FIVE   THE MELTING OF THE LIPID HYPOTHESIS

Nutritionism is good for the food business. But is it good for us? You might think that a national fixation on nutrients would lead to measurable improvements in public health. For that to happen, however, the underlying nutritional science and the policy recommendations (not to mention the journalism) based on that science would both have to be sound. This has seldom been the case.

The most important such nutrition campaign has been the thirty-year effort to reform the food supply and our eating habits in light of the lipid hypothesis—the idea that dietary fat is responsible for chronic disease. At the behest of government panels, nutrition scientists, and public health officials, we have dramatically changed the way we eat and the way we think about food, in what stands as the biggest experiment in applied nutritionism in history. Thirty years later, we have good reason to believe that putting the nutritionists in charge of the menu and the kitchen has not only ruined an untold number of meals, but also has done little for our health, except very possibly to make it worse.

These are strong words, I know. Here are a couple more: What the Soviet Union was to the ideology of Marxism, the Low-Fat Campaign is to the ideology of nutritionism—its supreme test and, as now is coming clear, its most abject failure. You can argue, as some diehards will do, that the problem was one of faulty execution or you can accept that the underlying tenets of the ideology contained the seeds of the eventual disaster.

At this point you’re probably saying to yourself, Hold on just a minute. Are you really saying the whole low-fat deal was bogus? But my supermarket is still packed with low-fat this and no-cholesterol that! My doctor is still on me about my cholesterol and telling me to switch to low-fat everything. I was flabbergasted at the news too, because no one in charge—not in the government, not in the public health community—has dared to come out and announce: Um, you know everything we’ve been telling you for the last thirty years about the links between dietary fat and heart disease? And fat and cancer? And fat and fat? Well, this just in: It now appears that none of it was true. We sincerely regret the error.

No, the admissions of error have been muffled, and the mea culpas impossible to find. But read around in the recent scientific literature and you will find a great many scientists beating a quiet retreat from the main tenets of the lipid hypothesis. Let me offer just one example, an article from a group of prominent nutrition scientists at the Harvard School of Public Health. In a recent review of the relevant research called “Types of Dietary Fat and Risk of Coronary Heart Disease: A Critical Review,”* the authors proceed to calmly remove, one by one, just about every strut supporting the theory that dietary fat causes heart disease.

Hu and his colleagues begin with a brief, uninflected summary of the lipophobic era that is noteworthy mostly for casting the episode in the historical past:

We can only wonder how in the world such crazy ideas ever found their way into the “public’s mind.” Surely not from anyone associated with the Harvard School of Public Health, I would hope. Well, as it turns out, the selfsame group, formerly in thrall to the lipid hypothesis, was recommending until the early 1990s, when the evidence about the dangers of trans fats could no longer be ignored, that people reduce their saturated fat intake by switching from butter to margarine. (Though red flags about trans fats can be spotted as far back as 1956, when Ancel Keyes, the father of the lipid hypothesis, suggested that rising consumption of hydrogenated vegetable oils might be responsible for the twentieth-century rise in coronary heart disease.)

But back to the critical review, which in its second paragraph drops this bombshell:

Say what?

The article then goes on blandly to survey the crumbling foundations of the lipid hypothesis, circa 2001: Only two studies have ever found “a significant positive association between saturated fat intake and risk of CHD [coronary heart disease]”; many more have failed to find an association. Only one study has ever found “a significant inverse association between polyunsaturated fat intake and CHD.” Let me translate: The amount of saturated fat in the diet probably may have little if any bearing on the risk of heart disease, and evidence that increasing polyunsaturated fats in the diet will reduce risk is slim to nil. As for the dangers of dietary cholesterol, the review found “a weak and nonsignificant positive association between dietary cholesterol and risk of CHD.” (Someone should tell the food processors, who continue to treat dietary cholesterol as a matter of life and death.) “Surprisingly,” the authors wrote, “there is little direct evidence linking higher egg consumption and increased risk of CHD”—surprising, because eggs are particularly high in cholesterol.

