The book has been a tremendous disappointment to me.
I read an article by Kolata in the New York Times a few days ago that was based on this book. I thought that the article was excellent, stressing the heritability component in obesity, and pointing to the failures of weight-control diets. I rushed to get the book, fully expecting fuller, more satisfactory explanations -- a truly book-length treatment of this important subject.
But the book here is actually no more than an article that has been heavily padded with cutesy anecdotes so as to achieve the physical corpulence of a book.
There are interesting (but not original) descriptions of diet fads throughout the ages. There are interesting (but depressingly familiar) accounts of failures of diets. There is an interesting account of animal studies on obesity. There are interesting accounts of twin studies that point to high heritability of obesity. And then there is endless prose that over-interprets all this: to wit, obesity is inherited, nothing can be done about it.
There is also an instance of gross malpractice of journalism. In the introduction, Kolata tells us that her book is the story of a high-science, two year long, carefully planned study of diets: Atkins versus LEARN. In chapter after boring chapter she gives us personality sketches of some of the participants and trivia about the progress of the study over the two year period. Then, at the end, while we wait for her to tell us the outcome, she tells us that, well, no, she can't say. The scientists haven't had the time to write up the results. Come on, Ms. K., if you don't know the outcome you shouldn't have bothered us with all that chatter about the wall color in the research room or what the weather was like on the first day of the study.
Journalistic malpractice isn't the worst thing about this book. The worst thing is that the author hasn't engaged with the intellectual problem that she posits. Her overall point is that obesity has very high heritability, i.e. that it is overwhelmingly determined by genetic factors. But then she also reports, as if this had nothing to do with her thesis, that numerous studies have shown that obesity is also strongly influenced by social class, the lower classes having higher rates. Now if that is true, what is the relationship to the high heritability ? Is lower class membership equally determined by genetic heritage ? Is it the same gene, or group of genes ? What, in other words, is the relationship between the claimed heritability of obesity and its correlation with class ? It doesn't seem to have occurred to Ms. K. to worry about such questions.