Childhood Trauma and Attachment: What Research Supports and What It Does Not
Author: Ukrainian Psychological Hub · Published: September 22, 2026 · Editorial Policy
Childhood trauma can matter for attachment, but the relationship is probabilistic rather than deterministic. Research consistently finds that childhood maltreatment and other adverse caregiving experiences are associated, on average, with greater attachment insecurity. The clearest recent quantitative evidence is a 2026 meta-analysis of 211 studies and 82,376 participants: childhood maltreatment was correlated with higher adult attachment anxiety and higher adult attachment avoidance. Those associations were meaningful, yet modest enough to rule out the popular idea that a traumatic childhood assigns a person one inevitable adult attachment style.
That distinction is the center of this article. Childhood trauma is an exposure-and-response concept. Attachment is a developmental and relational construct. Adult attachment research often describes variation along the dimensions of anxiety and avoidance. None of those constructs is a diagnosis, and none can be used as a retrospective test that reveals what happened in childhood. For a broader explanation of the adult constructs themselves, see Adult Attachment Theory and Attachment Styles in Adults.
The strongest scientific conclusion is therefore specific: adverse and maltreating childhood environments can be one pathway toward later attachment insecurity, but they are neither necessary nor sufficient for anxious, avoidant, fearful-avoidant, dismissing, preoccupied, or disorganized patterns. People with substantial childhood adversity can develop secure relationships. People with insecure attachment can have no known history of childhood trauma. Later relationships, social development, temperament and other individual differences, ongoing stress, and life events can reinforce or revise attachment expectations across time.
Childhood trauma and attachment: the short answer
The best current synthesis is Li and colleagues’ 2026 three-level meta-analysis, which pooled 1,304 effect sizes from 228 independent samples across 211 studies. Childhood maltreatment was associated with adult attachment anxiety at r = .23 and attachment avoidance at r = .20. Emotional maltreatment showed stronger associations than some other maltreatment categories, and effect sizes varied with measurement methods.
An r of .23 or .20 describes overlap between variables in groups of people. It does not mean that 20% or 23% of an individual person’s attachment style was “caused” by trauma. It does not allow a clinician or reader to infer maltreatment from a current relationship pattern. It does not mean that someone exposed to maltreatment will become insecurely attached. And it does not identify one specific style as the universal outcome of childhood trauma.
A 2021 systematic review and meta-analysis of 56 studies likewise found associations between recalled parent–child experiences and adult attachment anxiety and avoidance, including links involving neglect and psychological abuse. Much of that literature, however, depends on retrospective self-report, so it is especially important to distinguish association from proof of a developmental causal chain (Kim, Baek, & Park, 2021).
Prospective studies add important evidence because they establish temporal order more clearly. They still do not turn a complex developmental association into a one-cause explanation. The overall evidence supports increased probability, multiple pathways, and substantial individual variation.
Start with the definitions: trauma and attachment are different constructs
“Childhood trauma” is broad language. Depending on the study or clinical context, it can refer to exposure to potentially traumatic events, the psychological response to those events, or both. Childhood maltreatment is narrower and usually refers to abuse and neglect. Adverse childhood experiences are another partly overlapping framework. Because these categories are not identical, evidence from a maltreatment meta-analysis should not automatically be generalized to every bereavement, accident, medical event, disaster, family disruption, or other potentially traumatic experience.
Attachment refers to a system of seeking proximity, protection, comfort, and support in close relationships and to the expectations and regulatory strategies that develop around those relationships. In infancy and childhood, attachment is studied in relation to particular caregivers. In adult social-personality research, romantic attachment is commonly measured along two dimensions: attachment anxiety and attachment avoidance.
• Attachment anxiety concerns heightened concern about rejection, abandonment, or insufficient availability from an attachment figure.
• Attachment avoidance concerns discomfort with dependence, vulnerability, or closeness and a tendency to minimize reliance on attachment figures.
• Relative security is generally represented by lower anxiety and lower avoidance.
• Familiar four-style labels summarize combinations of these dimensions; they should not be treated as four diseases or four fixed biological types.
