PHS 398 (Rev. 9/04), Continuation Page

Principal Investigator/Program Director (Last, First, Middle): Allen, Joseph P.

This competing continuation proposal examines a developmental model of adolescents’ peer and family experiences as predictors of long-term psychosocial outcomes in adulthood. The ongoing study represents one of the most intensive extant observational assessments of adolescent social development within the family and peer group. It includes in-depth, multi-method assessments of 184 adolescents obtained annually from ages 13 to 22, including sociometric assessments, repeated observations of teen-parent and teen-peer social interactions, and interview data from teens, parents, close and distal peers, and teen romantic partners.

We propose to follow this original sample from ages 23 to 27 to examine theoretically-predicted links from adolescent socialization experiences with peers and parents to psychosocial functioning in adulthood. We focus upon two stage-salient tasks of adolescence—establishing autonomy and maintaining and building social bonds—and we hypothesize that the resolution of these tasks will be fundamental to successful adult psychosocial functioning. We also assess theoretically-salient cognitive, affective, and neural/structural qualities that may mediate continuities from adolescent to adult relationship quality and that may potentially serve as targets for intervention efforts.

This study is designed not only to advance our theoretical understanding of adult social development, but also to add an essential component to our understanding of adolescent development. We build on the vast body of research that has examined relational predictors of functioning within adolescence, but we now seek to extend and ground this research by determining which of the many qualities of social relationships that have been identified as important within adolescence are actually predictive of long-term, life outcomes in adulthood. We organize our efforts into three primary Aims identifying long-term relationship continuities, their mediators, and their ultimate functional outcomes:

Aim 1: Parent and Peer Predictors of Adult Relationship Quality: We begin by seeking to identify critical continuities in adolescent relationship qualities that persist into early adulthood. We focus upon identifying predictors of aspects of adult relationships—particularly social isolation and hostility—known to be closely linked to both mental and physical health. We will assess theoretically hypothesized continuities across time, across relationship types, and across stage-salient patterns of behavior, in an effort to identify the qualities of adolescent relationships with the greatest import for adult psychosocial functioning.

Aim 2: Cognitive, Affective, & Neural Mediators of Continuities from Adolescence to Adulthood: We next examine the processes within the individual that potentially mediate the continuities identified in Aim 1. We focus on three classes of mediators that each have the potential to register and carry the effects of critical adolescent experiences on future adult functioning. These include the individual’s attachment organization, rejection sensitivity, and pre-conscious neural biases toward avoidance in conflict situations.

Aim 3: Early Adult Adaptational Outcomes: Direct & Mediated Pathways from Adolescence: Ultimately, we are interested in the ways in which relationship continuities and internal mediators identified in the first two Aims “flow through” to predict broader qualities of adult mental health and functioning. In this Aim, we consider adolescent-era peer and family predictors of three key adult outcomes: internalizing symptoms, externalizing symptoms, and the development of functional autonomy (e.g., the capacity to manage one’s own career, financial, and residential needs). We assess both direct predictions from adolescent relationship qualities, as well as predictions mediated via adult relationship qualities (Aim 1) and intrapsychic factors (Aim 2).

We address all of these questions with observational and multi-reporter data from a socio-demographically heterogeneous final sample of 172 adolescents, their parents, peers, and romantic partners followed across a fifteen-year span. Overall, this study seeks to identify essential qualities of adolescents’ relationships with peers and parents that predict long-term success or failure in adult functioning so as to inform: a) parents, educators, and clinicians working with adolescents and early adults; b) interventions targeting parenting behaviors and/or peer influences (e.g., delinquency, aggression, and substance abuse prevention and treatment programs); and c) developing theories of the links between social relationships and functional outcomes across this critical portion of the lifespan.

Early adulthood appears as a critical choice point in social development. The decrease in normative role structures in this period is dramatic as individuals move from the universal schooling of adolescence to diverse paths that may include marriage, full-time employment, residence on one’s own, with parents, with peers or with a spouse, and/or continued educational pursuits (Arnett, 2000; Schulenberg et al., 2006). Residential and geographic instability soar during this period, bringing rapid changes in peer networks (Osgood et al., 2005; Schulenberg et al., 2005). Although these transitions are often positive, they become a source of stress for those who have reached this point less prepared, and not surprisingly the incidence of mental health problems increases steadily during this period (Gore & Aseltine, 2003; Osgood et al., 2005; Schulenberg et al., 2005; Schulenberg et al., 2006).

