The Diet Wars
On January 14, 1977, the Senate Select Committee on Nutrition and Human Needs, chaired by George McGovern, released a document called Dietary Goals of the United States. Dietary Goals heralded the arrival of a new consensus on nutrition: that dietary fat, by raising blood cholesterol, causes atherosclerosis and obesity, leading ultimately to premature death from coronary artery disease. The Committee aimed to lower the proportion of dietary fat to carbohydrates in American diets by recommending sharp reductions in meat and dairy consumption. This would presumably lower blood cholesterol levels and save lives.
The young Committee staffers who authored the report considered themselves to be confronting a serious public health problem by attempting to reform the most dangerous aspects of American eating habits. Of course, there were some marginalized skeptics in the scientific community who believed that, based on the existing evidence, it was premature to draw firm conclusions and make recommendations, but they were discounted as overcautious at best, or corrupt at worst. The media trumpeted the findings and recommendations, dismissing the dissenters as shills for the meat, egg, and dairy industries.
However, during the subsequent decades things didn’t quite work out as the Committee staff might have hoped. During the 1980s, rates of obesity and type-2 diabetes soared and have continued rising to this day. This occurred despite declining fat consumption and less cigarette smoking. How could this have happened?
As science journalist Gary Taubes argues in his controversial book Good Calories, Bad Calories, the scientific evidence supporting the fat-disease hypothesis reflected in Dietary Goals was always ambiguous, if not nonexistent, and since the release of the Dietary Goals, the case supporting it has actually become weaker, despite hundreds of millions of dollars subsequently spent on massive studies designed validate its assumptions. Taubes proposes an alternative hypothesis – that high rates of heart disease, type 2 diabetes, and obesity have nothing to do with the consumption of dietary fat, saturated or otherwise, and the predominant factor behind the prevalence of these diseases (and myriad others, including type 2 diabetes, hypertension, and even Alzheimer’s disease and cancer) is instead the excess consumption of carbohydrates, especially refined carbohydrates and sugar.
To make his case, Taubes marshals evidence from a wide range of sources. He evaluates historical data compiled by colonial and missionary doctors that show the striking appearance of chronic disease in indigenous populations when such Western staples as white rice and flour, sugar, and molasses are added to their traditional diets. He finds that pre-modern societies follow the same trajectory of having a large increase in a vast array of diseases (the “diseases of civilization”) including atherosclerosis, diabetes, stroke, and various cancers that were virtually unknown before contact with the modern world introduced significant amounts of refined carbohydrates to their diets. In some of these populations, such as the Inuit, diets were almost completely devoid of carbohydrates of any kind. Indeed, the advent of agriculture, Taubes argues, represented a radical change in the composition of the human diet, and that change is far too recent for us to have adapted through evolution.
The process behind the problem is straightforward if not obvious. When a person consumes a carbohydrate-rich meal, blood sugar spikes, and this triggers a responsive rise in insulin. When insulin levels become chronically elevated, the body’s tissues become increasingly resistant to it, and insulin levels rise yet further to compensate: a vicious circle. Additionally, blood levels of fructose are elevated following meals containing sugar or high-fructose corn syrup. These abnormal levels of blood sugar, fructose, and insulin, Taubes argues, are at the root of metabolic changes that presage diabetes, obesity, and heart disease.
The most mind-bending contention in the book has to be his rejection of the view that weight gain is caused by overeating and sedentary behavior. It seems so counterintuitive that it likely strikes most people as patently absurd. What advice could be more sensible than “eat less, and exercise?” Is this contrarianism gone too far? How can anyone claim that getting fat isn’t caused by too much eating and too little exercise?
The model of weight gain that most of us have in our minds, and the assumption that underlies most obesity research, goes something like this: If we eat more than we really need to and don’t move around enough, then we’re putting a surplus of calories in our bodies. Fat tissue is the depot where these extra calories get parked. Repeat this sequence enough and you gain weight. Reversing the process is as simple as having enough discipline to eat less and exercise more, thus creating a calorie deficit and burning off the fat reserves.
In contrast, in the model of fat accumulation that Gary Taubes advocates, the primary driver of the process is the fat tissue itself. (His hypothesis isn’t new; it’s been around since the early part of the 20th century but, despite a growing body of supporting evidence, fell out of favor in later decades because of its incompatibility with the dietary-fat-is-bad hypothesis.) Body fat accumulates when the fat tissue “pulls” in calories in response to insulin and blood sugar. In this formulation, sedentary behavior and overeating are the effects of the process, not the cause: When excessive amounts of calories are stored in fat tissue, it creates a shortage in energy available for other body tissues. This creates hunger and reduces energy expenditure. Exercising only increases hunger, thus resulting in the consumption of enough calories to compensate for the increase in calories expended. Moreover, people whose fat tissue is less prone to absorb excess calories have more nutrients available to other tissues, and therefore more energy, and thus are more likely to exercise. The apparent direction of causality from behavior to fat accumulation has been reversed.
Good Calories, Bad Calories is a tremendously integrative book: To support his contentions, Taubes marshals a vast array of evidence ranging from historical population-wide medical data to the biochemistry of blood lipoproteins. What is striking is that although his conclusions are scorned by much of the nutrition establishment, so much of the underlying science on which he bases his argument is completely uncontroversial, particularly with regard to the chain of causality from carbohydrate consumption to elevated insulin to fat accumulation.