The measurement distinction matters. Infant Strange Situation classifications, Adult Attachment Interview states of mind, and adult romantic self-report dimensions are related research traditions, but they are not interchangeable measurements of one hidden object. Our guide to What Causes Attachment Styles? explains this developmental and measurement problem in greater depth.
What the evidence supports with the most confidence
1. Childhood maltreatment is associated with adult attachment anxiety and avoidance
This is the clearest contemporary finding. The 2026 meta-analysis found positive pooled correlations between childhood maltreatment and both major adult attachment dimensions (Li et al., 2026). The association was not confined to only anxiety or only avoidance.
The effect sizes also show why deterministic language is misleading. If childhood maltreatment reliably assigned one adult attachment style, the correlations would be far larger and far more specific. Instead, the evidence describes increased risk distributed across multiple attachment outcomes and moderated by the kind of adversity, measurement strategy, and other features of the studies.
2. Caregiving quality contributes to attachment development
A 2024 meta-analysis of 174 studies, 230 effect sizes, 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 associations of similar order, and maternal sensitivity was negatively associated with avoidant, resistant, and disorganized classifications (Madigan et al., 2024).
This supports a core developmental proposition: repeated experiences of whether care is available, responsive, and reasonably predictable matter. It also shows that sensitivity is one contributor rather than a total explanation. An association around r = .25 leaves substantial variation unexplained and does not justify formulas such as “inconsistent parenting causes anxious attachment” or “emotionally distant parenting causes avoidance.”
3. Prospective maltreatment evidence strengthens temporal inference
The Minnesota Longitudinal Study of Risk and Adaptation prospectively tracked abuse and neglect and later assessed attachment states of mind in young adulthood. Raby and colleagues found that childhood abuse and neglect, particularly chronic or multiple forms and abuse by a primary caregiver, predicted some forms of insecure adult attachment representation in this high-risk sample. Importantly, the pattern was not uniform across every attachment dimension or every maltreatment subtype (Raby et al., 2017).
A separate prospective cohort study by Widom and colleagues used documented childhood physical abuse and neglect, matched controls, and adult follow-up. Documented neglect and physical abuse were associated with higher adult attachment anxiety; neglect was also associated with avoidance. The design reduces the problem of reconstructing childhood solely from adult memory, while still leaving room for confounding, selection, measurement limits, and multiple developmental pathways (Widom et al., 2018).
Together, prospective findings support more than a purely retrospective story: childhood caregiving adversity can precede later attachment differences. They do not establish that a given adult’s current style was caused by one childhood event, one parent, or trauma alone.
Association is not the same as individual causation
Questions about childhood trauma and attachment often collapse several levels of inference into one. A correlation says that two variables vary together across a sample. A prospective association adds temporal order because the childhood exposure is recorded before the adult outcome. A dose–response pattern can strengthen a causal hypothesis when more or more persistent exposure is associated with larger average differences. Mediation models test whether a measured variable statistically carries part of an association. None of those designs, by itself, proves an established causal relationship for an individual person.
Developmental research has additional complications. Childhood maltreatment can co-occur with poverty, family conflict, parental stress, caregiver mental illness, substance use, instability, neighborhood disadvantage, genetic liability, and other exposures. The same conditions can shape later mental health and relationships through routes that are not attachment-specific. Some studies rely on retrospective reports, which can be informative but are influenced by memory, current mood, interpretation, and the instruments used. Other studies rely on official records, which miss adversity that was never reported or substantiated.
Attachment measures also differ. A self-report scale completed about romantic relationships is not equivalent to an observational infant procedure or an interview coded for attachment states of mind. Pooling or comparing these measures requires care. The most defensible language is therefore that maltreatment is associated with, and in prospective studies can predict, later attachment insecurity. Causal inference is stronger for some caregiving processes than for the claim that “trauma causes attachment style” as a single general law.
Does childhood trauma cause anxious attachment?
Childhood maltreatment can be one pathway to higher attachment anxiety, and the 2026 meta-analysis found a pooled association of r = .23. Prospective studies also report later anxiety-related attachment differences after documented or prospectively recorded maltreatment. That makes an association with attachment anxiety well supported at the group level (Li et al., 2026; Widom et al., 2018).
