IN 1998, a book by Judith Rich Harris caused a sensation in the popular press-it made the cover of Newsweek-and generated a great deal of debate among scholars.1 This book, The Nurture Assumption, reviewed decades of social, behavioral, and medical research and concluded that parents have few lasting effects on their children. Instead, Harris argued, peers are the single greatest influence on the lives that adolescents lead and on the adults that they turn out to be. In the ensuing controversy, this thesis was effectively debunked, but few denied that the book had touched on a basic kernel of truth, namely that understanding the peer context is absolutely fundamental to understanding adolescence. Adolescents know this, and so do parents and teachers. The media has profitably embraced this basic idea. Scientists from virtually every discipline have proven it time and again. Of course, parents maintain an immense influence on adolescents and rightfully deserve attention, but understanding adolescence requires that peers also be given careful consideration.2,3

Given the centrality of peers in adolescent life, empirical research on the connections among various aspects of peer relations, health-related behavior, and health status is an important resource for adolescent health promotion efforts. Currently, the study of peer relationships, peer groups, and peer networks is still less common in adolescent-focused public health research than the study of families, neighborhoods, and other social contexts of adolescent development. As a single but telling example, less than 1% of all research articles published in the American Journal of Public Health last year that included adolescent samples focused on peer relations. Still, public health researchers are taking a closer look at adolescent peers and their implications for adolescent health, and the public health literature on this topic is growing more substantial every year. This literature documents and illuminates the ways that the health (eg, depression) and health-related behavior (eg, drinking, sexual risk taking) of young people are both supported and undermined by their friends, romantic partners, acquaintances, and schoolmates.4,5 Importantly, the basic conclusions of this growing body of research are informing public health practice, as evidenced by the proliferation of peer-based protocols in health programs and services targeting young people.6,7

In the following pages, we argue that this promising growth of peer-focused adolescent health research in public health should continue, and gain momentum, in the future. Further building this base of empirical knowledge will pay off in 2 ways. First, it should deepen, on a theoretical level, our understanding of the role of peer relations in adolescents' lives. Second, it will conceivably lead to improvements in the crafting of health services for and the delivery of health services to adolescents. With the goal of supporting this enterprise, we give an overview of some important themes in the extant multidisciplinary scientific literature on adolescents and their peers that are particularly relevant to the health and health-related behavior of young people and then identify and discuss several potentially valuable future avenues for peer-focused investigations that will contribute to adolescent health promotion.

SCIENTIFIC RESEARCH ON ADOLESCENT PEERS

What we know about adolescent peers

Reflecting the importance of peers in adolescent life, peer relations among young people have long been studied by psychologists, sociologists, economists, and other social and behavioral scientists and, increasingly, by medical researchers.2,3,8,9 Reviewing this rich literature would fill volumes. This article highlights some of the major points of this multidisciplinary research.

First, adolescent peer contexts consist of multiple, overlapping layers of social relations. In the United States, for example, adolescents typically develop within a complex system of peer relations ranging from the microlevel to the macrolevel.10 At the center of this developmental ecology are close friends, a category that includes adolescent dyads, the small cliques in which these dyads are clustered, and romantic partners. These small, intimate relations are then embedded within larger, more diffused peer crowds that include collections of peers that share common characteristics, social space, and identities.11 Crowds are basic units of organization in the elaborate networks of peer relations that crisscross middle schools and high schools and also connect different schools together.12,13 Finally, at the most macrolevel is something that is often casually referred to as youth culture, the intangible yet real systems of norms, values, and rituals that connect all of these disparate levels. This youth culture is often driven by the mass media.14

Second, although social relations-with all of their many layers-are developmentally significant in all stages of the life course, the link between peer relations and developmental trajectories is much stronger during adolescence than during childhood or adulthood.12 This phenomenon occurs because fulfilling one of the key developmental tasks of adolescence-individuating from parents-greatly elevates the importance of peers relative to family relationships. As a result, adolescents become more likely to internalize the views of peers into their own self-concepts, to place a higher premium on maintaining peer relations, and to be sensitive to the influences of their peers, all of which tend to increase pressures for conformity. Thus, what is ultimately a positive role of peer relations in youth development can initially create problems for adolescents' adjustment and functioning.15,16

