What Causes Attachment Styles? Childhood, Caregiving, Temperament, and Later Relationships
Updated: 6 days ago
Author: Ukrainian Psychological Hub · Published: September 22, 2026 · Editorial Policy
Attachment styles are shaped by development across time. Early caregiving matters, but there is no scientifically defensible formula in which one parent, one childhood event, or one age permanently assigns a person a secure, anxious, avoidant, or fearful-avoidant “type.” Adult attachment reflects a changing system built from repeated experiences of closeness, availability, protection, rejection, uncertainty, and responsiveness, together with individual differences in temperament and biology, broader family and cultural conditions, adversity, friendships, romantic relationships, and later life events.
The strongest evidence therefore supports a developmental model rather than a four-box origin story. In childhood, researchers study attachment relationships between children and particular caregivers. In adulthood, self-report research often measures two dimensions—attachment anxiety and attachment avoidance—and summarizes combinations of those dimensions with familiar style labels. Those constructs are related, but they are not interchangeable.
What causes attachment styles? The short answer
Adult attachment patterns emerge from multiple influences that accumulate and interact:
• caregiving that is more or less sensitive, responsive, predictable, and available;
• the child’s temperament and other individual characteristics;
• family stress, loss, conflict, instability, and the wider social environment;
• adverse childhood experiences, including maltreatment, which increase risk without determining one inevitable adult outcome;
• genetic differences and gene–environment processes;
• social development, friendships, and experiences outside the parent–child relationship;
• later romantic relationships and the perceived responsiveness of current partners;
• major life events and the meaning a person assigns to them;
• the fact that attachment can be relationship-specific, so the same adult may feel relatively secure with one person and more anxious or avoidant with another.
No single factor explains everyone. Even one of the most replicated childhood predictors—caregiver sensitivity—has a meaningful but moderate association with child attachment security. A 2024 meta-analysis covering 174 studies and 22,914 parent–child dyads found an overall association of r = .25 between caregiver sensitivity and child attachment security; maternal and paternal sensitivity showed similar associations in magnitude (Madigan et al., 2024). That is important evidence for caregiving, and it is also evidence against treating caregiving as a complete explanation.
For the broader distinction between potentially traumatic events, trauma responses, ACEs, clinical disorders, and population-level adult outcomes, see Childhood Trauma in Adults: Signs, Long-Term Effects, Relationships, and Treatment.
First, separate infant attachment from adult attachment style
Many popular explanations begin with an infant classification and end with an adult romantic label as though they were the same variable measured at two ages. Attachment research is more complicated.
The Strange Situation is an observational procedure developed for studying young children’s behavior with a caregiver during separation and reunion. It yields classifications such as secure, avoidant, resistant/ambivalent, and, in later work, disorganized. These classifications concern a child’s attachment relationship in a particular developmental context.
The Adult Attachment Interview, or AAI, is a different method. It evaluates the organization and coherence of an adult’s discourse about attachment experiences and includes classifications such as secure-autonomous, dismissing, preoccupied, and unresolved/disorganized. It is not simply a questionnaire asking whether someone is anxious or avoidant in romance.
Adult romantic attachment research often uses self-report measures organized around attachment anxiety and attachment avoidance. Anxiety concerns fears about rejection, abandonment, or insufficient availability. Avoidance concerns discomfort with dependence, closeness, or relying on others. Secure attachment is generally represented by lower anxiety and lower avoidance; familiar categorical labels are summaries of locations in this dimensional space.
This distinction matters for causation. A childhood Strange Situation classification cannot simply be read forward as an adult romantic style, and an adult self-report score cannot be read backward as a reconstruction of how a parent behaved. A review of the development of adult attachment concludes that early caregiving appears to contribute to adult attachment, while prospective longitudinal associations are often weak or inconsistent across different measurement traditions (Fraley & Roisman, 2019).
Fearful-avoidant adult self-report patterns deserve the same caution. High anxiety plus high avoidance is not automatically the same construct as infant disorganized attachment or an unresolved/disorganized AAI classification. Those distinctions belong to different measures and research literatures.
