Adult Attachment Theory: How Anxiety and Avoidance Shape Relationships
Updated: 1 day ago
Author: Ukrainian Psychological Hub · Published: September 22, 2026 · Editorial Policy
Adult attachment theory explains how people regulate closeness, security, dependence, and distress in important relationships. In contemporary research, the most useful starting point is not a set of four fixed personality types. It is a two-dimensional model: attachment anxiety and attachment avoidance. People can be relatively low or high on each dimension, and those positions can vary in degree, by relationship, and over time.
Attachment anxiety concerns fear of rejection, abandonment, or insufficient responsiveness. Attachment avoidance concerns discomfort with dependence, emotional closeness, or relying on another person. The familiar labels secure, anxious or preoccupied, dismissive-avoidant, and fearful-avoidant summarize regions of this two-dimensional space. They are research and explanatory constructs, not diagnoses, identities, or permanent boxes.
This distinction matters because adult attachment theory is frequently simplified online into statements such as “you are an avoidant” or “your childhood gave you one attachment style for life.” The scientific literature supports a more precise account. Adult attachment orientations show meaningful associations with relationship quality, emotion regulation, stress responses, mental health, and partner interaction, but they are dimensional, context-sensitive, measured in different ways, and not reducible to a single childhood classification. A major review of the field emphasizes both the explanatory power of adult attachment and the continuing debates about continuity, individual differences, and relationship processes (Fraley, 2019).
For a dedicated audit of which attachment claims are well supported, which remain probabilistic or contested, and which popular uses go beyond the evidence, see Is Attachment Theory Evidence-Based? What Research Supports, What It Does Not, and Common Misuses.
What Is Adult Attachment Theory?
Adult attachment theory is an extension of attachment theory to close relationships later in life. John Bowlby developed attachment theory to explain the human tendency to form strong affectional bonds and to seek security from particular others, especially under threat, distress, separation, or loss. Bowlby described attachment as an organized behavioral system rather than a simple preference for closeness. His account connected proximity seeking, separation distress, expectations about availability, and the regulation of felt security (Bowlby, 1977).
Research on adult romantic attachment took a decisive step when Cindy Hazan and Phillip Shaver proposed that romantic love could be studied as an attachment process. Their 1987 work translated ideas from the infant attachment literature into adult romantic relationships, including the possibility that expectations about closeness, trust, separation, and support become organized in patterned ways in adult love (Hazan & Shaver, 1987). That proposal launched a large social-personality research tradition on adult romantic attachment.
Adult romantic attachment is therefore related to the attachment system first studied in childhood, but it is not simply the infant system copied into adulthood. Adult romantic relationships are typically reciprocal: each partner may seek care and provide care, and the relationship also includes sexuality, mutual decision-making, shared identity, and long-term interdependence. Adult attachment research asks how attachment-related expectations and regulatory strategies operate within these more complex bonds.
From Bowlby and Ainsworth to Adult Romantic Attachment
Bowlby's theoretical work and Mary Ainsworth's observational research established the developmental foundation for attachment science. Infant research examined how children use caregivers during separation, reunion, exploration, and distress. Those findings generated classifications such as secure, avoidant, and resistant or ambivalent, later supplemented by a disorganized classification in separate work.
For the dedicated developmental evidence on caregiving, temperament, genetics, adversity, social context, and later relationships, see What Causes Attachment Styles? Childhood, Caregiving, Temperament, and Later Relationships
Adult romantic attachment research did not simply rename those infant classifications. Hazan and Shaver initially used three broad adult categories inspired by the earlier developmental literature, but subsequent research increasingly moved toward dimensional measurement. Bartholomew and Horowitz proposed a four-category model of adult attachment in which models of self and others could be combined to describe secure, preoccupied, dismissing, and fearful patterns (Bartholomew & Horowitz, 1991). Later psychometric work helped consolidate anxiety and avoidance as the dominant self-report dimensions used in social-personality research.
