Available online at www.sciencedirect.

com

ScienceDirect

The development of adult attachment styles: four lessons

R Chris Fraley1 and Glenn I Roisman2

Why are some adults secure or insecure in their relationships? that examines the implications of attachment styles for
The authors review four lessons they have learned from interpersonal functioning and mental health [1,2].
longitudinal research on the developmental antecedents of Research has shown, for example, that people who are
adult attachment styles. First, although adult attachment relatively secure are more likely to be committed to their
appears to have its origins in early caregiving experiences, relationships [3], are more supportive and engaged par-
those associations are weak and inconsistent across ents [4], report fewer depressive symptoms [5], and are
measurement domains. Second, attachment styles appear to able to adapt to stressful events in constructive ways [6].
be more malleable in childhood and adolescence than in
adulthood, leading to asymmetries in socialization and Given the relevance of attachment styles for a wide array
selection processes. Third, early experiences do not determine of outcomes, it is important to learn more about what
adult outcomes. Fourth, there is still a lot to learn, and future makes people secure or insecure in their relationships. In
research requires examining relationship-specific attachment this article we review four lessons we have extracted from
patterns, the distinction between distal and proximal factors, the research to date. These lessons are not intended to be
and interactions between relational and genetic vulnerabilities. absolute; they are necessarily provisional and will require
revision as new data emerge. Nonetheless, they represent
Addresses the current state-of-the-art in our efforts to understand
1
University of Illinois at Urbana-Champaign, USA why some people are more secure than others in their
2
University of Minnesota, USA attachment relationships.
Corresponding author: Fraley, R Chris ([email protected])
Lesson 1: Adult attachment styles appear to
have their origins, in part, in earlier
Current Opinion in Psychology 2019, 25:26–30 interpersonal experiences. But existing
This review comes from a themed issue on Attachment in adulthood longitudinal data have not produced a
Edited by [1_TD$IF]Jeffry [2_TD$IF]A Simpson and [3_TD$IF]Gery [4_TD$IF]Karantzas
consistent set of predictors of attachment
styles
For a complete overview see the Issue and the Editorial
One of the long-standing assumptions of attachment
Available online 24th February 2018 theory is that people’s attachment styles are reflections
https://doi.org/10.1016/j.copsyc.2018.02.008 of their previous caregiving experiences, beginning with
2352-250/ã 2018 Elsevier Ltd. All rights reserved. their parental relationships. Until recently, the only way
researchers have been able to evaluate this possibility is
through the use of retrospective reports — that is, asking
adults to reflect on what their early caregiving experi-
ences were like [7]. In the last few years, however, a
number of longitudinal studies of children have come of
There are vast individual differences in the ways in which age — studies that make it possible for researchers to
people relate to significant others in their lives. Some examine with prospective data the ways in which early
adults, for example, are relatively secure in their close caregiving experiences predict later attachment
relationships. They are comfortable depending on others, outcomes.
can turn to others for comfort when they are distressed,
and can use others as a secure base from which to explore For example, we and our colleagues have examined
the world. Other people, in contrast, are relatively inse- children who were participants in the NICHD Study
cure in their close relationships. They are uncomfortable of Early Child Care and Youth Development (SEC-
opening up to others and may worry that others will not be CYD) — a large sample of parents and their children
there for them when needed. who have been studied over time, beginning when the
children were 1 month of age [8
]. At age 18, we assessed
Social and personality psychologists refer to these kinds of the children’s attachment styles. We found that those
individual differences as self-reported ‘adult attachment who were insecure at age 18 were more likely than those
styles’ or ‘adult attachment orientations.’ Briefly, the term who were secure to have had less supportive parenting
‘adult attachment style’ refers to a constellation of knowl- over time (though not necessarily lower mean levels of
edge, expectations, and insecurities that people hold early supportive care), to come from families character-
about themselves and their close relationships. A large ized by instability (e.g., parental depression, father
body of research has accumulated over the past 30 years absence), and to have had lower quality friendships in

Current Opinion in Psychology 2019, 25:26–30 www.sciencedirect.com

 Development and attachment Fraley and Roisman 27

Table 1

Brief overview of longitudinal studies on the antecedents of adult attachment.

