About the SA Blog Network



Critical views of science in the news
Cross-Check Home

Journalist Gary Taubes Raises Bucks to Disprove His Diet Theory

The views expressed are those of the author and are not necessarily those of Scientific American.

Email   PrintPrint

There’s no such thing as objective science journalism, any more than there is objective science. Some journalists are just more overt about their biases.

Gary Taubes has been ferociously attacking conventional dietary wisdom for more than a decade. Conventional wisdom holds that consuming more calories than you burn off makes you fat. Wrong, Taubes insists: It’s not the calories, per se, but the kind of calories that matters. The chief culprit in obesity, he says, is carbohydrates. You can lose weight and keep it off, Taubes contends, on a diet with lots of fat and protein as long as you minimize your carbs.

Gary himself eats lots of meat, cheese, eggs, butter, oil and nuts and avoids bread, pasta, rice, cookies, soft drinks and even fruit and vegetables. He has laid out the evidence for his hypothesis in two books: the huge, data-dense Good Calories, Bad Calories (Knopf 2007) and the shorter, more user-friendly Why We Get Fat (Knopf 2010). I admire Taubes’s reporting and analytic skills, which is why I invited him to give a talk at Stevens Institute in 2009 and interviewed him on in 2011. I nonetheless have doubts about his dietary views, which I summarized on this blog.

My post annoyed my old friend Gary, but he acknowledged that the evidence for his viewpoint is still not clear-cut; almost all dietary studies are flawed, and better research is needed. This is such a truism in medical research that I didn’t take it seriously. In fact, it sounded like the kind of thing a journalist says when he’s getting ready to abandon an old, frustrating obsession.

But Taubes clearly meant what he said about the need for more research. He called me a couple weeks ago to tell me a remarkable story. John Arnold, a wealthy financier, recently contacted Taubes after hearing him talk on the radio about the need for better research on diet and obesity. Arnold, who with his wife oversees the Laura and John Arnold Foundation, wanted to help provide funds for such research.

That conversation led to the creation of the Nutrition Science Initiative, or NuSI, a nonprofit headed by Taubes and the physician Peter Attia and based in San Diego. NuSI describes itself in a statement released this week as “dedicated to dramatically reducing the economic and social burden of obesity and obesity-related diseases by significantly improving nutrition science.” The nonprofit plans to support research carried out by a “consortium of respected clinicians and scientists from the fields of endocrinology, metabolism, diabetes, obesity and nutrition.” The Arnold Foundation has put up money for NuSI’s first two years of operation.

NuSI collaborators—from Duke, the University of Massachusetts and other schools–are now planning rigorous tests of weight-reduction regimens. Most dietary studies, Taubes points out, rely on what subjects say about their eating and exercising rather than measuring what they’re actually doing. Some studies have observed subjects under tightly controlled conditions, with food and energy expenditures precisely monitored, but for obvious reasons such studies generally involve few subjects and don’t last long. Taubes hopes NuSI will carry out highly controlled experiment—perhaps involving special observation chambers that monitor energy expenditures—involving statistically significant numbers of subjects and lasting at least three or four months.

Taubes believes that the NuSI research will confirm his perspective on diet and obesity, but he realizes that the data could point in a different direction. Most of the scientists designing NuSI experiments, he says, do not share his views. But he founded NuSI, Taubes notes, not to prove he’s right but to help solve the obesity epidemic, which contributes as much as $150 billion a year to U.S. health-care costs.

I admire Taubes for stepping out of his role as a journalist-observer and becoming an active participant in a field he covers, especially when the results could end up undercutting his own claims. What he’s doing takes guts and initiative. To learn more about NuSI—perhaps because you want to participate in or support its research—check out its website,

By the way, just because there is no such thing as objective science or science journalism does not mean that there is no such thing as objective knowledge. This is the paradox of science, that flawed, biased, emotional creatures can over time arrive at the truth. In the realm of medicine, clear-cut, durable truths are hard to come by. But I have no doubt that with better research we can achieve much better understanding of and solutions to obesity. At the very least, NuSI seems to be a step in the right direction.