That history explains why several vocabularies now coexist. Infant attachment classifications, Adult Attachment Interview classifications, and adult romantic self-report dimensions all belong to the broader attachment tradition, yet they are not interchangeable measures of one identical variable. Treating them as interchangeable erases important differences in what is being assessed. For the dedicated longitudinal evidence on stability, life events, relationship-specific variation, and earned security, see Can Attachment Style Change? Stability, Life Events, and Earned Security.
The Two-Dimensional Model: Attachment Anxiety and Attachment Avoidance
The two-dimensional model describes adult romantic attachment using two continuous axes. Each person can score anywhere along each axis. The dimensions are conceptually distinct, which means someone can be low on both, high on one and low on the other, or high on both.
Attachment anxiety
Attachment anxiety refers to concerns about whether an attachment figure will remain available, caring, committed, and responsive. Higher anxiety is commonly associated with stronger sensitivity to signs of rejection or distance, greater worry about abandonment, and more intense efforts to restore closeness when security feels threatened. In stress-regulation models, this pattern is often described as hyperactivation of the attachment system: attention remains focused on threat, relationship cues, and obtaining reassurance or proximity.
Hyperactivation is not synonymous with being emotional, needy, jealous, or demanding. Those everyday labels are too broad. Attachment anxiety is a research dimension measured through patterns of expectations, concerns, and behavior across close relationships. A person's behavior at one moment cannot establish their score on that dimension.
Attachment avoidance
Attachment avoidance refers to discomfort with dependence, emotional intimacy, vulnerability, or relying on an attachment figure. Higher avoidance is commonly associated with efforts to maintain distance from attachment-related needs and with lower willingness to seek or accept support under some conditions. In stress-regulation models, this is often described as deactivation: attachment cues and dependency needs are downregulated, suppressed, or managed through self-reliance. For the style-specific definition, signs, triggers, and relationship patterns, see Avoidant Attachment Style: Signs, Triggers, Relationships, and the Dismissive-Avoidant Pattern.
Avoidance is not identical to introversion, needing solitude, setting boundaries, low emotional expressiveness, or ending an unhealthy relationship. The research construct concerns attachment-related regulation of closeness and dependence. A quiet partner, an independent partner, or a partner who asks for space after conflict cannot be classified as avoidant from that behavior alone.
Why two dimensions are stronger than four boxes
Dimensional models preserve information that categories discard. Two people who both receive the label “anxious” may differ substantially in how anxious they are and in their level of avoidance. Likewise, two people described as “secure” may both be relatively low in insecurity but still differ meaningfully in stress sensitivity, relationship history, and relationship-specific expectations.
Psychometric research on adult attachment has repeatedly examined how self-report measures represent these individual differences. Fraley, Waller, and Brennan used item response theory to analyze several major adult attachment instruments and showed why measurement quality and dimensional scoring matter for theoretical conclusions (Fraley, Waller, & Brennan, 2000). A broad review of adult attachment measures likewise stresses that researchers must choose instruments according to the construct, relationship focus, dimensional or categorical structure, and research purpose (Ravitz et al., 2010).
How the Four Familiar Attachment Styles Map Onto the Two Dimensions
The four familiar adult attachment styles can be understood as plain-language summaries of combinations of anxiety and avoidance. This mapping is useful for communication as long as the labels remain summaries rather than diagnoses.
Secure attachment: relatively low anxiety and low avoidance
A secure pattern describes relatively low concern about abandonment together with relatively low discomfort with closeness and dependence. Security in this dimensional sense is not perfection, constant calm, or the absence of conflict. A securely oriented person can feel jealous, become distressed, need reassurance, or withdraw temporarily. The defining idea is that attachment-related anxiety and avoidance are comparatively low, allowing more flexible use of closeness, support, and autonomy. For the dedicated low-anxiety, low-avoidance pattern page, see Secure Attachment Style: Signs, Relationships, and How Security Works.
Anxious or preoccupied attachment: higher anxiety, lower avoidance
The anxious or preoccupied summary describes higher attachment anxiety combined with lower avoidance. The attachment system is readily activated by ambiguity, distance, or perceived unresponsiveness, while closeness remains strongly desired. This label is commonly used in popular psychology, but the measurable construct is the degree of attachment anxiety rather than membership in a fixed personality class. For the dedicated signs, triggers, reassurance-seeking, and relationship-pattern page, see Anxious Attachment Style: Signs, Triggers, Relationships, and What Research Shows.
