Posts Tagged science

Dietary fat, weight, and the importance of context

In arguing against the low-fat campaign, Michael Pollan cites extensively a review out of the Harvard School of Public Health. The review focuses on the relationship between dietary fat and coronary heart disease, but it also addresses the supposed connection between low-fat diets and weight loss.

As Pollan tells it,

One other little grenade is dropped in the paper’s conclusion: Although “a major purported benefit of a low-fat diet is wieght 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. (45)

The official dietary advice, instead of promoting weight loss, would lead to weight gain. Damning, isn’t it? But as is too often the case in Pollan’s writing, the actual study tells a more complicated story. Here’s what the review says:

A major purported benefit of a low-fat diet is weight loss. But long-term clinical trials have not provided convincing evidence that reducing dietary fat can lead to substantial weight loss. On the contrary, there is some evidence that a diet containing a high amount of refined carbohydrates may increase hunger and promote overeating, which can lead to weight gain and obesity. It is now generally agreed that total energy intake, whether from fat or carbohydrate, relative to energy expenditure, is a more important determinant of body weight than dietary fat per se.

You’ll notice that the review quite specifically implicates refined carbohydrates like sugars, white flour, and white rice as a cause of weight gain. As it happens, these were not the kind of carbohydrates that the official advice told us to eat. Senator George McGovern’s Dietary Goals for the United States encouraged Americans to decrease consumption of sugars and increase consumption of fruits, vegetables, and grains. The National Academy of Sciences’ Diet, Nutrition, and Cancer did not make an explicit recommendation with respect to refined carbohydrates, but it specifically recommended accompanying the decrease in fat intake with an increase in consumption of fruits, vegetables, and whole grains.

Pollan is correct to say that official dietary advice encouraged Americans to replace fat with carbohydrates. His claim that the review implicates carbohydrate consumption in weight gain is more of a stretch, but perhaps still arguable. However, when he puts these claims together to argue that the review claims some evidence that would link the official dietary advice to weight gain, he misses the mark badly.


Leave a Comment

The Atkins diet and the “American paradox”

Michael Pollan suggests that rather than talk about the “French paradox” we ought to confront the “American paradox” which he defines as “a notably unhealthy population preoccupied with nutrition and diet and the idea of eating healthily” (9). We’re to make a connection and believe that our preoccupation with healthy eating is making us unhealthy.

The evidence he provides for for this, however, leaves much to be desired. As I wrote at the Berkeley Science Review blog, Americans didn’t follow much of the dietary advice which Pollan argues “bears direct responsibility for creating the public health crisis that now confronts us” (60). (James McWilliams has also pointed out that Pollan even refers in In Defense of Food to Americans’ increasing consumption of junk food when it suits his point.)

One piece of evidence which Pollan brings to the discussion underscores the limitations of his argument. He tells us that “when the Atkins diet storm hit the food industry in 2003, bread and pasta got a quick redesign (dialing back the carbs; boosting the proteins) while poor unreconstructed potatoes and carrots were left out in the carbohydrate cold” (38).

A tendency to go on weight loss diets is hardly evidence of an obsession with healthy eating. Attempts to lose weight are often motivated more by aesthetic concerns than health concerns, a distinction made particularly clear by the case of the Atkins diet, which many doctors and dietitians warned was unhealthy. Moreover, even when motivated by  health considerations, a weight-loss diet is an attempt to change an existing condition which is perceived to be a problem. Pollan seems to want us to identify the diet as the cause of the problem, which gets the chronology backward. Although Julie Guthman has argued that we need to look beyond food to understand obesity, to the extent that food is to blame, it seems more plausible that the problem lies with the soft drinks, fast food, and other junk food that were the villains of The Omnivore’s Dilemma.

Leave a Comment

Does nutritionism preclude pleasurable eating?

Michael Pollan devotes a chapter of In Defense of Food to defining the notion of “nutritionism,” which we’re to distinguish from nutrition. Whereas the latter is a scientific discipline, the professor of journalism explains, “As the ‘-ism’ suggests, [nutritionism] is not a scientific subject but an ideology.” This ideology rests on the premise that “Foods are essentially the sum of their nutrient parts” and several other premises which flow from this first one.

