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

Objectives: Loneliness 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.

Methods: We 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.

Results: Loneliness 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.

Discussion: Lonelier 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.