Dismissive-avoidant attachment: lower anxiety, higher avoidance
The dismissive-avoidant summary describes relatively lower attachment anxiety combined with higher attachment avoidance. The central pattern is deactivation of attachment needs and greater discomfort with dependence or closeness. Again, this is a dimensional region, not evidence that a person lacks feelings, cannot love, or is intentionally withholding.
Fearful-avoidant attachment: higher anxiety and higher avoidance
Fearful-avoidant is commonly used for the high-anxiety, high-avoidance quadrant in adult self-report models. The combination can produce tension between wanting closeness and simultaneously expecting danger, rejection, or difficulty depending on others. It should not automatically be equated with “disorganized attachment.” Infant disorganized attachment, unresolved or disorganized classifications in the Adult Attachment Interview, adult disorganization measures, and fearful-avoidant self-report patterns are related areas of research with different measurement traditions.
The distinction is especially important because a systematic review of adult disorganized attachment identified 18 different measures across 27 papers and concluded that the field still faces substantial measurement heterogeneity (Pollard et al., 2023). A single internet label cannot collapse these different constructs into one clinical category.
Internal Working Models: Expectations About Self, Others, and Relationships
A central concept in attachment theory is the internal working model: organized expectations about attachment figures, the self in relation to those figures, and what is likely to happen when closeness, support, separation, or distress become relevant. In adult relationships, working-model language is used to describe expectations such as whether a partner will be available, whether one's needs will be taken seriously, whether dependence is safe, and whether seeking support will lead to comfort, rejection, intrusion, or disappointment.
Internal working models are best understood as information-processing and expectation structures, not as hidden scripts that mechanically dictate every relationship. They can influence attention, interpretation, memory, emotion, and behavior, yet they operate in interaction with current partners and current circumstances. A partner who repeatedly responds with care provides different information than one who is chronically rejecting or unpredictable.
The idea also becomes more precise when relationship specificity is considered. People do not necessarily carry one identical working model into every relationship. The ECR-RS was developed specifically to assess anxiety and avoidance across different relational contexts, and research using it found that relationship-specific measures often predicted relational outcomes better than broad attachment measures (Fraley et al., 2011). A person can therefore feel relatively secure with one close other and considerably more anxious or avoidant with another.
Attachment Is Most Visible When Security Is Threatened
Attachment theory is fundamentally about regulation under conditions of threat, uncertainty, separation, or need. Differences between people often become especially visible when the relationship system is activated: after conflict, during illness, when a partner becomes less responsive, during prolonged separation, after a frightening event, or when commitment feels uncertain.
Simpson and Rholes summarize a large literature in which attachment anxiety and avoidance are linked to distinct patterns of emotion regulation in stressful romantic contexts (Simpson & Rholes, 2017). Higher anxiety is associated with intensified monitoring and attempts to restore closeness, whereas higher avoidance is associated with greater distancing and suppression of attachment-related needs. These are probabilistic tendencies rather than deterministic responses.
This stress-sensitive framing helps explain why someone may appear secure in calm periods but show stronger anxiety or avoidance when the relationship becomes uncertain. It also explains why attachment is not adequately assessed by asking whether a person “likes closeness.” The theory concerns the organization of security regulation, especially when closeness and support matter most.
How Attachment Anxiety and Avoidance Shape Romantic Relationships
The strongest evidence does not say that one attachment style causes every relationship problem. It shows reliable statistical associations between attachment dimensions and multiple aspects of romantic functioning.
A meta-analysis by Li and Chan synthesized 73 studies, 118 independent samples, and 21,602 participants. Both attachment anxiety and avoidance were associated with poorer cognitive, emotional, and behavioral indicators of romantic relationship quality. Avoidance showed stronger negative associations with satisfaction, connectedness, and general support, while anxiety showed a stronger positive association with general conflict (Li & Chan, 2012). These average patterns do not tell us how any specific couple will function, but they establish that the dimensions are meaningfully related to relationship outcomes.
