Attachment in the perinatal period
‘Children are not slates from which the past can be rubbed by a duster or sponge, but human beings who carry their previous experience with them and whose behaviour in the present is profoundly affected by what has gone on before.’Reference Bowlby3
Longitudinal follow-up studies convincingly show secure attachment is associated with positive emotional, social, cognitive, behavioural and developmental outcomes, whereas insecure attachment is associated with an increased risk of negative long-term mental, physical and social health outcomes.
A key part of attachment's relevance to adult mental health is mediated by the internal working model, which develops through the infant's cognitive and emotional experience of care. This becomes encoded as implicit-procedural memory and forms the basis of how we expect others to respond to us.Reference Bowlby1 Internal working models persist into adulthood and are particularly active during stress.Reference Adshead2 Mary Ainsworth elaborated on the importance of these models, conceptualising the secure base – if an infant feels safe to undertake tasks in a nurtured environment, this lays the foundation for exploration, self-assurance and self-efficacy.Reference Ainsworth, Blehar, Waters and Wall4
Contemporary attachment researchers emphasise the relevance of caregiver attunement to infant cues and reflective functioning. An attuned relationship facilitates the development of self-identity and worth, but critical within this is the development of mentalising abilities – our capacity to think about our own thoughts and the thoughts of others. Difficulties in mentalising have been linked to adult mental health conditions, including mood, psychotic, eating and personality disorders.
Intergenerational attachment
‘The attachment bond is unconsciously represented in cognitive structures, including words and images.’Reference Adshead2
Our childhood experience of attachment becomes an ‘attachment representation’ in our adult mind. A substantial body of work utilising Mary Main's Adult Attachment Interview suggests a parent's attachment representation predicts their own infant's attachment style, with transmission mediated via the effect of the attachment representation on caregiving behaviours, primarily parental sensitivity to infant cues.Reference van Ijzendoorn and Bakermans-Kranenburg5
This concept of intergenerational transmission could trigger feelings of guilt in parents. Adshead describes the perinatal period as ‘uniquely stressful’ in terms of social relating.Reference Adshead2 It is therefore crucial to discuss attachment with nuance. Not every child of a parent with an insecure attachment representation will display insecure attachment themselves. Furthermore, parental sensitivity does not entirely explain transmission – the so-called ‘transmission gap’.Reference van Ijzendoorn and Bakermans-Kranenburg5 Environmental factors and infant temperament are important. Viewing attachment theory within a bio-psychosocial model, we are left with an interactionist process – each shaping the other. Insecure attachment is therefore best seen as a risk mediator, in the context of other factors.Reference Adshead2
Attachment-informed interventions
‘Change continues throughout the life cycle so that changes for better or worse are always possible. It is this continuing potential for change that means that at no time of life is a person invulnerable to every possible adversity and also at no time of life is a person impermeable to favourable influence.’Reference Bowlby6
As with all risks in medicine, a critical question is whether the risk is modifiable. Although persistent, attachment style is thought to be modifiable. There is a developing evidence base around the use of attachment-informed psychological therapies for adult mental health conditions. We and others posit that, given the key relevance of attachment to adult mental health, targeted attachment-informed interventions deployed perinatally to optimise attachment in vulnerable families may have potential benefit not just for individual families, but for general public health.Reference Raouna, Malcolm, Ibrahim and MacBeth7 Definitions of the ‘perinatal period’ vary, but the term usually refers to pregnancy and the first postnatal year. This one-year cut-off is arbitrary – policymakers and researchers increasingly recognise the importance of the second postnatal year to parent–infant mental health, and Bowlby himself felt the first five years were critical to attachment formation.
Increased understanding of the complexities of attachment transmission has led to some uncertainty around the most appropriate routes and targets for interventions, and the needs of families can differ.Reference MacBeth, Christie, Golds, Morales, Raouna and Sawrikar8 Interventions can also be deemed expensive. However, we argue that these factors must be weighed against the potential long-term benefits.
There is a developing evidence base around attachment-informed parenting interventions for the perinatal period. Interventions take different forms. Some take a more behavioural approach, whereas others focus more on parental reflective functioning and mentalising.Reference MacBeth, Christie, Golds, Morales, Raouna and Sawrikar8,Reference Kohlhoff, Lieneman, Cibralic, Traynor and McNeil9 Some examples are described in Table 1.
Table 1 Examples of attachment-informed interventions
Going forward
‘We are moulded and remoulded by those who have loved us; and though the love may pass, we are nevertheless their work, for good or ill’ François Mauriac, quoted in.Reference Bowlby1
Attachment-informed interventions are not yet widely available. However, we argue that such research should be prioritised if we are to invest in primary prevention of adult mental health conditions and optimisation of perinatal well-being. From a public health perspective, the potential for long-term cost savings is surely considerable.Reference Raouna, Malcolm, Ibrahim and MacBeth7
Qualitative work with parents suggests an appetite for expansion of clinical provision in this area. Considering the sensitivity of the topic, co-production of interventions with families is essential. We also highlight the importance of considering the mental health of all members of the immediate family system. Bowlby's research is now over 60 years old, and societal structure and gender roles have changed considerably. However, even at the time of writing, Bowlby recognised that the mother may not always be the infant's ‘primary caregiver’. Equipping other caregivers in the family system with tools to support the primary caregiver will only add further value, benefitting the needs of both primary caregiver and child. We support an inclusive approach to attachment-informed interventions, with both recognition of the needs of mothers and increased understanding of the needs for all caregiver roles.
Finally, it is important to recognise that parent–infant relationships exist within a social context, and many face adverse circumstances such as substance misuse, domestic violence and socioeconomic deprivation.Reference Raouna, Malcolm, Ibrahim and MacBeth7 Tackling parental mental health without addressing concurrent social adversities does not necessarily lead to improvements in the parent–infant relationship, suggesting interventions that are not adapted for the needs of hard-to-reach populations are unlikely to be successful.Reference Raouna, Malcolm, Ibrahim and MacBeth7 Many trials focus on ‘at-risk’ populations, but we are not aware of any currently specifically targeting mothers with severe enduring mental health conditions (for example bipolar disorder), a significant evidence gap.
Conclusions
Despite being critical to our development and adult mental health, the current mainstay of attachment-informed interventions lies in the treatment of adult mental health problems. We argue that prevention is better than cure. Increased funding of research into attachment-informed interventions in the perinatal period may surely have public health benefits as well as potential to improve the well-being of infants and families. Given how attachment dynamics provide a foundation for caregiving and help-seeking in later life, as well as being evidence based, developing this aspect of mental health prevention simply makes good sense. Though there is some existing evidence, large, robust multicentre trials are lacking. Furthermore, there is a lack of longitudinal studies in this area. There are also research gaps around subgroups such as adoptive/foster dyads, single carer families and children of parents with severe enduring mental health conditions. Inclusive co-production with families is essential. It is time for us to move from theory to practice.
Data availability
The data that support this study are described within the publications cited within the text.
Author contributions
K.A. proposed the article. K.A. and C.M. contributed to analysis of the literature, manuscript preparation and drafting. K.A. submitted the manuscript.
Funding
C.M. receives funding from the EU Horizon 2020 Marie Curie Sklodowska programme grant number 95667.
Declaration of interest
None.
eLetters
No eLetters have been published for this article.