Predictors of Change in Attachment Security from Adolescence to Adulthood

PREDICTORS OF ATTACHMENT CHANGE1

Leah A. Grande

Distinguished Majors Thesis

University of Virginia

November 26, 2014

Advisor: Joseph Allen

Second Reader: Noelle Hurd

Abstract

This study examined factors that predicted relative changes in attachment security, as assessed via the Adult Attachment Interview, from ages 14 to 24 among 131 individuals. Adolescents whose parents demonstrated high levels of interparental hostility and who were verbally abusive toward the adolescent at the time of the first AAI collection experienced a relative decrease in attachment security over time. Conversely, youths who viewed their parents as accepting and who demonstrated engagement, assertiveness, and relatedness on tasks with parents experienced a relative increase in attachment security. Adolescents with high quality close friendships, who engaged effectively with a close peer on an advice task experienced relative increases in attachment security over time. Finally, adolescents who experienced broad dislike and were unaffected by and uninterested in social pressures became relatively less securely attached over this time period. These findings suggest certain qualities of functioning in familial and peer relationships that increase in importance over adolescence and have long-term implications for the attachment system.

Keywords: attachment security, AAI, adolescence, peer relations, parental relations

Predictors of Change in Attachment Security from Adolescence to Adulthood

Since the development of Mary Ainsworth’s Strange Situation, researchers have identified many of the positive outcomes of a secure attachment style, including lower levels of internalizing disorders, such as depression and anxiety, and lower reports of externalizing behavior, such as delinquency, aggression, and risky sexual behavior (Allen, Moore, Kuperminc, & Bell, 1998; Allen, Porter, McFarland, McElhaney, & Marsh, 2007; Allen et al., 2002; Dawson, Allen, Marston, Hafen, & Schad, 2014; Kobak, Zajac, & Smith, 2009). The securely attached individual is socially competent, well liked, and forms higher-quality, lower-stress peer and romantic relationships (Allen, Moore, et al., 1998; Allen et al., 2002; Allen et al., 2007; Dykas, Ziv, & Cassidy, 2008; Shomaker & Furman, 2009).

Although the outcomes and correlates of attachment security, including characteristics of peer and parental relationships, have been well researched, less work has been done examining factors that predict discontinuity in attachment security, especially discontinuity over a longer period of time. Although the longitudinal nature of attachment change research is not sufficient to support the inference of direct causal relationships, the identification of potential causal- factors is essential in our understanding of attachment and its therapeutic applications. Even though an understanding of infant attachment is important, early attachment security does not necessarily imply invulnerability to later attachment insecurity. For the insecurely attached child there is hope, especially if researchers can map out a pathway to future security.

Factors that predict change in attachment are theoretically distinct from those simply associated with attachment. Factors associated with attachment at multiple times could be causes or effects of attachment, or driven by third variables that also influence attachment. By contrast, factors predicting change are characteristics that ‘come online’ or become developmentally important over time. These are the strongest candidates for potential influences in the way models of attachment change over time.

In terms of potential predictors of the secure-to-insecure transition, theory has focused on family and peer factors that could overwhelm the attachment system (Allen, McElhaney, Kuperminc, & Jodl, 2004). Stressful events related to health, death of a family member, parental divorce, and conflicts with friends have been correlated with decreases in attachment security from infancy to adolescence (Hamilton, 2000; Van Ryzin, Carlson, & Sroufe, 2011). Researchers have identified the most serious challenges to an adolescent’s attachment system, and by extension the affect regulation system, as being threats to autonomy, relatedness, and future competence as an adult (Allen & Hauser, 1996; Ryan & Deci, 2000).

In terms of potential predictors of the insecure-to-secure transition, theory has focused on autonomy and relatedness in parental relationships and attachment-like qualities in peer relationships. Adolescence is a time during which substantial change in attachment organization is possible. Developing cognitive abilities, ‘epistemic’ distance from parents, and new peer relationships allow teens to re-evaluate attachment relationships with caregivers (Allen & Tan, in press). Conversely, failure in establishing autonomy with parents and lack of close peer relationships could prevent the continuation of a secure model of attachment. The familial and peer factors that could influence attachment in both a positive and negative direction are discussed below.

