Cluster‐Derived Psychosocial Phenotypes, Cognitive Outcomes, and AD Plasma Biomarkers of Black and Hispanic/Latino Older Adults Enrolled in HABS‐HD

Abstract

AbstractBackgroundAlthough Black and Hispanic/Latino older adults are at increased risk for Alzheimer’s disease (AD), there is considerable within‐group heterogeneity in psychosocial factors of risk and resiliency amongst these community members. The present study (1) used cluster analysis on measures of economic/social resources and psychiatric distress to identify distinct subgroups and (2) compared cognitive and AD biomarker outcomes of these groups in an effort to understand pathologic aging among racially/ethnically diverse older adults.MethodBaseline data from 816 Black and Hispanic/Latino participants without dementia enrolled in HABS‐HD were included. Variables in the cluster analysis included: annual income, state area deprivation index, social support, occupational complexity, chronic stress, and symptoms of depression and anxiety. A discriminate function analysis (DFA) tested the extent to which the individual variables predicted cluster‐group membership. ANOVAs and chi‐squared tests compared the groups on demographics. ANCOVAs controlling for age, education, sex, and race/ethnicity examined group differences in subjective cognitive concerns, objective memory and executive functioning composite scores, and plasma AD biomarkers (AB42/40, t‐tau, NfL).ResultThe cluster analysis identified 3 distinct groups: Group 1 was low resource/high distress (n = 215); Group 2 was high resource/low distress (n = 248); and Group 3 was low resource/low distress (n = 353; Figure 1). The DFA correctly classified 95.3% of participants. Results revealed significant group differences in education and the proportion of women and Black adults across the groups (see Table 1). ANCOVAs revealed significant group differences in subjective memory concerns, executive functioning composite, and plasma AD biomarkers; Group 1 consistently displayed poorer cognitive outcomes and more severe AD biomarker patterns relative to Groups 2 and 3, but there no differences in outcomes between Groups 2 and 3. No group differences on the memory composite were observed.ConclusionRisk for AD among racially/ethnically diverse older adults may differ as a function of current levels of resources and psychiatric distress. Group 1 represents individuals with low resources and high levels of distress that may be at increased risk for poor outcomes in late life. Future work should further clarify precise mechanisms of resiliency, such as social support, in Group 3 and explore longitudinal trajectories of cognitive decline across the groups.