The neurobiological basis of the attachment system
No single area was considered to be dedicated to attachment in the brain because it appeared to involve most regions. Indeed, so many areas were involved in human attachment that it was claimed that the brain was both built for and by affiliation (Allen,2013). However, at the simple level of description, a top-down bottom-up process may usefully be focused on to understand the attachment process.
At the base of the brain, the limbic system has often been referred to as the seat of emotions. Humans share this system with other mammals and it is strongly linked to the body. In a model defined by Damasio (2010), the limbic system gave humans current, ahistorical biofeedback on the self as an organism reacting to the objects (people and things) encountered. That bottom-up data generated a ‘proto-self’. Humans felt its effect powerfully as impulses to act in relation to the immediate environment, driven by hormonal changes in the body. This included both avoid and approach impulses, with clear relevance to the attachment process. Above the limbic system, the cortex system appeared exceptionally developed in humans. The quality of cortical functioning depended on prolonged interaction with more mature brains during childhood. The cortex system allowed immediate experience from the limbic system to be located within a time-based, self-referencing context. This generated reflection, mindfulness and abstract thought, resulting in what Damasio described as an ‘autobiographical self’. Much of these higher-level mental functions appeared to be mediated interpersonally via attachment throughout the lifespan, suggesting that thinking was a process that existed between people rather than simply in people. Bottom-up and top-down integration generated a ‘core self’ mediating between autobiographical and proto-self: ‘The constituents of this core self process are a set of inter-twined images: an image of the momentarily salient causative object (or person): an image of the organism as modified by the interaction with the object; and an image of object-related emotional responses’ (Allen, 2013, p. 244). The integration of top-down (cortex) with bottom-up (limbic) data has been described as the aim of mentalising: feeling, while thinking about feeling (Allen, 2013).
The attachment system mattered to psychological therapy because it was heavily implicated in the regulation of distress: ‘The brain's first and most powerful approach to affect regulation is via social proximity and interaction’ (Coan, 2008, p. 85). While humans could self-regulate, it was much more effective with another person. The neurobiological evidence strongly indicated that humans had been hardwired to affiliate and suffered greatly when this process, mediated by the attachment system, failed. The interpersonal activation of the attachment system and its role in affect regulation was visible in the ‘still face experiment’ (Tronick, Adamson, & Brazelton, 1975). The caregiver demonstrated their comprehension of the infant's emotional experience via facial/gestural mimicry (Gergely & Unoka, 2008). When the child detected contingent mirroring responses from the caregiver, they could down-regulate distress much more effectively. This implied that the attachment system facilitated neurobiological functions of individuals to work together, one cortex acting as a surrogate for the other to help them label their experience (Coan, 2008). This has been described as plugging in the attachment system to give your brain a rest (Allen, 2013). The activation of the attachment system facilitated affect regulation because it involved powerful but temporary chemical changes in the body, described as the ‘HOME’ response: oxytocin signalled both body and brain to temporarily lower its guard, leading to serotonin (a soothing opiate) plus dopamine, a reward hormone encouraging the organism to repeat the experience (Zak, 2012). Oxytocin, then, was a key marker of attachment and of great interest to those aiming to understand its role in therapy.
Oxytocin is a peptide and is both a neurotransmitter (affecting the brain) and a hormone (affecting the body). It is extremely old and its evolution started when organisms moved from reproduction via cell division to reproduction via sex. Sexual reproduction had the advantage of mixing genes from each parent, allowing the strongest DNA to be subjected to selection. However, sexual reproduction was also a risky strategy as it involved an appraisal of threat and a temporary lowering of guard. When most organisms were solely preoccupied with eating and killing other organisms or avoiding being eaten and killed, a mechanism was needed to signal trust. This was originally vasotocin, then isotocin and now oxytocin (Zak, 2012).
Over millions of years, oxytocin has been increasingly utilised in a wide range of relationships between animals. Most relevant to humans was the huge leap from reptilian reproduction, where the young hatched and were abandoned, to mammalian reproduction via birth and infant care. MacLean (1985) described mammalian care as having three distinctive components: nursing maternal care; a separation call (to signal distress at separation); and play (for social learning). All of these processes required a temporary suspension of defensiveness to allow trusting care, which was mediated by oxytocin. Numerous authors postulated that oxytocin was currently being utilised to drive novel forms of bonding beyond family groups, such as city building, trade and moral development that characterise the extraordinary human expansion of the last 100,000 years (Ridley,2010; Zak, 2012). These social phenomena demand powerful top-down processing structures, particularly involving the highly developed prefrontal medial cortex which appeared to be central in the ability to understand and differentiate (mentalise) the complex range of social intentions involved.
The dominance of the human species is a testament to the success of this evolutionary strategy. However, while evolution had determined that extending the attachment system was advantageous, that extension also conferred risk. Abandonment or an inability to attach to others had grave implications for the functioning and survival of mammals. Bowlby (1953) suggested that when mammals opted for family life they set the stage for the most distressing form of suffering: attachment trauma. Attachment trauma may be a better description for borderline personality disorder.
Given that so much of what defined the human brain had been driven by advances in the development of the attachment system, the breaking of trust in childhood attachment had marked consequences for the development of neurological functioning in the individual (Allen, 2013). It might be supposed that oxytocin could offer a remedy to those suffering attachment traumas. However, oxytocin appeared to be of little help to those who had not been supported by surrogate brains in childhood. This was demonstrated in studies where oxytocin was been administered to non-clinical and clinical populations. In the non-clinical population oxytocin increased co-operation but sadly it decreased co-operation for those diagnosed with borderline personality disorder (Bartz et al., 2011). Oxytocin had profound effects on the limbic system and could hyper-stimulate the attachment system. Feelings generated by oxytocin could be so powerful that they were perceived as a threat. Threat activated the amygdala, initiating the fight, flight or freeze responses and thereby overriding cortical functions, in effect ‘frazzling’ the brain (Arnsten, 1998). When discussing the difficulties in close relationships in an MBT psycho-education session, one participant elegantly summarised: ‘We with BPD have a lot of love but not a lot of empathy when we feel that love’. Therefore, the MBT approach devoted particular effort to deliberately applying explicit mentalising to the interpersonal exchanges between those involved in therapy.
The MBT model aimed to use the attachment system as a mechanism of change in therapy with adults with borderline personality disorder (Allen, Fonagy, & Bateman, 2008). Attachment was construed as complex, requiring a great deal from the cortex. It was highly dependent on one party being able to disclose emotionally (limbic system) to another, who was receptive and trusted to emotionally respond (limbic and cortex system). This demonstrated empathy and was most powerfully demonstrated in separation calls involved in the still face experiment. Mirroring was therefore heavily implicated in down-regulating distress through attachment and oxytocin was a useful marker of the activation of the attachment system. However, the way the brain processes oxytocin is not simple, requiring high levels of mentalising between individuals.