Exploring dynamics and mechanisms of attachment variability in middle childhood

While scholars have traditionally approached attachment as a rather stable, trait-like construct, there is an increasing awareness that attachment security can change over longer-term periods, and that state attachment at the level of expectations of trust in caregiver support can vary over shorter-term periods. The current dissertation followed up on these insights by means of two research lines. First, we aimed to further elucidate dynamics of state attachment variability in middle childhood, and their associations with trait attachment and psychological well-being. Second, we aimed to manipulate two factors that may affect attachment, namely attachment-related interpretation biases and oxytocin.

In the first line of research we explored state attachment variability across distressing situations and across days. Results indicated that children varied in their state attachment across distressing situations, and that there were two components underlying variability. One component reflected expectations of seeking and receiving support, and the other component reflected expectations of getting back-on-track following distress. State attachment also varied across days. There was partial temporal stability from day-to-day in state attachment. Reactivity effects reflected that state attachment was immediately sensitive to daily life experiences of caregiver support following distress. Concerning associations with trait attachment, children who were more securely attached at a trait-level seemed to vary less substantially in their state attachment across distressing situations. They also showed higher temporal stability in their state attachment from day-to-day. Additionally, these children’s state attachment was less susceptible to the positive effects of effective support experiences, but more susceptible to negative effects of experiencing no support during distress. Concerning associations with psychological well-being, a higher degree of variability in state attachment seemed related to less psychological problems; although there was some evidence that higher degree of state attachment variability in expectations of seeking- and receiving maternal support was less maladaptive when children’s trait attachment was accounted for.

In the second line of research we combined an attachment-related cognitive bias modification (CBM) training with intranasal oxytocin administration. We first assessed whether single intranasal oxytocin administration was related to side-effects. Results indicated that oxytocin was well-tolerated and did not result in more side-effects compared to a placebo nasal spray. We then explored the individual and joint effects of CBM and oxytocin on attachment-related interpretation biases and trust in maternal support. Results suggested that CBM training was only effective at an automatic level of processing. Oxytocin did not independently exert effects or enhance CBM effects.

All in all, the current dissertation adds to our understanding of attachment variability in middle childhood by illuminating how state attachment varies across similar situations and across time. Moreover, the results indicate that different internal and external factors may underlie attachment variability. Taken together, the findings suggest that attachment extends beyond a trait level and dynamically evolves over time and contexts. Therefore, an integrative model that incorporates state attachment (variability) may be needed to provide a more comprehensive picture of attachment in middle childhood.