Pollan argues that because we can’t see nutrients, in the nutritionist view, “it falls to scientists…to explain the hidden reality of foods to us. In form this is a quasireligious idea, suggesting the visible world is not the one that really matters, which implies the need for a priesthood.” The references to religion are a bit odd, given that Pollan will later take nutrition science to task for changing its mind. More importantly, though, the need for expert help doesn’t arise so much from the focus on nutrients as the decision to care about the health consequences of our dietary choices. If I wanted to know whether white rice or brown rice would make for a healthier meal, I suppose I could get all my friends together, assign half of them to eat white rice and half to eat brown rice and then record their health outcomes over the course of the next few years. No experts required! Of course, that would take a long time, I’d probably have a pretty small sample size, and people would stop being friends with me. So I’d prefer to direct the question to the “priesthood,” even though I’d be thinking about foods, which (unlike nutrients) I can see.

Another assumption of nutritionism, he explains, is “that the whole point of eating is to maintain and promote bodily health.” Perhaps this follows if we interpret his definition of nutritionism in the most literal sense, but I would bet that only a tiny minority of people who think about nutrition in terms of nutrients actually believe that we can understand everything there is to know about a food by studying its constituent nutrients. Some might believe that we can understand the relationship between diet and disease in terms of the nutrients, but that doesn’t mean they can’t also see a place for less healthful foods in the diet now and then, perhaps for reasons of culture or pleasure.

Pollan further argues, “It follows from the premise that food is foremost about physical health that the nutrients in food should be divided into the healthy ones and the unhealthy ones–good nutrients and bad.” This is another consequence that doesn’t really follow. Though Pollan has criticized the campaign to reduce fat intake, you’d be hard-pressed to find a nutrition researcher or dietitian who didn’t understand that some amount of dietary fat is essential. The Manichaean view, then, is more a product of a lack of attention to nuance than to the ideology of nutritionism itself.

And Pollan, of course, will sound every bit as Manichaean as any supporter of nutritionism when he contrasts fresh and processed foods. Indeed, in the same chapter, he argues that “the most troubling feature of nutritionism” is that it does not allow for “any qualitative distinction between whole foods and processed foods.” Quoting Gyorgy Scrinis, he argues that in the nutritionist view, “even processed foods may be considered to be ‘healthier’ for you than whole foods if they contain the appropriate quantities of some nutrients.”

To find this idea troubling, one has to take a position that is perhaps more radical than it might sound. One has to reject, for instance, the idea that iodized salt has played an important role in reducing brain damage in infants (or, I suppose, one could reject the idea that brain damage is unhealthy). Elsewhere, David Julian McClements, César Vega, Anne E. McBride, and Eric Andrew Decker have argued that fortification of flour has played a critical role in reducing deaths due to pellagra. To be sure, there are ways to process food that are less benign, but if Scrinis and Pollan want us to believe that processed foods cannot be healthier than whole foods under any circumstances, then they need to be able to answer to these examples.

By the end of the chapter, nutritionism has become something so big and bad that it seems to encompass just about everything that’s wrong with our way of eating. Now the time is ripe for Pollan to save us from this ideology which threatens not just our health but pleasure, common sense, and even food itself.

Comments (2)

Diet, Nutrition, Cancer, and Sloppy Journalism

Michael Pollan argues in In Defense of Food that the Dietary Goals for the United States, compiled by Senator George McGovern’s Senate Select Committee on Nutrition and Human Needs, ushered in an era in which dietary guidelines focused on nutrients to the exclusion of whole foods. He explains that McGovern’s committee initially recommended a reduction in red meat and dairy consumption before the interested lobbies complained. The result was a compromise, urging Americans to “choose meats, poultry, and fish that will reduce saturated fat intake.”

Pollan explains the significance of the change,

First, notice that the stark message to “eat less” of a particular food–in this case meat–had been deep-sixed; don’t look for it ever again in any official U.S. government dietary pronouncement. (24)

As evidence of this last point, Pollan trots out the National Academy of Sciences’ 1982 report Diet, Nutrition, and Cancer. He tells us,

The lesson of the McGovern fiasco was quickly absorbed by all who would pronounce on the American diet. When a few years later the National Academy of Sciences looked into the question of diet and cancer, it was careful to frame its recommendations nutrient by nutrient rather than food by food, to avoid offending any powerful interests. (25)

He further notes, “With each of its chapters focused on a single nutrient, the final draft of the National Academy of Sciences report, Diet, Nutrition and Cancer framed its recommendations in terms of saturated fats and antioxidants rather than beef and broccoli” (26).

