Health Disparities
Investigating the Role of Chronic Stress and Perceived Discrimination in Racial Disparities in Incident Dementia among US older adults: A Causal Mediation Analysis Mina Habib* Mina Habib Nicola Churchill Roch A Nianogo
Background: The rising prevalence of cognitive decline and dementia in the general population necessitates the need to identify and address factors that contribute to racial disparities in dementia outcomes.
Objective: This study investigates the role of ongoing chronic stress (OCS) and perceived everyday discrimination (PED) in the association between race and incident dementia (ID) among Non-Hispanic (NH) Black and White populations and aims to identify intervention pathways to reduce these disparities.
Methods: Utilizing the Health and Retirement Study—a longitudinal, nationally-representative study of Americans over 50—G-computation based parallel mediation analysis was performed on 5,602 participants (4,913 NH White, 689 NH Black). Multiple imputation (MI) and inverse probability of censoring weights (IPCW) addressed incomplete covariates and loss to follow-up. ID was identified using validated Lange-Weir Classification, and PED and OCS scores were obtained from established indices. Analysis included socio-demographic characteristics, baseline mediator values, and Apolipoprotein E4 status. Results from 100 MI/IPCW datasets were combined via Rubin’s rule, with 95% confidence intervals estimated via bootstrapping.
Results: NH Black participants demonstrated a higher dementia risk compared to NH Whites, with an adjusted risk ratio of 2.21 (95%CI: 1.69 – 2.73). Neither OCS or PED directly mediated the race-dementia association. However, we observed substantial proportions attributable to interaction (27% for OCS and 26% for PED) and proportions eliminated (27% for OCS and 26% for PED), indicating significant excess ratios due to interaction (ERINTref) between OCS and race (ERINTref: 0.32; 0.06 – 0.58) and PED and race (ERINTref: 0.3; 0.13 – 0.48).
Conclusions: Our findings, notably proportions attributable to interaction and eliminated, suggest racial disparities in ID due to OCS and PED can be eliminated through race-specific interventions targeting both pathways.
