SPECIAL ARTICLE
An attachment perspective on psychopathology
Mario Mikulincer1, PhilliP r. Shaver2
1School of Psychology, Interdisciplinary Center Herzliya, P.O. Box 167, Herzliya 46150, Israel
2Department of Psychology, University of California at Davis, One Shields Avenue, Davis, CA 95616-8686, USA
In recent years, attachment theory, which was originally formulated to describe and explain infant-parent emotional bonding, has been ap-
plied to the study of adolescent and adult romantic relationships and then to the study of psychological processes, such as interpersonal
functioning, emotion regulation, coping with stress, and mental health. In this paper, we offer a brief overview of the attachment perspective
on psychopathology. Following a brief account of attachment theory, we go on to explain how the study of individual differences in adult at-
tachment intersects with the study of psychopathology. Specifically, we review research findings showing that attachment insecurity is a major
contributor to mental disorders, and that the enhancement of attachment security can facilitate amelioration of psychopathology.
Key words: Attachment, psychopathology, emotion regulation, security, interpersonal relations, self, mental health
(World Psychiatry 2012;11:11-15)
Attachment theory (1-3) has proven to be a very fruitful one’s attachment history – what Hazan and Shaver (5) called
framework for studying emotion regulation and mental attachment style. Research clearly indicates that attachment
health. In particular, research on adult attachment processes styles can be measured in terms of two independent dimen-
and individual differences in attachment orientations has sions, attachment-related anxiety and avoidance (6). A per-
provided strong evidence for the anxiety-buffering function son’s position on the anxiety dimension indicates the degree
of what Bowlby (2) called the attachment behavioral system to which he or she worries that a partner will not be available
and for the relevance of attachment-related individual differ- and responsive in times of need. A person’s position on the
ences to coping with stress, managing distress, and retaining avoidance dimension indicates the extent to which he or she
psychological resilience (4). distrusts relationship partners’ good will and strives to main-
In this paper, we offer a brief overview of the attachment tain behavioral independence, self-reliance, and emotional
perspective on psychopathology. Following a brief account distance. The two dimensions can be measured with reliable
of attachment theory’s basic concepts, we review research and valid self-report scales (e.g., 6), and they are associated
findings showing that attachment insecurities – called attach- in theoretically predictable ways with relationship quality
ment anxiety and avoidance in the theory – are associated and adjustment (4).
with mental disorders, and that increases in attachment se- Mikulincer and Shaver (4) proposed that a person’s loca-
curity are an important part of successfully treating these tion in the two-dimensional conceptual space defined by
disorders. attachment anxiety and avoidance reflects both the person’s
sense of attachment security and the ways in which he or
she deals with threats and distress. People who score low
AttAchment theory: bAsic concepts on these dimensions are generally secure and tend to em-
ploy constructive and effective affect-regulation strategies.
Bowlby (2) claimed that human beings are born with an Those who score high on either the attachment anxiety or
innate psychobiological system (the attachment behavioral the avoidance dimension (or both) suffer from insecurity
system) that motivates them to seek proximity to significant and tend to rely on what Cassidy and Kobak (7) called sec-
others (attachment figures) in times of need. Bowlby (1) ondary attachment strategies, either deactivating or hyper-
also outlined major individual differences in the functioning activating their attachment system in an effort to cope with
of the attachment system. Interactions with attachment fig- threats.
ures who are available in times of need, and who are sensitive According to Mikulincer and Shaver (4), people scoring
and responsive to bids for proximity and support, promote a high on avoidant attachment tend to rely on deactivating
stable sense of attachment security and build positive mental strategies – trying not to seek proximity, denying attachment
representations of self and others. But when a person’s at- needs, and avoiding closeness and interdependence in rela-
tachment figures are not reliably available and supportive, tionships. These strategies develop in relationships with at-
proximity seeking fails to relieve distress, felt security is un- tachment figures who disapprove of and punish closeness
dermined, negative models of self and others are formed, and and expressions of need or vulnerability (8). In contrast,
the likelihood of later emotional problems and maladjust- people scoring high on attachment anxiety tend to rely on
ment increases. hyperactivating strategies – energetic attempts to achieve
When testing this theory in studies of adults, most re- proximity, support, and love combined with lack of confi-
searchers have focused on the systematic pattern of relation- dence that these resources will be provided and with resent-
al expectations, emotions, and behavior that results from ment and anger when they are not provided (7). These reac-
11
tions occur in relationships in which an attachment figure is ed expression of emotions, problems with intimacy, and so-
sometimes responsive but unreliably so, placing the needy cial avoidance.