Caregiving matters, especially responsiveness and sensitivity
Attachment theory gives caregiving a central role because children learn, over repeated interactions, what tends to happen when they seek proximity, comfort, protection, or help. A caregiver who notices signals, interprets them reasonably well, and responds in a timely and appropriate way creates more opportunities for the child to experience the relationship as a dependable source of support.
Meta-analytic evidence supports this basic proposition. An earlier synthesis of 66 studies involving 4,176 parent–child dyads found a modest association between parental sensitivity and infant attachment security (De Wolff & van IJzendoorn, 1997). The much larger 2024 meta-analysis by Madigan and colleagues found a pooled r = .25 across 174 studies and also showed that paternal sensitivity was positively associated with security, not merely maternal sensitivity (Madigan et al., 2024).
Intervention research strengthens the causal case. A meta-analysis of randomized early interventions found that programs targeting sensitive caregiving improved parental sensitivity and produced smaller but significant improvements in infant attachment security; interventions that changed sensitivity more successfully also tended to improve attachment outcomes more strongly (Bakermans-Kranenburg, van IJzendoorn, & Juffer, 2003).
The implication is precise: responsive caregiving is one real developmental influence. The evidence does not support a rule such as “an inconsistent mother causes anxious attachment” or “an emotionally distant father causes avoidant attachment.” Effect sizes are not large enough for that inference, children have more than one relationship and more than one developmental context, and adult romantic attachment is measured differently from infant attachment.
Why the one-parent, one-style explanation fails
A simple causal story is attractive because it turns a complex adult pattern into an origin narrative. It can also create false certainty.
First, children can form distinct attachment relationships with different caregivers. Security with one caregiver does not require an identical pattern with every other caregiver.
Second, caregiver sensitivity explains only part of the variation. Intergenerational attachment research also finds a “transmission gap”: parents’ own attachment representations are associated with their children’s attachment, but observed sensitivity does not fully explain that association. A meta-analysis of 95 samples and 4,819 participants found an association of r = .31 for secure-autonomous transmission and r = .21 for unresolved transmission, while confirming that sensitivity only partly accounted for transmission (Verhage et al., 2016).
Third, families exist inside larger systems. Economic strain, parental mental health, work conditions, social support, illness, migration, conflict, housing instability, and cultural caregiving practices can alter what responsiveness looks like and how consistently it is possible. Cross-cultural work has emphasized “universality without uniformity”: responsiveness to an infant’s needs can be expressed through different caregiving arrangements and behaviors across communities (Mesman et al., 2018).
Fourth, the adult outcome itself is not a single fixed trait. Adult attachment can include a general tendency plus relationship-specific expectations. A person can therefore carry a relatively stable global pattern while feeling more or less secure with particular attachment figures.
Temperament influences the system, but temperament is not attachment
Babies differ from one another from the beginning. Some are more reactive, more easily distressed, more behaviorally inhibited, or easier to soothe. These differences can affect how children signal needs, how caregivers respond, and how stressful particular interactions become. That makes temperament part of the developmental system.
It does not make attachment a disguised temperament test.
A meta-analysis of 109 independent samples involving 11,440 children found only a weak overall association between temperament and attachment security versus insecurity. Temperament was modestly associated with resistant attachment, but was not significantly associated with avoidant or disorganized attachment in the pooled analyses (Groh et al., 2017).
This distinction helps explain why two children raised in the same household can develop differently without implying that one child’s attachment was “caused by personality.” Temperament can influence the transactions that unfold between child and caregiver, while attachment also reflects experience within the relationship.
Genetics contribute to adult attachment differences
Adult attachment is neither purely learned nor reducible to genes. Behavioral-genetic evidence points to contributions from both.
A 2025 study using the Minnesota Twin Registry examined 1,377 older adult twins and measured both general attachment and relationship-specific attachment to mothers, fathers, partners, and best friends. The study estimated that about 36% of variation in adult attachment styles in that sample was attributable to genetic differences and about 64% to environmental influences not shared by twins. Heritability estimates were higher for parent-specific attachment, while nonshared environmental contributions were larger for partner- and best-friend-specific attachment (Dugan et al., 2025).