A later review in Nature Reviews Psychology describes how anxious expectations of rejection and avoidant expectations that partners cannot be trusted can generate cognitive, emotional, and behavioral responses that undermine couple functioning under stress. The review also emphasizes partner buffering: responsive partner behavior can sometimes reduce the expression or consequences of insecurity rather than merely mirror or amplify it (Overall, Pietromonaco, & Simpson, 2022).
This is one reason attachment theory works best as a relational process framework. Individual orientation matters, but so do partner behavior, dyadic patterns, stress, timing, and the specific issue being negotiated. The same attachment concern can produce different outcomes in a relationship that is consistently responsive than in one that is chronically unpredictable.
Support seeking and support reception
When distressed, people differ in whether they seek support, how clearly they signal need, and how easily they can use the support they receive. Higher anxiety can intensify proximity seeking and concern about whether support is sufficient. Higher avoidance can make dependence or receiving support uncomfortable, especially when support threatens autonomy or requires vulnerable disclosure. Partner responsiveness can alter these processes.
The broader relationship-science construct of perceived partner responsiveness is useful here. Feeling understood, valued, and cared for is not the same construct as attachment security, but responsive interactions can provide information relevant to felt security. Attachment theory and responsiveness research therefore intersect without becoming identical.
Conflict and the pursue-withdraw pattern
Attachment anxiety and avoidance can also become visible in conflict. A more anxious partner may press for reassurance, clarity, or immediate engagement, while a more avoidant partner may distance or disengage. This resembles the demand-withdraw pattern studied in relationship science, but the two constructs should not be conflated. Demand-withdraw is an interaction pattern; attachment anxiety and avoidance are individual differences that may contribute to some instances of that pattern but do not explain all of them.
Adult Attachment and Mental Health
Adult attachment dimensions are also associated with mental-health indicators, but they are not mental disorders. A meta-analysis covering 224 studies, 245 samples, and 79,722 participants found that higher attachment anxiety and avoidance were associated with more negative mental-health indicators and fewer positive indicators. The associations were moderate on average, with attachment anxiety showing somewhat stronger links than avoidance in that synthesis (Zhang et al., 2022).
These findings support attachment as a meaningful correlate of well-being and distress. They do not justify diagnosing someone from an attachment questionnaire or treating insecure attachment as a disorder. Depression, anxiety disorders, obsessive-compulsive disorder, trauma-related disorders, personality disorders, and other clinical conditions have their own diagnostic criteria and evidence bases. Attachment measures answer different questions.
How Adult Attachment Is Measured
Adult attachment has no single universal test because different research traditions measure different aspects of attachment. Understanding the method is essential before interpreting a result.
Self-report measures of adult romantic attachment
Social-personality research commonly uses self-report questionnaires that assess attachment anxiety and avoidance. The Experiences in Close Relationships family of measures is especially influential. These instruments ask people about expectations, feelings, and tendencies in close relationships and produce dimensional scores rather than a clinical diagnosis.
Self-report measurement has practical strengths: it is efficient, can target current relationships, and maps directly onto anxiety and avoidance. It also has limitations. Scores depend on item wording, reference relationship, self-knowledge, response tendencies, and the psychometric properties of the instrument. A 2022 systematic review of self-report measures of adult attachment security found substantial variability in the quality of psychometric evidence and did not identify a single instrument with sufficient evidence across all evaluated properties (Justo-Núñez, Morris, & Berry, 2022).
Relationship-specific self-report
Some measures ask about a specific partner, parent, friend, or other close relationship rather than asking people to summarize themselves globally. The ECR-RS is a prominent example. In a very large online sample and a second validation study, relationship-specific attachment orientations showed meaningful differentiation across relational contexts (Fraley et al., 2011). This is one of the clearest reasons to avoid telling readers that they possess one attachment style that applies identically everywhere.
The Adult Attachment Interview
The Adult Attachment Interview, or AAI, belongs to a different research tradition. It evaluates an adult's state of mind with respect to attachment through a structured interview about childhood attachment experiences, with scoring focused substantially on how the narrative is organized and discussed. Its classifications cannot simply be translated into the romantic self-report labels secure, anxious, avoidant, and fearful-avoidant.