The evidence does not support reversing the inference. High attachment anxiety in an adult does not demonstrate childhood trauma. Fear of abandonment, reassurance seeking, sensitivity to relational uncertainty, or distress during separation can develop through many pathways, including later relationship experiences, current relationship dynamics, temperament and personality, mental health, and repeated social learning.
It is therefore more accurate to say that childhood maltreatment may increase the probability of attachment anxiety, not that attachment anxiety proves a traumatic childhood. For the adult construct itself, see Anxious Attachment Style.
Does childhood trauma cause avoidant attachment?
Avoidance is also associated with childhood maltreatment in the current meta-analytic literature. Li and colleagues reported a pooled r = .20 between maltreatment and adult attachment avoidance. Prospective research adds examples in which neglect or chronic adversity predicted later dismissing or avoidant patterns, but the associations differ across samples and measurement systems (Li et al., 2026; Raby et al., 2017; Widom et al., 2018).
Avoidance should not be interpreted as evidence that someone learned “never to need anyone” because a caregiver rejected them. That story can fit some people’s histories, yet the same adult score can arise through different developmental routes. Current relationship conditions can also make someone more or less willing to depend on a particular partner.
For the evidence on the adult dimension and its behavioral correlates, see Avoidant Attachment Style.
Childhood trauma, disorganized attachment, and fearful-avoidant attachment
This is one of the areas where popular psychology most often merges distinct constructs. Disorganized attachment in infancy is an observational classification based on behavior in a caregiver–child procedure. Unresolved/disorganized states of mind in the Adult Attachment Interview are coded from discourse concerning attachment-related loss or trauma. Fearful-avoidant attachment in adult self-report research usually refers to the combination of high attachment anxiety and high attachment avoidance. These are not three names for the same thing.
A classic meta-analysis involving nearly 80 studies and more than 6,000 infant–parent dyads found that infant disorganized attachment was more common in high-risk contexts, including maltreatment, while also emphasizing that multiple precursors were involved and that frightening caregiving was not the only proposed mechanism (van IJzendoorn, Schuengel, & Bakermans-Kranenburg, 1999).
That evidence does not justify the online shortcut “childhood trauma equals disorganized attachment,” and it certainly does not justify diagnosing an adult as disorganized because they report contradictory feelings about intimacy. Different adult measures use different constructs, and the fearful-avoidant self-report category should not be treated as a direct adult continuation of infant disorganization.
For a focused explanation of these measurement boundaries, see Disorganized Attachment in Adults and Fearful-Avoidant Attachment Style.
Why there is no one-to-one map from trauma type to attachment style
Research sometimes finds stronger average associations for particular forms of maltreatment. In the 2026 meta-analysis, emotional maltreatment showed stronger links with adult attachment insecurity than some other categories. Other prospective studies have found different patterns for neglect, physical abuse, chronicity, caregiver-perpetrated abuse, or cumulative exposure.
These findings are scientifically useful as moderators: they help explain why effect sizes differ across groups and studies. They are not a decoding key for individuals. “Emotional neglect causes avoidant attachment,” “physical abuse causes fearful avoidance,” or “inconsistency causes anxiety” are stronger claims than the evidence can support.
Several reasons prevent a simple mapping. Maltreatment categories frequently co-occur. Severity, timing, chronicity, relationship to the perpetrator, availability of another safe caregiver, sibling and peer relationships, socioeconomic context, and later experiences differ widely. People also differ in temperament, interpretation, coping, social opportunity, and the partners they later encounter. The developmental system is many-to-many: one exposure can lead to different outcomes, and similar adult outcomes can emerge from different histories.
Can someone have childhood trauma and secure attachment?
Yes. Population-level risk does not erase individual variability. A person can have a history of maltreatment, loss, violence, or other childhood trauma and still show secure attachment in a current relationship or on an adult attachment measure. Security can reflect protective relationships in childhood, supportive relationships later, changing expectations, relationship-specific experiences, or other developmental resources.