Study Findings
Chopik et al. (2014) [9] Re-examined data on approximately 100 people from Block and Block longitudinal study and found that higher
caregiving nurturance at age 3 was associated with larger declines in attachment avoidance from age 14 to age 23.
Dinero et al. (2008) [10] Examined the quality of observed interactions between over 250 adolescents and their parents over a period of
10 years. They found that the quality of those interactions at age 15 predicted security at age 25, even after controlling
for age 15 assessments of security.
Fraley et al. (2013) [8
] In a sample of approximately 800 people, found that changes in caregiving (e.g., sensitivity and maternal depression)
were related to attachment at age 18. Also, developing social competence and friendship quality predicted later
attachment.
Salo et al. (2011) [11] Found in a sample of over 1000 people that maternal nurturance at age 10, as reported by mothers, was correlated
with the self-reported security of their children 21–27 years later.
Zayas et al. (2011) [12] The quality of maternal caregiving at 18 months predicted attachment 20 years later in a sample of approximately
35 people.

adolescence. A number of other studies have uncovered Namely, socialization processes should be more influen-
associations between early caregiving experiences and tial in early childhood relative to selection effects. But, as
attachment outcomes in young and later adulthood, many people develop, selection effects should begin to play a
of which are summarized in Table 1. larger role than socialization effects. We refer to the shift
in the dominance of these two processes as socialization-
In short, longitudinal investigations suggest that there are selection asymmetries (Figure 1).
associations between various aspects of children’s care-
giving environments and their attachment styles in adult- One of the implications of these asymmetries is that
hood. As we discuss in more detail below, however, these attachment styles should be more malleable in childhood
associations are relatively small in magnitude and thus far than they are in adulthood [16]. This pattern has now
are not particularly consistent across various ways of been observed across multiple studies. For example,
indexing early experience and later outcomes [13]. None- Jones and his colleagues [17
] recently examined attach-
theless, these studies are important because they provide ment styles in a sample of adolescents and their mothers
a longitudinal view into the developmental antecedents who were followed multiple times over five years. Impor-
of adult attachment styles — a view that was not possible tantly, the same assessment instrument was used to asses
until recently. both kids and their parents. They found that, on average,
there was more test–retest stability (r = .72) among
Lesson 2: Socialization and selection mothers over similar intervals than there was among their
processes are time-dependent teenagers (r = .42). These findings suggest that socializa-
Attachment is regarded as a prototypical socialization tion processes may be more influential in childhood
theory because of the emphasis it places on the role of (when people are more plastic) as opposed to adulthood
caregiving experiences in shaping individual differences (where selection dynamics dominate).
in attachment styles. But the theory also invokes a num-
ber of dynamics that are common to personality theories,
such as selection or person-driven effects. For example, Lesson 3: Foundations are not fate
the theory assumes that people’s attachment styles play a Socialization-selection asymmetries imply that attach-
role in how they interpret the behavior and intentions of ment styles are relatively open to environmental influ-
others, how they regulate their affect and behavior, and, ences in early childhood. This might lead to the conclu-
ultimately, how they experience their close relationships sion that that, not only will early experiences leave a
[14]. These person-driven processes are often invoked to persistent mark on social development, but the impres-
explain why insecure people tend to repeatedly experi- sions they leave will be powerful.
ence dysfunctional relationships [15].
But there is a flipside to early plasticity that is easy to
In short, attachment theory assumes both socialization overlook. Namely, if children’s attachment patterns are
(environment ! person) and selection (per- relatively malleable compared to those of adults, then
son ! environment) effects. But it is unlikely that these children’s attachment patterns can be shaped by multiple,
processes are equally influential. One of the ideas we potentially competing, experiences. In short, the pro-
have been exploring in recent years is that these pro- cesses that lead some adults to be more secure than others
cesses are differentially active at different points in the are not deterministic [18]. One consequence of this
lifespan — or even in the life course of a specific rela- conclusion is that, even if we can detect traces of early
tionship (e.g., the development of a romantic bond). experiences in later outcomes, the residue of those