Parental maltreatment is a significant correlate of attachment continuity and change with respect to infancy to late adolescence (Weinfield, Sroufe, & Egeland, 2000; Weinfield, Whaley, & Egeland, 2004). Paternal harsh conflict tactics are also associated with security in adolescence (Allen et al., 2007). However there have been mixed findings as to whether maltreatment creates change in attachment or instead is only associated with continuity in insecurity from infancy to adulthood (Hamilton, 2000; Schneider-Rosen, Braunwald, Carlson, & Cicchetti, 1985; Weinfield et al., 2000; Weinfield et al., 2004). Parental maltreatment can create stress to the adolescent’s affect regulation system, while preventing the teen from resolving the stress through attachment processes, given that the caregiver is the cause of the distress (Allen & Tan, in press). This can lead to defensive processing and prevent the adolescent from reflecting autonomously on attachment relationships. Research on maltreatment has focused on parental maltreatment of the adolescent, but parents’ verbal and physical abuse within their own relationship can also create emotional and physiological stress on the teen (Lucas-Thompson, 2012). When unresolved this stress could cause dysregulation in the teen’s attachment system. Interparental hostility could also model negative modes of relating to others in attachment relationships, which the adolescent may then go on to imitate with peers and romantic partners (Kouros, Cummings, & Davies, 2010; Miga, Gdula, & Allen, 2012).

Whereas parental maltreatment negatively predicts attachment security, parental relatedness and supportiveness appears to predict future security, at least in the short term. Displays of relatedness by mother and teen during a disagreement, including validating and agreeing, asking questions, and engaging in the conversation, were positively related to adolescent security (Allen et al., 1996; Allen et al., 2003; Allen et al., 2007). Relationship-maintaining behaviors in the midst of conflict are consistent predictors of adolescent security, especially when the behavior is that of the adolescent (Allen & Tan, in press). However, even though these factors are predictive of future security, research has not examined whether they are predictors of attachment change. Although maternal relatedness at age 14 may predict adolescent security at age 24, if maternal relatedness is similarly correlated with adolescent security at ages 14 and 24 and attachment is simply stable over time, then it may not be accounting for any of the change in attachment security over this time period. Adolescents’ perceptions of their mothers’ supportiveness is both an important marker of attachment security and predicts attachment security change from ages 16 to 18 (Allen et al., 2003; Allen et al., 2004). However, this is change over a relatively short time period, which may not be generalizable to time frames of a decade or longer (Allen et al., 2004). If these adaptive parental factors do predict change in attachment security over time, it is unlikely that this is because they are simply markers of a supportive parental relationship. Rather, a supportive parental relationship may serve as a secure base through which to safely explore autonomy, which becomes an increasingly important process throughout adolescence.

Secure adolescents can rely on parental relatedness as they safely strive for autonomy. Conversely, insecure adolescents’ autonomy strivings can threaten already weak parental relationships, disrupting the attachment system. Secure teens are able to maintain balanced assertiveness with their mothers; attachment security shows a curvilinear relation with maternal dominance in adolescence (Kobak, Cole, Ferenz-Gillies, & Fleming, 1993). Autonomy-undermining enmeshed and over-personalizing parent-teen interactions are cross-sectionally linked to insecurity in attachment (Allen & Hauser, 1996; Dozier & Kobak, 1992) and are also predictive of relative decreases in security over a two-year period in adolescence (Allen et al., 2004).

New research suggests approximately half of the variance in adolescent attachment security can be attributable to non-shared environmental factors, in other words, the unique environmental experiences that make even monozygotic twins different from one another (Fearon, Schmeli-Goetz, Viding, Fonagy, & Plomin, 2013). Surprisingly, the influence of shared environment is negligible. It is important to keep in mind that the parental characteristics predicting discontinuity in adolescent attachment may be ones that are specific to the relationship between the adolescent and parent, rather than more broad-based measures of family functioning. Peers are also a prime example of how non-shared environmental factors can influence development.

Even if an adolescent is undergoing autonomy and relatedness struggles with parents, new peer relationships can provide an avenue to learn about and re-evaluate attachment relationships. There are three different ways of assessing how peer relationships might invoke a change in attachment models. The first is to look at the quality of close peer relationships, as attachment models are most sensitive to functioning in intimate relationships. Developing attachment-like qualities in close relationships, such as intimacy and emotional support, could help a teen develop a more secure model of attachment relationships. Adolescent attachment security has been significantly associated with the quality of interactions with close friends (Black, Jaeger, McCartney, & Crittenden, 2000; Shomaker & Furman, 2009; Zimmermann, 2004). A meta-analysis found that the effect size of the correlation between attachment security and peer relations in school-age children and adolescents was largest for studies focusing on functioning in close friendships, as opposed to peer relations in large groups, although a follow-up study failed to replicate this (Pallini, Schneider, Baiocco, & Madigan, 2014; Schneider, Atkinson, & Tardif, 2001).

The second approach is to look at broad-based measures of peer relations. Acceptance in a peer group could increase a teen’s network of potential friends and also increase the likelihood that peers will be interested in forming a close relationship with the teen. Skills supporting acceptance into a large peer group, such as social competence, may also be the same skills supporting the formation of close friendships (Schneider et al., 2001). Adolescent security has been associated with broad-based measures of social competence, social acceptance, popularity, and prosocial behavior (Allen, Moore, et al., 1998; Allen et al., 2007; Dykas et al., 2008; Groh et al., 2014). Regardless of whether close friendships or broad peer relations are more strongly correlated to attachment security, both appear to be significantly, and perhaps uniquely, related to models of attachment. Both are important candidates for influencing change in attachment models over time.

The third perspective is to look at whether teens can form close relationships with peers while also maintaining their autonomy, similar to the process of establishing parental autonomy. Teens who are establishing attachment-like qualities in their peer relationships should also be able to use their peers as a secure base. These teens may then be better able to resist external pressures and may also experience less peer pressure within their close relationships (Allen et al., 2007).

In cross-sectional research over a three-year period in adolescence, security was best predicted by combining all three of these perspectives, including observations of adolescents’ calls for emotional support to a best friend, adolescents’ reports of overall quality of relationships with peers, a sociometric measure of popularity, and (inversely) peers’ reports of the amount of peer pressure experienced by the adolescent (Allen et al., 2007).

Although very little research has identified predictors of attachment change, especially over long time periods, there has been research to at least suggest potential predictors. Past research can also help suggest factors that are associated with the attachment system, but not predictive of change. For example, the behavioral correlates of attachment security, such as externalizing disorders, are more likely outputs of the attachment system rather than moderators of attachment change. Attachment insecurity has been shown to predict increasing levels of externalizing disorders and delinquency over short-term (two year) and long-term (up to eight year) time periods (Allen et al., 2002; Allen et al., 2007; Dawson et al., 2014; Kobak et al., 2009). This suggests that insecurity is not only correlated with externalizing behavior, but that it maintains and increases the strength of this correlation over time. Externalizing behavior thus appears to be strongly linked to the attachment system, but not a likely candidate for attachment change.

This study examined predictors of change in adolescent attachment security over a decade, from mid-adolescence to mid-twenties. An ethnically and socioeconomically diverse community sample of adolescents was selected to include a broad range of psychosocial and family functioning. This study independently assessed family (e.g., interparental conflict, parental verbal abuse, acceptance, engagement, relatedness, and autonomy) and peer factors (e.g,. qualities of close relationships, supportive behavior with a close friend, popularity, peer autonomy) for their capacity to predict changes in attachment security over time, relative to baseline levels of security.

It was predicted that parental hostility and maltreatment, whether directed towards the teen or to a spouse/partner, would predict relative decreases in attachment security over time. Second, the extent to which teens viewed their parents as secure base, including reports of parental acceptance, observed engagement and assertiveness when asking the parent for advice, and observed relatedness with parents while discussing a disagreement, would predict increases in attachment security. Finally, it was hypothesized that a teen’s ability to demonstrate autonomy while discussing a disagreement with parents would predict increases in attachment security.

Hypotheses for peer relationships were developed along a similar framework. It was hypothesized that a teen that could establish close friendships and use these friendships for attachment functions, such as effectively receiving support from a friend, would develop a more secure model of attachment over time. A teen that was well-liked by his or her peers would be more likely to form these close friendships and show increases in attachment security. Finally the ability to establish autonomy in peer relationships and resist peer pressure would positively predict attachment security.

In addition to identifying factors predictive of attachment change, it was also of interest which of these factors first demonstrated relationships to attachment security after age 14. In other words, which factors were uncorrelated with attachment security at age 14, but predicted attachment change and were correlated with attachment security at age 24. It was predicted that peer factors would be the most likely new factors, because peer relationships grow in importance and have the potential to serve attachment-like functions during adolescence.

This study was also interested in identifying factors that were correlated at both age 14 and 24, but not predictive of change. These factors only predict security at age 24 when security at age 14 is not taken into account. These qualities are markers of attachment, rather than predictors. They are very closely aligned with the attachment system and do not vary in importance from adolescence to adulthood. It was hypothesized that the behavioral manifestations of attachment security, such as externalizing behaviors (eg. delinquency, hyperactivity, hostility), might be associated with attachment at both ages, but not account for change in the attachment system

Method

Participants

This report is drawn from a larger longitudinal investigation of adolescent social development in familial and peer contexts. 175 seventh and eighth graders (82 male, 93 female) were assessed via the Adult Attachment Interview (AAI) at a mean age of 14.81 (SD = .55) and 131 (59 male, 93 female) were assessed again at a mean age of 23.46 (SD= 1.04). This study also drew on assessments from a four-year period roughly surrounding the first AAI collection (at ages 13.36, SD= .62, at Wave 1; 14.29, SD= .75, at Wave 2; 15.22, SD= .80, at Wave 3; 16.35, SD = .87, at Wave 4).

The sample was racially/ethnically and socioeconomically diverse: 107 adolescents identified themselves as Caucasian (58%), 53 as African American (29%), and 15 as being from other and/or mixed ethnic groups (13%). Adolescents’ parents reported a median family income in the $40,000-$59,999 range (M = 43,900, SD = $22,500).

At each wave, adolescents’ nominated their closest, same-gendered friend to be included in the study as well as an additional two peers from their extended circle of friends and acquaintances. Close friends reported that they had known the adolescents for an average of 4.39 years (SD = 3.24) at the second wave of data collection and an average of 5.26 years (SD = 3.45) at the third wave.

Adolescents were recruited from the seventh and eighth grades of a public middle school drawing from suburban and urban populations in the Southeastern United States. Students were recruited via an initial mailing to all parents of students in the school along with follow-up contact efforts at school lunches. Families of adolescents who indicated they were interested in the study were contacted by telephone. Of all students eligible for participation, 63% agreed to participate either as target participants or as peers providing collateral information. All participants provided informed assent before each interview sessions, and parents provide informed consent. Interviews took place in private offices within a university academic building.

Teen reports regarding parent behavior were available for 165 mothers and 150 fathers. In addition, observations of parent-teen interactions were obtained with 147 mothers and 78 fathers (primarily those fathers who resided with their teens). Observations of interactions with a close friend were also obtained for 155 teens, along with reports from that friend about the teen. Finally, peer sociometric data were obtained for 166 of the teens in the sample.

Procedure

In the initial introduction and throughout all sessions, confidentiality was assured to all study participants, and adolescents were told that their parents would not be informed of any of the answers they provided. Participants’ data were protected by a Confidentiality Certificate issued by the U.S. Department of Health and Human Services, which protected information from subpoena by federal, state, and local courts. Transportation and childcare were provided if necessary. Interviews with current closest peers were conducted between Waves 2 and 3 and again at Wave 4. Interviews with parents were collected at Wave 4. Paper measures completed by parents, close peers, and teens were collected Waves 1-4. Attachment data were obtained between Waves 2 and 3 and then again between Waves 12 and 16. Adolescents, their parents, and their peers were all paid for participation. Payment amounts increased over the course of the study but ranged from $20 to $50 per person for a two-hour visit.