Early Roots of Adult Competence: The Significance of Close Relationships from Infancy to Early Adulthood

Abstract

Social capital has traditionally been defined in terms of the amount of resources that one derives as a result of a diversity of interpersonal relationships. However, the quality of these relationships across development has not been examined as a contributor to social capital and few studies have examined the significance of various age-salient relationships in predicting adaptive functioning, especially testing for cumulative effects over time. Using data from the Minnesota Longitudinal Study of Risk and Adaptation, developmental models spanning from infancy to adulthood were tested via path modeling, linking quality of various age-salient relationships (e.g. infant-caregiver attachment, peer competence, friendship security, and effectiveness in romantic relationships) to global adaptive functioning at age 28. As hypothesized, quality of age-salient relationships during different developmental periods predicted the quality of subsequent relationships, but also showed links with adaptive functioning in early adulthood. Results also showed that the quality of infant attachment relationships not only was linked with more proximal relationships, but also had direct effects on global functioning, suggesting the potential significance of early relationship quality in adaption and well-being in adulthood.

The importance of concurrent relationships as resources for individual functioning has long been recognized (e.g., Cohen & Wills, 1985; Kessler, Price, & Wortman, 1985). However, the lasting effect that early relationships have on adult psychological adjustment and well-being has only recently received empirical attention (Ellis, Figueredo, Brumbach, & Schlomer, 2009). Attachment theory, with its emphasis on early caregiver-child relationships, is an ideal conceptual framework for understanding the enduring effects of early relationships, how relationships in early years interact with salient relationships in each subsequent life period, and how they influence adaptive functioning across the life-course (Boyce, 1985). This perspective proposes that early relationship experiences are integrated and provide the foundation for both future relationships and the negotiation of later developmental tasks (Sroufe, Egeland, Carlson, & Collins 2005; Sroufe & Waters, 1977). More specifically, the quality of early caregiving and subsequent important close relationships with, for example, romantic partners and peers form the basis of an individual’s belief in the availability of resources and the ability to utilize those resources when needed (Carlson, Sroufe, & Egeland, 2004). While attachment pertains, in part, to resource use in individual relationships (Main, Kaplan, & Cassidy, 1985), social capital refers to resources available as a function of involvement in a social group (Coleman, 1988). These individual and group orientations, in turn, undergird adaptation to varied roles and tasks in adult life. In other words, current adaptation results, in part, from the interplay between early relationship history and concurrent individual circumstances.

In the present study we investigated the contributions of significant close relationships across development to global functioning and well-being in early adulthood. In our view, the concept of social capital refers to the individual’s ability to utilize resources in the social environment to function effectively in early adult roles. The extent to which these resources are available, and the individual’s willingness to utilize them, can be predicted by relationship history across development (Coleman, 1988; Sroufe & Waters, 1977). We view social capital as multidimensional in that both the extent of group membership (i.e., breadth of relationships) and the quality of individual salient relationships (i.e., depth of relationships) play key roles in the available social capital from which an individual can benefit. Several lines of research suggest mechanisms through which prior relationships influence characteristics of current relationships. Research on attachment in adult relationships indicates that the quality of salient relationships in childhood and adolescence are related to the quality of relationships in adulthood (Allen & Hauser, 1996; Simpson, Collins, Tran, & Haydon, 2007). These and other similar findings imply that the depth of social capital is related to the individual’s history of interpersonal relationship functioning across development.

Individuals’ history of interpersonal relationship functioning can be traced to the infant-caregiver attachment relationship, the first salient relationship that develops in an individuals’ life. Attachment is not an individual characteristic but a characteristic of dyadic relationships (Ainsworth, 1989; Coleman, 1988). Broadly speaking, there are two classifications of infant-caregiver attachment relationships. Securely attached infants are those whose needs have consistently been met by their caregivers and who are able to effectively seek out and accept comfort. Within the context of a secure attachment relationship, these infants begin to develop the skills necessary to draw upon resources within the environment, including social relationships, to adapt and function effectively across development. In the alternative, insecurely attached infants are those whose needs have been consistently rejected or met in an inconsistent or inadequate fashion. These infants do not seek out comfort from their caregivers or are inconsolable when faced with a stressor (Ainsworth, Blehar, Waters, & Wall, 1978). Over time, these individuals are unlikely to successfully utilize environmental resources and are likely to be ineffective in their interactions with others. Internalized representations of relationships, developed from repeated interactions with caregivers in infancy, are drawn upon in the context of salient relationships later in development (Bowlby, 1973). In our view, these representations are what drive an individual’s ability to acquire and utilize social capital in adulthood.

