Race, Gender, and Late‐life Productive Activities and Dementia in the Health, Aging, and Body Composition Study

Background Scholars have not addressed if productive activities, like working and volunteering, can mitigate Black vs. White racial disparities in incident dementia among older adults. We hypothesized: working vs. not, and volunteering vs. not, will reduce dementia risk and decrease racial disparities in the Health, Aging, and Body Composition (Health ABC) study. Method Health ABC is a 17‐year epidemiologic, prospective cohort study developed by the National Institute on Aging, including community‐dwelling, well‐functioning, Black (42%) and White, men and women (52%), aged 70‐79 in 1997‐98 (n = 3,069 at baseline). Full sample and gender‐stratified logistic regression model incident dementia risk among older adults with adjustments for socio‐demographics (e.g., race, age, education, family income, and site) and Apolipoprotein E genotype (APOE). Result In a full sample, unadjusted model, working vs. not (Yr1) was significantly related to 22% lower odds of dementia (OR = 0.78, p0.05) and volunteering vs. not (Yr1) was borderline‐significantly related to 16% lower odds of dementia (OR = 0.84, p = 0.059). In respective socio‐demographic and full model adjustments, working and volunteering were not significantly related to dementia. Black vs. White older adults had 31% (OR = 1.31, p0.05) and 29% (OR = 1.29, p0.05) higher odds of dementia in working and volunteering models with socio‐demographic adjustments, respectively. Race was not significant after APOE adjustment. Among older men, no significant relationships existed between dementia with work or race in any models. Similarly, no significant relationships existed between dementia with volunteering or race in any men’s models. Among older women, in the unadjusted model, working vs. not (Yr1) was significantly related to 30% lower odds of dementia (OR = 0.70, p0.05), but it was not significant in other models. Similarly, in an unadjusted model for women, volunteering vs. not (Yr1) was significantly related to 24% lower odds of dementia (OR = 0.76, p0.05), but it was not significant in other models. Black vs. White women had 35% (OR = 1.35, p0.05) and 32% (OR = 1.32, p = 0.052) higher odds of dementia in working and volunteering models with socio‐demographic adjustments, respectively, but race was not significant after APOE adjustment. Conclusion This research provides insight about productive activities as possible contextual interventions to reduce racial disparities in incident dementia among older women.