By the end of the review, there is one strong association between a type of dietary fat and heart disease left standing, and it happens to be precisely the type of fat that the low-fat campaigners have spent most of the last thirty years encouraging us to consume more of: trans fats. It turns out that “a higher intake of trans fat can contribute to increased risk of CHD through multiple mechanisms”; to wit, it raises bad cholesterol and lowers good cholesterol (something not even the evil saturated fats can do); it increases triglycerides, a risk factor for CHD; it promotes inflammation and possibly thrombogenesis (clotting), and it may promote insulin resistance. Trans fat is really bad stuff, apparently, fully twice as bad as saturated fat in its impact on cholesterol ratios. If any of the authors of the critical review are conscious of the cosmic irony here—that the principal contribution of thirty years of official nutritional advice has been to replace a possibly mildly unhealthy fat in our diets with a demonstrably lethal one—they are not saying.

The paper is not quite prepared to throw out the entire lipid hypothesis, but by the end precious little of it is left standing. The authors conclude that while total levels of fat in the diet apparently have little bearing on the risk of heart disease (!), the ratio between types of fats does. Adding omega-3 fatty acids to the diet (that is, eating more of a certain kind of fat) “substantially reduces coronary and total mortality” in heart patients, and replacing saturated fats with polyunsaturated fats lowers blood cholesterol, which they deem an important risk factor for CHD. (Some researchers no longer do, pointing out that half the people who get heart attacks don’t have elevated cholesterol levels, and about half the people with elevated cholesterol do not suffer from CHD.) One other little grenade is dropped in the paper’s conclusion: Although “a major purported benefit of a low-fat diet is weight loss,” a review of the literature failed to turn up any convincing evidence of this proposition. To the contrary, it found “some evidence” that replacing fats in the diet with carbohydrates (as official dietary advice has urged us to do since the 1970s) will lead to weight gain.

I have dwelled on this paper because it fairly reflects the current thinking on the increasingly tenuous links between dietary fat and health. The lipid hypothesis is quietly melting away, but no one in the public health community, or the government, seems quite ready to publicly acknowledge it. For fear of what exactly? That we’ll binge on bacon double cheeseburgers? More likely that we’ll come to the unavoidable conclusion that the emperors of nutrition have no clothes and never listen to them again.

In fact, there have been dissenters to the lipid hypothesis all along, lipid biochemists like Mary Enig (who has been sounding the alarm on trans fats since the 1970s) and nutritionists like Fred Kummerow and John Yudkin (who have been sounding the alarm on refined carbohydrates, also since the 1970s), but these critics have always had trouble getting a hearing, especially after 1977, when the McGovern guidelines effectively closed off debate on the lipid hypothesis.

Scientific paradigms are never easy to challenge, even when they begin to crack under the weight of contradictory evidence. Few scientists ever look back to see where they and their paradigms might have gone astray; rather, they’re trained to keep moving forward, doing yet more science to add to the increments of our knowledge, patching up and preserving whatever of the current consensus can be preserved until the next big idea comes along. So don’t count on a scientific Aleksandr Solzhenitsyn to show up and expose the whole fat paradigm as a historical disaster.

The closest thing to such a figure we have had is not a scientist but a science journalist named Gary Taubes, who for the last decade has been blowing the whistle on the science behind the low-fat campaign. In a devastating series of articles and an important new book called Good Calories, Bad Calories, Taubes has all but demolished the whole lipid hypothesis, demonstrating just how little scientific backing it had from the very beginning.

Indeed. Wind the tape back to 1976, and you find plenty of reasons to doubt the lipid hypothesis even then. Some of these reasons were circumstantial, but nevertheless compelling. For instance, during the decades of the twentieth century when rates of heart disease were rising in America, Americans were actually reducing their intake of animal fats (in the form of lard and tallow). In place of those fats, they consumed substantially more vegetable oils, especially in the form of margarine, sales of which outpaced butter for the first time in 1957. Between the end of World War II and 1976 (the year of McGovern’s hearings), per capita consumption of animal fats from all sources dropped from eighty-four pounds to seventy-one, while fats from seed oils approximately doubled. Americans appeared to be moving in the direction of a “prudent diet” and yet, paradoxically, having more heart attacks on it, not fewer.*

As for the precipitous decline in heart disease during the years of World War II, that could just as easily be attributed to factors other than the scarcity of meat, butter, and eggs. Not just animal protein, but sugar and gasoline were also strictly rationed during the war. Americans generally ate less of everything, including, notably, refined carbohydrates; they did, however, eat more fish. And got more exercise because they couldn’t drive as freely thanks to gas rationing.