Research on intergenerational patterns illustrates the same principle. A 2025 pair of meta-analyses found only a weak association between a parent’s own childhood maltreatment history and offspring attachment insecurity, and no statistically significant pooled association with offspring disorganization. The findings are inconsistent with the idea that insecure attachment is transmitted automatically from one generation to the next (Sirparanta et al., 2025).
Protective factors should not be framed as moral toughness. Secure relationships, reliable adults, social support, material stability, opportunities for repair, and later relational experiences can alter developmental trajectories. A secure outcome does not make prior adversity less real, and an insecure outcome does not indicate personal failure.
Attachment shows continuity, but it is not fixed in childhood
A meta-analysis of 127 papers and 21,072 attachment assessments found moderate overall stability from infancy to early adulthood, r = .39, with stronger stability over shorter intervals and no significant stability in intervals longer than 15 years in that dataset (Pinquart, Feussner, & Ahnert, 2013).
Within early childhood, a later meta-analysis found moderate rather than complete stability: four-way classification stability was κ = .23, and secure-versus-insecure stability was r = .28. The authors also detected publication bias (Opie et al., 2021).
Longitudinal work on adult attachment reaches a similar conclusion. A review by Fraley and Roisman describes early caregiving as a contributor to adult attachment while emphasizing that prospective associations are often weak or inconsistent across measurement domains and that early experiences do not determine adult outcomes (Fraley & Roisman, 2019).
Adult attachment can also shift around later experiences. A multiwave longitudinal study of more than 4,000 people found that many life events were associated with immediate changes in attachment, and some event types were associated with more enduring average change, with substantial individual differences (Fraley, Gillath, & Deboeck, 2021).
This is why “your attachment style was set by age three” is not a research-based summary. Development shows both continuity and revision. Our dedicated article Can Attachment Style Change? covers stability, life events, and change in more detail, while Earned Secure Attachment examines the narrower concept of security developing despite earlier insecurity or adversity.
Plausible pathways between adverse caregiving and later attachment
The evidence supports several plausible developmental pathways, but these should be treated as mechanisms under study rather than universal scripts.
Expectations about availability and responsiveness
Repeated experiences can shape expectations about whether support will be available when needed. When care is reliably responsive, seeking help is more likely to be reinforced. When care is frightening, rejecting, inconsistent, or unavailable, other strategies may become more adaptive in that environment. Attachment theory provides a coherent framework for these learning histories, while empirical work shows that no single caregiving behavior explains all later attachment variance.
Emotion regulation within relationships
Close relationships provide opportunities to regulate distress with another person. Adverse caregiving may alter how readily someone seeks, accepts, mistrusts, or withdraws from interpersonal support. This is a plausible relational pathway, yet it should not be reduced to slogans about a permanently dysregulated nervous system. Attachment measures describe patterns of relational expectation and behavior; they do not diagnose a damaged autonomic system.
Chronicity, cumulative exposure, and caregiver involvement
Prospective findings suggest that chronic or multiple forms of abuse and neglect, and adversity involving primary caregivers, can matter for later attachment representations in some samples (Raby et al., 2017). These are moderator findings, not deterministic rules.
Later relationships as new information
Friendships, mentors, partners, and other close relationships can confirm earlier expectations or provide contradictory experience. This matters because adult attachment is partly relationship-specific. An adult may feel relatively secure with one partner and more anxious or avoidant with another. Current relational behavior therefore cannot be read as a simple fossil of childhood.
Prospective developmental evidence also points beyond a one-event childhood explanation. In the NICHD Study of Early Child Care and Youth Development, variation in age-18 attachment was linked with the quality of caregiving across development, emerging social competence, and the quality of a best friendship, illustrating a pathway that accumulates across multiple relationships (Fraley et al., 2013).
Intervention evidence supports modifiability in childhood
Randomized intervention meta-analysis provides causal support for one part of the developmental model: interventions that improve caregiver sensitivity can produce small improvements in infant attachment security. A 2003 meta-analysis of randomized interventions reported effects on both sensitivity and attachment insecurity, and a later systematic review of RCTs also found improvements in attachment security, though the later review included only four studies (Bakermans-Kranenburg, van IJzendoorn, & Juffer, 2003; Mountain, Cahill, & Thorpe, 2017).
