Attachment in romantic couples

Today, we’re looking at three studies of the beginning, middle and end of romantic relationships. First, Traut, Gander, Uhlich, Weidmann, Chopik & Grob (2025) published “Examining Change in Attachment in Romantic Couples: The role of relationship characteristics and codevelopment between partners” in Journal of Personality and Social Psychology. Here’s the edited abstract:

Attachment insecurity is associated with important relationship outcomes, well-being, and mental health. Attachment has long been considered a stable trait, but recent findings indicate that attachment insecurity decreases over time among adults in romantic relationships. Although theoretical considerations suggest that positive experiences within these relationships should contribute to changes in attachment, little is known about the specific characteristics that foster attachment security. Using a longitudinal dyadic sample of well-established couples (N = 1,036 couples; MRelationship Duration = 8.6 years, MAge = 32.9 years), this study aims to (a) examine changes in attachment anxiety and avoidance in adults in a romantic relationship over 20 months, (b) explore the codevelopment of attachment between partners, and (c) investigate actor and partner effects of relationship characteristics (i.e., relationship satisfaction, commitment, support, closeness, responsiveness, and disclosure) on changes in attachment. Dyadic growth curve models indicated small but significant decreases in anxiety over time in women. No codevelopment between partners was observed. Further, changes in attachment were largely unrelated to baseline relationship characteristics with few exceptions: Women with highly committed and responsive partners showed weaker declines in anxiety, likely due to higher initial anxiety levels. However, changes in anxiety were negatively associated with relationship duration. These findings suggest that avoidance remains stable and uninfluenced by relationship characteristics, while anxiety decreases slightly only in women, independent of partner codevelopment. This highlights that attachment insecurity partially changes over time during the course of a romantic relationship but is largely unaffected by relationship dynamics. 

This is a large longitudinal study that illustrates nicely that love can’t fix everything. Avoidant attachment is very hard to change so that isn’t surprising. The absence of codevelopment is interesting in that it’s easy to expect couples to become more similar over 20 months or more in a relationship. The finding that women with highly committed and responsive partners only reduce their anxiety a little and men not at all is a little surprising to me given the number of relationship qualities they examine. The next study looks at what happens when you become a parent. Lessard, Brassard, Gosselin, Brault-Labbé, Lafontaine & Péloquin (2025) published Attachment, Conflict, and Relationship Satisfaction of Couples Transitioning to Parenthood: The moderator role of intimacy” in Journal of Family Psychology. The edited abstract follows:

On average, couples experience increase in conflicts and decrease in relationship satisfaction during the transition to parenthood. How couples manage conflicts may improve or erode their relationship over time. While romantic attachment (anxiety, avoidance) is known to modulate couples’ adaptation during the transition to parenthood, the contribution of conflict resolution styles and intimacy to better understand these links seems promising. The goal of this study was to examine the mediator role of conflict resolution styles in the associations between romantic attachment and relationship satisfaction and the moderator role of intimacy in the conflict-satisfaction associations. In a dyadic prospective study involving 211 couples welcoming their first child, we assessed prenatal romantic attachment, conflict resolution style and intimacy at 4-month postpartum, and relationship satisfaction at 12-month postpartum. Results from structural equation modeling indicated that prenatal attachment insecurities predicted a higher use of conflict engagement, compliance, and withdrawal and a lower use of positive problem-solving conflict resolution styles in both partners at the actor and partner levels. Bearing mothers’ lower relationship satisfaction was explained by their own and their partners’ attachment insecurities through their lower use of positive problem solving and by their partners’ higher use of conflict engagement. Partners’ lower relationship satisfaction was explained by their own and the bearing mothers’ attachment insecurities through their own compliance and withdrawal styles, as well as the bearing mother’s conflict engagement. When partners perceived a high level of intimacy, their withdrawal and compliance were related to a higher level of relationship satisfaction. 

Here we see again that people who are insecurely attached don’t cope very well. While not a huge sample, I like the fact that they speak to the differences between moms and their partners. It makes perfect sense to me that, as a new mom, if you and your partner don’t know how to use positive problem solving or your partner uses conflict engagement, it’s a bad sign. On the flip side, when the partner complies or withdraws and the new mom uses conflict engagement, you’re also in trouble. The final study looks at the demise of relationships. Bühler & Orth (2025) published “Terminal Decline of Satisfaction in Romantic Relationships: Evidence from four longitudinal studies” in Journal of Personality and Social Psychology.  Here’s the edited abstract:

In this preregistered research, we tested whether there is a systematic, terminal decline in relationship satisfaction when people approach the end of their romantic relationship. Data came from four longitudinal studies with national samples. In the analyses, we used (piecewise) multilevel models with propensity score-matched event and control groups. Across studies, sample sizes ranged from 987 to 3,373 for event groups and from 1,351 to 4,717 for control groups. Relationship satisfaction systematically declined as a function of time-to-separation. The decline prior to separation was divided into a preterminal phase, characterized by a smaller decline, and a terminal phase, characterized by a sharp decline. Across studies, the onset of the terminal phase was estimated at 0.58–2.30 years prior to separation. For comparison purposes, we also examined relationship satisfaction as a function of time-since-beginning, showing that time-to-separation was a much better predictor of change than time-since-beginning. Moreover, for comparison purposes, we examined change in life satisfaction, showing that terminal decline was less visible in life satisfaction than in relationship satisfaction. Moderator analyses indicated that age at separation and marital status explained variance in the effect sizes. Moreover, individuals who were the recipients of the separation (in contrast to individuals who initiated the separation) entered the terminal phase later but then decreased more strongly. The findings support that ending relationships show a typical pattern of preterminal and terminal decline, which may have important implications for the timing of interventions aimed at improving relationships and preventing separation. 

This time we don’t have an assessment of attachment but we are looking at a huge sample with control groups. I thought these data were helpful in illustrating the transition from the preterminal to the terminal phase. I thought it was interesting, and absolutely reasonable, that the partner getting dumped realized it later but then decreased faster in relationship satisfaction (a kind of ‘you can’t fire me I quit’ response).