The costs of failures to establish positive social relationships in early adulthood are staggering. As the structured interactions of secondary schooling give way to the more diffuse social contacts provided by work or higher education, social isolation becomes a major (albeit rarely studied) risk. Remarkably, meta-analyses indicate that adult social isolation creates a greater overall risk of future mortality than does cigarette smoking (House et al., 1988), with effects ranging from endothelial injury to disturbances in endocrine and immune systems (Cacioppo et al., 2002; Knox & Uvnas-Moberg, 1998; Phillips et al., 2006; Uchino et al., 1996). Social isolation has been linked to numerous psychosocial problems as well, ranging from anxiety, depression, and risk of suicide, to long-term family problems (Barber et al., 2001; Kawachi, 2001; Newcomb & Bentler, 1988). Levels of adult social isolation have increased dramatically in the United States over the past two decades, with 1 in 4 adults now reporting they have no one with whom they can discuss important matters in their lives—a phenomenon that crosses both gender and racial/ethnic lines (McPherson et al., 2006).

Although lack of relationships can be debilitating, the presence of relationships characterized by high degrees of hostile, conflictual behavior has equally troubling long-term ramifications. Enduring patterns of hostility in relationships are linked not only to multiple DSM-IV Personality Disorder diagnoses, work difficulties, and risk of completed suicide (Duberstein et al., 1996; McCann et al., 1997; Rutter & Behrendt, 2004; Virtanen et al., 2005), but also to negative physical health outcomes ranging from coronary heart disease to immunological suppression (Kiecolt-Glaser et al., 1993; Smith et al., 2003; Whiteman et al., 2000). The magnitude of the problem is reflected in current estimates that 35% to 65% of young adults have been involved in romantic relationships marked by violence between partners—rates that increase steadily from late adolescence into the early twenties (Magdol et al., 1997).

Developmental Continuities from Adolescence to Adulthood

To fully understand and prevent these negative adult outcomes requires understanding their links to the developmental processes that preceded them. This proposed continuation is thus explicitly designed as a study of both adolescence and adulthood. Most of our theoretical understanding of the principles of adolescent development—from the push for autonomy to the intense focus on peer relationships—presumes that this development is ultimately oriented toward preparing the adolescent to meet the challenges of adulthood. Given the difficulties of conducting lifespan studies, however, the vast bulk of adolescent research does not examine these actual adult outcomes, but rather identifies predictors of functioning within adolescence. A critical challenge for our field, particularly in a world with limited resources for intervention, is to distinguish those adolescent behaviors that are temporarily disturbing or troublesome from those that predict serious dysfunction in adulthood. Research on adult outcomes thus becomes essential for understanding the long-term developmental import of many adolescent social experiences.

The proposed study will test a developmental conceptualization that considers relationship problems in early adulthood as linked to two critical tasks in adolescence: the need to establish autonomous decision-making ability and the importance of developing and maintaining supportive social bonds. Figure 1 below provides a broad heuristic overview of this conceptualization.)

The experience of managing conflict in relationships so as to develop personal autonomy while maintaining relationships has long been identified as a key feature of adolescent development (Allen et al., 1994a; Hill & Holmbeck, 1986). The primary reason that autonomy issues—assessed as verbal and emotional autonomy during adolescence—are considered so central to adolescent development is that the adolescent is in the midst of actively preparing to function autonomously as an adult. Remarkably, however, almost no research to date has examined the extent to which putatively critical adolescent verbal and emotional autonomy processes actually lead into autonomous adult functioning.

Similarly, establishing supportive social relationships beyond the family is a central task of adolescence that has also been theoretically linked to emotion regulation and social functioning across the lifespan (Berscheid, 2003; Collins, 1997). Ironically, the same factors that increase the need for autonomous decision-making and social support in early adulthood (e.g., geographic mobility, residential changes, major impending life decisions) also make it more challenging to form and utilize stable close relationships to provide that support. Thus, while the ability to establish supportive peer relationships is of great urgency during adolescence, the template that these relationships establish may become truly critical in early adulthood as peers become primary sources of support. Surprisingly, this transition from adolescence to adulthood has also been only minimally studied.

