A firestorm recently erupted over a paper in the Annals of Internal Medicine that found official advice limiting sugar in diets to be based on “low” or “very low” quality evidence. Because a food-industry group had funded the study, a slew of critics accused the authors of distorting the science to undermine nutrition guidelines and make sugar seem less harmful than it actually is. One prominent nutrition professor called the paper “shameful.” “It was really an attempt to undermine the scientific process,” said another.
Lost in this torrent of criticism was any significant discussion of the science itself. Regardless of its funding source, was the paper correct in saying that there is insufficient evidence to recommend limiting sugar? And do official guidelines even matter, since we pretty much know that sugar is bad for us?
The Annals paper examined a dozen guidelines on sugar passed by governments around the world since 2002, including the Dietary Guidelines for Americans, which last year recommended limiting sugar intake to 10 percent of calories. One would assume that such advice is based on an ample body of rigorous research. But the Annals study, which included all the papers listed in the various guidelines’ bibliographies themselves, claimed that reviews to date had overstated the evidence.
In the most rigorous review on sugar and weight, for instance, only five trials lasting six months or longer could be found, on a total of just 1,245 people. According to the Annals authors, this review portrayed the data as more consistent than it actually was and failed to adequately account for evidence indicating that studies in which sugar was shown to have no detrimental effect may have been suppressed from publication.
Moreover, less rigorous data from observational studies was widely found to be “inconsistent.” Sometimes sugar was associated with health problems—weight gain, Type 2 diabetes, and tooth decay—but sometimes it wasn’t.
“Overall, I would say the guidelines are not trustworthy,” Bradley Johnston, the study’s lead author and an assistant professor of clinical epidemiology and biostatistics at McMasters University told me.
The study’s finding should come as a surprise to anyone who has been avoiding sugar for years already. Sugar is a potent source of glucose, which, over time, does appear to wreak havoc on one’s metabolism and pave a direct path to obesity and diabetes. A large body of trial evidence has shown that when carbohydrate consumption is reduced, these diseases start to reverse themselves. Also, given all the recent headlines about sugar’s ill effects, from Katie Couric’s movie Fed Up to the passage of soda taxes in several cities, one could be forgiven for assuming that the evidence condemning sugar must be a done deal.
Yet here were the Annals authors saying it’s not. Reaction to the paper from nutrition experts and advocacy groups was swift, with criticism focused on the paper’s Achilles heel: It had been paid for by the International Life Sciences Institute, which receives 60 percent of its funding from 400 industry members, including some, like Coca-Cola, PepsiCo, and Mars, that very much stand to benefit from a study questioning caps on sugar.
“This is a classic example of industry-funded research aimed at one purpose and one purpose only: to cast doubt on the science linking diets high in sugars to poor health,” Marion Nestle, a prominent professor of nutrition at New York University, told National Public Radio. Dean Schillinger, the chief of the division of general internal medicine at San Francisco General Hospital, told the New York Times: “They’re hijacking the scientific process in a disingenuous way to sow doubt and jeopardize public health.”
Schillinger, with his colleague Cristin Kearns, also penned an editorial in Annals, which likened the sugar-review authors to lackeys hired by the tobacco industry to be “merchants of doubt” about the health hazards of smoking.
Industry manipulation of the science is obviously an ongoing, serious concern. It was, in part, why the editor-in-chief of Annals, Christine Laine, invited this editorial. “I wanted to show both sides of the issue,” she told me, although she said that she considered the editorial to be unusually “strident and hostile” for an academic journal. Indeed, Schillinger and Kearns both part-time advocates against sugar; they write articles and do other work for Sugar Science, a group devoted to educating the public about sugar’s health dangers. “It’s shown me that conflicts of interest are not only financial but also intellectual,” said Laine, who added disclosures about the authors’ Sugar Science affiliations to the editorial after a reader brought them to her attention, she says.
Ironically, undercutting a scientific paper by focusing on its funding source has mainly been used in the past to shoot down sugar skeptics. For instance, when the British nutrition professor John Yudkin suggested sugar as a dietary culprit in the early 1970s, the University of Minnesota researcher Ancel Keys, a key defender of the competing hypothesis, that dietary fat was responsible for chronic health issues, accused Yudkin of issuing “propaganda,” linked to “commercial backers [who] are not deterred by the facts.”
Now that the nation’s top nutrition authority, the U.S. Dietary Guidelines, has backed off caps on total fat and begun to condemn sugar instead, the public debate is also increasingly focusing on the sugar industry—indeed, so much so that other industry actors are escaping scrutiny. One has to ask, for instance, why there was no similar outrage over another recent paper, in The BMJ, with favorable findings for vegetable oils, nearly half of whose authors were actual employees of the giant vegetable-oil manufacturer Unilever. This would be like workers at Mars, Inc. publishing a study on the health benefits of sugar. Yet this sizable conflict of interest largely got a pass by the many journalists covering the story.
To be clear, industry funding absolutely can deter good science; tobacco promotion will always be the epitome of that. But the influence of funding isn’t invariable: While one meta-analysis found that funding sources do influence the conclusions of nutrition papers, another, by a fierce critic of industry funding, paradoxically did not. A healthy dose of skepticism over funding from all sources—including governments and other institutions, which may have their own pet hypotheses—is warranted, so long as it doesn’t sideline the science or shut down legitimate debate.