person on a partial reinforcement schedule that rewards per- Another related issue concerning the associations be-
sistence in proximity-seeking attempts, because they some- tween attachment insecurities and psychopathology is the
times succeed. extent to which attachment insecurities are a sufficient cause
Individual differences in attachment styles begin in inter- of mental disorders. In our view, beyond disorders such as
actions with parents during infancy and childhood (e.g., 9). separation anxiety and pathological grief, in which attach-
However, Bowlby (3) claimed that meaningful relational in- ment injuries are the main causes and themes, attachment
teractions during adolescence and adulthood can move a insecurities per se are unlikely to be sufficient causes of men-
person from one region to another of the two-dimensional tal disorders. Other factors (e.g., genetically determined tem-
conceptual space defined by attachment anxiety and avoid- perament; intelligence; life history, including abuse) are like-
ance. Moreover, a growing body of research shows that at- ly to converge with or amplify the effects of attachment expe-
tachment style can change, subtly or dramatically, depending riences on the way to psychopathology.
on current context, recent experiences, and recent relation- Consider, for example, the relation between attachment-
ships (e.g., 10,11). related avoidance and psychological distress. Many studies
of large community samples have found no association be-
tween avoidant attachment and self-report measures of glob-
AttAchment, mentAl heAlth, al distress (4). However, studies that focus on highly stressful
And psychopAthology events, such as exposure to missile attacks, living in a danger-
ous neighborhood, or giving birth to a handicapped infant,
According to attachment theory, interactions with incon- have indicated that avoidance is related to greater distress
sistent, unreliable, or insensitive attachment figures interfere and poorer long-term adjustment (4).
with the development of a secure, stable mental foundation; Life history factors are also important. For example, the
reduce resilience in coping with stressful life events; and pre- association between attachment insecurity and depression is
dispose a person to break down psychologically in times of higher among adults with a childhood history of physical,
crisis (3). Attachment insecurity can therefore be viewed as a psychological, or sexual abuse (e.g., 21). Stressful life events,
general vulnerability to mental disorders, with the particular poverty, physical health problems, and involvement in turbu-
symptomatology depending on genetic, developmental, and lent romantic relationships during adolescence also strength-
environmental factors. en the link between attachment insecurity and psychopa-
Mikulincer and Shaver (4) reviewed hundreds of cross- thology (e.g., 22).
sectional, longitudinal, and prospective studies of both clini- The causal links between attachment and psychopathol-
cal and non-clinical samples and found that attachment in- ogy are also complicated by research findings showing that
security was common among people with a wide variety of psychological problems can increase attachment insecurity.
mental disorders, ranging from mild distress to severe per- Davila et al (23), for example, found that late adolescent
sonality disorders and even schizophrenia. Consistently women who became less securely attached over periods of 6
compatible results have also been reported in recent studies. to 24 months were more likely than their peers to have a his-
For example, attachment insecurities (of both the anxious tory of psychopathology. Cozzarelli et al (24) found that
and avoidant varieties) are associated with depression (e.g., women who moved in the direction of insecure attachment
12), clinically significant anxiety (e.g., 13), obsessive-com- over a 2-year period following abortion were more likely
pulsive disorder (e.g., 14), post-traumatic stress disorder than other women who had an abortion to have a prior his-
(PTSD) (e.g., 15), suicidal tendencies (e.g., 16), and eating tory of depression or abuse. Solomon et al (25) assessed at-
disorders (e.g., 17). tachment insecurities and PTSD symptoms among Israeli
Attachment insecurity is also a key feature of many per- ex-prisoners of war (along with a matched control group of
sonality disorders (e.g., 18,19). However, the specific kind of veterans) 18 and 30 years after their release from captivity.
attachment insecurity differs across disorders. Anxious at- Attachment anxiety and avoidance increased over time
tachment is associated with dependent, histrionic, and bor- among the ex-prisoners, and the increases were predicted by
derline disorders, whereas avoidant attachment is associated the severity of PTSD symptoms at the first wave of measure-
with schizoid and avoidant disorders. Crawford et al (18) ment.
found that attachment anxiety is associated with what Lives- Overall, attachment insecurities seem to contribute non-
ley (20) called the “emotional dysregulation” component of specifically to many kinds of psychopathology. However,
personality disorders, which includes identity confusion, particular forms of attachment insecurity seem to predispose
anxiety, emotional lability, cognitive distortions, submissive- a person to particular configurations of mental disorders.
ness, oppositionality, self-harm, narcissism, and suspicious- The attachment-psychopathology link is moderated by a
ness. Crawford et al (19) also found that avoidant attachment large array of biological, psychological, and socio-cultural
is associated with what Livesley (20) called the “inhibited- factors, and mental disorders per se can erode a person’s
ness” component of personality problems, including restrict- sense of attachment security.
12 World Psychiatry 11:1 - February 2012
the heAling effects of AttAchment security mediAting processes
If attachment insecurities are risk factors for psychopa- According to attachment theory (3), the linkage between
thology, then the creation, maintenance, or restoration of a attachment insecurities (whether in the form of anxiety,
sense of attachment security should increase resilience and avoidance, or both) and psychopathology is mediated by sev-
improve mental health. According to attachment theory, in- eral pathways. In this section, we will review the most impor-
teractions with available and supportive attachment figures tant of these pathways.
impart a sense of safety, trigger positive emotions (e.g., relief,
satisfaction, gratitude, love), and provide psychological re-
sources for dealing with problems and adversities. Secure self-representations
individuals remain relatively unperturbed during times of
stress, recover faster from episodes of distress, and experi- According to attachment theory and research, lack of pa-
ence longer periods of positive affectivity, which contributes rental sensitivity and responsiveness contributes to disorders
to their overall emotional well-being and mental health. of the self, characterized by lack of self-cohesion, doubts
In some of our studies, we have examined the effects of about one’s internal coherence and continuity over time, un-
increased security on various indicators of mental health by stable self-esteem, and over-dependence on other people’s
experimentally activating mental representations of support- approval (e.g., 32,33). Insecure people are likely to be overly
ive attachment figures (e.g., 26,27). These research tech- self-critical, plagued by self-doubts, or prone to using de-
niques, which we (11) refer to as “security priming”, include fenses, such as destructive perfectionism, to counter feelings
subliminal pictures suggesting attachment-figure availability, of worthlessness and hopelessness (e.g., 34). These dysfunc-
subliminal names of people designated by participants as tional beliefs about oneself increase insecure people’s risk for
security-enhancing attachment figures, guided imagery high- developing mental disorders.
lighting the availability and supportiveness of an attachment Attachment research has also shown that attachment in-
figure, and visualization of the faces of security-enhancing securities are associated with pathological narcissism (e.g.,
attachment figures. 35). Whereas avoidant attachment is associated with overt
Security priming improves participants’ moods even in narcissism or grandiosity, which includes both self-praise
threatening contexts and eliminates the detrimental effects of and denial of weaknesses (36), attachment anxiety is associ-
threats on positive moods (e.g., 26). Mikulincer et al (28) ated with covert narcissism, characterized by self-focused
found that subliminal priming with security-related words attention, hypersensitivity to other people’s evaluations, and
mitigated cognitive symptoms of PTSD (heightened accessi- an exaggerated sense of entitlement (36).
bility of trauma-related words in a Stroop-color naming task)
in a non-clinical sample. Admoni (29) found that priming the
names of each participant’s security providers mitigated two emotion regulation
cognitive symptoms of eating disorders (distorted body per-
ception and heightened accessibility of food-related words in According to attachment theory, interactions with avail-
a Stroop task) in a sample of women hospitalized for eating able attachment figures and the resulting sense of attachment
disorders. security provide actual and symbolic supports for learning
There is also preliminary evidence that a sense of security constructive emotion-regulation strategies. For example, in-
provided by a psychotherapist improves a client’s mental teractions with emotionally accessible and responsive others
health. In a study based on data from the multi-site National provide a context in which a child can learn that acknowl-
Institute of Mental Health (NIMH) Treatment of Depression edgment and display of emotions is an important step toward
Collaborative Research Program, Zuroff and Blatt (30) found restoring emotional balance, and that it is useful and socially
that a client’s positive appraisals of his or her therapist’s sen- acceptable to express, explore, and try to understand one’s
sitivity and supportiveness predicted relief from depression feelings (37).
and maintenance of therapeutic benefits over an 18-month Unlike relatively secure people, avoidant individuals often
period. The results were not attributable to patient character- prefer to cordon off emotions from their thoughts and ac-
istics or severity of depression. In a one-year prospective tions. As a result, they tend to present a façade of security and
study of the effectiveness of residential treatment of high-risk composure, but leave suppressed distress unresolved in ways
adolescents, Gur (31) found that staff members’ provision of that impair their ability to deal with life’s inevitable adversi-
a sense of attachment security in the adolescents resulted in ties. This impairment is particularly likely during prolonged,
lower rates of anger, depression, and behavioral problems. demanding stressful experiences that require active coping
Although these preliminary findings are encouraging, there with a problem and mobilization of external sources of sup-
is still a great need for additional well-controlled research port (e.g., 38).
examining the long-term effects of security-enhancing thera- People who score high on attachment anxiety, in contrast,
peutic figures on clients’ mental health. often find negative emotions to be congruent with their at-
tachment-system hyperactivation. For them, “emotion regu-
13
lation” can mean emotion amplification and exaggeration of tion of the human mind and its vulnerability to pathologies.
worries, depressive reactions to actual or potential losses In a pioneering chapter on the social neuroscience of attach-
and failures, and PTSD intrusion symptoms following trau- ment processes, Coan (41) proposed what he calls social
mas. Attachment anxiety is also associated with socially de- baseline theory. According to this theory, the human brain
structive outbursts of anger and impulsive, demanding be- evolved in a highly social environment, and many of its basic
havior toward relationship partners, sometimes including functions rely on social co-regulation of emotions and phys-
violence (4). iological states. This means that, rather than conceptualizing
human beings as separate entities whose interactions with
each other need to be understood, it makes more sense to
problems in interpersonal relations consider social relatedness and its mental correlates as the
normal “baseline” condition. Using this as a starting point
According to attachment theory, recurrent failure to ob- helps us to see why experiences of separation, isolation, re-
tain support from attachment figures and to sustain a sense jection, abuse, and neglect are so psychologically painful,
of security, and the resulting reliance on secondary attach- and why dysfunctional relationships are often the causes or
ment strategies (hyperactivation and deactivation), interfere amplifiers of mental disorders.
with the acquisition of social skills and create serious prob-
lems in interpersonal relations. Bartholomew and Horowitz
(32), using as an assessment device the Inventory of Interper- references
sonal Problems (39), found that attachment anxiety was as-
sociated with more interpersonal problems in general. Se- 1. Bowlby J. Attachment and loss, Vol. 2. Separation: anxiety and an-
cure individuals did not show notable elevations in any par- ger. New York: Basic Books, 1973.
ticular sections of the problems circle, but avoidant people 2. Bowlby J. Attachment and loss, Vol. 1. Attachment (2nd ed.). New
generally had problems with nurturance (being cold, intro- York: Basic Books, 1982.
3. Bowlby J. A secure base: clinical applications of attachment theory.
verted, or competitive), and anxious people had problems London: Routledge, 1988.
with emotionality (e.g., being overly expressive). These prob- 4. Mikulincer M, Shaver PR. Attachment in adulthood: structure, dy-
lems seem to underlie insecure individuals’ self-reported namics, and change. New York: Guilford, 2007.
loneliness and social isolation (e.g., 40) and their relatively 5. Hazan C, Shaver PR. Romantic love conceptualized as an attach-
low relationship satisfaction, more frequent relationship ment process. J Pers Soc Psychol 1987;52:511-24.
6. Brennan KA, Clark CL, Shaver PR. Self-report measurement of
breakups, and more frequent conflicts and violence (4). adult romantic attachment: an integrative overview. In: Simpson
JA, Rholes WS (eds). Attachment theory and close relationships.
New York: Guilford, 1998:46-76.
conclusions 7. Cassidy J, Kobak RR. Avoidance and its relationship with other
defensive processes. In: Belsky J, Nezworski T (eds). Clinical impli-
cations of attachment. Hillsdale: Erlbaum, 1988:300-23.
Attachment insecurities are associated with a wide variety 8. Ainsworth MDS, Blehar MC, Waters E et al. Patterns of attach-
of mental disorders, ranging from mild negative affectivity to ment: assessed in the Strange Situation and at home. Hillsdale:
severe, disorganizing, and paralyzing personality disorders. Erlbaum, 1978.
The evidence suggests that insecure attachment orientations 9. Waters E, Merrick S, Treboux D et al. Attachment security in in-
(whether anxious or avoidant) are fairly general pathogenic fancy and early adulthood: a twenty-year longitudinal study. Child
Dev 2000;71:684-9.
states. Although many of the research findings supporting 10. Baldwin MW, Keelan JPR, Fehr B et al. Social-cognitive conceptu-
these ideas are correlational, several studies show a prospec- alization of attachment working models: availability and accessibil-
tive connection between attachment insecurities and vulner- ity effects. J Pers Soc Psychol 1996;71:94-109.
ability to disorders. From a therapeutic standpoint, we have 11. Mikulincer M, Shaver PR. Boosting attachment security to promote
reviewed preliminary evidence that situationally heightening mental health, prosocial values, and inter-group tolerance. Psychol
Inq 2007;18:139-56.
people’s sense of attachment security reduces the likelihood 12. Catanzaro A, Wei M. Adult attachment, dependence, self-criticism,
and intensity of psychiatric symptoms (e.g., PTSD, eating dis- and depressive symptoms: a test of a mediational model. J Pers
orders). This evidence underscores the soothing, healing, 2010;78:1135-62.
therapeutic effects of actual support offered by relationship 13. Bosmans G, Braet C, Van Vlierberghe L. Attachment and symp-
partners, including therapists, and the comfort and safety of- toms of psychopathology: early maladaptive schemas as a cognitive
link? Clin Psychol Psychother 2010;17:374-85.
fered by mental representations of supportive experiences 14. Doron G, Moulding R, Kyrios M et al. Adult attachment insecuri-
and loving and caring attachment figures. The research evi- ties are related to obsessive compulsive phenomena. J Soc Clin
dence causes us to be optimistic about the utility of clinical Psychol 2009;28:1022-49.
interventions that increase clients’ sense of attachment secu- 15. Ein-Dor T, Doron G, Solomon Z et al. Together in pain: attach-
rity. ment-related dyadic processes and posttraumatic stress disorder. J
Couns Psychol 2010;57:317-27.
In the long run, research on attachment security and inse- 16. Gormley B, McNiel DE. Adult attachment orientations, depressive
curity, and on the connections between insecurity and psy- symptoms, anger, and self-directed aggression by psychiatric pa-
chopathology, should contribute to a strongly social concep- tients. Cogn Ther Res 2010;34:272-81.
14 World Psychiatry 11:1 - February 2012
17. Illing V, Tasca GA, Balfour L et al. Attachment insecurity predicts sociation, 2006:77-99.
eating disorder symptoms and treatment outcomes in a clinical 29. Admoni S. Attachment security and eating disorders. Doctoral dis-
sample of women. J Nerv Ment Dis 2010;198:653-9. sertation, Bar-Ilan University, Ramat Gan, Israel, 2006.
18. Crawford TN, Livesley WJ, Jang KL et al. Insecure attachment and 30. Zuroff DC, Blatt SJ. The therapeutic relationship in the brief treat-
personality disorder: a twin study of adults. Eur J Personality ment of depression: contributions to clinical improvement and en-
2007;21:191-208. hanced adaptive capacities. J Consult Clin Psychol 2006;74:199-206.