A meta-analytic investigation by Roisman and colleagues found only trivial-to-small average convergence between AAI security and adult self-report attachment-style dimensions, with a mean correlation of .09 in the first study of their report (Roisman et al., 2007). That weak overlap is scientifically important: the methods capture partly distinct constructs and should not be treated as interchangeable versions of one test.
Why measurement method changes the meaning of “attachment style”
When a study says that participants were “secure” or “avoidant,” the next question should be: measured how? A forced-choice category, a continuous self-report score, a relationship-specific questionnaire, the AAI, and an adult disorganization instrument can produce labels that sound similar while operationalizing different constructs. Ravitz and colleagues' review of 29 adult attachment instruments highlights precisely this heterogeneity (Ravitz et al., 2010).
Adult Attachment Is Both Person-Level and Relationship-Specific
Adult attachment research often finds stable individual differences, which is why broad measures can be useful. At the same time, attachment representations can vary across relationships. These two statements are compatible. A person may have a general tendency toward higher anxiety while feeling markedly safer with one partner than another. They may also feel secure in friendship and more avoidant in romance, or the reverse.
The relationship-specific evidence matters for everyday interpretation. If a person's anxiety rises sharply in one relationship, that pattern may reflect an interaction between their broader vulnerability and the current relationship environment. It does not automatically prove that the partner “caused” the orientation, nor that childhood alone predetermined it.
For researchers and clinicians, the practical implication is to specify the target. Are we measuring a general romantic orientation, a particular partnership, attachment to parents, or a broader state of mind about early attachment experiences? Without that specification, the phrase “attachment style” can become too vague to be scientifically useful.
What Adult Attachment Theory Does Not Mean
Adult attachment theory is powerful partly because it is easy to overextend. Several boundaries keep the construct accurate.
Attachment style is not a clinical diagnosis
Secure, anxious, avoidant, dismissive-avoidant, and fearful-avoidant are not DSM or ICD diagnoses. “Insecure attachment” is not an attachment disorder. Reactive Attachment Disorder and Disinhibited Social Engagement Disorder are separate clinical constructs primarily associated with childhood and specific histories and diagnostic criteria. They should not be used as labels for ordinary adult relationship patterns.
Attachment anxiety is not the same as anxiety disorder
Attachment anxiety concerns attachment-related expectations and regulation in close relationships. An anxiety disorder is a clinical condition defined by its own symptom pattern, duration, impairment, and diagnostic criteria. The constructs can be associated without being identical.
Attachment avoidance is not emotional unavailability
Avoidance is a dimensional attachment construct. “Emotionally unavailable” is an everyday relational description that can refer to many different situations: stress, depression, low commitment, conflict, poor communication, limited emotional awareness, incompatible expectations, or attachment-related avoidance among them. One label cannot substitute for assessment of the actual relationship process.
Fearful-avoidant is not automatically disorganized attachment
Fearful-avoidant usually refers to a high-anxiety, high-avoidance self-report pattern. Disorganized attachment has multiple measurement traditions, including infant behavior, AAI unresolved or disorganized classifications, and adult disorganization measures. The adult disorganization literature itself remains methodologically heterogeneous (Pollard et al., 2023). Using the terms as automatic synonyms creates a false sense of precision.
Attachment is not destiny
Attachment theory does not require the claim that early caregiving fixes a lifelong adult style. Developmental history can matter, but later relationships, stressful events, corrective experiences, personality, culture, and current partner behavior also contribute to how attachment is expressed. The scientific question is continuity and change, not permanence.
Does Childhood Determine Adult Attachment?
Childhood attachment is part of the developmental history of adult attachment, but a straight line from an infant classification to one adult romantic style is too simple. Measurement changes across development, attachment figures change, and later experiences add new information.
A meta-analysis of 127 papers including 21,072 attachment assessments found a moderate overall association in attachment security across time, but stability declined across longer intervals and was not significant across intervals longer than 15 years in that synthesis (Pinquart, Feussner, & Ahnert, 2013). That pattern is compatible with both continuity and change. It does not support a claim that infancy mechanically determines adult romantic attachment.