Schillinger and Kearns were right to raise doubts. Sugar defenders have, since the early part of the 20th century, worked diligently to promote their product, such that President Franklin Roosevelt, in the mid-1930s, was quoted as saying the sugar lobby was “the most powerful pressure group that had descended on the national capitol” during his lifetime. The extent of industry manipulation, through ad campaigns and efforts to twist the science are described by the journalist Gary Taubes in his new book, The Case Against Sugar.
Yet Taubes believes that any industry with a PR budget has attempted pretty much the same. And he is upfront about the lack of rigorous evidence against sugar, stating in the introduction of his book, “I’m going to concede in advance a key point that those who defend the role of sugar in our diet will invariably make. … [I]t cannot be established definitively, with the science as it now stands, that sugar is uniquely harmful.”
Clinical trials on sugar are possible; it’s just that very few have been done. Emerging evidence suggests that the sugar industry may have stifled those inquiries, but Taubes believes more evidence supports the explanation that for decades, a monolith of nutrition scientists has just genuinely and obsessively had a preoccupation with fat and cholesterol which simply blotted out everything else. The National Institutes of Health (NIH) spent billions of dollars on large clinical trials, all trying to pin chronic disease on dietary fat and cholesterol. In fact, sugar was such a non-suspect for so many years that the major, NIH-funded observational studies took few pains even to measure it.
While the evidence to date shows zero benefit from sugar and a clear signal of harm, there hasn’t been enough time to fund and conduct definitive trials. Meanwhile, governments naturally feel they can’t wait. Facing panic over the continued, relentless climb in obesity and diabetes rates with no solution in sight, they’ve gone ahead and passed sugar guidelines pinned to exact thresholds, of 10 percent or 5 percent of calories. This advice is clearly well-intentioned. Yet if, as the Annals paper concludes, experts are skirting scientific norms by passing guidelines based on weak evidence, the whole process of guideline-making is effectively watered down. And the need for reliable guidance is no abstract question; indeed, everything from our waistlines to whether we might eat eggs for breakfast depends upon it.
As Americans well know, there have been many reversals in our guidelines in recent years—on dietary cholesterol, on total fat, on whether to eat breakfast to maintain a healthy weight. These were all official guidelines based on weak evidence that, when actually tested in clinical trials, were found to be unjustified. It turned out that people had been avoiding egg yolks, lobster, and fat, generally, to no avail, and that skipping breakfast altogether might actually be the best option for weight loss.
Instances of flip-flopping on nutritional advice not only erode the public trust, but make people think that the basic science itself is flawed—which, for the most part, it’s not. Instead, the central problem has been that experts and policy makers have passed judgment before that good science was done. And once a judgment is codified as policy, it’s hard to repeal. This was the case, for instance, with the low-fat diet, which although adopted as a U.S. guideline in 1980, wasn’t actually studied in trials for another decade-plus. This kind of mistake, at its very worst, is potentially deadly: Indeed, the low-fat advice, by shifting consumption to carbohydrates such as grains and sugar, is now regarded as a probable cause of the obesity and diabetes epidemics.
When the Senate first passed the government’s warnings against fat and cholesterol in the late 1970s, officials argued that the urgency of responding to public-health crises overrode any concerns about insufficient scientific evidence. “Undoubtedly there will be people who have said we have not proven our point,” said Harvard’s Mark Hegsted, an advisor to the report, at the time of its release. Yet, citing the epidemics of heart disease, cancer, diabetes, and hypertension, he stated, “We cannot afford to temporize. We have an obligation to assist the public in making correct food choices. … To do less is to avoid our responsibility.”
These are the same arguments being made today, on sugar. It makes sense to have a strong hunch that sugar is bad. Sugar has no nutritional value. It’s a direct shot of glucose to the blood stream and fructose to the liver. The historical evidence against it presented by Taubes in his book is compelling. Personally, I try hard to avoid it. But I also tend to avoid refined carbohydrates, such as white bread and cereals. Based on the existing data, I suspect that too much fructose from today’s highly sweetened fruit crops is bad, and that the 40-plus percent increase in our consumption of grains since 1970 have simply overloaded us in carbohydrates altogether.
The NIH should fund rigorous trials to sort out these issues. Meanwhile, in the absence of more definitive science, it’s worth at least considering criticism of the potentially “low” quality evidence behind existing nutrition advice. Maybe the government should be issuing a strong cautionary note, based on the existing, emerging evidence, rather than a specific formal “Guideline”—since basing guidelines on hunches that are probably right unavoidably opens up the possibility for guidelines based on hunches that are wrong.
An educated guess is not enough, warned the late Senator Charles Percy, in objecting to the government’s original dietary advice, 35 years ago. He thought it paved over limitations in the data with excessive confidence. “The best way to [provide dietary guidance] is to fully inform the public not only about what is known but also what remains controversial,” he said.
He was talking about fat and cholesterol; today’s Annals paper is talking about sucrose, glucose, fructose. We’ve been down this road before, with experts, pressed into urgency on behalf of the public health, convincing themselves that insufficient evidence could suffice. Therefore, in the matter of national guidelines, it’s worth being cautious—and not immediately dismissing those who send up cautionary flags.