In support of this perspective, previous research has shown that attachment style in infancy predicts the nature and quality of social interactions later in childhood and into adulthood. For instance, infants who were classified as securely attached, in contrast to insecurely attached infants, were more likely to be socially competent with their peers in toddlerhood, preschool, middle childhood, and adolescence (Sroufe et al., 2005). Attachment theory also postulates that the quality of the infant caregiver relationship influences the quality of romantic relationships in adulthood (Collins, Welsh, & Furman, 2009). Indeed, there is a large body of research in the social psychology literature linking attachment representations in adulthood to the quality of romantic relationships (see Feeney, 2008 for a review). Thus, as the individual progresses through development, he/she develops the capacity to form new, qualitatively different relationships, each building off of a skill gained previously. In childhood, attainment of a basic level of social competence allows for the cultivation of deeper, more secure friendships in adolescence. The egalitarian and reciprocal nature of friendships then sets the stage for romantic relationships in early adulthood. All of these relationships are preceded by and often co-exist with the individual’s relationship with his/her caregivers. It is this nexus of high quality relationships that can be alternatively viewed as social capital and, as such, confers on the individual the capacity for better or worse adaptive functioning in adulthood.

Competence and adaptation manifest in many different ways across developmental periods. However, competence and adaptation during one period does not arise de novo but is often preceded by adaptation in previous periods. Thus, the establishment of high-quality, age-appropriate relationships can be seen as both an indicator of successful adaptation as well as a contributor to future adaptation. For example, the developmental tasks that are salient in early adulthood include the establishment and maintenance of a romantic relationship, continuous gainful employment, maintenance of mental and physical health, and, in some cases, competent parenting (McCormick, Kuo, & Masten, 2011). Closer examination of these tasks reveals that all hinge on interpersonal relationships. While competence in romantic relationships and parenting are self-evident in their relational nature, the other two areas are less so. Work competence depends on one’s ability to amicably co-exist with co-workers, and social support plays a key role in the development of and recovery from mental and physical illness (House, Landis, & Umberson, 1988; Kessler et al., 1985). Thus, many of the salient developmental tasks of early adulthood have a relational component and as such are expected to be influenced by an individual’s social capital as manifest in his/ her history of social functioning.

The Current Study

Very little research has examined the role of social capital in predicting overall early adult functioning from a developmental perspective. The current study examines the role of social capital as represented by age-salient interpersonal relationships across development in predicting adjustment in adulthood. Our perspective on social capital derives from developmental theory and, rather than examining social capital at one point in time, is based on the assumption that social capital is accumulated across development and has its roots in salient relationships, with a special role for attachment relationships in infancy.

Method

Participants

Data are drawn from the Minnesota Longitudinal Study of Risk and Adaptation (MLSRA; Sroufe et al., 2005); an on-going, prospective longitudinal study of at-risk children and their families began in 1976. The original sample of 267 primiparous mothers was receiving prenatal care from Minneapolis Public Health clinics. The current analyses focused on a subset of the child participants who completed the age 28 assessment (N = 157; 79 males, 78 females). At the time of recruitment the mothers in the current sample (M_age = 20.7 years, age range: 15-34 years) were low income, 58% of the mothers were single, and 33% of the mothers did not have a high school degree. Non-significant t-tests indicated no differences between the attrition group and this subsample on mother’s SES, age, marital status, educational level, or risk status at the time of the child’s birth. Frequent data collection and multi-methods, including individual assessments and dyadic observations with significant relationship partners have provided information about relationship experiences and individual functioning. When considered in toto, these measures of significant close relationships across multiple developmental periods tap into both the breadth and the depth of social capital that the participants have accumulated across childhood and into early adulthood. Specifically, the breadth of social capital was assessed through examining multiple relationships, including parent-child, peer, and romantic relationships at salient developmental time points. Likewise, qualitative aspects of these close relationships were assessed to examine the depth of social capital.

Measures of Social Capital

Infant attachment security (12 and 18 months)

Parent-infant relationships were assessed using the Strange Situation laboratory procedure (Ainsworth et al., 1978) when participants were 12 and 18 months old. This procedure was designed to assess individual differences in the quality of infant-parent attachment relationship (i.e., secure or insecure) by activating infant attachment behaviors with a series of stressful infant-parent separations and reunions. A combined score for ages 12 and 18 months was derived based on whether participants were classified as insecure at both time points (0), secure at one time point (1), or secure at both time points (2). Forty-seven percent of the participants were secure at both 12 and 18 month assessments, 22% were insecure at both time points, and 31% were secure at one time point and insecure at the other time point.

Peer competence (7, 8, and 9 years)

When participants were 7, 8, and 9 years old, their teachers ranked participants’ peer competence within their class based on a description of a socially competent child. Specifically, a socially competent child was considered sociable (fairly frequent social contact with peers), accepted by their peers, having friendships, displaying social skills and leadership qualities, and understanding another child’s perspectives and desires. Teachers ranked all children in their class from the most socially competent to the least competent. Each participant’s percentile rank was calculated (taking the number of children below that child’s ranking divided by the number of children in the class and multiplied by 100) and used in subsequent data analyses. For the current study, an average of the child’s rank for all three years was used. Averaged across 7, 8, and 9 years old, twenty-two percent of the participants were in the lowest quartile in relation to their classmates in peer competence while 13% were in the highest quartile (M = 47.67, SD = 23.25).

Parent-child relationship (13 years)

At age 13 years, the quality of the parent-child relationship was coded from videotapes of participants interacting with their primary caregiver (usually their birth mother) in a series of four developmentally appropriate problem solving tasks. Trained coders rated the general functioning of the dyads in terms of the degree of balance between the needs of the relationship and the needs of the individual on a 7-point scale (M = 4.34, SD = 1.05). Higher scores indicated that the development of the individual family members was supported by the relationship, and enthusiasm, elaboration, and enjoyment in the tasks as well as a differentiation of roles was observed. Lower scores were assigned to dyads that displayed high levels of loneliness and estrangement as well as those where it appeared that maintaining the relationship took the energy of the individuals without enhancing their experiences. In other words, lower scores were assigned to those interactions that were destructive to participants’ personal development. Interrater reliability was .63 on this scale.

Friendship security (16 years)

At age 16 years, participants completed a semi-structured interview about their current, non-romantic closest friend relationships. The interview, which was audiotaped, asked participants to describe the qualities of their current closest friend in response to a set of questions. Participants were, for example, asked how easy they could share things (good and bad) with their closest friend and also how close they felt toward the closest friend. Trained raters coded each interview on a friendship security scale. This 7-point scale reflected the extent to which each adolescent reported feeling as if s/he could be her/himself in that friendship and expected his/her closest friend to be available and supportive (M = 4.32, SD = 1.45). Higher scores were assigned to participants who reported they could be wholly themselves without hiding or withholding information and who expected that their closest friend would be available and supportive in both good and bad times. Lower scores were given to those who said their closest friend was relatively unavailable, unresponsive, unreliable, or even rejecting and cruel. The interrater reliability for this scale was .59 (Spearman-Brown correction, .74).

Relationship effectiveness (23 years)

At age 23 years, participants were interviewed about their current romantic relationship and past relationship(s). The relationship effectiveness scale was a trained coder rated overall evaluation (5-point scale) of the degree to which participants were competently engaged in romantic relationships between ages 21-23 years based on their response to a set of interview questions about history of romantic relationships and the quality of those relationships (M = 2.68, SD = 1.26). Effectively engaged participants demonstrated a record of forming and maintaining high quality relationships that appeared to contribute to a positive sense of self. Higher scores were assigned to participants with a track record of relationships in which there was mutual caring, trust, and emotional closeness. Lower scores were assigned to participants who were unable to maintain romantic relationships for more than a short period of time or were involved in relationships that were emotionally distant and distrusting, or were characterized by victimization, active rejection, or chronic intense conflict. The interrater reliability for this scale was .94.

Measures of Adult Functioning

Global adjustment (26 years)

At age 26 years, global adjustment, a measure of adult functioning, was rated based on participants’ overall adjustment in adulthood. Specifically, trained interviewers rated participants’ global adjustment on a 5-point scale based on semi-structured interviews conducted with participants at age 26, including questions regarding relationships, work, mental health, and self-evaluations (M = 3.66, SD = 1.12). Higher scores were assigned to participants who meaningfully engaged in the world of work and close relationships, which were an important part of their lives. Moreover, a sense of purposefulness, goal-directedness, and a capacity for self-evaluation and reflectivity was shown. Lower ratings were assigned to participants who were functioning poorly in all areas, problems were chronic and pervasive, and there was a lack of self-evaluation, purposefulness, and reflectivity.

Global adaptive functioning (28 years)

Global adaptive functioning at age 28 was assessed as part of the Clinician’s Version of the Structured Clinical Interview for DSM-IV, Axis I Disorders (First, Gibbon, Spitzer, & Williams, 1997) administered when participants were 28 years old. Trained coders rated the Global Adaptive Functioning scale for the month prior the interview on a 1-100 scale (First, Spitzer, Gibbon, & Williams, 1997; M = 72.38, SD = 14.46). The scale was based on considerations of psychological, social, and occupational functioning on a hypothetical continuum of overall functioning. Higher scores indicated better overall adaptive functioning. The interrater reliability for this scale was .83.

Control Measures

I.Q. (9 years)

Participants were administered the abbreviated short form of the Wechsler Intelligence Scale for Children – Revised (WISC-R; Wechsler, 1974) when they were 9 years old. The abbreviated short form is comprised of the vocabulary, similarities, and block design sub-tests. Sattler (1982) reports a validity coefficient of .931 with the full Scale IQ and reliabilities of each subtest at .79 or greater. Sattler’s (1982) method of computing deviation IQ’s for short forms was employed. Derived IQ’s are included in analyses (M = 105.09, SD = 14.86).

Life stress (28 years)

A concurrent measure of life stress was used as a control variable for the outcome measure of adult global functioning at age 28 years. Life stress at age 28 was assessed using the Life Events Schedule (LES; Egeland, Breitenbucher, & Rosenberg, 1980), adapted from the Life Events Inventory (Cochrane & Robertson, 1973). The LES consists of 41 interview questions that ask about the presence and severity of various life events and problems that may have occurred in the past year (e.g., job changes, financial stressors, arrests, illnesses, marriage, divorce). Trained coders rated responses to the LES questions with weighted scores based on (a) the number of stressful items checked and (b) the intensity of disruption stemming from each endorsed stressor (M = 9.61, SD = 6.86). The interrater reliability was .98.

Data Analysis Plan

We developed and tested four nested path models of the interrelation between salient relationships earlier in development and adaptive functioning in early adulthood using MPlus version 6.1 (Muthén & Muthén, 1998-2010). In all of the proposed models, the interrelations between salient relationships across development are the same—that is, later relationships build on earlier relationships. Hypothesized paths from these relationships to adult functioning vary by model, however. Furthermore, IQ was included as a control variable in all of the models, and life stress at age 28 was included as a concurrent control measure for global adaptive functioning at age 28.

The following models were run and analyzed. First, we examined a model wherein all of the relationships predict directly to age 26 and age 28 functioning (direct model). By examining this model, we considered whether earlier relationships directly influence later adult functioning. The second model included direct paths from earlier relationships to age 26 functioning, with age 26 acting as a mediator leading to age 28 functioning (mediated through 26 model). This second model allowed us to investigate whether age 26 functioning mediates the impact of earlier relationships on adult functioning at age 28. A third model was analyzed wherein early adult romantic relationship effectiveness served as a mediator of the relation between earlier relationships and later adult functioning, with early romantic relationships affecting age 26 functioning and age 26 functioning directly influencing age 28 functioning (mediated through romantic relationships model). Our intent with this model was to examine whether earlier parent-child and peer relationships influence later adult functioning through romantic relationships in earlier adulthood. Our fourth, and final model, was a modification of model 3 but also included a direct path from attachment security in infancy to adaptive functioning at age 28 (direct from attachment model). With this model we tested whether attachment security in infancy not only affects adult functioning indirectly through various other earlier relationships, but also whether it has direct effects on adaptive functioning in adulthood. We also compared these models to the base-line (null) models that included all the variables under investigation but did not include any paths.

The Comparative Fit Index (CFI), the Root Mean Square Error of Approximation (RMSEA), and the Standardized Root-Mean Square Residual (SRMR) were evaluated to determine whether each model was an acceptable fit of the data. As a rule of thumb, a good fit is indicated by a CFI at or above .95 and an RMSEA (OR SRMR) of .05 or less, and an acceptable fit is indicated by a CFI at .90 to .94 and a RMSEA (or SRMR) at .06 to .08 (Hoyle, 1995; Hoyle & Panter, 1995; McDonald & Ho, 2002). Chi-square difference testing (Muthén & Muthén, 1998-2010) was used for model comparison by comparing the fit of the nested models.

Missing data

Participants were included in analyses if they had outcome data at age 28 (N = 157; 79 males, 78 females). Amount of missing data ranged from 0% to 9.55%, with a mean of 4.29% over all variables and all time points. To allow analysis of the sample of 157 participants with outcome data, full information maximum likelihood estimation (FIML) using Mplus 6.1 (Muthén & Muthén, 1998-2010) was used.

Results

Correlations

Bivariate correlations (Table 1) indicate that both 26 year global adjustment and 28 year global adaptive functioning are significantly correlated with all other variables with the exception of parent-child relationships at 13 years.

Table 1. Zero-Order Correlations.

	1	2	3	4	5	6	7	8
1. GAF 28 yrs
2. Global Adj. 26 yrs	49***	--
3. Rel. Effect. 23 yrs	32***	.47***	--
4. Parent-Child Rel. 13 yrs	.08	.12	.20*	--
5. Friend. Security 16 yrs	.18*	.19*	.23**	.20*	--
6. Peer Comp. 7-9 yrs	.20*	.35***	.25**	.19*	.18*	--
7. Attachment 12-18 mos	.41***	.22**	.21**	.03	.20*	.15f	--
8. I.Q.	.16*	.29***	.05	.06	.18*	.26**	.10	--
9. Life Stress 28 yrs	−.48***	−.32***	−.36***	−.12	−.08	−.17*	−.22**	−.00

Note. GAF = Global Adaptive Functioning

^*p < .05

^**p < .01

^***p < .001

Path analyses

Model 4, the model wherein earlier relationships are mediated through romantic relationships in early adulthood and infant attachment directly affects age 28 global adaptive functioning is a significant improvement over both the baseline (null) model and model 3 (mediated through romantic relationships model), and fit the data as well as the other two models (see Table 2). Given that model 4 (direct from attachment model) is more parsimonious than either the direct model or the model where relationships are mediated through age 26 adaptation, and was a better fit than either the null model or the model where earlier relationships are mediated entirely through romantic relationships, model 4 was determined to be the most plausible model. Results of the analysis for model 4 are presented in Figure 1.

Table 2. Goodness of Fit Indices for Alternative Models.

Model	df	χ2	χ2/df	Δ χ2(Δ df)	CFI	RMSEA	SRMR
Model 4 (Final Model)	12	12.26	1.03		.999	.012	.041
Null Model	35	227.84	6.51	215.58(23)***
Model 1 (Direct)	4	3.58	.90	-8.68(8)	1.00	.000	.021
Model 2 (Mediated	9	13.31	1.48	-1.05(3)	.978	.055	.043
through 26 Years)
Model 3 (Mediated through Romantic Relationships)	13	22.22	1.71	9.96(1)**	.952	.067	.059

Note.

^*p < .05

^**p < 01

^***p < .001

Figure 1.

Path-analysis examining social capital predicting to global adaptive functioning at age 28 years. Note. χ²/df = 12.26/12 (p = .425), RMSEA = .012, CFI = .999, SRMR = .041, Adjusted BIC = 4213.159 †p < .10, * p < .05, **p < .01, ***p < .001

Model 4 fit the data well, χ²/df = 1.02 (p = .425), root-mean-square error of approximation (RMSEA) = .012 (90% CI = .000 - .083), comparative fit index (CFI) = .999, standardized root-mean-square residual (SRMR) = .041, and accounts for 39.5% of the variance in global adaptive functioning at age 28.