Third, the strong link between peer relations and individual development in adolescence is rooted in 2 primary the selection of adolescents into peer groups and the influence of peer groups on adolescents.2^,17 Beginning with selection, an adolescent typically enters into a peer group because she or he is attracted to the group's norms, values, and activities. Alternatively, peer group members, who serve as gatekeepers to the group, are attracted to the adolescent's own characteristics and qualities and invite or deny the adolescent's entry into the group. The subsequent developmental trajectories of the adolescent, therefore, may be a manifestation of why she or he joined-or were allowed to join-the peer group in the first place.18,19 Turning to influence, peer group members model appropriate and desired behaviors for each other and coerce and cajole each other into following established, agreed-upon expectations. Thus, once an adolescent has entered a peer group, she or he is likely to be influenced by the members of that group, especially if the peer group members have strong emotional bonds or if the adolescent fears isolation.20,18 Although these 2 mechanisms are often viewed as competitive forces, they are not mutually exclusive. For example, just because an adolescent selects or is selected into a party-oriented peer group because of her or his own drinking behavior does not mean that associating with such peers will not then lead to increases in that same behavior over time.

Fourth, peer relations are associated with many different aspects of adolescents' development, both positive and negative, but they appear to be especially salient to the kinds of adolescent behaviors that are prohibited by parents, school personnel, the police, and other adults as dangerous, inappropriate, or immoral. This general category of risky adolescent behavior includes many health-related behaviors, such as unprotected sex, drinking, smoking, drug use, reckless driving, delinquency, and violence. In the face of strong legal and social constraints, peer relations are often the deciding factor in whether adolescents initiate or maintain such behaviors.15,18,19,21

Adolescent peers and public health research

The study of adolescent peers has not been as common in public health as it has been in developmental psychology, sociology, and other fields of research, but this is changing. In recent years, public health research has devoted more attention to peers. Scientists in this field have linked peer relations to a domain of adolescent development-health-that was traditionally underrepresented in those other fields of research as they laid the foundation of the peer literature. In this way, public health research has begun to make an important contribution to an already rich literature.6,9,17 Again, comprehensively reviewing this growing body of public health research on adolescents and their peers is not possible in this short essay. Instead, we discuss a few ways in which peer-focused activity among public health researchers has advanced the promotion of adolescent physical and mental health.

One way that peers are significant for adolescent health is that they can lead adolescents into a variety of behaviors that have short- and long-term consequences for physical and mental functioning. In other words, peers can influence adolescent behaviors in the present that have clear implications for adolescent health in both the present and the future. Prominent examples of the kind of behaviors that are influenced by peers on one hand and that influence health status on the other are substance use (eg, alcohol use, drug use, smoking), sexual activity (eg, unprotected sex), diet and eating, exercise, and physical and social aggression.4,9,17,21-24 Mirroring epidemiological approaches to disease transmission,25 public health research on this topic has studied peers as social vectors of health and health behavior transmission. These vectors can be proximal, in the form of personal relationships and face-to-face interactions with friends or romantic partners. They can also be more distal, in the form of the general culture that is often created by large groups of youth who come together for long periods of time in a common setting (eg, the school).6,17,26.

This kind of approach to adolescent health-peers as agents or facilitators of adolescent health-related behavior-is typified by the substantial body of public health research utilizing peer network models. In these models, adolescents' health-related behavior is viewed as arising from, managed by, and contributing to direct and indirect ties that young people have with peers within their schools, neighborhoods, or other networks of social relations.4,5,22,27 In this line of research, peer relations provide the context in which adolescents make decisions about health and health-related behavior.