Heritability has a technical meaning. It describes variation among people in a particular population under particular conditions; it does not mean that 36% of an individual person’s attachment style is “genetic,” and it does not imply immutability. Genetic influences can operate partly through temperament, personality, emotion regulation, sensitivity to social cues, partner selection, and the environments people encounter or help create.
Prospective developmental research also warns against simplistic gene stories. A longitudinal study following children from infancy to age 18 found that caregiving quality, emerging social competence, and best-friend relationship quality predicted later adult attachment, while most candidate genetic variants and temperament measures examined in that study showed little or no direct association (Fraley et al., 2013). Taken together with newer twin evidence, the broader picture is gene–environment interplay rather than a single “attachment gene.”
Adversity and maltreatment can increase risk without determining destiny
Severe or chronic adversity can affect expectations about safety, trust, availability, and closeness. Childhood maltreatment is therefore an important risk factor in adult attachment research.
The newest large synthesis is a 2026 three-level meta-analysis of 211 studies, 228 independent samples, 1,304 effect sizes, and 82,376 participants. Childhood maltreatment was associated with higher adult attachment anxiety at r = .23 and higher avoidance at r = .20, with stronger associations for emotional maltreatment (Li et al., 2026).
Those correlations matter, but they do not allow a clinician, partner, or online quiz to infer maltreatment from an adult attachment pattern. Many people with adverse childhood histories do not show the same adult attachment profile, and adults with elevated anxiety or avoidance do not necessarily have a history of maltreatment.
A 2021 systematic review and meta-analysis of 56 studies likewise found associations between adult insecure attachment and multiple reported parent–child experiences, including neglect and psychological abuse, but much of this literature relies on retrospective reports of childhood experience (Kim, Baek, & Park, 2021). Retrospective association is informative, yet it cannot reconstruct a single causal pathway for an individual.
The phrase “attachment trauma” is widely used in therapy and popular psychology, but it is not itself an official DSM or ICD diagnosis. It should not become a catch-all explanation for every insecure adult relationship pattern. Maltreatment, traumatic events, disrupted caregiving, attachment classifications, post-traumatic symptoms, and adult attachment anxiety or avoidance are related topics that require separate definitions and evidence.
Early attachment shows continuity, not destiny
If attachment develops through experience, some continuity is expected. If development remains open to new experience, perfect continuity is not.
A meta-analysis of 127 papers encompassing 21,072 attachment assessments from infancy to early adulthood found an overall stability coefficient of r = .39. Stability was stronger over shorter intervals and no significant stability was detected for intervals longer than 15 years in that dataset (Pinquart, Feussner, & Ahnert, 2013).
Within early childhood itself, a later meta-analysis found moderate rather than complete stability. Across infancy, toddlerhood, and preschool/early-school periods, the pooled four-way classification stability was κ = .23 and secure-versus-insecure stability was r = .28; the authors also found evidence of publication bias (Opie et al., 2021).
These findings are incompatible with the claim that an attachment pattern is permanently “set” in infancy. Early relationships can create expectations and regulatory strategies that persist, yet subsequent development can consolidate, revise, or contextualize them.
Friendships and social development also matter
Parents are not the only interpersonal input between infancy and adult romance. Peers, friendships, mentors, siblings, and other close relationships provide repeated experiences of acceptance, rejection, dependence, trust, repair, and support.
The longitudinal study by Fraley and colleagues found that the quality of the caregiving environment, emerging social competence, and the quality of the adolescent’s best friendship all contributed to variation in attachment at age 18 (Fraley et al., 2013). That result matters because it places adult attachment development on a pathway, not at a single childhood moment.
Developmental models therefore make more sense when they ask how relationship expectations are updated across many social contexts. A child who had inconsistent caregiving can later encounter dependable friends and partners. A child who experienced security can later encounter betrayal, coercion, repeated rejection, or major loss. Neither pathway guarantees a particular outcome, but both can provide new information to the attachment system.
Later romantic relationships can reinforce or revise attachment expectations
Adult relationships are not merely places where a childhood style is displayed. They can also become environments in which attachment expectations are maintained, challenged, or reorganized.