As valuable as Taubes’s exposition of the scientific issues is, his examination of the controversy, the key players, and the way the scientific consensus developed might be even more illuminating. Indeed, anyone interested in the history and philosophy of science will find much food for thought in Good Calories, Bad Calories. Assuming his assessment of the science is correct, how could the scientific establishment have been (and continue to be) so disastrously wrong?
One reason behind the fat hypothesis’s ascendancy was the tendency of researchers to draw premature conclusions from epidemiological studies rather than waiting for the results of controlled trials. An epidemiological study is an observational study of a human population that uses statistics to determine the cause of a given health outcome. Although they can be extremely valuable in certain contexts, such as determining the source of an infectious disease, drawing reliable conclusions can be tricky, especially with regards to complex chronic diseases. Controlled clinical trials are necessary to draw more solid conclusions.
Nevertheless, clinical trials can take many years and millions of dollars to complete. But some researchers and government officials, like the Senate Committee staffers who drew up Dietary Goals of the United States, felt that with lives at stake, they didn’t have time to wait for more definitive results. That definitive evidence would never come, but the message of the government recommendations, amplified by the media, fed back into the research community, helping to reinforce the developing scientific consensus into dogma.
Taubes also cites selection bias as a key factor—the tendency to accept any and all evidence that supports one’s hypothesis, while explaining away or dismissing data that refutes it. Indeed, the story that emerges from his book is that of reckless indifference to contrary evidence and hostility on the part of these scientists towards researchers who advocated alternatives to the prevailing consensus. There is a natural human tendency to rationalize away data that contradicts our beliefs and a natural inclination to ostracize and punish those who threaten our social or professional status. Good science is supposed to correct for these self-protective instincts by systematically subjecting hypotheses to rigorous attempts to falsify them. I suspect that much of the problem here can be attributed to a certain hubris on the part of many scientists who perhaps believe they are not subject to these pitfalls.
Since the publication of his book, scientific evidence and opinion continues to turn toward Taubes’s views, albeit at a glacial pace. In December 2007, the American Diabetes Association recognized low-carbohydrate diets as safe for weight loss, modifying its previous position discouraging such diets. Also, a recent, well-publicized Israeli study published in the New England Journal of Medicine compared the followers of an Atkins-inspired low-carbohydrate diet, a low-fat diet based on American Heart Association recommendations, and a “Mediterranean” diet, rich in whole grains, olive oil, wine, and fish. The study, which followed the dieters for two years, found that the Atkins group lost the most: 58 percent more than those on the AHA’s low-fat regimen, which was the poorest performer of the three diets. More importantly, the Atkins group’s blood cholesterol levels were the most favorable in terms of heart disease risk, with the greatest increases in HDL (“good”) cholesterol and declines in triglycerides.
The story of the diet wars has implications and lessons for journalists and policymakers as much as it does for scientists. Policymakers need to be more aware of the uncertainty that attends any scientific controversy. On issues of public health, libertarians have sometimes pointed out that the government has no business trying to get us to adopt healthy habits. But even those who don’t buy into philosophical objections to the government’s tinkering with our health decisions should nevertheless counsel a more restrained, humble attitude towards policymaking. The global warming controversy is of course the most obvious crossroads of science and public policy today. In that debate it is sometimes argued that, even if the risk of a catastrophic rise in global temperatures is very small, its actual occurrence would be so devastating as to warrant dramatic action today. But the experience of the diet wars shows that there’s often no telling what results hasty government action based on uncertain scientific understanding will produce – up to and including making the problem worse. This is not a call for total government inaction in response to any scientific controversy; yet governments should always act with caution, restraint, and the recognition that there can be serious risks to action as well as inaction.
Scientists and policymakers haven’t been the only ones to err in the diet wars. The media, too, lacked the proper skepticism to serve the public interest. Journalists often let science-based stories slide into simplified templates like “science versus religion,” “science versus Big Business,” or “science versus the scaremongers.” Stories are framed with a background narrative of noble researchers (and brave journalists) defending Truth against dark forces of backwardness, ignorance, or greed.
In some cases, the narratives fit. The most obvious example is the tobacco industry’s obfuscation and dishonesty during the 1950s and 1960s, as the link between smoking and lung cancer became increasingly clear. But most controversies do not fit neatly into a satisfying moral framework. Shortly after the Dietary Goals report was published, meat and dairy producers were cast as the villains of the story. (Purveyors of highly refined, sugary manufactured foods were conveniently ignored.) The story of unhealthy, dietary-fat-laden modern lifestyles also resonated with the cultural narrative of American overconsumption: Meat production puts a heavy demand on the environment, and sedentary lifestyles are associated with cars and suburban living. But if the vast energy expended in the media towards goading Americans to change their gluttonous, sedentary lifestyles had instead been directed towards looking critically at what the nutrition establishment was saying, public health would have been better served.
That so many could get the diet question as wrong as they did for so long suggests that a spirit of healthy institutional skepticism was lacking in nutrition establishment, and to some extent the scientific community as a whole. It certainly wouldn’t hurt if the government, media, and public adopted some of the skeptical instincts of good science as well.
–Mark Nugent is a Web designer and a recent law school graduate. He has a Web site as spinline.net.