www.sciencedirect.com Current Opinion in Psychology 2019, 25:26–30

28 Attachment in adulthood

Figure 1

Socialization effects have primacy Selection effects have primacy

People are relatively plastic People play an active role

and environments play in selecting and shaping
a greater role in their interpersonal
shaping the person environments

Time
Current Opinion in Psychology

Socialization-selection asymmetries in development. Early in life (or in a relationship), people are relatively malleable and are shaped by their
caregiving environments. As people develop, they come to play a greater role in shaping their environments, by evoking specific responses from
others or by selecting themselves into relationships or contexts that are consistent with their expectations and preferences. The implications of
these asymmetries are that (a) people are more malleable early in life than later and (b) stability will be higher later in life than earlier.

experiences will not be large in an absolute sense. In First, people can develop relationship-specific attach-
statistical terms, the effect sizes may be small. ment styles that are capable of accommodating to the
divergent interpersonal experiences they have had (e.g.,
Figure 2 illustrates the typical effect size we see in the [19,20]). Even if a person has a secure foundation with his
kinds of longitudinal studies summarized in Table 1 — a or her parents, experiences with peers may lead the
correlation of about .15. As can be seen, people with person to feel insecure in the context of non-familial
positive early experiences are more likely to have secure relationships. This can lead a person to be insecure in,
attachment styles — and people with negative early say, romantic relationships, even if the person was and
experiences are more likely to have insecure attachment continues to be secure in parental relationships. In short,
styles (see the dark green points in Figure 2). But there sometimes what appears to be discontinuity over time is
are plenty of exceptions to this trend. There are many not “change” in the traditional sense (i.e., a person’s
people who had supportive caregiving experiences, for attachment style has evolved from secure to insecure),
example, who are relatively insecure as adults. Similarly, but is a shift in which working models or relational
there are many people who had less than ideal caregiving schema are most active for a person in any given context
experiences who nonetheless report high levels of secu- (e.g., [21]). The field desperately needs more research on
rity (see the gray points in Figure 2). how changes in specific relational contexts do or do not
translate into changes in other relational contexts.
Lesson 4. There is much yet to be understood
about the etiology of adult attachment styles Second, a person’s attachment styles in adulthood may be
The fact that these kinds of exceptions exist suggests that better understood with respect to recent interpersonal
we still have a lot to learn about why some people are experiences rather than distal ones alone [22]. Even if
secure and others are insecure in their relationships. We early caregiving experiences can sometimes leave a trace
would like to close with three ideas that we think might in adult attachment patterns, there is abundant evidence
help clarify why these exceptions exist. Hopefully these that on-going attachment experiences are important for
ideas can serve as guide posts for future research. understanding interpersonal functioning (e.g., [23
]).

Current Opinion in Psychology 2019, 25:26–30 www.sciencedirect.com

 Development and attachment Fraley and Roisman 29

Figure 2 [27,28], or the right additive and nonadditive combination

of genes [29,30]. Regardless, there is a lot to learn by
bringing a genetically informed perspective to close rela-
3

tionships research more generally, and adult attachment

research in particular.
2
Attachment in adulthood

Conflict of interest statement

1

Nothing declared.
0

References and recommended reading

Papers of particular interest, published within the period of review,
-1

have been highlighted as:

of special interest
-2

1. Gillath O, Karantzas G, Fraley RC: Adult Attachment: A Concise

Guide to Theory and Research. Academic Press; 2016.
-3

2. Mikulincer M, Shaver PR: Attachment in Adulthood: Structure,

-3 -2 -1 0 1 2 3 Dynamics, and Change. edn 2. Guilford Press; 2016.
Early experience 3. Segal N, Fraley RC: Broadening the investment model: an
intensive longitudinal study on attachment and perceived
Current Opinion in Psychology
partner responsiveness in commitment dynamics. J Soc Pers
Relatnsh 2016, 33:581-599.
An illustration of the association between early caregiving experiences 4. Jones JD, Cassidy J, Shaver PR: Parents’ self-reported
and adult attachment styles. Even when positive associations exist, attachment styles: a review of links with parenting behaviors,
there are many people who exhibit incongruent patterns: cases in emotions, and cognitions. Pers Soc Psychol Rev 2015, 19:44-76.
which early caregiving environments were positive, but later outcomes 5. Bradford AB, Burningham KL, Sandberg JG, Johnson LN: The
are not, for example. association between the parent–child relationship and
symptoms of anxiety and depression: the roles of attachment
and perceived spouse attachment behaviors. J Marital Fam
Ther 2017, 43:291-307.
Because attachment theorists believe that attachment 6. Jayamaha SD, Girme YU, Overall NC: When attachment anxiety
styles have their origins in early experiences, it is easy impedes support provision: the role of feeling unvalued and
unappreciated. J Fam Stud 2017, 31:181-191.
to mistakenly assume that early experiences are the most
7. Hazan C, Shaver PR: Romantic love conceptualized as an
important ingredient in shaping adults’ self-reported attachment process. J Pers Soc Psychol 1987, 52:511-524.
attachment styles. This is unlikely to be the case.
8. Fraley RC, Roisman GI, Booth-LaForce C, Owen MT, Holland AS:
Understanding early experiences is critical for maintain-
Interpersonal and genetic origins of adult attachment styles: a
ing a lifespan perspective on attachment and close rela- longitudinal study from infancy to early adulthood. J Pers Soc
Psychol 2013, 104:817-838.
tionships, but early experiences per se might not be Examines multiple developmental and genetic antecedents of attach-
essential for answering proximal questions about adult ment styles at age 18, using longitudinal data from the Study of Early Child
attachment security. Care and Youth Development.
9. Chopik WJ, Moors AC, Edelstein RS: Maternal nurturance
predicts declines in attachment avoidance in emerging
Finally, there may be value in studying of possible adulthood. J Res Pers 2014, 53:47-53.
genetic contributions, along with gene–environment pro-
10. Dinero RE, Conger RD, Shaver PR, Widaman KF, Larsen-Rife D:
cesses (including but not limited to gene by environment Influence of family of origin and adult romantic partners on
interactions), to the development of adult attachment romantic attachment security. J Fam Psychol 2008, 22:622-632.
styles [8
,24]. On the one hand, behavior-genetic research 11. Salo J, Jokela M, Lehtimäki T, Keltikangas-Järvinen L: Serotonin
involving twin samples suggests that attachment styles receptor 2A gene moderates the effect of childhood maternal
nurturance on adulthood social attachment. Genes Brain Behav
are heritable. For example, in a behavior genetic analysis 2011, 10:702-709.
of self-report measures of attachment, Donnellan 12. Zayas V, Mischel W, Shoda Y, Aber JL: Roots of adult
et al. [25] estimated that approximately 45% of the varia- attachment: maternal caregiving at 18 months predicts adult
tion in attachment-related anxiety can be attributed to attachment to peers and partners. Soc Psychol Pers Sci 2011,
2:289-297.
additive genetic variation. On the other hand, efforts to
13. Steele RD, Waters TEA, Bost KK, Vaughn BE, Truitt W, Waters HS,
identify the specific polymorphisms and gene–environ- Booth-LaForce C, Roisman GI: Caregiving antecedents of
ment processes has proven challenging [8
]. As we have secure base script knowledge: a comparative analysis of
noted elsewhere, this ‘mystery of missing heritability’ young adult attachment representations. Dev Psychol 2014,
50:2526-2538.
[26] is becoming more salient in psychological research. It
14. Collins NL, Ford MB, Guichard AC, Allard LM: Working models of
might be the case that researchers simply need more time attachment and attribution processes in intimate
to identify the right genes, the right epigenetic processes relationships. Pers Soc Psychol Bull 2006, 32:201-219.

