Attachment in Infancy and Early Childhood: What Security Means
Attachment is the organized relationship through which an infant or young child seeks protection, emotional reassurance, and access to a familiar caregiver when tired, frightened, ill, uncertain, or overwhelmed. An attachment figure becomes a safe haven during distress and a secure base from which the child can explore people, objects, and unfamiliar situations. Security describes the child's emerging expectation that this particular caregiver will be available and helpful when support is needed. These expectations develop within repeated interactions and are expressed through behavior long before children can describe them in words. The central scientific question is how children organize these relationships over time, how researchers measure their quality, and what attachment evidence can reasonably tell us about development (Benoit, 2004; National Research Council and Institute of Medicine, 2000).
Most young children form attachments to familiar caregivers, including when the relationship is stressful or inconsistently responsive. An attachment bond is therefore a relationship, while attachment security refers to one important quality of its organization. Secure attachment is expressed through a flexible combination of seeking comfort, accepting reassurance, returning to play, and keeping track of a caregiver's availability. The precise visible behaviors vary with age, mobility, temperament, the situation, familiarity, and cultural practices. A securely attached infant can cry intensely during a separation; a child who approaches strangers with ease is not automatically securely attached. Researchers infer attachment organization from patterns across structured observations, rather than from a single moment of affection or separation distress (Benoit, 2004; American Academy of Pediatrics).
Attachment theory provides an important developmental model of how protective relationships support children's regulation and exploration. Its strongest empirical base concerns observed caregiver-child behavior, the association between caregiving sensitivity and security, and the effects of selected early interventions. Longer-range links with cognition, friendships, and emotional or behavioral difficulties are generally probabilistic and of modest average size. A child's early attachment classification cannot establish a diagnosis or determine an adult personality or relationship destiny (Madigan et al., 2024; Greenwood et al., 2020; Groh et al., 2014).
Origins and Core Concepts of Attachment Theory
Bowlby's protection system and Ainsworth's secure base
John Bowlby proposed that young humans have evolved behavioral systems that increase proximity to protective adults when danger or distress is perceived. Crying, following, reaching, checking back, clinging, and seeking proximity can serve that goal. During conditions of relative safety, exploration becomes more prominent. Mary Ainsworth developed the secure-base construct through detailed observation and later devised the Strange Situation Procedure to study how infants coordinate exploration, separation, and reunion with familiar caregivers. This historical framework remains useful because it predicts observable changes in the child's behavior as perceived safety and caregiver availability change (National Research Council and Institute of Medicine, 2000; Benoit, 2004).
Attachment theory also describes internal working models: developing expectations about whether a caregiver can be relied upon, whether seeking help is effective, and how relationships operate under uncertainty. In infancy these expectations are inferred from repeated interaction patterns, not measured as fully articulated beliefs. During the preschool years, growing language, memory, symbolic representation, and social understanding make attachment-related expectations more expressible in stories, conversations, and behavior. Internal working models are theoretically useful summaries of experience, but the same observable behavior can reflect several interacting mechanisms; a caregiver-child relationship should not be reduced to an unseen mental script.
Attachment, bonding, affection, dependence, and temperament
Attachment must be distinguished from the adult's emotional bond with the infant, sometimes called bonding. Caregiver affection, delight, skin-to-skin contact, and enjoyment can support relationships, yet one early contact episode does not establish the infant's attachment classification. Likewise, feeding, entertaining, instruction, and discipline contribute to caregiving without defining attachment security. Children can use a caregiver for comfort while showing different degrees of closeness during play, and independent exploration is compatible with secure attachment. The attachment system and the exploration system work together (Benoit, 2004).
Temperament refers to relatively early-emerging individual differences in reactivity and regulation, such as sensitivity to novelty, activity, irritability, or ease of soothing. A highly reactive infant can form secure attachments, and an outwardly calm infant can have difficulty relying on a caregiver. Temperament influences what the child displays in an assessment and how caregivers perceive or respond to signals. Research using the observer Attachment Q Sort found a substantially weaker association between attachment security and temperament than between security and caregiver sensitivity, supporting their conceptual distinction (van IJzendoorn et al., 2004).
How Attachment Develops Across the Early Years
Early months: familiarity, expectation, and co-regulation
During the first months, infants recognize recurring voices, smells, faces, handling, and rhythms of care. They are dependent on adults for physiological and emotional regulation, but a recognizable attachment organization emerges gradually. Comforting an upset infant, responding to hunger and fatigue, supporting sleep and safe physical contact, and sharing attentive face-to-face interactions create recurring opportunities for the infant to learn what happens after a signal. There is no precise birthday on which attachment suddenly appears. Security is better understood as a developing organization of relationship behavior, with infants showing increasingly specific preferences as sensory, cognitive, motor, and social capacities expand (National Research Council and Institute of Medicine, 2000).
