Changes in Attachment Orientation Over a 59-year Period: Determinants of Change and Implications for Health.

Abstract

More than forty years have passed since John Bowlby (1969/1982) proposed that the attachment system is influential “from the cradle to the grave” (p. 208). During those years, researchers have documented many important links between individual differences in attachment and behavior, emotion, and cognition in close relationships. Yet the majority of this work has focused on early childhood and young adulthood. Relatively little attention has been paid to attachment processes from a lifespan perspective—one that includes middle and older adulthood. Methodological limitations also prevent researchers from measuring how attachment changes over long intervals of time. In this dissertation, I developed a measure of attachment orientation using existing measures of personality and then used this method to examine changes in attachment orientation from age 3 to 62 using data from the Block and Block Longitudinal Study, the Intergenerational Studies, and the Radcliffe College Class of 1964 Sample. I also tested whether relationship status and satisfaction moderated changes in attachment orientation among adults. Finally, I examined how individual differences in change were related to subjective health across the lifespan. My findings demonstrate that attachment anxiety increased during childhood and adolescence before decreasing in adulthood. Attachment avoidance increased slightly until middle age before declining in older adulthood. Being in a relationship and having higher marital satisfaction predicted lower levels of anxiety and avoidance across adulthood, particularly in old age. Finally, anxiety was consistently associated with poorer health across adulthood. Taken together, these findings provide much-needed insight into how attachment orientations change over long stretches of time, as well as information about what predicts these changes and the implications of these changes. My dissertation also serves as an illustrative example of how observer-reports of personality can be used to create measures of a construct that was not previously included in data collection. Considerations for measuring longitudinal changes in attachment with observer-based measures of personality are discussed, along with future directions and implications for studying changes in relationships over time.