Controlling for IQ, global adjustment at age 26 (β = .32, p = .000), life stress at age 28 (β = −.34, p = .000), and attachment at 12-18 months (β = .26, p = .001) had significant direct effects on global adaptive functioning at age 28. In addition to these direct effects, significant indirect effects on adaptive functioning at age 28 were also found. Significant indirect effects from attachment in infancy (12-18 months; β = .055, p = .023), peer competence in middle childhood (7-9 yrs; β = .064, p = .008), and romantic relationship effectiveness in early adulthood (23 years; β = .27, p = .000) on global adaptive functioning at age 28 were found. Furthermore, the model indicated significant paths from attachment in infancy to friendship security in adolescence (β = .17, p = .045), peer competence in middle childhood to romantic relationship effectiveness in early adulthood (β = .20, p = .013), romantic relationship effectiveness to both global adjustment at 26 years (β = .46, p = .000) and life stress at 28 years (β = −.27, p = .002), and global adjustment at 26 years to life stress at 28 years (β = −.22, p = .012). Marginal effects were also found in the model as indicated on figure 1.

Discussion

The current study examined the effects of social capital on adjustment in early adulthood from a developmental perspective. Based on attachment theory, we put forward the perspective that social capital has its roots in salient interpersonal relationships across development, with an especially important role for infant-caregiver attachment relationships. We specifically investigated whether the quality of age-salient relationships during different developmental periods predicted not only the quality of subsequent relationships, but also adaptive functioning in early adulthood.

Four alternative models were run to examine the associations between interpersonal relationships across development and later adult adaptation. Results indicated that the best fitting and most parsimonious model was the model wherein early interpersonal functioning was mediated through the quality of romantic relationships to early adult adjustment at age 26, and then to global adaptive functioning at age 28, with a direct effect from the security of attachment relationships in infancy to 28 year functioning. These results suggest that individuals with a history of high quality interpersonal relationships across development are more likely to be better adjusted in adulthood. In particular, controlling for global adjustment at age 26, IQ, and concurrent level of life stress, individuals who had secure attachment relationships with their primary caregivers in infancy were more likely to be better adjusted at age 28 (as measured by global adaptive functioning).

The influence of the quality of attachment in infancy on adult adjustment is particularly noteworthy. Developmental theory emphasizes the role of early experiences as potent predictors of later outcomes. It is theorized that early experiences are foundational in nature and set in motion specific developmental pathways. As suggested by developmental theory, indirect effects for infant attachment were found, such that attachment influenced relationships in middle childhood, adolescence, and early adulthood and through its effects on these relationships, attachment security indirectly affected adult functioning. In the current study, however, attachment not only was related to more proximal relationships, but also had direct effects on global adaptive functioning in adulthood, even when considering the effects of concurrent stressful events and earlier adaptation in work and relationships. These results indicate that infant attachment relationships may play a special role in adult functioning.

Although we obtained frequent, comprehensive information on our participants due to the prospective, longitudinal nature of our study, this also limited our sample size. We, therefore, did not have the power to detect significant differences by gender. We did, however, attempt to examine potential gender differences via a multiple group analysis using Mplus 6.1 (Muthén & Muthén, 1998-2010). Although paths appeared to be different based on gender, chi-square difference testing indicated that the model where paths were allowed to be free by group was not significantly different from the model where paths were constrained to be the same across gender. Studies including a larger number of participants are warranted. Furthermore, given that the original sample of mothers was low-income, results should not be generalized to a larger population with a more diverse SES. It must also be remembered that the results of this study, as all path and structural equation analyses, are correlational, and do not imply causation. Furthermore, the broad patterns that we have identified in this study suggest that closer examination of the processes underlying these results is necessary in order to identify the mechanisms by which early infant-caregiver attachment relationships affects global adaptive functioning in adulthood. Future research should not only include a larger, more diverse sample, but should also investigate these underlying mechanisms.

Our results support the perspective that social capital is cumulative across development, deriving from age-salient interpersonal relationships that have as their foundation early infant-child attachment relationships. It is our contention that secure infant attachment relationships set in motion a process whereby individuals are able to successfully derive resources from the environment to function effectively in important life roles. Secure attachment relationships not only provide the framework for successful interpersonal relationships later in development, but also interact with current individual circumstances (illustrated in our study by current life stress) and earlier adaptation to predict functioning in adulthood. Early experiences in infant-caregiver relationships have a special and enduring role for later developmental adaptation, even into adulthood.