After a few minutes of skimming the report, I couldn’t help but wonder whether Pollan and I had looked at the same document. He is correct to point out that the recommendations referenced fat, rather than meat. He’s not quite correct to say that each chapter focuses on a single nutrient. Some of the chapters discuss multiple nutrients (for instance, Chapter 10, titled “Minerals,” discusses selenium, zinc, iron, copper, iodine, molybdenum, cadmium, arsenic and lead), but on that point he’s close enough for me.

Where Pollan misses the mark badly is in failing to note that the report’s recommendations weren’t found in these nutrient-focused chapters. These chapters summarized the existing science on the various nutrients, without saying what we should do about it. The report did include a section called “Interim Dietary Guidelines,” but this was found in the report’s Executive Summary. Summarized, the guidelines were as follows:

  1. “The Committee recommends that the consumption of both saturated and unsaturated fats be reduced in the average U.S. diet.”
  2. “The committee emphasizes the importance of including fruits, vegetables, and whole grain cereal products in the daily diet…Results of laboratory experiments have supported these findings in tests of individual nutritive and nonnutritive constituents of fruits (especially citrus fruits) and vegetables (especially carotene-rich and cruciferous vegetables).”
  3. “[T]he committee recommends that the consumption of food preserved by salt-curing (including salt-pickling) or smoking be minimized.”
  4. “The committee recommends that efforts continue to be made to minimize contamination of foods with carcinogens from any source. Where such contaminants are unavoidable, permissible levels should continue to be established and food supply monitored to assure that such levels are not exceeded. Furthermore, intentional additives (direct and indirect) should continue to be evaluated for carcinogenic activity before they are approved for use in the food supply.”
  5. “The committee suggests that further efforts be made to identify mutagens in food and to expedite testing for their carcinogenicity. Where feasible and prudent, mutagens should be removed or their concentration minimized when this can be accomplished without jeopardizing the nutritive value of foods or introducing other potentially hazardous substances into the diet.”
  6. “[T]he committee recommends that if alcoholic beverages are consumed, it be done in moderation.”

Thus, of the six interim guidelines, only one was expressed in terms of nutrients, two phrased their recommendations in terms of foods (one of those even recommended eating less of something!), two were seemingly commonsense encouragements to keep dangerous substances out of the food supply, and one was a reminder not to drink too much.

Although the recommendation pertaining to vegetables did briefly mention carotenes, the committee included a clarification between the second and third guidelines:

These recommendations apply only to foods as sources of nutrients–not to dietary supplements of individual nutrients. The vast literature examined in this report focuses on the relationship between the consumption of foods and the incidence of cancer in human populations. In contrast there, is very little information on the effects of various levels of individual nutrients on the risk of cancer in humans. Therefore, the committee is unable to predict the health effects of high and potentially toxic doses of isolated nutrients consumed in the form of supplements.

They were serious about this point, too. In his critique of reductionism (which Pollan quotes), T. Colin Campbell (a co-author of the report) wrote that the Federal Trade Commission sued General Nutrition Centers for using the report’s guidelines to advertise supplements.

Although Pollan has argued that nutrition science “knows much less than it cares to admit” (14), the report shows that the science admits its limits more than Pollan cares to admit. The lack of recommendations on so many of the nutrients considered is an implicit acknowledgment of the lack of evidence. Reading the chapters focusing on the various nutrients, one sees the point made more explicitly. For instance, the researchers wrote in the conclusion of their section on selenium, “firm conclusions cannot be drawn on the basis of the present limited evidence.”

