Overweight friends as social contagion (II)

Yesterday, we reported about Christakis’ and Fowler’s top-rate study of how obesity spreads in social networks.

In today’s NYT, Gail Collins has a humorous Op-Ed (“Fat comes in on little cat feet“) about the ramifications of such a study.  Included are the following:

“8 p.m. — ‘Friends.’ In a much-anticipated reunion special, the gang has all bought condos in the same strangely affordable Manhattan apartment building. Tension mounts as Phoebe and Rachel notice that Monica is putting on weight. Well aware of the new study showing that obesity travels through friendship networks, they evict her. ‘The body mass of the many is more important than the survival of the one,’ says a saddened Ross. ‘Even if she is my sister.’ Later, the rest of the group reminisces about good times past with their now-shunned buddy. Nicole Richie guest stars as Chandler’s new love interest.”

Collins notes that the study found that obesity spreads through social networks and Christakis and Fowler “believe this is true even if said friend lives in Bangkok. The far-away friend has far more influence on your weight than relatives in the same house. And your neighbors can gain or lose the equivalent of several persons without it having any impact whatsoever.”

Collins doesn’t blame the researchers for their findings but notes that this is unlikely to be the “kind of information that’s going to brighten up anybody’s day. I’ve been overweight my entire life, and although I’ve had a lot of friends, I can’t think of one who got fat while hanging around with me. But if there’s anybody out there, I really do apologize. I’d have dropped you ages ago if only I’d known.”

Given the social contagion finding, Collins speculates:” Can you imagine how mean the high school mean girls are going to get if they think they have scientific evidence that ostracizing the chubby kids is a blow for physical fitness?….And now that his theory about leprosy-bearing Mexicans sneaking across the border has been completely debunked, Lou Dobbs will be hyperventilating about obese illegal immigrants ingratiating themselves and their fat into American communities.”

Christakis and Fowler are clear to indicate that they do not recommend dropping fat friends from one’s social network since friendship has many other health benefits.  A Slate magazine article (“Maybe fat people should be stigmatized“) thinks that they authors are being too PC and avoiding people taking responsibility for their decisions.    And Christakis and Fowler compare having an overweight friend to having no friend and conclude that the former confers more health benefits (even with the accompanying increased risk of obesity);  if their choice set was a fat friend or a thin friend, they would have gotten a different result since the thin friend would still confer the health benefits of friends without the obesity risk.  To this, Christakis said in a phone interview with Collins that “The network of fat-influencing relationships are so dense that in the end ‘your weight status might depend on the weight difference of your sister’s brother’s friend.’ ”   Sounds like a bit of a cop-out.  Like arguing that someone shouldn’t avoid risk factors because there are so many other risks out there.

But maybe in our days of increased social isolation, people should hang on the friends they have since so many Americans are losing their close friends.

David Lazer has an interesting post about how Christakis et al. deftly handle the issue of causality in their paper using longitudinal data.

And Ellen Goodman had an interesting post on this called Obesity Contagion (Boston Globe, 8/3/07)

The NEJM article by Christakis et al. on the spread of obesity through networks is available here.