This is evidence that attachment-relevant processes can change. It is not evidence that every adult with childhood trauma requires attachment-focused therapy, nor that an intervention developed for infant–caregiver dyads can be assumed to work for adult romantic attachment.
What research does not support
• Childhood trauma does not automatically produce an insecure adult attachment style.
• Anxious, avoidant, or fearful-avoidant attachment does not prove a history of abuse, neglect, or trauma.
• One attachment label cannot identify which parent behaved in which way.
• A specific maltreatment category does not map reliably onto one specific adult attachment style.
• Infant disorganized attachment, Adult Attachment Interview unresolved/disorganized states, and fearful-avoidant romantic self-report are not interchangeable constructs.
• Attachment insecurity is not a psychiatric diagnosis and should not be treated as a disorder score.
• Attachment theory does not explain every adult mental health symptom, relationship problem, or physical health condition.
• Current distress, emotional numbness, memory gaps, people pleasing, conflict avoidance, jealousy, or difficulty trusting do not by themselves demonstrate hidden childhood trauma.
• Attachment language cannot validate a recovered memory or establish that an unremembered traumatic event occurred.
For a broader review of the theory’s evidence base and common misuses, see Is Attachment Theory Evidence-Based?.
Childhood trauma, “attachment trauma,” attachment disorders, and trauma bonding
The phrase “attachment trauma” is used in psychotherapy, trauma discourse, and popular psychology to describe harmful experiences occurring in attachment relationships. It can be a clinically meaningful descriptive phrase, but it is not a standalone DSM or ICD diagnosis. Our article Attachment Trauma explains what the term can usefully mean and where it becomes too broad.
Reactive attachment disorder is a different concept. It is a childhood clinical diagnosis associated with extreme insufficient care and a particular pattern of inhibited attachment behavior. Adult attachment anxiety or avoidance should not be renamed “attachment disorder.” See Attachment Disorder in Adults for the diagnostic distinction.
Trauma bonding is different again. In common clinical and advocacy usage, it refers to powerful attachment within an abusive or coercive relationship, often discussed in relation to intermittent reinforcement and cycles of harm and reconciliation. It is not a synonym for childhood trauma, insecure attachment, or “attachment trauma.”
PTSD and complex PTSD are also separate diagnostic concepts. Trauma exposure does not equal PTSD, and an attachment pattern does not establish PTSD or complex PTSD. A person can have trauma-related symptoms without a particular attachment style, or attachment insecurity without a trauma-related disorder.
Why self-assessment has limits
It can be useful to notice patterns in closeness, dependency, reassurance seeking, withdrawal, conflict, trust, and help-seeking. Those observations are best treated as descriptions of current experience, not as a forensic reconstruction of childhood.
An attachment questionnaire can estimate where someone falls on a self-report dimension. It cannot determine whether childhood trauma occurred, identify the cause of a current pattern, or diagnose a trauma-related disorder. Likewise, an ACE questionnaire counts selected categories of adversity; it is not an attachment test and cannot tell a person how “traumatized” they are.
If memories of childhood are incomplete, ordinary forgetting, childhood amnesia, attention, family narrative, current interpretation, and many other factors can contribute. Attachment patterns, dreams, bodily sensations, emotional reactions, or relationship difficulties should not be used as evidence that a hidden traumatic event must have occurred.
What this evidence means in practice
For someone who knows they experienced childhood maltreatment or another traumatic childhood event, attachment concepts can provide one framework for understanding how expectations of closeness and support developed. The most useful questions are usually present-focused: What happens when I need reassurance? What happens when someone depends on me? Which situations increase fear of rejection? When do I withdraw? Do these responses vary across relationships? What helps me feel safe enough to communicate more directly?
For someone who recognizes an anxious or avoidant pattern but does not know of childhood trauma, there is no scientific reason to search for a hidden traumatic explanation. Current relationship dynamics, later experiences, personality, social learning, stress, and other mental-health processes may be more relevant. Therapy should be organized around current distress, impairment, goals, and any established diagnosis rather than around the assumption that an attachment label proves childhood trauma.