The infant's contribution matters from the start. A premature infant, a child with sensory differences, or a baby whose signals are difficult to interpret may require adjustments in timing, stimulation, positioning, and soothing. Caregivers learn through repeated attempts, and infants gradually communicate in increasingly differentiated ways. The resulting relationship is transactional: each partner's behavior changes the circumstances to which the other responds. Parenting sensitivity is best understood in relation to the child's needs and context rather than through a rigid checklist of behaviors.
Later infancy: selective attachment and separation responses
During the second half of the first year, many infants become more selective about familiar adults and increasingly use them to regulate distress and support exploration. As mobility, recognition of familiar people, and memory of separations develop, a baby may protest when a caregiver leaves the room, become wary of unfamiliar people, or seek reassurance before exploring. These behaviors vary markedly between children and situations. Separation protest by itself is neither proof of insecurity nor a screening test for pathology (American Academy of Pediatrics; National Research Council and Institute of Medicine, 2000).
The quality of reunion is often especially informative. Some infants, after distress, seek proximity, accept comfort, and resume exploration. Others minimize open displays of need or have difficulty settling even when contact is available. These responses should be interpreted as relationship-specific strategies observed in a particular setting, with careful attention to what the child actually does over repeated episodes. A brief goodbye at child care can be shaped by illness, sleep deprivation, familiarity with staff, recent family changes, or stress without revealing a stable attachment classification.
Toddlerhood: proximity, exploration, and emerging autonomy
Toddlers increasingly walk, communicate needs, test boundaries, anticipate routines, and negotiate brief separations. Attachment does not fade when autonomy grows. It becomes expressed through checking back, requesting help, seeking emotional repair after conflict, and returning to activity after reassurance. A toddler may insist on doing something independently while still wanting the caregiver nearby. Such behavior demonstrates how security and exploration can develop together. When a toddler is overwhelmed, caregiver help with identifying feelings, reducing demands, and restoring safety can support a return to action.
Development also changes the meaning of apparent dependence. Temporary increases in clinginess during illness, transitions, the arrival of a sibling, bereavement, travel, or a change in care arrangements may be adaptive signals of increased support needs. Patterns that persist with broad functional impairment deserve assessment in their own right, but routine fluctuations do not constitute an attachment disorder.
Preschool and early school years: negotiated goals and relationships
From approximately age three onward, language, autobiographical memory, and understanding of other people's intentions permit more sophisticated coordination. Children can discuss departures and reunions, predict who will collect them, seek reassurance by words or gestures, and use familiar routines to regulate distress. Studies adapt attachment assessments for this period because the behavior of a five-year-old cannot be interpreted using the same codes as that of a twelve-month-old. A meta-analysis of 97 preschool and early-childhood samples demonstrates the continuing usefulness of attachment classifications while also showing substantial variation by sample context and measurement procedure (preschool attachment meta-analysis, 2023).
Security during these years involves confidence that a caregiver will respond when needed, while day-to-day cooperation, boundaries, peer relationships, and verbal problem-solving acquire larger roles. A child's emotional life becomes distributed across several contexts, including child care, playgroups, extended family, and school. Attachment remains one developmental resource among many; social development cannot be explained by a single relationship classification.
The Four Research Patterns of Infant Attachment
Secure attachment
In a typical Strange Situation assessment, a securely attached infant uses the caregiver as a base for exploration and, if distressed by separation, seeks and uses reunion contact for comfort. The child may cry, cling, or require time to recover. Security is reflected in the organization and flexibility of the response, including the capacity to resume engagement as distress falls. Secure attachment does not require continuous happiness, instant soothing, or an absence of conflict. The evidence supports associations between security and sensitive caregiving, with many other child and contextual factors also relevant (Benoit, 2004; Madigan et al., 2024).
Insecure-avoidant attachment
An avoidant classification describes a pattern in which the infant shows relatively limited overt seeking of comfort or avoidance of the caregiver on reunion, sometimes while maintaining attention toward the environment. This pattern is interpreted as an organized strategy for managing attachment-related stress in that relationship. Limited visible crying cannot reliably be read as emotional indifference, inner calm, or a preference for independence. Nor does the classification imply that the child has no attachment. It identifies one observed organization of reunion behavior under structured conditions (Benoit, 2004).
Insecure-resistant or ambivalent attachment
A resistant or ambivalent classification describes behavior characterized by pronounced proximity seeking combined with difficulty settling or returning to exploration during reunion. An infant may seek contact while resisting aspects of it, becoming angry, or remaining preoccupied with the caregiver's availability. The pattern reflects the organization of behavior during attachment-related stress. A single prolonged tantrum, demanding phase, or difficult handoff does not establish this classification. Temperament and immediate circumstances can influence the intensity of the infant's distress independently of the relationship pattern.
