lots of information on the Web indicating that there is a link between cancer and sugar. One such study, published in 2019, was even carried out by Sloan Kettering researchers.

Sloan Kettering were kind enough to reply, which was unusual because I got in touch with lots of organizations while doing this research and very few replied. This is not surprising. In my experience, most organizations are either unwilling or unable to respond to feedback. Once something gets published, it’s finished. It might as well be set in stone. Reviewing it, responding to feedback on it, let alone taking it down, would be a truly exceptional activity. We must change that. We must make review and removal as important as publishing. This is how we will get control of the publishing flood and ensure quality, accurate information.

Sloan Kettering responded that what the blog post was trying to communicate was that there was not enough evidence of a direct link (causation) between sugar and cancer as and when it was written. However, they stated that they were “actively studying how sugar relates to cancer risk, treatment, and outcomes, so we certainly don’t consider the possible link between sugar and cancer a ‘myth’.” What I concluded from my correspondence with Sloan Kettering is that while the link is not fully proven and more research needs to be done, there is enough evidence of a connection between sugar and cancer for further research.

Yet, as I looked at lots and lots of other sources from reputable organizations, I kept coming across the word “myth” in connection with sugar and cancer. For example, a Breastcancer.org page stated in 2019: “MYTH: Consuming too much sugar causes breast cancer. FACT: There is no evidence that sugar in the diet causes breast cancer.”

A myth is a false belief or idea. A myth is the original fake news. Even in an age of zettabytes, words matter. When you say something is a myth you are using very strong, definitive language. In medicine, myths are associated with fraud, fallacies, fads, quacks. But is it really a myth that there is a link between sugar and breast cancer? A 2010 study by Universidade de Lisboa linked obesity and breast cancer, with sugar being a prime cause of obesity. “Epidemiological studies have shown that dietary sugar intake has a significant impact on the development of breast cancer,” a 2016 study published by the National Center for Biotechnology Information stated. A study from Tufts University published in May 2019 linked excessive sugar intake to breast cancer.

Again and again, I kept coming across the word “myth.” Again and again, I kept finding research that showed there was indeed a link between sugar and cancer. “A myth says: ‘Sugar feeds cancer.’ But the truth is that sugar doesn’t make cancer grow faster,” Cancer.org stated in October 2018. Cancer.org was kind enough to reply to me when I questioned the use of the word “myth.” “Overall, there isn’t enough good evidence to suggest that sugar itself can cause cancer or feed/help cancers,” was their reply. “This doesn’t mean that there aren’t any studies which do suggest a link, but we base our advice on a review of all the available research and give more weight to the most rigorous scientific studies. We do try and highlight the indirect link between cancer risk and sugar. Eating lots of sugary foods over time can cause you to gain weight and there is strong evidence to show that being overweight or obese increases the risk of 13 different types of cancer.”

That sort of reply is balanced, logical and reasonable. So why isn’t that sort of balanced content published online? Why are so many eminent and respectable organizations still claiming that the sugar–cancer link is a myth? They’re not stating that the link is unproven, that further evidence is required. They are instead stating that it is a myth, a fallacy. Is it simply because they have not updated their content? Perhaps.

Another reason may have to do with deeply embedded beliefs in nutritional science that have for the past 50 years downplayed the potential dangers of sugar as they pursued other theories and hypotheses for promoting healthy eating. Connected with that downplaying is an inevitable lack of research on the sugar–cancer link over the last 50 years because to do such research would have been frowned on by the medical establishment. It would not have been a good career move. Why? Because, behind the scenes, Big Sugar corporations have wielded huge power and influence over such research.

Data is political. There is always a story behind why certain data exists and why other data doesn’t. Data is always imperfect because it reflects our imperfect society. Always be skeptical about data.

Corrupt data

“Can sugar cause cancer? It seems that evidence pointing this way was discovered in a study funded by the sugar industry nearly 50 years ago—but the work was never published,” an article in Medical News Today stated in November 2017. Big Sugar, it seems, was repressing information that linked sugar to cancer and other diseases. For every set of Big Data, be sure there is a Big Sugar lurking in the background, spinning, lying, manipulating, hiding, distorting.

In 1972, John Yudkin, the founder of the Department of Nutrition at Queen Elizabeth College, London, stated that, “If only a small fraction of what we know about the effects of sugar were to be revealed in relation to any other material used as a food additive, that material would promptly be banned.” On the other side of the Atlantic sat the king of US nutrition, US physiologist Ancel Keys. He relentlessly attacked Yudkin, calling his work “flimsy indeed,” a “mountain of nonsense,” “discredited propaganda.” Keys and his cohorts silenced and discredited Yudkin. The message was very clear to up-and coming-nutritionists: If you want a career, don’t touch sugar.

Some were brave, though. In 1976, Harvard nutritionist Jean Mayer linked sugar not just to tooth decay but also to obesity and diabetes. Big Sugar immediately got to work, describing Mayer as a “scientific farce and a journalistic disgrace.” According to Big Sugar, she was one of these “persuasive purveyors of nutritional rubbish,” “opportunists dedicated to exploiting the consuming public.” All who opposed Big Sugar were quacks.

In 2016, The New York Times reported that new evidence had emerged that in 1967 Big Sugar had paid three Harvard scientists to publish a review of research on sugar, fat and heart disease. The studies used in the review were handpicked by Big Sugar, and, of course, told a sweet story.

