A study published in the October issue of Journal of Adolescent Health reports that kids who are more popular are more likely to initiate smoking (see: Valente TW, Unger JB, Johnson CA. Do popular students smoke? The association between popularity and smoking among middle school students. Journal of Adolescent Health. Vol. 37, No. 4, October 2005).
The study followed 1,486 sixth-graders for one year and found that students who were more popular (defined as being named often as a friend by other students in their class) were more likely to experiment with smoking and more likely to become susceptible to smoking (defined as not ruling out smoking in the future) when re-interviewed as seventh-graders.
The study suggests an explanation for this finding: "Researchers theorize that popular sixth-graders may believe that being among the first to experiment with smoking will help them stay popular. Popular students try to set trends without deviating very far from the norms of the community, according to the study."
The study also found that isolated students (those who no other students in the class listed as being a friend of theirs) were also more likely to start smoking.
The study explains this finding: "The authors surmise that teen-agers who are isolated in the classroom may be connected to older friends who are more likely to smoke. These friends provide role models for smoking."
The Rest of the Story
The interpretation of these results suggested in the press release seems to lead to the conclusion that one way to prevent youth smoking experimentation is to encourage kids not to become popular (i.e., not to have too many friends). However, I don't think parents who don't want their kids to smoke need to start advising them to avoid having too many friends.
While one possible interpretation of these results is that being popular is an independent risk factor for smoking experimentation, there is another possible explanation. It is possible that the number of classmates who list a kid as being a friend is not so much a measure of popularity as it is of the type of social (peer) network or group of which a kid is a part. And the type of peer group that kids associate with may actually be the causative factor in increasing susceptibility to smoking and increasing smoking experimentation risk.
Kids who were defined as isolated may be part of a distinctive peer group that is itself quite isolated. As suggested in the study, these kids may be prone to smoking because of their increased connection to older kids who smoke.
But isn't it likely that kids who were defined as popular because many other kids named them as friends are also part of a distinctive peer group? This group may be characterized by a greater need to be perceived as being part of the "in" group at school, and therefore may have self-identity and self-esteem issues that would make its members more susceptible to the influences of parental, peer, and societal influences to smoke.
From my experience with this age group, the kids who tend to hang out with a very large peer group are more susceptible to smoking not because they are popular, but because the social group they hang with tends to itself be more susceptible to societal influences, especially those from advertising and entertainment media, and these groups likely have greater exposure to older kids - role models - who smoke.
So I think a reasonable interpretation of the findings of this study is not that popularity increases smoking risk, but that the number of classmates who report a kids as being a friend of theirs is an indication of the type of social group that a child is a part of, and it is the nature of these social groups that predispose to smoking.
I think that kids who have a higher level of self-esteem and a more established self-identity that they are comfortable with see less of a need to surround themselves in a larger peer group, and are also less likely to become completely isolated. This would explain why it is precisely this group, whose members have a few friends in their class, that is at lowest risk for smoking initiation and smoking susceptibility.
So I don't think parents need encourage their children not to make friends. What I think this study really points to is the role of peer groups in the smoking initiation process and the role of self-esteem and self-image issues in the choice of peer groups. And that is something that parents, schools, and communities can and should do something about.