Disorganized attachment
Disorganized attachment describes particular lapses or contradictions in the organization of attachment behavior during assessment, such as interrupted, conflicted, disoriented, or apprehensive behaviors in the caregiver's presence under stress. In preschool research, some coding systems describe controlling patterns as developmentally modified manifestations. Disorganization has been associated with elevated risk in some adversity-exposed samples, but the classification is heterogeneous and should be interpreted within the validated assessment context. It is not a psychiatric diagnosis, cannot by itself establish maltreatment, and cannot identify an abusive parent. Treating it as an individual risk verdict is inconsistent with the scientific limits of the measure (Granqvist et al., 2017; Facompré et al., 2018).
The four-way scheme is a research classification of behavior in a relationship, under specific conditions, and during a particular developmental period. It is not a permanent personality typology. The term attachment styles is often used in popular discussion for both infant classifications and adult self-report relationship patterns, although those instruments measure different constructs. Scientific interpretation requires naming the child's age, caregiver, assessment method, and observed behaviors rather than treating all such labels as interchangeable.
How Scientists Measure Attachment
The Strange Situation Procedure
The Strange Situation Procedure is a standardized sequence of short episodes involving the child, familiar caregiver, unfamiliar adult, exploration, separations, and reunions. It was designed principally for the second year of life, with modifications for older children. Trained coders consider the organization of exploration, proximity seeking, contact maintenance, avoidance, resistance, and reunion behavior; disorganization coding adds attention to contradictory or interrupted patterns. It is a research procedure requiring standardized administration, developmentally appropriate inclusion criteria, and coding expertise. Parents should not recreate distressing separations at home to determine a child's classification (Benoit, 2004; Greenwood et al., 2020).
Laboratory procedures provide standardization but also have ecological limits. Some children are more distressed by unfamiliar laboratories than others. Prior experience with separations, cultural practices, language, developmental differences, and the assessment's precise episode sequence can affect responses. The fact that an instrument predicts meaningful group-level outcomes does not make it an all-purpose clinical test. Reliability depends on training and coding procedures, and results should not be interpreted without the observational context.
The Attachment Q Sort and naturalistic observation
The observer Attachment Q Sort assesses the extent to which a child's everyday secure-base behavior resembles a criterion profile, generally through trained observation across ordinary settings. A large meta-analysis of 139 studies and 13,835 children found that observer-based security scores correlated with Strange Situation security at r = .31 and with caregiver sensitivity at r = .39; the relation with temperament was weaker at r = .16. The investigators distinguished the observer method from self-reported versions and cautioned that the latter lacked comparable evidence of validity (van IJzendoorn et al., 2004).
For preschoolers and older children, investigators also use age-adapted observation, story-completion tasks, interviews, and caregiver-child interaction measures. These measures are not interchangeable, and the evidence attached to one age band or coding system cannot simply be transferred to another. Research scores are most meaningful when assessed in populations for which the instrument has been validated. A child's bedtime request, preference for one caregiver, or manner of greeting a parent is not a standardized attachment assessment.
Research classification, clinical evaluation, and developmental screening
Attachment security is a developmental research construct. Routine developmental monitoring follows progress in communication, movement, learning, social interaction, and adaptive skills; standardized screening uses validated age-appropriate instruments to identify children who may benefit from further evaluation. Clinical diagnosis integrates history, functioning, observation, differential explanations, and established diagnostic criteria. A research attachment classification does not replace any of these processes. Persistent changes in functioning, pronounced relationship difficulties, developmental loss of skills, or concerns about safety should prompt a discussion with a pediatric or qualified child mental-health professional using the relevant clinical pathway. Professional evaluation considers the entire child and caregiving environment.
Why Some Attachment Relationships Are More Secure
Caregiver sensitivity and reliable emotional availability
Caregiver sensitivity means noticing a child's signals, interpreting them in context, and responding in a timely, appropriate, and developmentally attuned way. It may involve comforting a frightened infant, allowing a curious toddler manageable space to explore, following the child's focus during play, providing predictable transitions, and making an effort to repair missed or mistaken responses. Sensitivity depends on the infant's signal and the situation; the same response is not appropriate for every child or every occasion. A sensitive caregiver may sometimes misunderstand a signal and then adjust through renewed attention.