As for those chapters on nutrients, one finds plenty of references to whole foods there, too. For instance, though Pollan writes, “The language of the final report highlighted the benefits of the antioxidants in vegetables rather than the vegetables themselves” (26), the report’s summary of the epidemiological evidence on vitamin A reads,

A growing accumulation of epidemiological evidence indicates that there is an inverse relationship between the risk of cancer and consumption of foods containing vitamin A (e.g., liver) or its precursors (e.g., some carotenoids in dark green and deep yellow vegetables). Most of the data, however, do not show whether the effects are due to carotenoids, to vitamin A itself, or to some other constituents of these foods.

Of course, none of this is directly relevant to the question of how much we can learn about human nutrition by studying nutrients (instead of foods), and it’s on that question that Pollan seems to make his main argument. However, what we can observe here is a pattern of misrepresenting the claims that science has made. This, of course, is helpful to somebody trying to discredit a scientific discipline. But it should be much less helpful if Pollan wants to convince the reader that he is a careful researcher and a credible source of information.

Comments (8)

Science, culture, and nutrient deficiencies

Michael Pollan writes in the first chapter of In Defense of Food,

Vitamins did a lot for the prestige of nutritional science. These special molecules, which at first were isolated from foods and then later synthesized in a laboratory, could cure people of nutritional deficiencies such as scurvy or beriberi almost overnight in a convincing demonstration of reductive chemistry’s power. (21)

Pollan doesn’t dwell on the topic of nutrient deficiencies, but I will. In their article “In Defense of Food Science,” David Julian McClements, César Vega, Anne E. McBride, and Eric Andrew Decker point out that fortified foods have played important role in alleviating a number of diseases, namely pellagra, spinal tube defects, beriberi, goiter, and rickets.

The example of beriberi, which is detailed in Kenneth J. Carpenter’s book Beriberi, White Rice, and Vitamin B: A Disease, A Cause, and a Cure, is particularly instructive. Beriberi, which is often fatal when left untreated, was so prevalent in late nineteenth century Japan that one western doctor dubbed it “the national disease of Japan.” Moreover, western doctors visiting Japan after the Perry Expedition brought about the country’s opening to the West in 1854 found evidence that the disease had been around for a long time.

In the late 1800s, it was determined that beriberi was diet-related, and shortly thereafter it was discovered that beriberi was caused by a deficiency of the nutrient now known as thiamine (vitamin B1). In the case of the Japanese sufferers of beriberi, the thiamine deficiency was related to a diet based heavily on white rice. Although rice naturally contains thiamine, most of that thiamine is found in the husk, which is removed to produce white rice. Carpenter explains the preference for white rice, “In practice, nearly everyone who eats rice as their staple food, that is, the centerpiece for each of their daily meals, prefers white rice if it is available, and eating it can be a status symbol.”

Thus, the example of white rice is one in which longstanding cultural preferences resulted in dietary patterns that were demonstrably unhealthy. This should be reason to doubt an argument Pollan will make later, that the continued existence of a diet is proof of its healthfulness. Furthermore, the subsequent success of nutrition science in curing beriberi shows that science can help improve on a diet that tradition recommends.

Comments (6)

In defense of expertise

Michael Pollan writes in In Defense of Food,

[I]t does seem to me a symptom of our present confusion about food that people feel the need to consult a journalist, or for that matter a nutritionist or doctor or government food pyramid, on so basic a question about the conduct of our everyday lives as humans. I mean, what other animal needs professional help in deciding what it should eat? (2)

Pollan is, of course, correct to point out that no other species consults experts in deciding what to eat, but this is hardly the only way in which our lives are different from those of other animal species. Some differences–say, replacing candle light with incandescent and later fluorescent lighting–probably don’t heavily influence our dietary needs. Others, such as our tendency to live longer lives and spend more time sitting than our ancestors, seem more likely to factor into our nutritional needs.

That isn’t to say that experts will never give bad advice. However, the idea that Pollan hints at here — that it’s strangely unnatural to seek professional dietary advice — deserves to be placed into context. There are many things that are unnatural about our lives. I’d guess that most of us (at least, most people who are reading this) are far enough removed from nature that if we wanted to eat a “natural” diet, we’d have to consult an expert just to figure out what that even meant.

