Saturday, July 19, 2008

Paul Campos' "The Obesity Myth"

I recently read Paul Campos' book "The Obesity Myth," whose thesis is that America's obsession with thinness and body fat is crazy.

Much of the book is concerned with American culture -- from supermarket magazines to a case where authorities tried to take a little girl away from her parents because she was too fat. In all of this, I generally agree with Campos that getting hysterical over a few pounds of fat is silly.

What concerns me, however, was the portion of the book where Campos argues that dietary and obesity studies actually show that being overweight is OK. In Campos' words:
Being heavier than average may be a sign of the presence of other factors that are health risks, especially a lack of physical activity, but there is no real evidence that weight itself causes undue health problems, at least not until one reaches a level of body mass that is more than 100 pounds [overweight]. (p. 137).
Campos supports this contention by pointing to numerous studies showing that the risk of mortality has a U-shaped curve when compared to Body Mass Index (BMI): People in the medium range of BMI (with a BMI of 25 to 29, technically "overweight") have the lowest risk, while people who are either underweight or obese have a higher risk.

Interesting point. But there are several things that bother me:

  • Elsewhere in the book, Campos himself recognizes that BMI is not a very good indicator of whether someone is actually obese, and that people with any significant amount of muscle mass will probably have a BMI in the 25-29 range and be classified as "overweight." (See chapter 11). In a study that Campos doesn't cite (as far as I can tell from the extremely poor endnote system), "BMI correctly identified about 44% of obese men, and 52% of obese women" compared to bodyfat percentage. For everyone else -- about half of all people -- BMI apparently isn't an accurate indicator of bodyfat.

    But Campos never puts two and two together. As far as I saw, Campos never mentioned the obvious fact that if people who exercise often have a higher BMI, that would contaminate all of the findings that he trumpets -- i.e., the findings that people with medium range BMIs have a decreased risk of mortality. If people with medium range BMIs have lower mortality, maybe it's because that group includes anyone who works out.

    Thus, the U-shaped curve might be explained as follows: people who are underweight have higher mortality (because they never exercise and have no muscle mass); people with medium BMIs have lower mortality, because that group includes all of the millions of people who get some exercise; and obese people who rarely exercise have higher mortality. That seems to me a parsimonious explanation for the U-shaped curve, as compared to Campos's notion that being fat is just plain irrelevant.

    Indeed, that explanation is consistent with at least one study that Campos does not discuss, which found a linear relationship between body fat and mortality (more body fat = more mortality), even while finding as to the same group of people that the "lowest risk was observed for men belonging to the middle fifth of BMI." Campos would like the latter conclusion, but -- because BMI is such a bad indicator of fatness -- it can still be true that more body fat is bad.

  • Campos at times seems to be taking both sides of an issue. On page 27, he criticizes the "studies that are most often cited by anti-fat warriors," noting that they often fail to "control for any of the variables" that affect mortality beyond just body fat. But back on page 15, he criticized one famous study precisely because it controlled for smoking: He claims that it was a "questionable tactic" to compare "thin non-smoking women to fatter non-smoking women," and that in doing so, the study authors had been "manipulating the data" and "exaggerating the meaning of the data."

  • As noted above, Campos says that fatness in and of itself isn't a risk to mortality. But he does seem to admit that how healthily a person eats and how much he/she exercises does matter. Campos doesn't give me any reason to think that there are significant numbers of Americans who manage to be substantially overweight (in terms of body fat) even while exercising regularly and eating a modest amount of healthy food.

In short, I liked much of Campos' book, but I'm not convinced of his reading of the medical literature.


Post a Comment

Subscribe to Post Comments [Atom]

<< Home