Aging
Examining the Joint Effects of Positive Spousal Social Support and Closeness on Dementia Risk using Causal Survival Analysis Nicola Churchill* Nicola Churchill Churchill Churchill University of California, Los Angeles
Little is known about how different sources of social support work together to modify dementia risk, or how the quality of spousal relationships may amplify or constrain the cognitive benefits of broader social networks. Aspects of spousal relationship quality may help protect against dementia risk, including via mental health-related mechanisms. We evaluated the individual and joint effects of two measures of spousal relationship quality – perceived positive support and closeness – in association with dementia risk. We used data on respondents 50 years+ and their spouses to the U.S. Health and Retirement Survey (HRS), 2006-2020, focusing on those dementia/cognitively impaired but no dementia (CIND)-free and who were married or partnered at analytic baseline. Reporting closeness with spouse was associated with a greater risk of dementia (RR:1.07 95% CI:1.01, 1.14) and reporting positive support from a spouse was associated with a lower risk of dementia RR:0.87 95% CI:0.71, 1.06), with no meaningful variation by sex/gender. Among men, reporting both positive social support from and closeness with their spouse was associated with a greater risk of dementia (RR:1.06 95%CI:0.94, 1.19). Conversely, among women, reporting both positive social support from and closeness with their spouse was associated with a lower risk of dementia (RR:0.93 95%CI:0.84, 1.02). Our results suggest nuanced links between aspects of the spousal relationship with dementia risk, including by sex/gender and consideration of individual vs. joint exposures. Findings of association between spousal support and greater dementia risk could be influenced by reverse causation, whereby greater spousal closeness reflects caregiving dynamics in the context of early cognitive decline. Further research should address additional sources of bias (e.g. marital selection) and evaluate potential mechanisms to interrogate plausibility of observed associations.