In assessing continuities from adolescence to adulthood, we adopt a critical developmental tasks perspective on long-term continuities (Sroufe, 1997). This perspective recognizes that it is under developmental stress—facing a conflict, a new social situation, or a personal challenge—that profound difficulties in functioning are most likely to become manifest and lead to long-term negative outcomes. We also recognize that to optimally inform both intervention and theory development, we need to not only identify long-term continuities but also to begin to understand the processes that mediate these continuities.

In the proposed study, we begin by seeking to identify several specific, fundamental continuities in autonomy and connection processes in relationships across a fifteen-year span from adolescence to adulthood (Aim 1). We then examine affective, cognitive, and neural mediators of the continuities we observe (Aim 2). Ultimately, we consider both relationship processes and intrapsychic mediators as predictors of early adult mental health and functioning (Aim 3).

Aim 1: Parent and Peer Predictors of Early Adult Relationship Quality

Strikingly, almost all prospective studies of continuities in relationship quality from adolescence onward leave off by age 23. This is the starting point of this continuation and an age at which many adults are just beginning to make autonomous decisions with lifelong ramifications, and to form close relationships that will be potentially lifelong in nature. Our first Aim is designed to clarify the primary types of continuity that exist in relationship qualities from adolescence into adulthood. As outlined in Figure 2 below, we propose a series of theoretically-derived hypotheses about continuities across relationship types, across different types of behavior, and across developmental epochs.

Hypotheses for Aim 1:

1.1 Continuities in relationship qualities will be observed: from parent-teen relationships in adolescence to peer relationships in adulthood; and from non-romantic peer interactions in adolescence to romantic partner interactions in adulthood.

In addition, specific behavioral continuities will be observed from:

1.2 Adolescent failure to establish autonomy in conflicts to hostility in adulthood;

1.3 Adolescent failure to obtain social support to both hostility and social isolation in adulthood.

Continuities Across Relationship Types. The transition from adolescence to adulthood is accompanied by dramatic changes in both peer and parent relationships. Peer relationships evolve from adolescent chumships to lifelong romantic relationships, while relationships with parents are transitioning from dependent relationships to equal (or even caretaking) relationships. We thus do not expect continuities across this period necessarily to be homotypic in nature. Rather, consistent with our focus on the critical developmental tasks of this period, we expect to find heterotypic continuity in which success meeting salient social developmental tasks in adulthood is best predicted by success meeting developmentally analogous tasks of adolescence—even if these play out in different relationships with different challenges across these two eras.

We are increasingly finding that within adolescence, for example, peer relationship qualities do not emerge de novo, but rather display striking continuities with parent-teen relationship qualities in domains ranging from warmth in interactions to management of conflicts (see Progress Report). We hypothesize that this transfer from parent to peer interaction qualities will continue into adulthood.

Similarly, we do not expect adult romantic relationship qualities necessarily to be best predicted by qualities of early adolescent romantic relationships. In part, romantic relationships in early adolescence are likely to lack the longevity necessary to allow true give-and-take intimacy and autonomy negotiations to fully emerge, and the onset of these relationships appears more determined by local school and neighborhood norms than by developmental status (Dornbusch, 1981; Roisman et al., 2004). We do, however, expect that qualities of intimacy and autonomy in non-romantic close friendships and in parent-teen relationships will tap developmental processes with great relevance for adult romantic relationship quality. Initial evidence suggests links between parenting and later romantic relationship functioning (Donnellan et al., 2005). Although peer predictors of this adult relationship functioning have not been examined, this proposed link finds support from studies within adolescence in which peer interactions outside of a romantic context predict qualities of behavior within romantic relationships (Conger et al., 2001; Connolly et al., 2000a; Furman & Wehner, 1994; Prinstein & La Greca, 2004). Within early adulthood we also recognize that for many individuals, this is appropriately a time of exploration. Thus, again using our developmental perspective, we focus not on outcomes (e.g., marriage) but rather on the task of engagement in relationships characterized by significant intimacy and support and by absence of severe levels of hostile conflict and aggression.