In a study of 236 couples, people who perceived their partners as more responsive reported lower partner-specific attachment anxiety and avoidance, with especially strong benefits among people who were generally more insecure (Rice, Kumashiro, & Arriaga, 2020). This finding fits the broader idea that feeling understood, valued, and cared for can support security within a particular relationship. Our separate guide to perceived partner responsiveness explains that construct in more depth.
Partner-specific security can also move in coordinated ways within couples. A longitudinal study of 172 couples assessed five times over one year found coordinated changes in partners’ relationship-specific security after accounting for their prototypical levels (Hudson et al., 2014).
At the same time, current relationships should not be treated as omnipotent “reprogramming” devices. A 2025 study of 1,036 established couples followed across 20 months found only small average change: attachment anxiety decreased modestly among women, avoidance was comparatively stable, and most measured relationship characteristics were not reliable predictors of change (Traut et al., 2025). The evidence supports both stability and change, depending on the person, dimension, relationship, and time scale.
Life events can change attachment, but effects vary
Major experiences can alter expectations about whether other people will remain available, whether closeness is safe, and whether dependence will be met with care or disappointment. Starting or ending a relationship, bereavement, betrayal, illness, parenthood, relocation, or other events may therefore coincide with changes in attachment.
A multiwave study of more than 4,000 people examined attachment trajectories before and after naturally occurring life events. About half of the event types studied were associated with immediate change. On average, people often moved back toward the security level predicted by their prior trajectory, yet a quarter of the event types were associated with enduring average changes. Individual differences were substantial, and people’s interpretation of events as positive or negative helped predict change (Fraley, Gillath, & Deboeck, 2021).
Relationship stage can matter too. In a 2024 longitudinal study of 1,741 adults, people in newer romantic relationships showed greater fluctuations in partner-specific attachment anxiety and stronger anxiety responses to attachment-relevant events than people in more established relationships. The same pattern was not found for avoidance (Dugan, Fraley, Gillath, & Deboeck, 2024).
Adult attachment is therefore better understood as a structured tendency with room for context and updating than as a permanent personality category.
Relationship-specific attachment explains why one person can look “different” with different people
A person can have a general tendency toward greater security, anxiety, or avoidance and still show different patterns in different close relationships. This is one reason anecdotes such as “I was secure until I dated this person” or “I am anxious with partners but not with friends” can contain psychologically meaningful information without proving that a partner single-handedly created a new global style.
The Minnesota Twin Registry study found that the balance of genetic and nonshared environmental influences differed across parent-, partner-, and friend-specific attachment (Dugan et al., 2025). Couple research likewise shows that perceived partner responsiveness is associated with security toward that particular partner (Rice et al., 2020).
This relationship-specificity is also why a current romantic relationship should not be used as a diagnostic test of childhood. What happens with one partner may reflect longstanding expectations, characteristics of the current relationship, the partner’s behavior, current stress, recent events, or interactions among all of these factors.
What tends to contribute to secure, anxious, avoidant, and fearful-avoidant patterns?
The familiar four labels are most useful as plain-language summaries of anxiety and avoidance, not as four diseases with four separate causes.
Secure patterns, characterized by relatively low anxiety and low avoidance, are more likely when close relationships repeatedly provide credible experiences of availability, responsiveness, repair, and support. Early sensitive caregiving contributes to security, and later responsive relationships can support partner-specific security.
Anxious or preoccupied patterns emphasize attachment anxiety. Developmentally, anxiety can be understandable when availability feels uncertain or difficult to predict, but no single caregiving behavior uniquely produces adult anxiety. Temperament, stressful experiences, peer and romantic relationships, loss, current partner responsiveness, and individual interpretation of events can all matter.
Dismissive-avoidant patterns emphasize attachment avoidance. Avoidance can function as a strategy of minimizing dependence or attachment needs when closeness is expected to be unrewarding, intrusive, unsafe, or unreliable. Again, an adult avoidance score does not reveal which childhood event—or which parent—caused it.
Fearful-avoidant adult self-report patterns combine high anxiety with high avoidance. They can involve simultaneously wanting closeness and anticipating danger, rejection, or disappointment. A history of adversity can increase risk for both anxiety and avoidance at the population level, but fearful-avoidant scores are not proof of trauma and should not automatically be relabeled “disorganized attachment.”