In spite of Pollan’s appeal to the rest of the animal kingdom, a few sentences later he makes it clear that he’s really lamenting the decline of food culture. This is a surprising turn, considering that so many aspects of food culture are also unique to humans. One wonders if Pollan knows of other animal species that cook their food or have the “[d]eep cultural taboos against gluttony” which Pollan mentioned in The Omnivore’s Dilemma. If not, then his appeal to nature does not seem to fit with his argument.

As for culture, I won’t deny that it can teach us a few things. But it happens that our lives are not only increasingly unnatural, but increasingly different from those of our relatively recent ancestors. It’s not obvious that the diet our ancestors ate will work for our more sedentary lifestyles. For that matter, Pollan tends (as I’ll write about soon) to overstate the extent to which traditional diets have really worked for people. It’s perfectly reasonable to ask whether we can do better, and if we’re going to ask the question, I hope the answer will be based on a careful look at the available evidence.

Comments (1)

Beginning “In Defense of Food”

I will soon commence my third reading of In Defense of Food, and this time I’ll be blogging it. I haven’t exactly decided how much I’ll say. It won’t be a daily thing, but I’ll aim for a couple of posts a week.

For now, here are a few pieces of criticism which I’ve found interesting:

  • Daniel Engber critiques “Unhappy Meals,” the New York Times Magazine piece that grew into In Defense of Food. He argues, “Modern nutrition may be more of an ideology than a science, but so is Pollan’s nutritional Darwinism.”
  • Registered Dietitian Ginny Messina writes,

    But for the most part, Pollan’s reasoning about nutrition and research was pretty unsophisticated and uninformed. He carefully describes all of the reasons why nutrition research is flawed, and then employs some of the worst examples of research (animal studies and completely uncontrolled observational approaches) to support his own arguments. He quotes “nutrition professionals” whose credentials and opinions are questionable at best. Almost without exception, his observations on nutrition are wrong—sometimes subtly so, sometimes overtly so, and sometimes in ways that are actually dangerous.

    Pollan defends his right to provide nutrition advice because he speaks on the authority of “tradition and common sense.” But, tradition and common sense will get you about 90% of the way to a healthy diet. The other 10% can have devastating effects and Pollan really has no sense of this.

  • James McWilliams argues,

    Every book is allowed an inconsistency or two. But In Defense of Food contains so many logical contradictions that it eventually leaves the impression of having been cobbled together in a mad rush to meet a publication deadline. Pollan laments on page 9 that “we are becoming a nation of orthorexics: people with an unhealthy obsession with healthy eating.” But by page 186, as if lacking a culinary care in the world, “we” are consuming calories “found in convenience food-snacks, microwavable entrees, soft drinks, and packaged food of all kinds-which happens to be the source of most of the 300 or so calories Americans have added to their daily diet since 1980.” Suddenly, and without explanation, a nation of obsessive nutrient-counting orthorexics has become a nation of careless, Twinkie-gorging anti-orthorexics.

  • If you have institutional access, food scientists David Julian McClements, César Vega, Anne E. McBride, and Eric Andrew Decker took to the pages of Gastronomica last summer to defend their profession in an article titled “In Defense of Food Science.” They argue, “Pollan’s rules are closer to clever catchprases that advocate a particular point of view rather than offering genuinely practical advice.” They go on to write,

    We…propose that foods be judged on the basis of their final relevant attributes (e.g., quality, nutritional profile, and environmental responsibility), rather than strictly on how they are produced (e.g., at home or in a factory). Although it is generally assumed that homemade foods are better than processed foods, this is not necessarily true. Homemade croissants, cakes, or french fries, may be free of additives, but the nutritional consequences of eating them are similar to those of their factory–or restaurant–produced counterparts. We therefore encourage food activists to advocate not only for high-quality fresh foods, but also for the development of more responsible processed foods.

Also, I’m introducing a new comments policy which delineates which kinds of comments are allowed and which I’ll respond to. In short, so long as you’re civil and stay on topic, you should be fine.

Update (1/18): In the comments, Signe Rousseau reminds me of food scientist Gregory R. Ziegler’s piece criticizing Pollan for “selective use of science to support his opinions.” I might as well also mention that I’ve previously blogged Julie Guthman’s criticisms, as well as a few of my own.

Comments (2)

« Newer Posts · Older Posts »
%d bloggers like this: