Longitudinal Dyadic Associations Between Loneliness and Cognition Among Older Couples in the United States

ObjectivesLoneliness is associated with diminished health and cognition for older individuals. However, little research has examined dyadic loneliness—that is, loneliness of both partners in a relationship—and its potential consequences for cognitive functioning among both spouses, nor whether one partner’s cognition may affect both partners’ loneliness over time.MethodsWe analyze 3-wave dyadic Health and Retirement Study data (2010–2020; N = 1,061 dyads) to determine (a) whether loneliness predicts participants’ own and/or their partners’ episodic memory and verbal fluency over 8 years, and (b) whether cognitive functioning predicts older spouses’ own or their partners’ loneliness over the same period.ResultsLoneliness predicted participants’ own and their partners’ loneliness at follow-up, at both time points. Loneliness was also associated with own episodic memory at follow-up, but not with verbal fluency. Episodic memory and verbal fluency predicted one another over time. Neither episodic memory nor verbal fluency predicted loneliness at follow-up. Significant dyadic mediation was established such that Time 1 loneliness was linked with partner’s Time 3 episodic memory via that partner’s Time 2 loneliness.DiscussionLonelier older adults displayed worse trajectories of episodic memory over time, yet poor memory did not precede changes to loneliness. Further, having a lonely partner was linked with poorer episodic memory 8 years later, indicating that both one’s own and—to a lesser extent—a partner’s emotional well-being may be consequential for maintaining cognitive functioning with age. Associations were more clearly established with episodic memory than with verbal fluency, suggesting potential domain-specific effects of loneliness.