The link between sensitivity and attachment security is one of the field's most replicated findings. An earlier meta-analysis reported an association of about r = .24 in a set of observational studies using comparable methods. A 2024 synthesis examined maternal and paternal sensitivity as predictors of attachment security, and a newer umbrella review combining earlier meta-analyses estimates an overall sensitivity–attachment correlation near r = .25 across 253 primary studies and more than 37,000 participants. These magnitudes indicate a meaningful average relationship alongside substantial unexplained variation; they do not imply that an individual parent's single mistake creates insecurity (de Wolff and van IJzendoorn, 1997; Madigan et al., 2024; caregiver sensitivity umbrella review).
Stress, mental health, support, and ecological conditions
The ability to respond sensitively is shaped by sleep, caregiver health, relationship support, employment demands, material security, discrimination, safety, access to child care, and social policy. Parents coping with economic adversity or mental-health difficulties may have fewer resources to sustain predictable interaction, even when they care deeply for their children. Interventions and public policy can help by providing psychosocial support, material stability, safe services, and relationships that support caregivers. Scientific explanations of insecure attachment must account for these conditions rather than reducing all variation to personal parenting qualities (World Health Organization, 2023).
Child behavior also influences adult responses. An easily overwhelmed infant may elicit different support from a caregiver than a quiet, observant infant; the caregiver's adaptation then changes the child's subsequent behavior. Families face different circumstances as children grow, and caregivers themselves learn new interaction skills. Such bidirectional processes complicate causal inference from naturally occurring associations. A correlation between caregiver behavior and child security can reflect a mixture of caregiver influence, child effects, shared environment, and circumstances that researchers did not measure.
Multiple caregivers and the specificity of attachment
Children can develop attachment relationships with several familiar and reliably available adults. Mothers, fathers, grandparents, foster parents, adoptive parents, and other consistent caregivers may provide protection and serve as secure bases in distinctive ways. The child's attachment classification with one person need not be identical to the classification with another. A 2022 meta-analysis of 95 studies comparing behaviorally measured attachment to mothers and fathers found a correlation of r = .32 between the two and only a very small average difference in security. These findings support studying each relationship and its context rather than inferring a child's global security from one dyad (Pinquart, 2022).
Regular child care, shared parenting, and caregiving by relatives are compatible with the development of attachment relationships. What matters analytically is the experience of particular relationships over time, including continuity, availability, and the quality of responses under stress. Family configurations themselves are not attachment classifications. A child may turn to one person during illness and another during a different kind of challenge; such preferences can reflect familiarity, division of roles, or experience rather than a rank ordering of love.
Adversity and protection
Maltreatment, frightening caregiver behavior, severe disruption, and ongoing stress may threaten a child's expectation of safety and are associated with greater rates of some insecure or disorganized classifications in group research. These associations demand attention to context, measurement, selection, and the child's other protective relationships. Disorganization is more frequent in certain risk samples, yet many disorganized classifications occur without documented maltreatment, and the classification alone cannot be used to confirm harm. The scientific and ethical response to possible maltreatment is a comprehensive, safety-centered evaluation informed by direct evidence, rather than assigning blame on the basis of attachment coding (Granqvist et al., 2017; preschool attachment meta-analysis, 2023).
Stability, Change, and the Limits of Early Prediction
An early attachment classification captures the child's relationship organization during a defined period. It can be informative without being permanent. A 2020 meta-analysis found moderate stability across infancy, toddlerhood, and the preschool or early-school period: the four-way classification agreement was approximately kappa = .23, while secure versus insecure classifications showed an association of r = .28. Security appeared relatively more stable than some other classifications. The review also identified publication-bias concerns. These findings indicate both continuity and substantial opportunity for change as caregiving, family circumstances, and the child's developmental capacities change (Greenwood et al., 2020).
A child who has experienced inconsistent responses can encounter new patterns of availability and care. Conversely, acute stress, caregiver loss, or major disruptions may change how a child organizes contact and exploration. Children construct relationship expectations through repeated experience, and the diversity of developmental pathways prevents deterministic inference from early observation. Attachment is one probabilistic contributor to later functioning; later peer experiences, family relationships, health, schooling, structural conditions, and interventions also matter. The phrase early years are important describes developmental sensitivity, while the evidence for moderate stability describes meaningful plasticity.
Intergenerational patterns and their limits
Researchers have examined associations between adults' representations of attachment and the organization of their children's attachment relationships. A major synthesis found intergenerational associations but also a transmission gap: measured caregiver sensitivity explained only part of the relationship, and many sources of variance remained. Family histories, present relationships, mental-health resources, social conditions, and differences in measurement may shape what is observed. A parent's retrospective account of their own childhood cannot establish that their infant will receive a particular attachment classification. Supportive experiences in adulthood and changes in current caregiving provide plausible routes for change (Verhage et al., 2016).
