A few weeks ago, I wrote a Top of Mind about Gary Taubes' Good Calories, Bad Calories. Now, a professor of kinesiology at U. Va named Glenn Gaesser has responded with his critique, which is below, along with my initial e-mail response. After this and the New York Times' story today showing fat people live longer, I'm as confused as ever, though I still think it makes intuitive sense that the best way to remain lean is to eat as little sugar and white flour as possible.
BTW, this is an e-mail exchange, so please read from the bottom up.
Todd,
I have attached a copy of the article I wrote for JADA. Taubes’ book addresses many diet issues, but the most glaring inaccuracies in my view pertain to the issue carbohydrates and body weight.
For example, if carbohydrates make us fat and sick, as Taubes claims, then there should be virtually unassailable evidence to support his thesis. There is not. In fact, with regard to diet, body weight and obesity, there is compelling evidence to suggest that the truth is entirely opposite to what Taubes contends. Among the key findings in my JADA review:
(1) Virtually every major epidemiologic study shows an inverse relationship between carbohydrate intake and BMI. Potential confounders, such as energy intake, smoking, and physical activity, do not appear to undermine this relationship.
(2) The majority of epidemiologic studies also show that glycemic load is inversely related to body mass index;
(3) As for glycemic index (GI), most studies show either no relationship or an inverse relationship between GI and BMI. Several large cohort studies, including the Nurses’ Health Study II, the Women’s Health Study, and the Health Professionals Follow-up Study, revealed that higher-GI diets were associated with lower BMIs. Far from being a contributing cause of obesity, these studies actually suggest that high-GI diets may better for weight control. Although epidemiologic studies cannot prove cause-effect, at the very least these findings need to be addressed with regard to the underlying carb-obesity link. Taubes does not discuss or mention a single one of these studies.
(4) The epidemiological findings are consistent with a large number of interventions studies showing modest “spontaneous” weight loss with non-energy-restricted low-fat diets. I could not find a single study in the literature that showed that non-energy-restricted low-fat diets lead to weight gain, no matter how long the intervention. For example, in the CARMEN trial (Am J Clin Nutr, 2002; 75: 11-20), 6 months of non-energy-restricted consumption of a diet that required the subjects to replace 25% of the fat calories in the diet with simple carbs (which, according to Taubes’ hypothesis, should have resulted in weight gain), subjects’ weights remained constant over the 6 months. When the fat was replaced with complex carbs, the subjects lost about 9 pounds during the 6 months (even though the subjects were instructed not to actively try to lose weight). A number of similarly designed studies (brieftly described in my review in JADA), show the same thing.
(5) While whole-grain intake is associated with lower BMIs, refined grain intake is not consistently linked to higher BMIs.
A recent study published on PLOS Medicine (www.plosmedicine.org; August 2007, vol 4) on a combined analysis of the both Nurses’ Health Studies revealed that the nurses with the highest glycemic load had the lowest BMIs. They also had the lowest risk of diabetes. These results obviously fly in the face of the carbophobic view that high-glycemic load diets cause obesity and diabetes. The most likely explanation for these findings is that the nurses with the highest glycemic load consumed 8-10 times as much whole grains as the nurses with the lowest glycemic load. This also suggests that judging a diet solely on the basis of its carbohydrate content or glycemic properties is not justified.
If sugar is as bad as Taubes contends, how does he explain the data from the National Health and Nutrition Examination Survey, which shows an inverse relationship between sugar intake and BMI? (see Am J Clin Nutr 2003; 77: 1426-1433). The high sugar intake did not appear to adversely affect glucose and insulin because the high-sugar consumers had the same average fasting glucose and insulin as low-sugar consumers, and C-Peptide concentrations (an index of insulin secretion) were actually lower in the high-carb/high-sugar consumers. These NHANES data are consistent with the findings of most epidemiologic studies–sugar consumption is generally associated with lower body weights. (I am not advocating loading up on sugar, but rather just pointing out research findings that do not support the hypothesis that sugar consumption is tightly linked to obesity.)