Can you identify the childhood cause from your current attachment style?
Usually, no.
An adult self-report pattern is an assessment of current or typical expectations and behavior in close relationships. It is not a forensic record of childhood. Two adults with similar anxiety scores may have arrived there through different developmental pathways. Two siblings from the same family can differ. The same person can vary across relationships and periods of life.
Research can identify probabilistic associations across groups. It can test whether caregiving sensitivity predicts later security, whether maltreatment is associated with anxiety and avoidance, or whether current partner responsiveness relates to relationship-specific security. Those findings do not let us reverse the causal arrow and say, for example, “because you are avoidant, your caregiver must have rejected you.”
This is an important limit of attachment-style quizzes and social-media checklists. They may help organize self-reflection, but they cannot determine the cause of a person’s attachment pattern or establish a diagnosis.
Does attachment style form in the first two years of life?
Attachment relationships begin early, and infancy is developmentally important. The claim that an adult attachment style is fully formed and fixed by age two goes beyond the evidence.
Research on early childhood finds only moderate stability of attachment classifications, and studies spanning infancy to later developmental periods show partial rather than perfect continuity (Opie et al., 2021; Pinquart et al., 2013). Adult attachment also changes in response to some later life events and varies across relationships (Fraley et al., 2021).
Infancy matters because it is the beginning of the learning history, not because development stops there.
What does this mean for parents and caregivers?
Attachment research supports a practical conclusion that is more useful than parent-blame: repeated responsiveness matters, and relationships are built across many interactions.
Sensitive caregiving involves noticing signals, interpreting them in context, and responding in ways that help a child use the caregiver as a source of safety and support. Parents do not have to respond perfectly to every cue, and attachment cannot be reduced to a scorecard of isolated mistakes. The association between sensitivity and security is moderate, not deterministic, and randomized intervention evidence shows that caregiving behavior itself can change (Bakermans-Kranenburg et al., 2003).
It also makes sense to view parenting in context. Caregivers function under real constraints, including sleep deprivation, illness, trauma histories, poverty, discrimination, family conflict, limited social support, and cultural expectations about child care. A scientifically serious account of attachment includes those conditions rather than treating one caregiver as the sole source of a child’s future relationship life.
How researchers can get closer to causal answers
Different research designs answer different parts of the question.
Longitudinal studies follow people over time and test whether earlier caregiving, friendships, stressors, or relationship experiences predict later attachment. They are stronger than asking adults to remember childhood after measuring current insecurity, although they still cannot control every confounding factor.
Randomized caregiving interventions provide more direct evidence that changing parental sensitivity can change probabilities of child attachment security.
Twin studies estimate how much population variation is associated with genetic and environmental differences. They do not identify a single gene or establish a fixed genetic destiny.
Dyadic couple studies examine both partners and can test whether one person’s experience or change is coordinated with the other’s.
Relationship-specific measures distinguish a general attachment tendency from security with a particular parent, friend, or partner.
Meta-analyses pool results across many studies, giving a more stable estimate of associations and revealing heterogeneity that a single study can hide.
The convergence of these methods supports a multi-causal developmental account. Caregiving is important, biology matters, environments beyond the family matter, and later close relationships can add new learning. The relative contribution of each differs across people and across relationships.
For a broader audit of the research base, measurement limits, and common popular misuses, see Is Attachment Theory Evidence-Based? What Research Supports, What It Does Not, and Common Misuses
Frequently asked questions
What causes insecure attachment styles?
Insecure attachment is associated with multiple developmental conditions, including lower or less consistent caregiver sensitivity, adverse experiences, family stress, and later relationship experiences. Individual differences and genetics also contribute. No single factor is necessary or sufficient, and an insecure adult attachment pattern does not identify one specific childhood cause.
Is attachment style genetic?
Partly, at the population level. A 2025 twin study estimated about 36% heritability for adult attachment styles in its sample, with the remaining variation attributed mainly to nonshared environmental influences (Dugan et al., 2025). Heritability is not a percentage of an individual person and does not mean the trait cannot change.
Does temperament cause anxious or avoidant attachment?