A full account of what causes attachment styles requires its own developmental analysis. For the present theory article, the central point is narrower: adult orientations are informed by history but expressed in current relational systems, and their measurement should not be treated as a retrospective diagnosis of childhood.
Can Adult Attachment Change?
Adult attachment orientations show both stability and change. Relationship-specific differences alone demonstrate that attachment is not a single immutable essence. Longitudinal research also documents shifts across time. A person's scores can change following relationship experiences, life transitions, or changes in the social environment.
The change question deserves separate treatment because “becoming secure” can mean several different things: lower self-reported anxiety or avoidance, greater security in one relationship, changes in narrative state of mind, or a broader developmental pattern sometimes discussed as earned security. These are not interchangeable outcomes. Adult attachment theory therefore supports plasticity without promising that a style can be “rewired” on a fixed schedule.
The Role of Partners: Attachment Is Not a One-Person Story
A common popular-psychology error is to explain a couple entirely through one person's style. Adult attachment research increasingly studies dyadic processes. What one partner fears, expects, or does affects the other partner's responses, which then feed back into the interaction.
Partner buffering research is especially important because it shows that insecurity does not operate in isolation. Overall and colleagues review evidence that particular forms of partner response can reduce attachment-related threat and help prevent insecurity from spilling over into broader couple and family functioning (Overall, Pietromonaco, & Simpson, 2022). In practice, this means a high-anxiety or high-avoidance score does not produce a predetermined relationship outcome.
The relational environment also matters for interpretation. Reassurance seeking may look very different in a relationship characterized by repeated disappearances than in one characterized by reliable contact. Withdrawal may have a different meaning during overwhelming conflict than during ordinary bids for closeness. Attachment theory becomes more informative when the behavior is placed back into its dyadic context.
Secure Base and Safe Haven in Adult Relationships
Two classic attachment functions remain useful in adult relationships. Safe haven refers to turning toward an attachment figure for comfort, protection, or regulation when distressed. Secure base refers to the way felt security can support exploration, autonomy, learning, work, creativity, and engagement with the wider world.
In adult partnerships, these functions are reciprocal and dynamic. One partner may be the source of comfort in one situation and need comfort in another. Security is therefore compatible with autonomy: the theory does not define healthy attachment as constant proximity. The secure-base idea is specifically about the freedom to explore when dependable connection is available.
These functions also help distinguish attachment from simple liking or relationship satisfaction. Someone can enjoy a partner without using them as a central source of security, while another relationship can become attachment-relevant because the person turns to that figure under threat and relies on the bond as a secure base.
Adult Attachment Theory and Human–AI Relationships
Adult attachment theory now provides a framework for studying a new question: whether and how artificial conversational systems can become involved in human attachment processes. The conceptual bridge is legitimate because attachment theory specifies functions such as proximity seeking, safe haven, secure base, separation distress, anxiety, and avoidance. The evidence base, however, must remain separated by domain.
Research on human adult attachment does not prove that an AI has attachment feelings or subjective experience. Conversely, emerging studies of human–AI attachment cannot be used as evidence that human romantic attachment works in a particular way. The relevant claim is narrower: a human user can potentially organize attachment-related expectations and regulation around an artificial relationship.
For a focused analysis of the criteria, evidence, and limits of that possibility, see Can AI Become an Attachment Figure? What Attachment Theory Can and Cannot Tell Us. For the historical concepts of safe haven and secure base as they are applied to AI-mediated relationships, see Bowlby, Ainsworth, and AI Attachment. The separate AI Attachment Styles article examines anxiety, avoidance, and security in the emerging human–AI literature without treating AI findings as evidence for ordinary human romantic attachment.
How to Use Attachment Theory Without Turning It Into a Labeling System
The most scientifically useful way to apply adult attachment theory is to ask process questions. What happens to a person's sense of security when closeness becomes uncertain? Do they intensify proximity seeking, reassurance, monitoring, or protest? Do they distance, suppress need, or rely more heavily on self-sufficiency? Does this happen broadly or mainly with one partner? What does the partner do next? Which behaviors reduce threat, and which amplify it?
This approach preserves the theory's explanatory value while avoiding personality essentialism. “High attachment anxiety in this relationship” communicates more information than “an anxious person.” “Higher avoidance under dependency-related stress” is more testable than “an avoidant never wants intimacy.” Dimensional language also leaves room for mixed patterns, situational shifts, and change.
It is equally important not to diagnose absent partners through attachment terminology. Attachment questionnaires are designed to measure the respondent, not to convert one person's frustrations into a diagnosis of somebody else. A pattern of delayed texting, reluctance to discuss feelings, jealousy, or frequent reassurance seeking can have many possible explanations. Attachment is one framework for investigation, not a shortcut around evidence.
A focused application of these dimensions to emerging limerence research is available in Limerence and Attachment Styles: Anxiety, Avoidance, and Relationship Uncertainty, which separates direct limerence evidence from broader attachment and relationship-uncertainty science.
Frequently Asked Questions
What are the two dimensions of adult attachment?
The two main self-report dimensions are attachment anxiety and attachment avoidance. Anxiety concerns fear of rejection, abandonment, or inadequate responsiveness. Avoidance concerns discomfort with dependence, vulnerability, and closeness. They are continuous dimensions, so a person can be relatively high or low on each rather than belonging to one absolute type.
What are the four adult attachment styles?
The familiar four-style summary maps onto the two dimensions: secure is relatively low anxiety and low avoidance; anxious or preoccupied is higher anxiety and lower avoidance; dismissive-avoidant is lower anxiety and higher avoidance; fearful-avoidant is higher anxiety and higher avoidance. The labels are useful summaries of dimensional patterns, not diagnoses.
Is adult attachment theory scientifically supported?
There is a large empirical literature linking adult attachment anxiety and avoidance with relationship functioning, stress regulation, support processes, and mental-health indicators. Meta-analyses and reviews show reliable associations, including the relationship-quality synthesis by Li and Chan and the mental-health meta-analysis by Zhang and colleagues. Scientific support for the broad framework does not make every popular attachment claim valid. Specific claims still require specific evidence.
Is attachment style a personality type?
No. Adult attachment dimensions can show trait-like stability, but they are relationship-relevant orientations rather than a complete personality taxonomy. They can vary by partner and context, and they explain only part of why people think, feel, and behave as they do in relationships.
Is fearful-avoidant attachment the same as disorganized attachment?
Not automatically. Fearful-avoidant usually refers to a high-anxiety, high-avoidance quadrant in adult self-report models. Disorganization is measured through several other traditions, including infant behavior, AAI unresolved or disorganized classifications, and adult disorganization instruments. The terms should be specified rather than treated as synonyms.
Can someone have different attachment patterns with different people?
Yes. Relationship-specific measurement shows that people can hold different attachment orientations in different relationships. A general tendency may still exist, but it does not erase partner-specific variation (Fraley et al., 2011).
Does attachment anxiety mean someone has an anxiety disorder?
No. Attachment anxiety is a relationship construct. Anxiety disorders are clinical diagnoses with different criteria. They can coexist and may be statistically associated, but one does not establish the other.
Does attachment avoidance mean someone does not love their partner?
No. Avoidance concerns regulation of closeness, dependence, and vulnerability. It does not measure whether love exists, whether a partner has empathy, or whether a relationship should continue.
Can attachment patterns change?
Yes, research supports both continuity and change. The degree and meaning of change depend on what is being measured: a general self-report orientation, a relationship-specific score, or another attachment construct. Claims that a style can be permanently “healed” in a fixed number of days go beyond the evidence.
The Core Idea
Adult attachment theory is best understood as a theory of security regulation in close relationships. Its modern self-report tradition is organized primarily around two continuous dimensions: attachment anxiety and attachment avoidance. The four familiar styles are summaries of combinations of those dimensions, not fixed identities.
The framework becomes most informative when it preserves distinctions. Adult romantic attachment is not the same as infant Strange Situation classification. The Adult Attachment Interview is not simply a longer attachment-style quiz. Fearful-avoidant self-report patterns are not automatically equivalent to disorganized attachment. Insecure attachment is not an attachment disorder. And no attachment score can replace clinical diagnosis, relationship context, or direct evidence about another person's motives.
