Causal mediation of the association between educational attainment and late-life cognitive function by trajectories of social participation

BACKGROUND: Education is a well-established predictor for late-life cognition. Moreover, social participation is associated with late-life cognition and could mediate the association between education and cognitive function. Yet, few studies have formally tested this question within a causal mediation framework. METHODS: We used data from the Taiwan Longitudinal Study in Aging. Participants were 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 techniques to define clustered patterns of social participation and work trajectories. We then used a four-way decomposition to identify causal mediation of the association between educational attainment and cognitive function by social participation. Cognitive function was measured by the Short Portable Mental Status Questionnaire. RESULTS: The mediating pathway through social participation accounted for 5% or less of the total relation of education on late-life cognitive function for both younger and older subgroups. The interaction between education and social participation accounted for 22% (middle/high school) and 34% (college+) of the total effect, contrasting with no formal education. We estimated that 36% (95% C.I. = 0.09, 0.69) of the educational disparity in late-life cognition among the younger cohort could be eliminated if the least educated group's social participation matched that of the college or above group. Social participation was not a significant mediator in the older subgroup. CONCLUSION: To mitigate educational disparities in late-life cognitive function, further exploration of encouraging more active social participation or stable employment should be undertaken.