Another way that peers are significant for adolescent health is that they often engage in actions that directly harm the health and well-being of young people. In other words, adolescents actually hurt-either physically or emotionally-each other. These direct effects occur when adolescents are bullied by their peers, are the victims of crime and violence at the hands of their peers, are subjected to social aggression from their peers, or are socially ostracized by their peers.28 One striking example of this phenomenon concerns obese and overweight youth. Ample evidence has documented that these youth are often victimized, both physically and emotionally, by peers at school because of their weight. Even when they are not being explicitly bullied or mistreated, they are at risk for being socially excluded. Such social isolation has implications for depression, self-esteem, and disordered eating.29-31

Gender clearly plays a role in peer victimization. Although boys are more likely than girls to be physically victimized by their friends and schoolmates, girls are more likely to be physically, sexually, and psychologically victimized by their romantic partners.32,33 A good deal of empirical research has been devoted to determining which girls are at risk for physical abuse, sexual coercion, and psychological trauma in their dating relationships and why their partners are at risk for perpetrating such violence.28,32,33 Another strand of public health research has extended this inquiry by examining the peer contexts that encourage and support boys' engagement in aggressive and violent behavior in their romantic relationships. In other words, these researchers are considering how violence in one important kind of peer relationships (eg, romantic or sexual partnership) in adolescence is embedded in larger contexts of peer relations.33-35 This research drives home the point that the importance of peers to adolescent health is not solely a function of their influence over each other but also of their actions toward each other.

A third major area of public health research involves the assessment of the utility of peer-based strategies for adolescent health programs and services. Some of the most common foci of this research are peer counseling, peer education, and peer-led intervention protocols. All 3 have become increasingly prominent in health education and intervention programs, especially those targeting risky adolescent behaviors like sex, drinking, and drug use.9,36 Although some studies have documented that these peer-based strategies increase program impact, others have reported that having peer educators does not benefit the young people receiving the services any more than they would benefit from having adult leaders.37,38 Another question that has generated mixed empirical results is whether the peer leaders themselves reduce their engagement in the targeted risky behaviors because of their enactment of this special role in the program.39-41 This inconclusive evidence could reflect inconsistencies in the rigor of evaluations, defects in design and implementation, or problems with the effectiveness of health education more generally.9 Regardless, this line of applied research and evaluation indicates that public health researchers are attempting to incorporate what is known about adolescent-peer dynamics into adolescent health promotion.

In sum, peer studies represent a growing subfield within adolescent-focused public health research. Below, we suggest 3 ways that this extant public health literature linking peer relations to adolescent physical health, mental health, and health behavior may be expanded in the future. In doing so, we have tried to consider research topics that can realistically-not just in theory-be of use to public health policy makers, practitioners, and interventionists in ways that are acceptable to stakeholders.

FUTURE AVENUES FOR PUBLIC HEALTH RESEARCH ON PEERS

Viewing peer networks as information channels

A peer network is a complex collection of social ties among individuals. These ties represent far more than social relationships. They also serve as channels through which various resources, such as emotional support and practical assistance, can be transmitted. At all stages of the life course, one important resource that flows through network ties is the information that helps individuals manage their lives. Thus, a person in a large, nondense social network will have more diverse, nonredundant sources of information on which to draw when making important life decisions. For example, sociologists and economists have demonstrated that individuals have better chances of getting jobs or getting into college when they can gather knowledge, tips, and inside secrets from friends and, especially, friends of friends in their social networks.42,43 This basic model of peer networks as information channels is echoed in US-based and international public health research on the vectors of human immunodeficiency virus (HIV) transmission. This body of research has documented that peer networks can help block HIV transmission and encourage HIV treatment among diverse groups of young adults when they transmit accurate information about the disease in an emotionally and instrumentally supportive way.25,44-46

To benefit adolescent-focused public health research in the future, the lessons of this multidisciplinary research on the role of peer networks as information channels cannot be applied too narrowly. Information does not simply refer to facts about the dangers of some health risk behavior or the merits of some healthy behavior. Moreover, the sharing of information among peers does not simply include one adolescent repeating the health education literature to a friend.47 Instead, the information channeled through peer networks that is most likely to influence adolescent health concerns modes of access to desired commodities and opportunities. These modes of access include, but are not limited to, invitations to parties where health risk behaviors are likely to occur, tips on where to buy or obtain alcohol if one is underage, cues about prevailing norms for high-status behavior, and advice on how to avoid being caught and punished for engaging in behavior that is prohibited by parents and other adults.47-49 The same peers who share health education information in the classroom may facilitate risk behavior outside of class.

