The present study explored processing biases resulting from manipulating the temporal accessibility of relational schemas. By priming relational schemas, relationship-specific attachment styles were activated and their biasing effect on relevant information processing (namely recall for attachment-relevant words versus other words, interpersonal expectations, and affect) was examined. It was found that participants primed with a secure-style relational schema recalled more positive attachment words than those primed with an avoidant style. Although pre-priming endorsements of interpersonal expectations were influenced by global attachment style, once primed, participants showed primed-style-congruent responses. That is, primed secures showed higher endorsement of positive and lower endorsement of negative interpersonal expectations relative to the other primed style groups. Finally, primed secures reported more positive and less negative affect than the other primed style groups. Implications for understanding the way differential attachment experiences influence close relationships through life are considered.

Research shows that priming attachment security results in positive relationship expectations and affect (Rowe & Carnelley, 2003). We examined whether repetitive priming of attachment security (e.g., experimentally activating cognitive representations of attachment security) would have more lasting effects on relationship‐ and self‐views. Participants provided baseline measures at Time 1. On 3 occasions (across 3 days), we primed participants with attachment security or a neutral prime (Times 2–4). Two days later (Time 5), participants completed trait‐level measures not preceded by a prime. As expected, those repeatedly primed with attachment security reported more positive relationship expectations, more positive self‐views, and less attachment anxiety at Time 5 than at Time 1; those primed with neutral primes showed no change with time. These priming effects last longer than those typically found.

Background Although existing research has suggested an association between insecure attachment styles and mental health difficulties, these studies often have small sample sizes, use cross-sectional designs, and measure attachment as a discrete variable at a single point. It is also unclear whether these associations persist into late adolescence. In this prospective study we aimed to determine whether insecure attachment style in early childhood is associated with depression and self-harm at 18 years. Methods We used data from the Avon Longitudinal Study of Parents and Children cohort. Mothers completed attachment related questionnaires when their child was 18, 30, and 42 months old. Offspring depression and lifetime self-harm was assessed at 18 years using the Clinical Interview Schedule-Revised. Attachment was derived as a latent variable in a structural equation modelling framework. Logistic regression was performed on participants with complete attachment data (n=7032) to examine the association between attachment style and depression and self-harm, with adjustment for potential confounders. Differential dropout was accounted for using multiple imputation. Results We found some evidence for an association between attachment quality and depression and self-harm. In the fully adjusted imputed model, a one standard deviation increase in insecure attachment was associated with a 13% increase in the odds of depression at age 18 (OR=1.13; 95%CI=1.00 to 1.27) and a 14% increase in the odds of self-harm at age 18 (OR=1.14; 95%CI=1.02 to 1.25). Conclusions Our findings strengthen the evidence suggesting that an insecure attachment style assessed in early childhood is associated with mental health difficulties in late adolescence. If this association is causal, policies to support parenting or attachment-based interventions to improve attachment quality could help reduce mental health difficulties in adolescence/young adulthood. Keywords ALSPAC, attachment, depression, self-harm, longitudinal, structural equation modelling Abbreviations ALSPAC: Avon Longitudinal Study of Parents and Children; SEM: Structural Equation Modelling