As for the carb-disease connection, the published data are mixed. Some studies show a link between carb intake and/or glycemic properties of the diet and incidence rates for vascular disease diabetes. Many studies, by contrast show no relationship (some of this is discussed in my JADA review). And, with few exceptions, the vast majority of studies show no link between the glycemic properties of the diet and cancer. In fact high-carb diets that are rich in whole grains have been reported to be associated with reduced risk of several chronic disease, most notably vascular disease and diabetes. This beneficial effect of whole grains (cereal fiber seems to be the key link) appears to be independent of the glycemic properties of the diet. (See Am J Clin 2004; 80: 348-356, for example, and the JADA review, mentioned above, for discussion.)
In the Epilogue of his book Taubes concludes that “certain conclusions seem inescapable” to him. Among them is the assertion that “the fewer carbohydrates we consume, the leaner we will be.” This conclusion has little scientific support. Based on the bulk of published evidence, the opposite appears to be the case.
Best regards,
Glenn A. Gaesser, Ph.D.
Professor and Director
Kinesiology Program
University of Virginia
210 Emmet St., So.
PO Box 400407
Charlottesville, VA 22904
Phone: 434-924-3543
Fax: 434-924-1389
email: gag2q@virginia.edu

From: Wasserman Todd [mailto:twasserman@brandweek.com]
Sent: Tuesday, November 06, 2007 3:31 PM
To: Gaesser, Glenn A *HS
Subject: Re: Good Calories Bad Calories
Dear Glenn:
Thanks for the note. Since I’m a magazine editor and you are a professor of kinesiology, I’m inclined to defer to your judgment, but that would be easier if you provided proof. What facts, in particular, did Mr. Taubes get wrong?
Can you send your recent article?
Thanks.
-Todd
On 11/6/07 2:55 PM, "Gaesser, Glenn A *HS" wrote:
November 6, 2007
Mr. Todd Wasserman
Editor
Brandweek
770 Broadway, Fl 7
New York, NY 10003
Dear Todd:
I am writing in response to your article about Gary Taubes new book, Good Calories, Bad Calories in the October 22, 2007 edition of Brandweek. You wrote that the book provides “enough data to support the hypothesis that what’s really making Americans fatter is not fat, but rather carbohydrates in the form of pasta, bread, sugar and all those other goodies long ago deemed ‘fattening’.”
For the record, Mr. Taubes has his facts wrong. There are volumes of scientific evidence that suggests far from being the “food of the devil,” high-carb diets are generally associated with higher diet quality, better weight control and reduced risk of degenerative diseases. In fact, my recent article published in the October issue of the Journal of the American Dietetic Association discusses how the government’s ongoing NHANES studies show that people who consume the highest percentage of carbohydrates have the lowest BMI, as well as having similar levels of blood glucose and insulin as compared to their counterparts who consume few carbs.
A balanced diet that incorporates all food groups – including grains – is optimum to maintain a healthy weight and deliver vitamins and nutrients to the body. It is also important to note that the 2005 Dietary Guidelines recommend at least half of all grain products consumed should be whole grain products and the remaining from other grain sources, including enriched. In fact, folic acid is added to enriched bread at twice the amount of that found in whole grains.
Contrary to popular belief, bread and carbohydrates—even those with a high glycemic index--can have health benefits. Eating in moderation, in combination with regular physical activity are the most important things when maintaining health – not cutting out specific foods. His stance on exercise is ludicrous as has been proven by numerous scientific studies which he seems to have completely ignored.
Please let me know if you have questions or would like more data or clinical studies.
Sincerely,
Glenn A. Gaesser, Ph.D.
Professor and Director
Kinesiology Program
University of Virginia
210 Emmet St., So.
PO Box 400407
Charlottesville, VA 22904
Phone: 434-924-3543
Fax: 434-924-1389
email: gag2q@virginia.edu