www.sciencedirect.com Current Opinion in Psychology 2019, 25:26–30

30 Attachment in adulthood

15. Brumbaugh CC, Fraley RC: Transference and attachment: how 22. Fraley RC: A connectionist approach to the organization and
do attachment patterns get carried forward from one continuity of working models of attachment. J Pers 2007,
relationship to the next? Pers Soc Psychol Bull 2006, 75:1157-1180.
32:552-560.
23. Girme YU, Agnew CR, VanderDrift LE, Harvey MS, Rholes SW,
16. Fraley RC, Brumbaugh CC: A dynamical systems approach to
Simpson JA: The ebbs and flows of attachment: within-person
understanding stability and change in attachment security. In variation in attachment undermine relationship well-being
Adult Attachment: Theory, Research, and Clinical Implications. across time. J Pers Soc Psychol 2017. in press.
Edited by Rholes WS, Simpson JA. Guilford Press; 2004:86-132. Reports a set of studies which show that within-person variation in
attachment is related to relationship functioning over time.
17. Jones JD, Fraley RC, Ehrlich KB, Stern JA, Lejuez CW, Shaver PR,

Cassidy J: Stability of attachment style in adolescence: an 24. Roisman GI, Fraley RC: The limits of genetic influence: a
empirical test of alternative developmental processes. Child behavior-genetic analysis of infant–caregiver relationship
Dev 2017. in press. quality and temperament. Child Dev 2006, 77:1656-1667.
Examines the test–retest stability of attachment styles in a sample of
parents and their teenagers. Results show that stability tends to be higher 25. Donnellan MB, Burt SA, Levendosky A, Klump K: Genes,
among parents than it is among their children. personality, and attachment in adults: a multivariate behavioral
genetic analysis. Pers Soc Psychol Bull 2008, 34:3-16.
18. Sroufe LA, Egeland B, Carlson EA, Collins WA: The Development
of the Person: the Minnesota Study of Risk and Adaptation from 26. Zuk O, Hechter E, Sunyaev SR, Lander ES: The mystery of
Birth to Adulthood. Guilford Press; 2009. missing heritability: genetic interactions create phantom
heritability. Proc Natl Acad Sci 2012, 109:1193-1198.
19. Collins NL, Guichard AC, Ford MB, Feeney BC: Working models
of attachment: new developments and emerging themes. In 27. Bagot RC, Meaney MJ: Epigenetics and the biological basis of
Adult Attachment: Theory, Research, and Clinical Implications. gene  environment interactions. J Am Acad Child Adolesc
Edited by Rholes WS, Simpson JA. Guilford; 2004:196-239. Psychiatry 2010, 49:752-771.
20. Fraley RC, Heffernan ME, Vicary AM, Brumbaugh CC: The 28. Bosmans G, Young JF, Hankin BL: NR3C1 methylation as a
experiences in close relationships–relationship structures moderator of the effects of maternal support and stress on
questionnaire: a method for assessing attachment insecure attachment development. Dev Psychol 2018, 54:29-38.
orientations across relationships. Psychol Assess 2011,
23:615-625. 29. Yang J et al.: Common SNPs explain a large proportion of the
heritability for human height. Nat Genet 2010, 42:565-569.
21. Baldwin MW, Keelan JPR, Fehr B, Enns V, Koh-Rangarajoo E:
Social cognitive conceptualization of attachment working 30. Sauce B, Matzel LD: The paradox of intelligence: heritability
models: availability and accessibility effects. J Pers Soc and malleability coexist in hidden gene–environment interplay.
Psychol 1996, 71:94-104. Psychol Bull 2018, 144:26-47.

Current Opinion in Psychology 2019, 25:26–30 www.sciencedirect.com