19. Meyer B, Pilkonis PA. An attachment model of personality disor- 31. Gur O. Changes in adjustment and attachment-related representa-
ders. In: Lenzenweger MF, Clarkin JF (eds). Major theories of per- tions among high-risk adolescents during residential treatment: the
sonality disorder. New York: Guilford, 2005:231-81. transformational impact of the functioning of caregiving figures as
20. Livesley WJ. Classifying personality disorders: ideal types, proto- a secure base. Doctoral dissertation, Bar-Ilan University, Ramat
types, or dimensions? J Pers Disord 1991;5:52-9. Gan, Israel, 2006.
21. Whiffen VE, Judd ME, Aube JA. Intimate relationships moderate 32. Bartholomew K, Horowitz LM. Attachment styles among young
the association between childhood sexual abuse and depression. J adults: a test of a four-category model. J Pers Soc Psychol 1991;61:
Interpers Violence 1999;14:940-54. 226-44.
22. Davila J, Steinberg SJ, Kachadourian L et al. Romantic involvement 33. Park LE, Crocker J, Mickelson KD. Attachment styles and contin-
and depressive symptoms in early and late adolescence: the role of gencies of self-worth. Pers Soc Psychol B 2004;30:1243-54.
a preoccupied relational style. Pers Relationship 2004;11:161-78. 34. Wei M, Heppner PP, Russell DW et al. Maladaptive perfectionism
23. Davila J, Burge D, Hammen C. Why does attachment style change? and ineffective coping as mediators between attachment and future
J Pers Soc Psychol 1997;73:826-38. depression: a prospective analysis. J Couns Psychol 2006;53:67-79.
24. Cozzarelli C, Karafa JA, Collins NL et al. Stability and change in 35. Dickinson KA, Pincus AL. Interpersonal analysis of grandiose and
adult attachment styles: associations with personal vulnerabilities, vulnerable narcissism. J Pers Disord 2003;17:188-207.
life events, and global construals of self and others. J Soc Clin Psy- 36. Wink P. Two faces of narcissism. J Pers Soc Psychol 1991;61:590-7.
chol 2003;22:315-46. 37. Cassidy J. Emotion regulation: influences of attachment relation-
25. Solomon Z, Dekel R, Mikulincer M. Complex trauma of war captiv- ships. Monogr Soc Res Child Dev 1994;59:228-83.
ity: a prospective study of attachment and post-traumatic stress 38. Berant E, Mikulincer M, Shaver PR. Mothers’ attachment style,
disorder. Psychol Med 2008;38:1427-34. their mental health, and their children’s emotional vulnerabilities:
26. Mikulincer M, Hirschberger G, Nachmias O et al. The affective a seven-year study of children with congenital heart disease. J Pers
component of the secure base schema: affective priming with rep- 2008;76:31-66.
resentations of attachment security. J Pers Soc Psychol 2001;81: 39. Horowitz LM, Rosenberg SE, Baer BA et al. Inventory of Interper-
305-21. sonal Problems: psychometric properties and clinical applications.
27. Mikulincer M, Shaver PR. Attachment theory and intergroup bias: J Consult Clin Psychol 1988;56:885-92.
evidence that priming the secure base schema attenuates negative 40. Larose S, Bernier A. Social support processes: mediators of attach-
reactions to out-groups. J Pers Soc Psychol 2001;81:97-115. ment state of mind and adjustment in late adolescence. Attach Hum
28. Mikulincer M, Shaver PR, Horesh N. Attachment bases of emotion Dev 2001;3:96-120.
regulation and posttraumatic adjustment. In: Snyder DK, Simpson 41. Coan JA. Toward a neuroscience of attachment. In: Cassidy J, Shav-
JA, Hughes JN (eds). Emotion regulation in families: pathways to er PR (eds). Handbook of attachment: theory, research, and clinical
dysfunction and health. Washington: American Psychological As- applications (2nd ed.). New York: Guilford, 2008:241-68.
15