Future public health research, therefore, should seek to identity the specific ways that peers transmit health-related information to each other in their everyday lives, either through face-to-face contact or through social networking activities on the Internet (eg, Myspace). Such a knowledge base could then be used to rethink peer-mentoring programs as well as health education more generally.

Locating powerful and vulnerable adolescents in peer networks

Every adolescent in some given peer context (eg, a friendship network, a club, a school, a neighborhood) is neither equally powerful in setting the health-related norms of that context nor equally vulnerable to the influence of these health-related norms.4,5 Studying such individual variability within peer contexts is important not only for theoretical reasons but also for practical reasons.

For example, consider 2 adolescents in a neighborhood peer context. The first has many friends in that neighborhood. The second has far fewer friends in that neighborhood but is desired by nearly everyone as a friend. Which of these 2 adolescents will play the biggest role in defining the norms of health-related behavior in that neighborhood peer context: the one with the greatest number of social ties or the one with the greatest amount of social status? As another example, consider a third adolescent who is on the margins of the same neighborhood and is repeatedly ostracized by other youth. Will this marginalized adolescent be more likely to internalize and follow prevailing norms about health behavior in that neighborhood than the centrally located adolescent or the high-status adolescent mentioned above? The first adolescent may be motivated to conform by a desire to fit in or avoid victimization, but the latter 2 have much more to lose from bucking neighborhood peer conventions about health-related behavior.

These questions concern the intersection of centrality, status, and isolation in adolescent-peer networks.50 Researchers are starting to address this intersection more thoroughly.4,48,51 This research, however, needs to be extended more explicitly from general behavior to actual health-related behavior. Such an extension should facilitate the identification of the types of adolescents who will be the most effective agents of peer-based health promotion efforts, those who will be most in need of these programs and interventions, and those who will be most affected by these programs and interventions.

Linking peer relations to larger settings and structures

One logical conclusion that can be drawn from reviewing the extant literature on peer relations and adolescent health is that manipulating the norms and values of peers-or the sensitivity of adolescents to these norms and values-should produce changes in adolescent health-related behavior. The problem with this conclusion, of course, is that adolescent-peer relations and adolescent-peer orientations are both extremely difficult to manipulate from the outside.52 In other words, just because peer relations are important to adolescent health does not mean that they can be easily incorporated into public health policy and practice. The challenge for public health researchers, then, is to identify useful, doable ways that adolescent-peer relations can be targeted by public health policy.

One model for meeting this great challenge can be found in educational research. The power of peers to influence adolescent academic engagement, effort, and achievement has long been recognized. Yet, historically, evidence-based educational policy has rarely dealt with peer relations for precisely the reasons laid out above.52 Eventually, however, many policy-oriented educational researchers concluded that peer relations were too important to the educational process to ignore and began to search for ways that they could be incorporated into policy interventions. A particularly influential strategy involved identifying the classroom and school characteristics that were most closely associated with academically oriented and supportive peer contexts. Using policy to encourage and replicate these classroom and school characteristics can lead to changes in the peer contexts to which young people are exposed on a daily basis. Research on Catholic schools, small schools (eg, schools within schools), and classroom cooperative learning offer vivid examples of the utility of this strategy.53-55

