Attachment behaviors embody what an infant does to promote proximity to a caregiver and ensure safety, security, and survival. Warm, sensitive, and responsive parenting generally contributes to secure attachment. In order for an infant to be attached, an attachment figure or object must, of course, be available. The attachment figure needs to have an ongoing emotional investment in the child.[1,2] In some situations caregivers other than parental figures, often out of necessity, act as temporary attachment figures; not all of these will become attachment objects. Certain situations may preclude the development of relational stability and secure attachment. Children placed in emergency foster homes, children placed in orphanages with multiple caregivers, and children displaced by environmental and other disasters are at risk. When attachment falls outside of a normative range and into extremes, mental health clinicians commonly use the term “attachment disorder.” Unfortunately, this term is sometimes used in an imprecise manner to refer to general relationship disturbances in early infancy. Empirical data through longitudinal and cross-sectional studies is emerging but is fraught with methodological constraints. Here we will address the current clinical understanding of what an attachment disorder represents and what to do when you encounter a clearly disturbed attachment relationship.

What is not an attachment disorder?

In identifying attachment disorders we look at what is happening to individual partners and what is happening within their relationship. An attachment disorder is not defined solely in terms of an infant’s behavioral disturbances. Day-to-day parenting errors are commonplace and do not contribute to an attachment or relationship disorder. A typical parent-infant relationship is not always optimal; repairing mismatches or interactive errors is an essential part of the ongoing negotiation within the dyad.[3]

Secure attachment facilitates infant mental health and can mitigate the effects of other social-emotional risk factors, thus operating as a protective factor. However, not all atypical attachment patterns (insecure, disorganized) described previously in this theme issue give rise to an attachment disorder or future psychopathology. Nor do social deprivations necessarily give rise directly to an attachment disorder. However, research findings indicate that those in the social high-risk sample who also have insecure attachment often develop psychopathology when followed to adolescence.[4] Even disorganized attachment is best viewed as a vulnerability factor for adverse social and emotional outcomes.[5] 

Parents who plan to adopt may approach physicians with questions regarding attachment disorders. They should be told that attachment security has been related to the timing of the placement and adoptive caregiving quality. Research suggests that infants adopted early in the first year of life are just as likely as non-adopted infants to develop healthy attachment relationships[6] and are not always destined to experience disordered attachment. 

Attachment cannot be quantified and every dyad has unique qualities. Measurement of attachment security is complex and includes variables that are sometimes difficult to decipher on clinical observation alone.[7,8] Traditionally, clinicians have depended upon observable attachment behaviors in relation to a particular attachment figure, the caregiver’s accessibility and responsiveness to the child, and the constancy of this relationship over time, especially in situations where the infant is distressed.

Diagnostic classification

For the formal diagnosis of an attachment disorder (reactive attachment disorder) in 1980, earlier versions of the International Statistical Classification of Diseases[9] and the Diagnostic and Statistical Manual of Mental Disorders[10] (ICD-9 and DSM-III) required an age of onset before 8 months, failure to thrive, and a pervasive lack of social responsivity. Despite revisions, newer versions of both manuals continue to be less than adequate in that they do not describe the full range of observed difficulties. Also, children with attachment difficulties may not show problems across all relationships. A child may experience an attachment difficulty with the primary caregiver but not with other adults or peers.[11] Finally, attachment disorders are relational disorders and do not conform to nosological systems that characterize the disorder as residing in the individual. The DSM-IV-TR describes reactive attachment disorder that can manifest as either inhibited or disinhibited behavior.

The DC: 0-3R[12] is somewhat more useful when classifying attachment disorders but also limited. The advantage of this system over that of the DSM-IV-TR is the introduction of a relationship measure on axis 2, with no relationship disorder at one end and an abusive relationship disorder at the other end. In between these two poles, we find over-involved, under-involved, anxious, angry/hostile, and mixed relationship disorders. There are problems, however, in that parental behaviors are emphasized to the exclusion of the child’s attachment behaviors.[13]

Clinical presentations that can resemble attachment distrubances include an infant’s withdrawal due to depression or autism. Although the significant social impairments associated with autism may resemble an attachment disorder, autistic children are capable of forming attachment relationships with their caregivers.[14] 

Zeanah and Boris[15] suggest revised criteria for attachment disorders. According to their classification system (Table 1), there are three main types of disorders: disorder of nonattachment, secure base distortion, and disrupted attachment.[16] 

Assessment

In spite of having an improved understanding of attachment disorders, we still have difficulties with classification and validation of attachment disorder, and assessment can be challenging. In reality, it is often difficult to sort out whether early attachment disturbances constitute a risk for disorder, or are in themselves a clinical problem requiring treatment. 

Assessment for an attachment disorder is best done by a clinician trained in mental health problems in early childhood. An assessment in the clinical setting includes the following:

• Focusing on the child’s relationship with attachment figures, especially the caregiver’s ability to provide a secure base and a safe haven for the infant.

• Directly observing relevant infant behaviors through structured observational paradigms that capture attachment to a caregiver, especially reunion experiences and various attachment-seeking behaviors, including controlling behaviors.