Attachment and Other Developmental Outcomes
Emotion regulation and stress responses
A responsive attachment figure can help a young child recover from distress through voice, touch where welcomed, proximity, predictability, and interpretation of confusing situations. With development, children learn to anticipate support, request assistance in more differentiated ways, and contribute increasingly to their own regulation. This developmental account connects attachment to co-regulation, but the constructs have different scopes: co-regulation describes interactive support for regulation across relationships and circumstances, whereas attachment concerns a particular protective bond. Children differ in sensory needs and soothing preferences, and evidence about broad regulatory skills should not be treated as a direct measure of security.
Peer competence and social understanding
A meta-analysis of 80 independent samples involving 4,441 children found that early attachment security was associated with better social competence with peers at an average standardized effect of d = .39. The estimate supports an association at the group level, while observational evidence cannot isolate a single causal pathway. Caregiver relationships may contribute to expectations about reciprocity, opportunities for exploration, emotion regulation, and conflict repair, while child temperament, language, neighborhood resources, and peer experiences also influence social adjustment (Groh et al., 2014).
Cognitive development, language, and exploratory learning
Secure-base behavior permits exploration under conditions of perceived safety, which provides opportunities for play, interaction, and learning. Still, attachment security and cognitive development are distinct constructs. A 2023 meta-analytic study incorporating 125 studies from 107 samples and 9,213 children found small associations between mother-child attachment security and child cognition (r = .17) and language (r = .16). A modeled indirect pathway from caregiver sensitivity through attachment to these outcomes was also small. These results require attention to measurement, shared family factors, and the predominance of North American and European samples; they do not justify telling parents that a child's vocabulary score directly reveals attachment quality (attachment, cognition, and language meta-analysis, 2023).
Emotional and behavioral difficulties
Meta-analytic evidence links insecure attachment with elevated average risk of some later emotional and behavioral difficulties. A 2010 synthesis of 69 samples with 5,947 children found a small-to-moderate association between insecurity and externalizing behaviors (d = .31). A separate 2012 synthesis of 42 samples with 4,614 children found a smaller association with internalizing symptoms (d = .15). The magnitude of the associations varied across classifications and study methods. These are population-level associations with overlapping distributions, rather than tests capable of predicting which individual child will develop a disorder (Fearon et al., 2010; Groh et al., 2012).
A developmental pathway can involve several influences at once. Early insecurity may coexist with temperament-related reactivity, family conflict, socioeconomic adversity, chronic stress, and limited access to support. Some studies adjust for such variables; others cannot. The possibility of bidirectional effects and unmeasured confounding must remain visible in interpretation. Neither anxiety nor aggression establishes an attachment classification, and an attachment classification is not itself an anxiety, behavioral, or trauma diagnosis.
What Intervention Studies Establish
Behavioral sensitivity interventions
Randomized interventions offer stronger causal evidence than observational associations because the assignment of support can reduce pre-existing differences between groups. A widely cited 2003 meta-analysis of 70 studies found that interventions aimed at parental sensitivity and attachment produced improvements in caregiving sensitivity and smaller improvements in infant attachment security. Programs with a focused behavioral component and a manageable number of sessions often performed better than interventions relying primarily on extensive theoretical education. Even in randomized trials, results describe average effects within studied populations and delivery conditions (Bakermans-Kranenburg et al., 2003).
Video-feedback approaches record brief caregiver-child interactions and use them to help adults recognize children's signals and respond sensitively. A Cochrane review of 22 studies, with 1,889 randomized dyads or family units, found moderate-certainty evidence that video feedback improved parental sensitivity (standardized mean difference approximately .34). Evidence for attachment security was less certain and differed by assessment method, with very low certainty for those comparisons. This separation of outcomes matters: improving observed parental behavior does not automatically establish a comparably strong or durable change in a child's attachment classification (O'Hara et al., 2019).
Targeted attachment-focused programs
Attachment and Biobehavioral Catch-up is one intervention studied in families exposed to elevated adversity. In a randomized trial of 120 children, the intervention group showed a higher proportion classified as secure and a lower proportion classified as disorganized than an active comparison group. The result provides causal support for benefit under those trial conditions, while generalization requires attention to target population, implementation quality, and independent replication (Bernard et al., 2012).
A meta-analysis of 25 randomized trials of Video-feedback Intervention to promote Positive Parenting and Sensitive Discipline reported average effects on parenting behavior, child attachment security, and parental attitudes, while the pooled association with changes in child externalizing behavior did not demonstrate the same strength. Other reviews of interventions for disorganized attachment describe heterogeneous methods and evidence certainty. These differences argue for evaluating each intervention against its specific target and outcome, rather than describing all attachment programs as equally established (VIPP-SD meta-analysis, 2022; Wright et al., 2017).