Credit: Alfred A. Knopf.

John Horgan About the Author: Every week, hockey-playing science writer John Horgan takes a puckish, provocative look at breaking science. A teacher at Stevens Institute of Technology, Horgan is the author of four books, including The End of Science (Addison Wesley, 1996) and The End of War (McSweeney's, 2012). Follow on Twitter @Horganism.

The views expressed are those of the author and are not necessarily those of Scientific American.

Rights & Permissions

Comments 8 Comments

Add Comment
  1. 1. rickilewis 7:59 pm 09/13/2012

    I eagerly await the results of Gary Taubes research. I’ve loved his articles because I agree with him 100%. I know it is anecdotal, but once I stopped eating the bad white foods, I quickly dropped 40 pounds, and gained back only a few over the past 8 years. Low-carb is a way of life, not a diet. Yes, many studies are flawed, but over the years I’ve noted more in favor of low-carb than low-fat. There’s a flicker of interest, then everyone goes back to pushing low-fat. Go Gary!

    Link to this
  2. 2. Johnay 8:11 am 09/14/2012

    The thing about the efficacy of diets is that even anecdotal evidence, even leaving aside its anecdotal nature, is suspect in itself for at least one of the same reasons larger dietary studies are less-reliable than we would wish: lack of reliable data. How many people track their calories & food choices for a significant period *before* going on a diet, to get a good comparison of what they actually changed and what effect it had?

    I suspect many low-carb and low-fat dieters may also be eating fewer calories, either from less inattentive eating, or from initial difficulties finding allowed foods they like, or both. And I suspect that many people would find themselves losing weight just from paying more attention to what they eat, even with no effort or intention to change their diet. It’s hard to isolate the effects of any proposed dietary adjustment.

    Link to this
  3. 3. Johnay 8:18 am 09/14/2012

    Oh, and a big “duh!” moment for me: when I read the headline I initially interpreted it to mean Mr. Taubes was doing dietary experiments with male deer as the subjects, and while reading the article was wondering why there was no further mention of said deer, how he was getting them to eat low-carb, and what insights into the efficacy of low-carb diets could be gained from a study on herbivores.


    Link to this
  4. 4. VivaLaEvolucion 4:47 pm 09/14/2012

    Another thing to consider is that the bacteria in your gut consume a lot of the calories you eat. And different types of foods cause different types and numbers of bacteria in the guts. I am surprised that this article didn’t mention that fact as it is totally relevant.

    Link to this
  5. 5. cvictorg 4:05 pm 09/19/2012

    “Americans at the beginning of the 21st century are consuming more food and several hundred more calories per person per day than did their counterparts in the late 1950s (when per capita calorie consumption was at the lowest level in the last century), or even in the 1970s. The aggregate food supply in 2000 provided 3,800 calories per person per day, 500 calories above the 1970 level and 800 calories above the record low in 1957 and 1958.

    Of that 3,800 calories, USDA’s Economic Research Service (ERS) estimates that roughly 1,100 calories were lost to spoilage, plate waste, and cooking and other losses, putting dietary intake of calories in 2000 at just under 2,700 calories per person per day. ERS data suggest that average daily calorie intake increased by 24.5 percent, or about 530 calories, between 1970 and 2000. Of that 24.5-percent increase, grains (mainly refined grain products) contributed 9.5 percentage points; added fats and oils, 9.0 percentage points; added sugars, 4.7 percentage points; fruits and vegetables together, 1.5 percentage points; meats and nuts together, 1 percentage point; and dairy products and eggs together, -1.5 percentage point.

