It seems to me that people who are obese seem to spend lots of time around with other people who are obese.  Smokers tend to pal around with other smokers.  This raises an important question: Do friends cause each other to pick up bad habits?  Not that I’m claiming it to be intentional, but do people pass bad habits to their friends through some form of social osmosis?

Allow me to begin with a story that embarrasses me.   When I was 18, I met a guy named “Ray” who was smart, funny and an accomplished athlete.  Ray also had a noticeable tic .  He sporadically jerked his head whenever he talked with others—he did this several times per minute.  I spent some time with Ray while visiting my then-girlfriend at college back in the 70’s.  After a few days up at her college, I noticed that I was starting to exhibit the same tic.  I can assure you that I didn’t do this intentionally.  When I noticed the problem I consciously forced myself to stop doing it, lest it became an ingrained habit.  Did Ray’s bad habit cause me to pick up my new bad habit?  Based on the timing, there’s not much doubt in my mind.  

Similarly, I’ve noticed that when I like someone and I’ve spent considerable time with them, I sometimes start talking like them, picking up their dialect, their expressions, their gestures and their vocabulary; the clues are usually subtle but often undeniable.  I’ve caught myself doing this dozens of times over my life.  Consider, also, the many speeded-up videos demonstrating that people who like each other imitate each others’ body postures during their conversations.  http://ione.psy.uconn.edu/~cespaweb/SoCorr.html  Here’s a description of this phenomenon from a 2008 article entitled “Synchrony and swing in conversation: coordination, temporal dynamics, and communication,” by Daniel C. Richardson, Rick Dale, and Kevin Shockley:

In conversation too, there are multiple levels of coordination. When two people exchange words, they share a great many things besides. For example, conversants will spontaneously converge upon dialect, speaking rate, vocal intensity, and pausing frequency. Even without interacting with them, people will spontaneously imitate the speech patterns of others. But conversational partners do not limit their behavioral coordination to speech. They spontaneously move in synchrony with each other’s speech rhythms and match one another’s postures. LaFrance (1982), for example, demonstrated that listeners tend to mirror a speaker’s posture whom they find engaging. Imitation can be found throughout human interaction: neonates imitate facial gestures, infants imitate vocalic sounds, and adults spontaneously imitate facial expressions.

Corporate logos and fashions spread like wildfire through populations; people often urge– sometimes blatantly and sometimes subtly–their friends to join the rest of the herd by buying a particular style or brand of clothing or eating a certain type of food.  When a celebrity hits it big, fans adopt their hair styles and dress styles;  they name their children after the celebrities. As indicated in the Richardson article, the urge to imitate is even common between young infants their mothers.

There is little doubt in my mind that we tend to imitate those we like and admire.  Do we also imitate the bad habits of people we like and admire?  This question was addressed in the January 23, 2009 edition of Science (available online only to subscribers), in an article called “Friendship as a Health Factor.”  The point was summarized as follows at the top of the article: “In a string of hot articles, two social scientists report that obesity, smoking, and other facets of health “spread” in networks.”  Those two scientists are Nicholas Christakis (a social scientist and hospice physician) and James Fowler (a political scientist).

Christakis and Fowler made use of tracking sheets maintained by participants in the famous Framingham Heart Study, begun in 1948. Over the course of the study, the tracking sheet form asked the participants “name a close friend who can find you in case we can’t.”  Christakis and Fowler crunched this data every which way and documented that “every facet of health examined so far appears to spread from person to person. Obesity spreads. Happiness and unhappiness spread. Smoking habits spread.” These conclusions have made Christakis and Fowler notable throughout the scientific world as well as in the pop-culture.  Why are those articles about Christakis and Fowler spreading virally through the media? Perhaps because all of those media guys have been palling around with each other and subtlely influencing each other to address the same topic!

How strong are in the conclusions of Christakis and Fowler? Consider their conclusions regarding obesity:

Christakis and Fowler found that an individual’s chance of becoming obese increased 57% if someone [they] named as a friend became obese in the same time interval. More surprisingly, the effect surfaced, but to a lesser degree, even when a direct friend wasn’t involved: obesity and a friend’s friend (or any social contact) boosted the chance of obesity by 20% and in a friend’s friend’s friend, by 10%. There was no effect beyond 3 degrees of separation-a pattern that the two have seen in subsequent studies of other health effects. The impact was also weaker among friends of the opposite sex, and there was no effect among neighbors.

They found similar patterns with regard to smoking and the quitting of smoking (in a paper published in May, 2008). “If a spouse quit smoking, an individual’s chance of quitting increased 67% and for a friend the figure was 36%.”

Critics have argued that Christakis and Fowler have confused correlation with causation. They argue that “identifying patterns among friends is not the same as proving that one friend causes another to do something.” What might actually be causing these effects? The critics have offered two possibilities.  The first alternative cause, ‘homophily,’ is the tendency of individuals to associate with people of similar to themselves. The second is a shared environment.

For example, a fast food restaurant might pop up in the neighborhood, contributing to weight gain among people living nearby. In both cases, one person’s weight gain is not the reason for the others.

Christakis and Fowler claim that they have already addressed and ruled out both of these criticisms. They argue that the effects they have documented are not caused by homophily because the effects show up over time. They also deny that shared environment is a cause because some clustering appears across geographic distances, and clustering does not appear among neighbors. They further point to the fact that “the strength of an effect depends on the strength of a friendship.”

The research of Christakis and Fowler will be discussed for years to come; they plan to publish a book on their findings in early 2010. Their findings, if true, should bear on the way doctors treat certain conditions. For example, if children get fatter because their friends are getting fatter, the cause is not simply a shared environment or the fact that the children “just happened” to make friends with similar-sized children.  Should we put warning stickers on our friends? Not likely.  Exactly how doctors would take the causal power of a friendship network into account raises numerous issues that need to be discussed.

The Christakis and Fowler article repeatedly notes that the effect they observed would also account for the spread of good habits across friendships. That is my common observation as well: we are (consciously or unconsciously) inspired by the good habits and conduct of our friends.  I cannot count the number of times that I have caught myself imitating (or, at least, tempted to imitate) the actions of a friend who I admired, and this temptation is not necessarily because I have consciously decided to imitate that friend.  I find myself strongly motivated by an exemplar version of morality rather than any rule-based system, a point made persuasively by Jonathan Haidt in his superb 2007 book, The Happiness Hypothesis.

It would seem that Christakis’s and Fowler’s findings would work in reverse when considering the anti-imitative behavior of people when they strive to avoid imitating their perceived “enemies.” Consider, for instance, how rare it is that fundamentalist Christians wear Islamic clothing .  How often do you see a Red Sox fan where a Yankee jersey?

I’m convinced that Christakis and Fowler are on to something big; I will be watching to see how this discussion continues to unfold.  The more I think about it, yes, maybe we do need warning stickers for friends . . .