The Washington PostDemocracy Dies in Darkness

What this scathing exchange between top scientists reveals about what nutritionists actually know

Placeholder while article actions load

The new Dietary Guidelines for Americans, the federal government’s influential advice compendium, took more than a year to compile. It cites hundreds of scientific papers.  It is built on the recommendations of an expert panel. The book shapes what millions of people eat.

Yet ever since it was published last month, a very public, very caustic spat has been waged among experts over the quality of the science behind the recommendations and the extent to which any of the advice can be trusted.

That much was interesting enough. But in recent weeks one exchange has gone even farther, at times approaching Trumpian levels of invective, over whether nutrition scientists... well, actually know what they’re talking about.

This broader debate began in earnest when one of the nation’s leading medical journals published a scathing commentary by one of the nation’s most prominent cardiologists. The piece was entitled “U.S. Dietary Guidelines: An Evidence-Free Zone,” and it got more pointed as it went along, daring to liken nutrition advice to the pronouncements of "cult-like advocates."

“Unfortunately, the current and past U.S. dietary guidelines represent a nearly evidence-free zone,” said Steven E. Nissen of the Cleveland Clinic, who is noted for, among other things, detecting heart attack risks in a once-popular diabetes drug.

Too often, Nissen said, nutrition advice is based, not on true scientific experiments, known as randomized controlled trials, but on observational studies, which are generally considered a weaker form of evidence.

The lack of high-quality experimental evidence, Nissen said, “has left dietary advice to cult-like advocates.” The problem, Nissen asserted, is particular to the field of nutrition and the lower standards those scientists follow.  In the journals of  any other field, he said, “these types of poorly controlled observational studies would be difficult to publish.”

Not surprisingly, many in the nutrition world took umbrage.

David Katz,  the founding director of Yale University's Yale-Griffin Prevention Research Center, responded by questioning whether Nissen knows what he’s talking about. Nissen, after all, is a cardiologist. How would he know anything about the science of nutrition?

He compared Nissen’s criticism to “a dermatologist specializing in acne [writing] an expert critique of the latest advances in neurosurgery.”

“The result was entirely predictable," Katz wrote. "The commentary was overwhelmingly wrong, criticizing aspects of the Dietary Guidelines the author implied were current that had, in fact, been abandoned years ago- or never present at all in some cases.” 

And more than that, Katz complained that the field of nutrition gets too little esteem. Katz bemoaned the “unique disrespect our culture... has for nutrition.” Indeed, Katz fretted, the danger is that people will just stop listening to nutrition advice - that they would adopt a kind of nutritional nihilism: no one knows anything. He blamed some shortfalls in the Dietary Guidelines not on bad science, but on the nefarious influence of big food companies.

“The commentary as part of the larger context of nutrition disrespect is very worrisome. If we carry this tendency to its logical conclusion, we will succeed in convincing the public that there are no experts and there is no expertise in nutrition, and so- they (you) shouldn’t listen to any of us.”

Others in the nutrition realm cheered this response.

"A sound, level-headed rebuttal to the handful of nutrition science naysayers," tweeted Dariush Mozaffarian, dean of the nutrition school at Tufts University, who, it should be noted, has previously expressed his own doubts about the Dietary Guidelines.

What does this all mean?

First, to be fair to both sides, it should be noted that Nissen's criticism is shared by many within the nutritional field. Many already have complained that nutritional advice is too often based on the flimsier sorts of evidence.

Indeed, many on both sides of the debate believe that, due to incomplete evidence, previous versions of the Dietary Guidelines for Americans went too far in warnings about saturated fats and cholesterol. (The new Dietary Guidelines dropped its fixed limit on cholesterol and, contrary to some, this wasn't done merely because so many people were complying with the old limit. Most men, for example, weren't. The limit was dropped at least in part because the evidence for the specific dangers of cholesterol had been found wanting.)

So how is it that dietary pronouncements get made when the evidence isn't yet all there?

While the reasons are manifold, one of the strongest forces in play is public pressure. People, and the public health officials they look to, want to know what to eat. So even when the science is incomplete, the demand for answers tends to promote information that may not be fully verified, some scientists say.

The other primary reason is that questions about human diet aren't easily answered.

To study medical issues, scientists prefer to arrange randomized controlled trials, or RCTs. This method of experiment is frequently referred to as the "gold standard" in research, credited with yielding the most reliable results. In randomized controlled trials, test subjects are randomly assigned to various treatments.

Unfortunately, when it comes to long-term diet questions, RCTs can be difficult, and sometimes, all but impossible, to conduct.

Conceptually, an RCT to examine one diet or another would be simple: test subjects would be randomly assigned different meal plans. Their health outcomes would be recorded.

But the logistics of that experiment would be staggering. Thousands of patients would have to be enrolled, their diets carefully regulated. Would people actually stick to the prescribed diet? Moreover, because it takes so long for health effects to develop in people, the experiment would have to be carried out for years. By one estimate, one such proposed study on the health effects of salt would have to enroll 28,000 patients for at least five years. It has not been done.

The expense and logistics of randomized controlled trials leaves nutrition scientists to consider the other forms of evidence. This is where observational studies come in. In observational studies, subjects are not assigned randomly to different diets as in a traditional experiment. Instead, scientists merely record what kinds of foods each person consumes - they eat whatever they like - and then observe what happens to their health.  For example, the scientists might notice that people who eat lots of one specific food, say broccoli, get cancer more often. The scientists might then report that broccoli is "associated" with heart disease.

These kind of associations are reported quite often - and for no good reason. As a paper published in the American Journal of Clinical Nutrition lamented: "Is everything we eat associated with cancer?" The paper reviewed the academic studies conducted on 50 common foods. Of those, 40 had been studied for their impact on cancer. Individually,  most of those studies found that consumption of the food was  correlated with cancer. Yet when all of the research on the food was considered collectively those  effects typically shrank or disappeared.

Many scientists agree that more careful randomized controlled experiments in nutrition would be helpful - and some have been done. Nissen points, for example, to a study funded by the Spanish government regarding the effects of the Mediterranean diet. He and others argue that the U.S. government, through the National Institutes of Health ought to put up more money on key nutrition questions.

Until then, however, we might simply have to get used to some uncertainty.

Given the volume of diet advice out there, it seems probable that some of it is sound. But which items to believe?

Experts differ.

One of the leading skeptics in the field, David Allison, associate dean for science at University of Alabama-Birmingham, has asserted that only a handful of notions are proven beyond a reasonable doubt.

"There are a few things we are certain about," he has said. "We know that you can’t live without food, and that if you eat too much, you get fat. There are certain essential nutrients - vitamins and minerals - that you need to have. You should make sure there is no lead or mercury or other toxins in your food. After that the knowledge base gets thinner and thinner."

Other scientists, on the other hand, endorse a broader array of current nutritional guidelines. Mozaffarian, for example, has graded items of common nutritional advice in terms of how much scientific consensus surrounds them. He sees broad consensus on some items in the Dietary Guidelines and significant controversy around others.

EAT BUT VERIFY: More of the Post's coverage of the new Dietary Guidelines for Americans.