One of the scientists was Dr. Frederick Stare, then head of Harvard University’s department of nutrition. Dr. Stare claimed that people got as much or more food value from processed foods as they did from natural food, which he called a “food fad.” He advised people to eat “additives—they’re good for you.” He thought that Coca-Cola was a wonderful “healthy between-meals snack.” He loved sugar, calling it “a quick energy boon and pleasant to take.” You guessed it. Big Sugar was pumping donations into his department.

Another scientist who wrote the Big Sugar article was Mark Hegsted, who would go on to draft the forerunner to the US federal government’s dietary guidelines. In 1980, the US issued its first set of dietary guidelines. 15% of US citizens were obese when these guidelines were published. By 2016, it was 40%. Correlation is not causation, though it is a bit strange that after the first dietary guidelines were published, US obesity bulged. The committee writing the 2020 US nutrition guidelines was well sprinkled with former lobbyists and those who have been funded by Big Sugar and Big Food.

Bias in data is rampant, as we will see in the chapter on artificial intelligence. Many organizations are either unable or unwilling to ensure that their data is fair and objective. If you use dodgy, corrupt data, you get dodgy, corrupt results. Personalizing crap data gives you personalized crap.

More data, less light

One of the reasons organizations may give for claiming something is a myth when in fact the reality is more complex and subtle, is that they need to make things simple, to “dumb down” for the general public. There is some merit to this argument. Another reason is that many organizations have still not adapted to how the Web has changed communications, from a one-way, controlled channel to a messy mix of diverse publishing and feedback. Historically, such organizations were often the only source of information on the subject within a country.

There is also a culture in traditional publishing that believes that that which is published is to be revered and that what has gone before sets the scene for the future. If your organization has been calling something a myth for decades, it’s hard to change the language, the tone. As well, organizations everywhere are notoriously bad at being able to review, update and where appropriate remove or archive. Review and maintenance are so essential when it comes to data. We must become much better at taking care of what we have rather than constantly creating new stuff.

A counter-argument is that we should allow everything to be published. That the flood of new data will ultimately drive science and society forward. If we look back in history to other revolutions in communication, we find that this was not always the case. With information comes misinformation. We see how in the United States and Great Britain, for example, Facebook et al. are accelerating the development of misinformation societies.

“There is no evidence that, except in religion, printing hastened the spread of new ideas… In fact, the printing of medieval scientific texts may have delayed the acceptance of Copernicus,” Elizabeth Eisenstein wrote in her book, The Printing Revolution in Early Modern Europe.

We must manage our data much better. We must establish processes to root out as much as possible of what is wrong, what has been deliberately manipulated, what is prejudicial, what is fake. Otherwise, the digital world—whose building blocks are data—will become a world of crap and lies.

Back around 2011, the Norwegian Cancer Society had a 5,000-page website, with 45 part-time publishers. The Society carried out a Top Tasks survey, which is a research method I developed to help understand what really matters to people. The results showed that a very small set of tasks, centering around treatment, symptoms and diagnosis, were vastly more important to people than a whole range of other content. A comprehensive review of the content on the website began. Lots of duplicate content was discovered. This was mainly because the Society worked in departmental silos, each silo creating its own content, unaware that similar content existed in another silo.

It was found that having 45 part-time publishers was unmanageable and ineffective. Because publishing content was a small part of these people’s jobs, they could never find time to properly review, to collaborate with others who were publishing, or to do training and improve their skills. It was decided that the team should be reduced to six people who would be able to dedicate a substantial portion of their time to content. These people would also actively collaborate with each other.

The result was that the site was reduced from 5,000 to 500 pages that were consistently reviewed and managed. The Society is a charity and needs to get donations from the public. On the old site, there were calls for donations everywhere. However, the Top Tasks results clearly showed that donating to the Society was in no way a top task. A brave decision was made to focus on the citizens’ top tasks of treatment, symptoms and prevention, and to remove lots of content connected with donations. The results? Nurses reported that when interacting with people who had been to the website, they were clearly better informed. And donations? Donations doubled.

During the Ebola crisis, the Ebola factsheet page on the World Health Organization (WHO) website was a vital resource for doctors, nurses and other interested parties. Yet it was a real challenge to get this page reviewed and updated. The reason was that the WHO was so focused on publishing new information about Ebola that it struggled to review and update essential content that was already published. It seemed that everybody within the WHO wanted to publish something on Ebola, to show what they or their division was doing to combat the disease. The WHO knew how important the factsheet page was, how it was infinitely more important than the vast majority of other pages on Ebola, but it too was paralyzed by a tsunami of internal publishing.

The Web is great. All the challenges data faces can be overcome. Let me tell you a story about a fellow named Tom. It was 1993 and Tom was living in Washington DC. Tom had a serious hip issue. He was aware of research about a novel approach to hip surgery and he was constantly asking his doctor to get him a copy of the research. After months and months of trying to get his doctor to give him the research, Tom got frustrated and went to a medical library. There, he was arrested for attempting to get the research. Arrested.

A Web full of data, for all its drawbacks, is infinitely better than a world where some high priests guard the knowledge. We must learn to navigate the data, to filter and interpret it. If we create data and information, we must take responsibility for it, from the day it’s published until the day it’s removed—should that day need to arrive. We must review what we publish with a regularity that reflects its likelihood to go out of date.

Key actions

Make continuous review central to all publishing activities. Only publish the content you have the capacity to professionally manage on an ongoing basis. Review means re-editing where appropriate and removing where necessary.

Buy now

Links