• Taking the infant’s developmental level into consideration—for instance, the later development of attachment relationships in a child with Down syndrome (between 12 and 24 months).[14] 

Examples of cases where these principles have been followed and children have been diagnosed with an attachment disorder can be found in the box, “Cases of attachment disorder seen at the Infant Psychiatry Clinic at BC Children’s Hospital.”

Preventive intervention 

Family physicians and pediatricians are well positioned to begin the process of alleviating the suffering arising from relational disturbance. Some interventions do not need a child therapist or a child psychiatrist. Watchful monitoring of the parent-child relationship can be very useful and can help prevent relational disturbance in some cases. 

One effective form of preventive intervention involves home visiting by a professional (e.g., community health nurse) or paraprofessional. Home visiting prevents abusive parenting in vulnerable families by providing a therapeutic holding environment.[17] The physician may also need to advocate regarding psychosocial issues, such as financial assistance and housing. In general, policies that promote economic security, enhance the parent-child relationship (e.g., through maternity and paternity leave), strengthen parenting skills, and reduce the isolation of the nuclear family can indirectly and positively affect the parent-child relationship.[18] 

Another preventive strategy involves using psychoeducational models to encourage parents to respond promptly to their child’s distress. Several well-known approaches are available. For example, the Keys to Caregiving program[19] provides educational material (videotapes, pamphlets) for parents to learn how to read infants’ cues appropriately. The materials can be used on their own or with the support of a professional. 

Where a mother presents with signs of depression, timely referral to an appropriate community resource is important. Maternal depression and anxiety are known risk factors for the offspring’s development,[14] particularly in the social and emotional domains. Physicians are encouraged to pay attention to the dyadic relationship in this situation and refer the dyad, and not only the mother, when necessary. In the Lower Mainland, physicians can refer patients to the Infant Psychiatry Clinic at BC Children’s Hospital (604 875-2719) and the Alan Cashmore Centre (604 454-1676). Outside the Lower Mainland, community health nurses and local child and youth mental health centres are a good source of information about community resources. 

Specific interventions 

Interventions that rely on a developmental rather than psychiatric classification approach have proven useful. In this approach, disturbed attachment relationships are seen as initiating processes that compromise development, whereas secure attachment increases resistance to stress and promotes resilience.[15] 

In attachment disorders the patient is neither the parent nor the infant alone, but both together. Stern describes several “ports of entry” where intervention may occur.[20] These may be through the infant[21] or mother.[22,23] Intervention models may be based on dyadic interactions[24-27] , triadic interactions,[28] or group work. 

More detailed information about different intervention models is provided in Table 2 [10] .

Child maltreatment and neglect

In situations involving child endangerment, attachment therapy takes a different form. Physicians are mandated to report their concerns to the appropriate authorities, even though they may sometimes feel that they cannot continue to be helpful to the child and family if an adversarial relationship ensues. These situations usually occur within multiproblem families and need multimodal interventions. A collaborative approach involving legal and social welfare systems is necessary. The intervention approaches often include a combination of dyadic therapies such as therapeutic visitation, interaction guidance, modified interaction guidance, and parent-infant psychotherapy.[29] Taking steps to stop therapist burnout in these challenging cases is crucial.

Effectiveness of therapeutic modalities 

More unconventional attachment interventions (e.g., “holding therapy”), have not been shown to be effective.[6] Evidence suggests that programs that directly address the desired behavioral changes (e.g., improving maternal sensitivity and infant security) tend to be more effective than those that address more general issues.[30] The external component, that is, interactions that can be observed, appears to be a necessary component of effective intervention. 

When Bakermans-Kranenburg and colleagues completed a metaanalysis of 70 studies of sensitivity and attachment interventions in early childhood in families with and without multiple problems, they found that the most effective interventions used a moderate number of sessions and a clear-cut behavioral focus. Increased parental sensitivity was associated with increased attachment security, supporting the causal role of sensitivity in the development of attachment.[31] 

Conclusions

Attachment difficulties hamper the developmental course of young children and are stressful for caregivers. In order to help troubled families, clinicians need to understand that they have an important role to play. Nearly all infant and dyadic interventions emphasize the therapeutic alliance with the parent and indirectly with the infant. Interventions do not have to be lengthy but they do have to be timely. Moreover, a combination of therapies is recommended with multiproblem families, provided team building occurs so that multiple therapies are synchronized. Currently, there is a major thrust to evaluate all the modalities and we hope to gain insight soon into what works best for whom. 

The psychoanalyst Michael Balint referred to the doctor herself or himself as a powerful “medication.”[32] The power of relationships is undeniable, and makes the patient-doctor relationship an important ingredient in promoting healthy parent-child interaction.

Table 1. Types of attachment disorders.

Pratibha Reebye, MBBS, FRCPC, Terry Kope, MD, FRCPC

Dr Reebye is a clinical professor at the University of British Columbia and clinical head of the Infant Psychiatry Clinic at BC Children’s Hospital. Dr Kope is a clinical assistant professor at the University of British Columbia and a psychiatric consultant to the Alan Cashmore Centre, Vancouver Community Mental Health Services.