In the July 26, 2007, edition of the New England Journal of Medicine, a report was published entitled, "The Spread of Obesity in a Large Social Network over 32 Years." This report has generated hundreds of breathless media reports, the theme of which is that obesity appears "contagious" - that is, fatness is spread among friends.
The authors of this study used information from the Framingham database, combing through records for information about body weight, relatives, and social contacts of individuals who were enrolled in this famous study. They then used esoteric computer modeling software to create various "animations" depicting evolving social relationships and the development of obesity over time. They concluded that subjects whose friends became obese during the course of the study had a significantly increased chance of becoming obese themselves; or, as the much of the media seems to have concluded, being friends with fat people can make you fat. Obesity is contagious.
DrRich comments:
I must admit I'd read five or 10 reports from various print and online media before I sat down to read the study myself. So I was surprised to find out what a thin soup this rather turgid study actually serves up. Here are my observations about this study:1) The list of "friends" used in this study came from administrative records used in the Framingham study for follow-up purposes, in which subjects were asked to list relatives and a "close friend" who would know their whereabouts at all times. I don't know about you, but when I'm asked for such a thing I often give the name of somebody I went to high school with 40 years ago who famously keeps copious records of classmates' addresses, but whom I haven't seen in person since our 25th reunion. Somebody who can reliably provide your contact information often may be a very good friend; but perhaps not. The baseline assumption made in this study - that the researchers know who the subjects' close friends were - can at least be questioned.
2) The kind of computer modeling used in this study is not for mere mortals to understand. The description of the statistical maneuverings the authors engaged in - complete with numerous simulations using differing assumptions and manipulation of variables - is enough to make your head spin. While it may all be perfectly legitimate, there are only a handful of humans who claim to understand this kind of complex computer modeling, the results of which have been said by some to resemble "received knowledge." It's akin to what the medieval Church used to hand out when most people were illiterate and there were no Bibles in the vernacular.
3) When presented with received knowledge, all you can do is test this new information against what you already know to be true. In this case, we are asked to believe that having a fat friend is more influential on your own chance of developing obesity than either genetics or on the obesity of your spouse (with whom, in many cases, one quite often will share meals). Somehow, it doesn't make sense. If I have to choose between, on one hand, the mountain of data tying obesity to genetic influences and the eating habits within one's own home, and on the other hand, the results of this study, I'll pass on the study.
4) None of the media reports I have read appeared to have noticed the following: the influence of fat friends on obesity in this study did not hold up for the following combinations of friends: man-woman, woman-man, woman-woman. It only held up when both friends were men. I was totally stunned to discover this finding buried in the body of the results section, but apparently, whether you're a man or woman, go ahead and enjoy your obese female friends.
In sum, there is less to this study than meets the eye. The most disturbing thing is the apparent readiness of the media (and even of medical commentators) to "buy" the received knowledge that falls out of this sort of complex computer modeling. The next thing you know, people will want to begin making sweeping policy decisions based on such stuff.
Oh, and if you don't want to get fat, cut back on your food intake and increase your exercise.
Sources:
Christakis NA and Fowler JH, The Spread of Obesity in a Large Social Network over 32 Years. New England Journal of Medicine, July 26, 2007;357:370-379. Available on-line at:
http://content.nejm.org/cgi/content/full/357/4/370#F2