For someone with a known trauma history and clinically significant post-traumatic symptoms, evidence-based trauma assessment and treatment can be appropriate. For someone whose main difficulty is a relationship pattern without PTSD, therapy may focus more directly on communication, emotion regulation, beliefs about closeness and dependence, interpersonal behavior, and the present relationship context. The treatment target should follow the clinical problem, not a viral attachment label.
A history of childhood adversity does not create an obligation to pursue trauma therapy. Some people want to explore developmental history; others benefit more from work on present relationships, symptoms, or practical stressors. The existence of an association in research does not prescribe one treatment for every individual.
How strong is the science? An evidence-status map
Established evidence
Caregiving quality, including sensitivity and responsiveness, is associated with child attachment security. Adult attachment anxiety and avoidance are measurable dimensions with a large research literature. Childhood maltreatment is associated with higher adult attachment anxiety and avoidance at the group level. Attachment shows both stability and change across development.
Longitudinal evidence
Prospective cohorts show that documented or prospectively recorded childhood abuse and neglect can predict later attachment differences. This strengthens temporal inference and reduces some forms of recall bias. Results differ by sample, subtype, chronicity, perpetrator relationship, and attachment measure.
Plausible mechanisms
Repeated experiences of caregiver availability, rejection, fear, inconsistency, and repair plausibly shape expectations about closeness and support. Social competence, friendships, later partners, and current relationship responsiveness can also help maintain or revise these expectations. No single pathway explains all people.
Mixed or context-dependent evidence
Claims that one kind of trauma produces one specific adult style are mixed and highly context-dependent. Findings about “disorganized attachment” require particular caution because child observation, adult interview classifications, and adult romantic self-report categories measure different constructs.
Insufficient evidence
There is insufficient evidence to infer childhood trauma from a current attachment label, to use attachment style as a diagnostic trauma test, to determine a perpetrator or event from adult relationship behavior, or to claim that trauma permanently fixes attachment for life.
Methodological limits readers should know
Meta-analysis improves precision by combining studies, but it does not erase weaknesses in the studies being combined. The maltreatment–attachment literature includes retrospective self-report, cross-sectional designs, varying definitions of maltreatment, different age ranges, different attachment instruments, and populations with different levels of clinical or social risk. Heterogeneity is therefore substantive, not merely statistical.
Retrospective measures can capture experiences that official records miss, but they are also influenced by memory and current interpretation. Official records establish that authorities documented or substantiated an event, but they miss unreported abuse and can reflect unequal surveillance across communities. Prospective cohorts improve temporal ordering but may have attrition, historical cohort effects, and limited generalizability.
Cumulative adversity is another problem. Maltreatment rarely occurs in an experimental vacuum. Family violence, caregiver instability, economic hardship, housing insecurity, illness, and social disadvantage can cluster together. Statistical adjustment helps but cannot guarantee that every confounding pathway has been removed.
Finally, adult attachment measures are often self-reports about romantic relationships, whereas developmental attachment research may use behavioral observation or coded interviews. A correlation between “childhood maltreatment” and “adult attachment” can therefore mean somewhat different things across studies. Good interpretation follows the population, measurement method, and design of each study rather than treating all attachment terminology as interchangeable.
Frequently asked questions
Does childhood trauma cause attachment issues in adults?
Childhood maltreatment is associated with higher adult attachment anxiety and avoidance, and prospective studies support temporal links in some cohorts. The evidence supports increased risk, not a universal causal rule for every person. “Attachment issues” is also informal language rather than a diagnosis.
Which attachment style is most associated with childhood trauma?
There is no single style that uniquely identifies childhood trauma. The 2026 meta-analysis found maltreatment associated with both attachment anxiety and attachment avoidance. Specific subtype-to-style mappings are not reliable enough to use as individual predictions.
Can childhood trauma cause fearful-avoidant attachment?
Trauma and maltreatment can be associated with both high anxiety and high avoidance, so fearful-avoidant patterns can occur among people with traumatic histories. But fearful-avoidant self-report is not a trauma diagnosis, and it is not equivalent to infant disorganized attachment. Many developmental routes can lead to high anxiety plus high avoidance.
Does disorganized attachment mean a child was abused?