Temperament can influence interactions and emotional regulation, but it does not explain attachment by itself. A large meta-analysis found only a weak overall association between temperament and attachment security and no significant pooled association with avoidant or disorganized attachment (Groh et al., 2017).
Can trauma cause fearful-avoidant attachment?
Adversity and maltreatment are associated with higher adult attachment anxiety and avoidance, so they can be part of some developmental pathways. A fearful-avoidant self-report pattern cannot be used as proof that trauma occurred, and it should not be treated as synonymous with infant disorganized attachment or unresolved/disorganized AAI classification.
Can you have different attachment patterns with different people?
Yes. Adult attachment can include both general tendencies and relationship-specific patterns. Research on partners, parents, and friends shows meaningful relationship-specific variation (Dugan et al., 2025; Rice et al., 2020).
Can a romantic partner change your attachment style?
A partner can contribute to experiences that increase or decrease security, particularly within that relationship, but effects vary. Perceived partner responsiveness is associated with lower partner-specific anxiety and avoidance, while longitudinal research finds that many attachment tendencies remain fairly stable and that average change can be small (Rice et al., 2020; Traut et al., 2025).
Can a secure childhood become insecure later?
Yes. Security shows continuity, not permanence. Later losses, betrayals, relationship experiences, stress, and life transitions may alter attachment trajectories for some people. The reverse is also possible: people with earlier insecurity can develop greater security over time.
Can an online attachment quiz tell me why I am anxious or avoidant?
No. A self-report quiz can, at best, summarize current responses to questions about anxiety and avoidance. It cannot establish childhood causation, identify a particular caregiver’s behavior, diagnose trauma, or determine a clinical disorder.
Is attachment style a diagnosis?
No. Secure, anxious/preoccupied, dismissive-avoidant, and fearful-avoidant are research and popular-language summaries of attachment patterns. They are not DSM or ICD diagnoses. Reactive Attachment Disorder and Disinhibited Social Engagement Disorder are distinct clinical diagnoses with their own criteria and should not be used as names for ordinary adult relationship patterns.
The bottom line
Attachment styles come from developmental histories rather than single causes. Early caregiving and sensitivity matter, but their effects are probabilistic. Temperament contributes modestly to early attachment differences. Genetic influences account for part of adult variation, while nonshared environments remain substantial. Adversity raises risk without dictating one adult outcome. Friendships, social development, romantic relationships, partner responsiveness, and life events continue to supply information that can reinforce or revise attachment expectations.
The most accurate answer to “what caused my attachment style?” is therefore usually a developmental one: a pattern emerged through repeated interactions between the person and several environments across time. Adult attachment is structured enough to show continuity and flexible enough to vary by relationship and to change.
For current romantic relationships, one mechanism that deserves special attention is how reliably a partner feels understanding, validating, and caring. See Perceived Partner Responsiveness: Why Feeling Understood, Valued, and Cared For Builds Intimacy. For the broader relational context, see Emotional Intimacy in Marriage: What It Is, What It Looks Like, and How Couples Build It.