Used with those distinctions intact, adult attachment theory offers a powerful way to understand why closeness can feel safe in one context and threatening in another, why stress brings different regulatory strategies to the surface, and why relationships themselves can either intensify insecurity or support greater security.
Related Articles
References
Bartholomew, K., & Horowitz, L. M. (1991). Attachment styles among young adults: A test of a four-category model. Journal of Personality and Social Psychology, 61(2), 226–244. https://doi.org/10.1037/0022-3514.61.2.226
Bowlby, J. (1977). The making and breaking of affectional bonds. I. Aetiology and psychopathology in the light of attachment theory. British Journal of Psychiatry, 130, 201–210. https://doi.org/10.1192/bjp.130.3.201
Fraley, R. C. (2019). Attachment in adulthood: Recent developments, emerging debates, and future directions. Annual Review of Psychology, 70, 401–422. https://doi.org/10.1146/annurev-psych-010418-102813
Fraley, R. C., Heffernan, M. E., Vicary, A. M., & Brumbaugh, C. C. (2011). The Experiences in Close Relationships–Relationship Structures questionnaire: A method for assessing attachment orientations across relationships. Psychological Assessment, 23(3), 615–625. https://doi.org/10.1037/a0022898
Fraley, R. C., Waller, N. G., & Brennan, K. A. (2000). An item response theory analysis of self-report measures of adult attachment. Journal of Personality and Social Psychology, 78(2), 350–365. https://doi.org/10.1037/0022-3514.78.2.350
Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as an attachment process. Journal of Personality and Social Psychology, 52(3), 511–524. https://doi.org/10.1037/0022-3514.52.3.511
Justo-Núñez, M., Morris, L., & Berry, K. (2022). Self-report measures of secure attachment in adulthood: A systematic review. Clinical Psychology & Psychotherapy, 29(6), 1812–1842. https://doi.org/10.1002/cpp.2756
Li, T., & Chan, D. K.-S. (2012). How anxious and avoidant attachment affect romantic relationship quality differently: A meta-analytic review. European Journal of Social Psychology, 42(4), 406–419. https://doi.org/10.1002/ejsp.1842
Overall, N. C., Pietromonaco, P. R., & Simpson, J. A. (2022). Buffering and spillover of adult attachment insecurity in couple and family relationships. Nature Reviews Psychology, 1, 101–111. https://doi.org/10.1038/s44159-021-00011-1
Pinquart, M., Feussner, C., & Ahnert, L. (2013). Meta-analytic evidence for stability in attachments from infancy to early adulthood. Attachment & Human Development, 15(2), 189–218. https://doi.org/10.1080/14616734.2013.746257
Pollard, C., Bucci, S., & Berry, K. (2023). A systematic review of measures of adult disorganized attachment. British Journal of Clinical Psychology, 62(2), 329–355. https://doi.org/10.1111/bjc.12411
Ravitz, P., Maunder, R., Hunter, J., Sthankiya, B., & Lancee, W. (2010). Adult attachment measures: A 25-year review. Journal of Psychosomatic Research, 69(4), 419–432. https://doi.org/10.1016/j.jpsychores.2009.08.006
Roisman, G. I., Holland, A., Fortuna, K., Fraley, R. C., Clausell, E., & Clarke, A. (2007). The Adult Attachment Interview and self-reports of attachment style: An empirical rapprochement. Journal of Personality and Social Psychology, 92(4), 678–697. https://doi.org/10.1037/0022-3514.92.4.678
Simpson, J. A., & Rholes, W. S. (2017). Adult attachment, stress, and romantic relationships. Current Opinion in Psychology, 13, 19–24. https://doi.org/10.1016/j.copsyc.2016.04.006
Zhang, X., Li, J., Xie, F., Chen, X., Xu, W., & Hudson, N. W. (2022). The relationship between adult attachment and mental health: A meta-analysis. Journal of Personality and Social Psychology, 123(5), 1089–1137. https://doi.org/10.1037/pspp0000437