Aging
Association between Social Participation Over Time and Cognitive Function Yu-Tien Hsu* Yu-Tien Hsu Ichiro Kawachi Jarvis T. Chen Hanno Hoven
Background
Greater social participation has been related to less cognitive decline with aging in many studies. However, there is limited understanding of how trajectories in social participation over time be associated with older adults’ cognitive function in later life.
Methods
We used data from the Taiwan Longitudinal Study in Aging (TLSA), a nationally representative sample. The sample comprised participants aged 50 or above and recruited in 1989 (n = 4,400), stratified by age group (aged 50-64 years versus 65 or above). We used sequence analysis and optimal matching technique to define clustered patterns of social participation and work history in 1996, 1999, 2003, 2007, and 2011. We applied linear regression analysis to model the associations between cluster membership and level of cognitive function in 2015, measured with the Short Portable Mental Status Questionnaire (SPMSQ). We used multinomial regression analysis to explore the factors related to belong to a cluster.
Results
We identified six clusters for each subgroup. The clusters presented in the younger subgroup included: (1) socializing and helping; (2) low social participation; (3) working; (4) active in multiple states; (5) socializing; and (6) Helping. In the older subgroup, we identified the following clusters: (1) socializing; (2) helping; (3) low social participation; (4) working & low social participation; (5) helping & socializing; and (6) active in multiple states. Most participants reported changes in their pattern of social participation during the study period. However, once an individual transitioned into the low social participation cluster, they tended to stay in that cluster. In both the younger and older subgroups, we found that the working dominant cluster was associated with higher cognitive function compared to the low social participation cluster (younger subgroup: b = 0.52, 95% C.I. = 0.03 – 1.01; older subgroup: b = 2.33, 95% C.I. = 1.31 – 3.36). We also found that the multiple active social participation and working cluster was related to higher cognition (younger subgroup: b = 0.58, 95% C.I. = 0.09 – 1.06; older subgroup: b = 1.15, 95% C.I. = 0.05 – 2.26), after controlling for important covariates and accounting for censoring weights. More active patterns of social participation were found for men, at younger ages, and among non-urban residents and those without chronic disease.
Conclusion
Stable involvement in multiple types of social participation activities over time was associated with better cognitive function in older adults. Gender and urbanicity of residence were significant predictors of more active patterns of social participation.