Related Articles
References
Bakermans-Kranenburg, M. J., van IJzendoorn, M. H., & Juffer, F. (2003). Less is more: Meta-analyses of sensitivity and attachment interventions in early childhood. Psychological Bulletin, 129(2), 195–215. https://doi.org/10.1037/0033-2909.129.2.195
De Wolff, M. S., & van IJzendoorn, M. H. (1997). Sensitivity and attachment: A meta-analysis on parental antecedents of infant attachment. Child Development, 68(4), 571–591. https://pubmed.ncbi.nlm.nih.gov/9306636/
Dugan, K. A., Fraley, R. C., Gillath, O., & Deboeck, P. R. (2024). Testing the canalization hypothesis of attachment theory: Examining within-subject variation in attachment security. Journal of Personality and Social Psychology, 126(3), 511–541. https://doi.org/10.1037/pspp0000488
Dugan, K. A., Kunkel, J. K., Fraley, R. C., Briley, D. A., McGue, M., Krueger, R. F., & Roisman, G. I. (2025). Genetic and environmental contributions to adult attachment styles: Evidence from the Minnesota Twin Registry. Journal of Personality and Social Psychology, 128(3), 639–669. https://doi.org/10.1037/pspp0000516
Fraley, R. C., Gillath, O., & Deboeck, P. R. (2021). Do life events lead to enduring changes in adult attachment styles? A naturalistic longitudinal investigation. Journal of Personality and Social Psychology, 120(6), 1567–1606. https://doi.org/10.1037/pspi0000326
Fraley, R. C., & Roisman, G. I. (2019). The development of adult attachment styles: Four lessons. Current Opinion in Psychology, 25, 26–30. https://doi.org/10.1016/j.copsyc.2018.02.008
Fraley, R. C., Roisman, G. I., Booth-LaForce, C., Owen, M. T., & Holland, A. S. (2013). Interpersonal and genetic origins of adult attachment styles: A longitudinal study from infancy to early adulthood. Journal of Personality and Social Psychology, 104(5), 817–838. https://doi.org/10.1037/a0031435
Groh, A. M., Narayan, A. J., Bakermans-Kranenburg, M. J., Roisman, G. I., Vaughn, B. E., Fearon, R. M. P., & van IJzendoorn, M. H. (2017). Attachment and temperament in the early life course: A meta-analytic review. Child Development, 88(3), 770–795. https://doi.org/10.1111/cdev.12677
Hudson, N. W., Fraley, R. C., Brumbaugh, C. C., & Vicary, A. M. (2014). Coregulation in romantic partners’ attachment styles: A longitudinal investigation. Personality and Social Psychology Bulletin, 40(7), 845–857. https://doi.org/10.1177/0146167214528989
Kim, S.-H., Baek, M., & Park, S. (2021). Association of parent–child experiences with insecure attachment in adulthood: A systematic review and meta-analysis. Journal of Family Theory & Review, 13(1), 58–76. https://doi.org/10.1111/jftr.12402
Li, S., Ma, Y., Cao, X., Liu, J., Qi, Y., & Chen, X. (2026). Childhood maltreatment and adult attachment: A three-level meta-analytic review. Trauma, Violence, & Abuse. Advance online publication. https://doi.org/10.1177/15248380261455767
Madigan, S., Deneault, A.-A., Duschinsky, R., Bakermans-Kranenburg, M. J., Schuengel, C., van IJzendoorn, M. H., Ly, A., Fearon, R. M. P., Eirich, R., & Verhage, M. L. (2024). Maternal and paternal sensitivity: Key determinants of child attachment security examined through meta-analysis. Psychological Bulletin, 150(7), 839–872. https://doi.org/10.1037/bul0000433
Mesman, J., Minter, T., Angnged, A., Cissé, I. A. H., Salali, G. D., & Migliano, A. B. (2018). Universality without uniformity: A culturally inclusive approach to sensitive responsiveness in infant caregiving. Child Development, 89(3), 837–850. https://doi.org/10.1111/cdev.12795
Opie, J. E., McIntosh, J. E., Esler, T. B., Duschinsky, R., George, C., Schore, A., Kothe, E. J., Tan, E. S., Greenwood, C. J., & Olsson, C. A. (2021). Early childhood attachment stability and change: A meta-analysis. Attachment & Human Development, 23(6), 897–930. https://doi.org/10.1080/14616734.2020.1800769
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
Rice, T. M., Kumashiro, M., & Arriaga, X. B. (2020). Mind the gap: Perceived partner responsiveness as a bridge between general and partner-specific attachment security. International Journal of Environmental Research and Public Health, 17(19), 7178. https://doi.org/10.3390/ijerph17197178
Traut, A. C., Gander, F., Uhlich, M., Weidmann, R., Chopik, W. J., & Grob, A. (2025). Examining change in attachment in romantic couples: The role of relationship characteristics and codevelopment between partners. Journal of Personality and Social Psychology, 129(3), 602–617. https://doi.org/10.1037/pspp0000558
Verhage, M. L., Schuengel, C., Madigan, S., Fearon, R. M. P., Oosterman, M., Cassibba, R., Bakermans-Kranenburg, M. J., & van IJzendoorn, M. H. (2016). Narrowing the transmission gap: A synthesis of three decades of research on intergenerational transmission of attachment. Psychological Bulletin, 142(4), 337–366. https://doi.org/10.1037/bul0000038