The lesson to be learned from this line of educational research, therefore, is that changing settings and structures that organize adolescent peer relations may indirectly affect adolescent behavior. Perhaps this same phenomenon applies to adolescent health-related behavior. Certainly, schools differ in the prevailing health-related norms and values of their peer contexts.51,56 These differences can be exploited by public health researchers. For example, observational research measuring the correlations among various school structures, student body compositions, programs, curricula, and services on one hand and different patterns of peer relations on the other could be followed by qualitative studies to identify mechanisms of social influence, which could be followed by experimental studies intended to enhance causal. The focus need not be solely on schools. Indeed, a similar model could be developed for after-school programs, community-based sports teams, neighborhood associations, and other organized settings for young people that tend to double as adolescent-peer contexts. This stepwise process systematically linking larger settings to peer relations to adolescent health-related behavior will undoubtedly take a long time. Ultimately, however, such a course of investigation will prove valuable if it allows the power of peer relations to be more effectively harnessed in public health practice.

Importantly, the body of evidence arising from this stepwise investigation could be valuable even if large-scale policies organizing and reforming schools and other social settings are not enacted. Specifically, it could serve as a source of advice and information for interventionists designing adolescent health programs. A large portion of these programs are school-, community-, or activity-based. Consequently, practitioners could be helped by knowing something about the peer cultures and network structures of these settings for their interventions and programs. Yet, the costs-in money, time, and stakeholders' patience-of gathering this information are invariably prohibitive. If these practitioners could draw on a sound literature identifying the kinds of peer networks most likely to be found in different types of schools, neighborhoods, or activities, they could then make scientifically supported probabilistic assumptions about how to adapt their programs.

Drawing links between adolescent-peer relations and the larger social structures and settings in which they are embedded, therefore, could support adolescent health promotion on many levels. This activity could point to new, large-scale policies or help improve existing small-scale programs.

Illuminating unintended consequences of adolescent health intervention

One important role of public health research is the identification of potential problems that may undermine or disrupt well-meaning, well-designed health promotion policies and programs. Three themes from contemporary research on peer contexts of adolescent health-related behavior demonstrate why this role will be so important as adolescent health promotions increase in number.

First, grouping together multiple at-risk youth in need of some behavioral intervention can actually make things worse for these youth by creating a new peer culture organized around the very behaviors that the intervention was trying to change.57,58 In other words, incorporating peer-based strategies into adolescent health programs may make things worse, not better, if done without sufficient consideration of peer dynamics.

Second, the effectiveness of adolescent sex interventions, including the abstinence programs that have garnered such attention in recent years, can be undermined when participation becomes a mechanism for adolescents to engage with peers, enter peer networks, or gain status among peers.4,59 In other words, the mission of these adolescent health promotions is undercut when adolescents join them for the wrong (eg, social) reasons.

Third, in some schools and communities, a risky adolescent health-related behavior, like drinking, can also be a highly valued activity among peers.51 In such circumstances, programs aiming to reduce this behavior will be hamstrung from the beginning if they do not recognize that quitting this behavior could result in social marginalization.

A more careful explication of how considerations about peer relations factor into adolescents' decisions about health-related behaviors and enrollment in programs designed to curtail those behaviors has the power to inform adolescent health promotions. Importantly, the programs that benefit may not even use peer-based strategies at all. Worth stressing is that public health researchers must do more than conduct studies to create this knowledge base. They must learn how to disseminate the findings of these studies to practitioners and stakeholders in accessible ways.

CONCLUSIONS

Our objective in writing this essay was to demonstrate the importance of the emergence of peer relations in adolescent-focused public health research and to advocate that this activity continues into the future. On a theoretical level, such research will cultivate better understanding of why adolescents engage in both healthy and unhealthy behaviors. On a more applied level, it will allow practitioners and policy makers to more explicitly utilize the power of peers in health interventions and to implicitly recognize this power when designing non-peer-focused interventions. These goals are doable. The methodologies for studying adolescent peer relations (eg, experience-sampling method, network techniques) are well developed, the pros and cons of different measurement techniques (eg, self-reports vs sociometrics) are well known, and appropriate data are available (eg, National Longitudinal Study of Adolescent Health). Thus, public health research is clearly in a good position to harness the power of peers to advance efforts for adolescent health promotion.

REFERENCES