Public-health support and choosing care
The World Health Organization recommends supporting responsive care and opportunities for early learning during the first three years of life and integrating support for maternal mental health into early childhood services. Its broader parenting guidelines advocate accessible evidence-based programs to strengthen parent-child relationships and reduce harsh parenting or maltreatment. For a family, appropriate help may include pediatric guidance, caregiver mental-health care, practical assistance, support with routines, and an evidence-based parenting program. Intervention choice should be individualized to the child, caregiver, cultural context, and presenting concern (World Health Organization, 2023).
Intervention evidence is strongest when a specified program, population, comparator, and outcome have been evaluated using randomized methods and sufficient follow-up. It is less informative when a program simply claims to build secure attachment without defining a measure or providing transparent study methods. Clinical recommendations require the assessment of actual difficulties and strengths, rather than selection solely by an online attachment-style label.
Culture, Family Diversity, and Research Generalizability
Attachment relationships appear across a wide range of cultural settings, while their behavioral expression and the meaning of separations, physical closeness, independence, and shared caregiving vary. Cross-cultural work has documented both common secure-base patterns and variation within and between societies. Early meta-analysis found substantial differences among samples within the same country, a warning against treating national culture as a single parenting template. A nine-country observational study found secure-base behavior across all settings it examined, while the organization of particular behaviors differed (van IJzendoorn and Kroonenberg, 1988; cross-cultural secure-base study, 2013).
A central methodological problem is whether a laboratory separation has the same novelty, social meaning, and stress level across caregiving systems. Researchers must consider sample composition, minority and migrant experiences, translation, observational context, and measurement invariance before comparing classification frequencies. Large contemporary meta-analyses remain heavily weighted toward North American and European samples. This limits the confidence with which estimates can be generalized to all children. Cultural variation should be investigated rather than pathologized.
Foster, adoptive, single-parent, multigenerational, same-sex-parent, and community-supported households can organize consistent caregiving in different ways. Attachment theory identifies the quality and availability of particular protective relationships as scientifically relevant; household labels do not determine that quality. Children and caregivers may also communicate needs through culturally patterned forms of gaze, touch, proximity, and verbal exchange. Sensitive response requires understanding those patterns in context.
Practical Implications for Caregivers
Supporting secure-base experiences begins with an ordinary, repeated pattern of being available when a child needs help. Notice the child's signals, distinguish fear from frustration or tiredness when possible, and offer comfort in forms the child can accept. Build predictable routines for departures and reunions, remain engaged during exploration, and adapt the degree of assistance as the child's capabilities expand. These are developmental principles, not a formula requiring constant proximity or uninterrupted attention. Caregivers also need sleep, physical and mental-health support, and practical assistance; a stable network of trusted adults can support both adult and child needs (World Health Organization, 2023).
When the caregiver misunderstands a signal, repair is useful: slow down, re-engage, acknowledge distress when the child can understand, and restore predictable interaction. During separations, familiar rituals and an honest explanation appropriate to age help establish a comprehensible sequence. Avoid disappearing without warning when a child expects a goodbye, and coordinate with familiar child-care adults to provide continuity. An infant's protest is a request for support and information about the situation, while an older child's persistent avoidance of ordinary activities or severe impairment warrants a more specific assessment (American Academy of Pediatrics).
As children become more independent, provide a reliable base for experimentation and self-correction. Let a toddler try a manageable task while staying available to help. During conflict, set safe and developmentally appropriate limits and return to connection after the immediate problem is addressed. Allow different trusted caregivers to develop their own predictable ways of comforting and communicating. The relevant standard is responsive, safe, age-attuned interaction over time; isolated moments of crying, frustration, or disagreement are insufficient evidence about the security of an attachment relationship.
When Professional Evaluation Is Appropriate
Seek pediatric or qualified child mental-health advice when there are persistent and substantial difficulties with comfort-seeking or daily relationships, marked distress across settings, sustained loss of previously acquired skills, significant developmental concerns, or a suspected threat to safety. A clinician will consider health and developmental history, language and sensory abilities, child-care arrangements, relationship continuity, family stress, and the possibility of other explanations. Standardized developmental screening and clinical assessment answer different questions from attachment research instruments. An online quiz, brief home video, or interpretation of crying cannot provide an infant attachment diagnosis.
When a child has experienced neglect, maltreatment, traumatic separation, or multiple placement disruptions, timely attention to safety and stable, responsive care is particularly important. Specialist assessment may consider distinct clinical conditions, developmental disorders, or trauma-related symptoms using their own criteria. Caregivers seeking help can ask what has been measured, whether the program has controlled trial evidence, which outcomes are expected, and how the plan will account for the whole family's circumstances. The purpose of evaluation is to identify actionable support for the child and caregivers.