No. Maltreatment is associated with higher rates of disorganized attachment in children, but disorganization does not prove maltreatment. Multiple caregiving, contextual, and child-related factors have been studied, and the classification cannot establish that abuse occurred.
Can a person with childhood trauma have secure attachment?
Yes. Associations describe probabilities across groups. Protective relationships, later experiences, current partner responsiveness, individual development, and other factors can support security. Trauma history and current attachment security can coexist.
Can attachment style tell me whether I have repressed childhood trauma?
No. An attachment pattern cannot verify an unremembered event. Memory gaps, emotional reactions, dreams, bodily sensations, or relationship difficulties are not proof of hidden trauma. Questions about memory require their own careful clinical and neurological assessment when impairment is significant.
Is insecure attachment a mental disorder?
No. Adult attachment anxiety and avoidance are research constructs describing variation in close relationships. They are not psychiatric diagnoses. Reactive attachment disorder and disinhibited social engagement disorder are distinct childhood diagnoses and should not be confused with adult attachment styles.
Can attachment change after childhood?
Yes. Research shows moderate continuity alongside meaningful change, and adult life events and relationships can be associated with shifts in attachment. See Can Attachment Style Change? for the longitudinal evidence.
Do I need trauma therapy if I have an insecure attachment style?
Not automatically. Treatment should follow the person’s actual symptoms, impairment, goals, trauma history, and diagnoses. An attachment style by itself does not establish PTSD, complex PTSD, or a need for trauma-focused treatment.
Can childhood trauma explain all of my relationship problems?
No single developmental factor explains an adult relationship in full. Childhood experience can matter, while current partner behavior, communication, compatibility, stress, mental health, social context, and later learning also shape relationship functioning. A useful formulation keeps both developmental history and present conditions in view.
The bottom line
Childhood maltreatment and adult attachment are linked. The newest large meta-analysis finds small-to-moderate correlations with both attachment anxiety and avoidance, and prospective cohorts show that childhood abuse and neglect can precede later attachment differences. Caregiver sensitivity is also a replicated contributor to attachment security. At the same time, attachment is not a childhood verdict. Stability is incomplete, later relationships and life events matter, and early experiences do not determine a single adult outcome (Li et al., 2026; Madigan et al., 2024; Fraley & Roisman, 2019).
The scientifically defensible formulation is therefore neither “childhood trauma has nothing to do with attachment” nor “trauma determines your attachment style.” Childhood adversity can change probabilities within a developmental system that remains open to other influences. Attachment patterns can carry history without functioning as a diagnostic record of that history.
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
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
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
Mountain, G., Cahill, J., & Thorpe, H. (2017). Sensitivity and attachment interventions in early childhood: A systematic review and meta-analysis. Infant Behavior and Development, 46, 14–32. https://doi.org/10.1016/j.infbeh.2016.10.006
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
Raby, K. L., Labella, M. H., Martin, J., Carlson, E. A., & Roisman, G. I. (2017). Childhood abuse and neglect and insecure attachment states of mind in adulthood: Prospective, longitudinal evidence from a high-risk sample. Development and Psychopathology, 29(2), 347–363. https://doi.org/10.1017/S0954579417000037
Sirparanta, A., Bureau, J.-F., Yurkowski, K., Poirier, A.-A., Lavoie, H., & Lafontaine, M.-F. (2025). Parental history of childhood maltreatment and offspring attachment insecurity and disorganization: Two meta-analyses. Trauma, Violence, & Abuse, 26(3), 420–435. https://doi.org/10.1177/15248380241282995
van IJzendoorn, M. H., Schuengel, C., & Bakermans-Kranenburg, M. J. (1999). Disorganized attachment in early childhood: Meta-analysis of precursors, concomitants, and sequelae. Development and Psychopathology, 11(2), 225–249. https://doi.org/10.1017/S0954579499002035
Widom, C. S., Czaja, S. J., Kozakowski, S. S., & Chauhan, P. (2018). Does adult attachment style mediate the relationship between childhood maltreatment and mental and physical health outcomes? Child Abuse & Neglect, 76, 533–545. https://doi.org/10.1016/j.chiabu.2017.05.002