    Some of the observed increase in caloric intake may be associated with the increase in eating out. Data from USDA’s food intake surveys show that the foodaway-from-home sector provided 32 percent of total food energy consumption in 1994-96, up from 18 percent in 1977-78. The data also suggest that, when eating out, people either eat more or eat higher calorie foods—or both—and that this tendency appears to be increasing.

    Although multiple factors can account for weight gain, the basic cause is an excess of energy intake over energy expenditure. In general, Americans’ activity levels have not kept pace with their increase in calorie consumption. Many people apparently are oblivious to the number of calories they consume.”

    “The scientists started by testing 1,399 adults and 963 children to determine how many calories their bodies burn in total under free-living conditions. The test is the most accurate measure of total calorie burning in real-life situations.

    Once they had determined each person’s calorie burning rate, Swinburn and his colleagues were able to calculate how much adults needed to eat in order to maintain a stable weight and how much children needed to eat in order to maintain a normal growth curve.

    They then worked out how much Americans were actually eating, using national food supply data (the amount of food produced and imported, minus the amount exported, thrown away and used for animals or other non-human uses) from the 1970s and the early 2000s.

    The researchers used their findings to predict how much weight they would expect Americans to have gained over the 30-year period studied if food intake were the only influence. They used data from a nationally representative survey (NHANES) that recorded the weight of Americans in the 1970s and early 2000s to determine the actual weight gain over that period.

    “If the actual weight increase was the same as what we predicted, that meant that food intake was virtually entirely responsible. If it wasn’t, that meant changes in physical activity also played a role,” Swinburn said. “If the actual weight gain was higher than predicted, that would suggest that a decrease in physical activity played a role.”

    The researchers found that in children, the predicted and actual weight increase matched exactly, indicating that the increases in energy intake alone over the 30 years studied could explain the weight increase.

    “For adults, we predicted that they would be 10.8 kg heavier, but in fact they were 8.6 kg heavier. That suggests that excess food intake still explains the weight gain, but that there may have been increases in physical activity over the 30 years that have blunted what would otherwise have been a higher weight gain,” Swinburn said.”


    Link to this
  6. 6. XcCyclist 1:33 pm 09/26/2012

    The experience of registered dietician Diane Kress may be of interest in this debate. Originally no fan of low-carb diets (as advocated by Taubes), she describes in her Metabolism Miracle how a stubborn subset of her clients could not lose weight following the standard low-fat, count calories approach; she assumed their food logs inaccurate. But then she herself began putting on belly fat and her lipids and blood pressure went south; she developed diabetes 2 despite following the nutritional principles she had been taught. It was only when she sharply curtailed carbs did she regain health — and the approach worked for that problem subset of clients too.

    It seems to me BOTH Taubes and his critics might be right (which is perhaps why the argument is so fierce) — they are targeting different types of metabolism.
    For her part, Kress observes most of her obese patients seem to be of the “carbohydrate- intolerant” type.

    Link to this
  7. 7. sk0769 8:19 pm 12/5/2012

    I ate a starchy plant based diet, mostly vegetarian, for about 20 years. I couldn’t lose weight after pregnancy, and that is when I encountered “Good Calories, Bad Calories.” I did track nutritional intake and physical output via “May My Run”, and between my pregnancies I ran a marathon and two half marathons, but never succeeded in dropping the last ten pounds after my first pregnancy. Cutting out carbohydrates from grains and potatoes and cutting out sugar after my second pregnancy was what allowed me to lose weight, and keep it off. It was strange for me to include so much more animal protein and also more animal fat (in the form of butter, cream and cheese), and I told myself that I would go back to a vegetarian diet if my lipid profile got worse, but it has actually improved. Even though this is an anecdotal story, it fits with the larger pattern that Taubes wrote about and speaks about.

    Link to this
  8. 8. sk0769 8:19 pm 12/5/2012

    MaP My Run, that is: a website that allows you to count calories eaten and also burned.

    Link to this

Add a Comment
You must sign in or register as a member to submit a comment.

More from Scientific American

Email this Article