Common Questions About Early Attachment
Does responding to a crying baby create dependence?
Responding sensitively to distress provides information that support is available. The secure-base model describes how a dependable source of comfort can coexist with, and help organize, exploration. The appropriate response depends on safety, health, developmental stage, and the infant's signals. Scientific evidence does not support interpreting responsive comfort in ordinary infancy as the cause of pathological dependence (Benoit, 2004; World Health Organization, 2023).
Can an infant be securely attached to more than one person?
Yes. Multiple familiar adults can serve as attachment figures. Security is evaluated within particular relationships, so results for one caregiver cannot automatically be generalized to all others. Research on mother-child and father-child attachments finds meaningful but incomplete correspondence across relationships (Pinquart, 2022).
Is a child who cries at daycare insecurely attached?
No such conclusion follows from a daycare departure. Separation distress is common in early childhood and varies with temperament, fatigue, illness, familiarity, and change. Attachment studies examine organized patterns of behavior across standardized episodes, especially reunion, rather than treating one separation reaction as diagnostic (American Academy of Pediatrics).
Can attachment security change after the first year?
Yes. Meta-analytic findings indicate moderate continuity and substantial change across early childhood. New caregiving experiences, reduced stress, supportive relationships, and some evidence-based interventions can affect developmental trajectories. Early classification is informative at the group level but does not fix an individual's future (Greenwood et al., 2020; Bakermans-Kranenburg et al., 2003).
Does disorganized attachment mean that abuse occurred?
No. Disorganization is an observational classification and cannot establish a history of abuse, prove that a particular caregiver caused harm, or substitute for a child protection or clinical evaluation. Safety concerns require independent evidence and appropriate professional investigation (Granqvist et al., 2017).
Is attachment security the same as having an attachment style as an adult?
No. Infant attachment studies typically use observed behavior in caregiver-child relationships, while much adult attachment-style research uses self-report measures about close relationships. The constructs are related by theory and research questions, yet their measures, developmental contexts, and predictive claims differ. The scope of the present article is infant and early-childhood attachment, including security and its developmental consequences.
Scientific Synthesis
Attachment in infancy and early childhood is an organized system of seeking protection and using familiar caregivers as secure bases for exploration. Its emergence reflects accumulating interactions between developing children and their caregiving environments. Secure, avoidant, resistant, and disorganized patterns are research classifications derived from age-appropriate observations; their interpretation requires trained methods and attention to context. Converging meta-analyses support associations between sensitivity and security, moderate rather than rigid stability, and generally small-to-moderate links with later outcomes. Randomized intervention studies show that some changes in caregiving and security are possible, while findings vary by program and measure. Scientific and clinical practice follows from this evidence by supporting safe, responsive caregiving, accessible family resources, and individualized evaluation when difficulties are persistent or safety is in question.
Related Articles
For the larger developmental framework and its evidence base, read Cluster Child Development and Parenting. The distinction between infant attachment and adult attachment explanations is developed in What Causes Attachment Styles? Childhood, Caregiving, Temperament, and Later Relationships. The roles of supportive interaction and emotional recovery are explored in Co-Regulation: How People Help Each Other Manage Emotions. For interpretation of separation behavior, see Separation Anxiety vs Normal Separation Distress: What Is the Difference?.
References
American Academy of Pediatrics. How to Ease Your Child's Separation Anxiety. HealthyChildren.org. Official resource.
Bakermans-Kranenburg, M. J., van IJzendoorn, M. H., and Juffer, F. (2003). Less is more: Meta-analyses of sensitivity and attachment interventions in early childhood. Psychological Bulletin, 129, 195–215. https://doi.org/10.1037/0033-2909.129.2.195.
Benoit, D. (2004). Infant-parent attachment: Definition, types, antecedents, measurement and outcome. Paediatrics and Child Health. PubMed Central.
Bernard, K., Dozier, M., Bick, J., Lewis-Morrarty, E., Lindhiem, O., and Carlson, E. (2012). Enhancing attachment organization among maltreated children: Results of a randomized clinical trial. Child Development, 83, 623–636. https://doi.org/10.1111/j.1467-8624.2011.01712.x.
De Wolff, M. S., and van IJzendoorn, M. H. (1997). Sensitivity and attachment: A meta-analysis on parental antecedents of infant attachment. Child Development, 68, 571–591. https://doi.org/10.1111/j.1467-8624.1997.tb04218.x.
Deneault, A.-A., Bureau, J.-F., Duschinsky, R., Fearon, R. M. P., and Madigan, S. (2023). A meta-analysis of the distribution of preschool and early childhood attachment as assessed in the Strange Situation Procedure and its modified versions. Attachment and Human Development, 25, 322–351. https://doi.org/10.1080/14616734.2023.2187852.