But the lipid hypothesis would not be deterred. Researchers in the 1950s and 1960s had studied populations in other countries that had substantially lower rates of heart disease, which could be explained by their lower consumption of saturated fat. That it could just as easily be explained by other factors—fewer total calories? fewer refined carbohydrates? more exercise? more fruits and vegetables or fish?—did not disturb the gathering consensus that fat must be the key.

The consensus hinged on two suggestive links that were well established by the early sixties: a link between high rates of cholesterol in the blood and the likelihood of heart disease and a link between saturated fat in the diet and cholesterol levels in the blood. Both these links have held up, but it doesn’t necessarily follow from them that consumption of saturated fat leads to heart disease, unless you can also demonstrate that serum cholesterol is a cause of heart disease and not, say, just a symptom of it. And though evidence for a link between cholesterol in the diet and cholesterol in the blood has always been tenuous, the belief that the former contributed to the latter has persisted, perhaps because it makes such intuitive sense—and perhaps because it has been so heavily promoted by the margarine makers.

Despite these gaps, it seemed a short easy step for McGovern’s committee to link the links, as it were, and conclude that eating meat and dairy (as important sources of both saturated fat and cholesterol) contributed to heart disease. After all, the American Heart Association had already taken the same short link-linking step and had been advocating a prudent diet low in fat and cholesterol since 1961. Still, the committee was not unaware of the controversy surrounding the research on which it was basing its recommendations. It had received a strongly worded letter of dissent from the American Medical Association, arguing that “there is a potential for harmful effects for a radical long-term dietary change as would occur through adoption of the proposed national goal.”

Still, the national goal was adopted. Never before had the government endeavored to change the diet of the whole population. In the past nutritional policies had targeted particular populations at risk for particular deficiencies. But as Taubes has documented, the attitude on the committee was that even if all the data weren’t hard as rock quite yet, what would be the harm in getting Americans to cut down on dietary fats? At the press conference introducing the Dietary Goals, Mark Hegsted, the Harvard School of Public Health nutritionist who helped to shape them, put it this way: “The question to be asked is not why we should change our diet, but why not?”

At least one good answer to that question was apparently overlooked. Perhaps because fat was in such bad repute in 1977, Dr. Hegsted and his colleagues must not have stopped to consider how a change in the levels or ratios of the various lipids, and the promotion of a biologically novel fat like trans fat, might affect human physiology. It bears remembering that the human brain is about 60 percent fat; every neuron is sheathed in a protective layer of the stuff. Fats make up the structure of our cell walls, the ratios between the various kinds influencing the permeability of the cells to everything from glucose and hormones to microbes and toxins. Without adequate amounts of fat in the diet, fat-soluble vitamins like A and E can’t pass through the intestinal walls. All this was known in 1977. But the Hippocratic oath—“First do no harm”—evidently does not apply to official dietary advice, which at least in 1977 followed a very different principle: “Why not?”

So potentially much was at stake for our health and well-being when the government threw its weight behind a wholesale change in the American diet. True, it was entirely possible that the nation would have chosen simply to ignore the Dietary Goals and go on eating as it had. But that’s not what happened. Instead, the goals were taken seriously, and one of the more ambitious nutritional experiments in our history got under way. Authority over the national menu, which in the past had rested largely with tradition and habit (and mom), shifted perceptibly in January 1977: Culture ceded a large measure of its influence over how we ate and thought about eating to science. Or what passes for science in dietary matters; nutritionism would be a more accurate term. “Premature or not,” The New York Times’ Jane Brody wrote in 1981, “the Dietary Goals are beginning to reshape the nutritional philosophy, if not yet the eating habits, of most Americans.”