Deneault, A.-A., Duschinsky, R., van IJzendoorn, M. H., Roisman, G. I., Ly, A., Fearon, R. M. P., and Madigan, S. (2023). Does child-mother attachment predict and mediate language and cognitive outcomes? A series of meta-analyses. Developmental Review, 70, 101093. https://doi.org/10.1016/j.dr.2023.101093.
Facompré, C. R., Bernard, K., and Waters, T. E. A. (2018). Effectiveness of interventions in preventing disorganized attachment: A meta-analysis. Development and Psychopathology, 30, 1–11. https://doi.org/10.1017/S0954579417000426.
Fearon, R. P., Bakermans-Kranenburg, M. J., van IJzendoorn, M. H., Lapsley, A. M., and Roisman, G. I. (2010). The significance of insecure attachment and disorganization in the development of children's externalizing behavior: A meta-analytic study. Child Development, 81, 435–456. https://doi.org/10.1111/j.1467-8624.2009.01405.x.
Granqvist, P., et al. (2017). Disorganized attachment in infancy: A review of the phenomenon and its implications for clinicians and policy-makers. Attachment and Human Development. PubMed Central.
Greenwood, C. J., et al. (2020). Early childhood attachment stability and change: A meta-analysis. Attachment and Human Development. https://doi.org/10.1080/14616734.2020.1800769.
Groh, A. M., Fearon, R. P., Bakermans-Kranenburg, M. J., van IJzendoorn, M. H., Steele, R. D., and Roisman, G. I. (2014). The significance of attachment security for children's social competence with peers: A meta-analytic study. Attachment and Human Development, 16, 103–136. https://doi.org/10.1080/14616734.2014.883636.
Groh, A. M., Roisman, G. I., van IJzendoorn, M. H., Bakermans-Kranenburg, M. J., and Fearon, R. P. (2012). The significance of insecure and disorganized attachment for children's internalizing symptoms: A meta-analytic study. Child Development, 83, 591–610. https://doi.org/10.1111/j.1467-8624.2011.01711.x.
Madigan, S., Deneault, A.-A., Duschinsky, R., et al. (2024). Maternal and paternal sensitivity: Key determinants of child attachment security examined through meta-analysis. Psychological Bulletin, 150, 839–872. https://doi.org/10.1037/bul0000433.
National Research Council and Institute of Medicine. (2000). Nurturing relationships. In From Neurons to Neighborhoods: The Science of Early Childhood Development. National Academies Press. NCBI Bookshelf.
Nivison, M. D., Fearon, P., Jenkins, J. M., and Madigan, S. (2026). Annual Research Review: The role of caregiver sensitivity in children's developmental outcomes—an umbrella review. Journal of Child Psychology and Psychiatry, 67, 486–507. https://doi.org/10.1111/jcpp.70087.
O'Hara, L., Smith, E. R., Barlow, J., Livingstone, N., Herath, N. I. N. S., Wei, Y., Spreckelsen, T. F., and Macdonald, G. (2019). Video feedback for parental sensitivity and attachment security in children under five years. Cochrane Database of Systematic Reviews, CD012348. https://doi.org/10.1002/14651858.CD012348.pub2.
Pinquart, M. (2022). Attachment security to mothers and fathers: A meta-analysis on mean-level differences and correlations of behavioural measures. Infant and Child Development, 31, e2364. https://doi.org/10.1002/icd.2364.
van IJzendoorn, M. H., and Kroonenberg, P. M. (1988). Cross-cultural patterns of attachment: A meta-analysis of the Strange Situation. Child Development, 59, 147–156. ERIC.
van IJzendoorn, M. H., Vereijken, C. M. J. L., Bakermans-Kranenburg, M. J., and Riksen-Walraven, J. M. (2004). Assessing attachment security with the Attachment Q Sort: Meta-analytic evidence for the validity of the observer AQS. Child Development, 75, 1188–1213. https://doi.org/10.1111/j.1467-8624.2004.00733.x.
Verhage, M. L., Schuengel, C., Madigan, S., et al. (2016). Narrowing the transmission gap: A synthesis of three decades of research on intergenerational transmission of attachment. Psychological Bulletin, 142, 337–366. https://doi.org/10.1037/bul0000038.
World Health Organization. (2023). WHO guidelines on parenting interventions to prevent maltreatment and enhance parent-child relationships with children aged 0–17 years. Official guideline.
Wright, B., et al. (2017). Decreasing rates of disorganised attachment in infants and young children, who are at risk of developing, or who already have disorganised attachment. PLOS ONE, 12, e0180858. https://doi.org/10.1371/journal.pone.0180858.