Investigating Potential Racial Differences in Dementia‐Related Plasma Biomarkers

Abstract Background Black self‐identified individuals are disproportionately affected by dementia, a phenomenon best attributable to differences in social determinants of health (SDOH). Plasma protein levels can serve as antemortem indicators of dementia pathology; recent findings indicate potential race‐specific differences in these biomarkers of tau and non‐tau pathologies between Black self‐identified or White individuals. Methods This sample included 568 participants (77, 13.56% Black self‐identified, Table 1 ) from the Boston University Alzheimer’s Disease and Research Center (BUADRC). Participants completed a battery of neuropsychological tests and cognitive diagnoses were made during multidisciplinary diagnostic consensus conferences. Plasma samples were analyzed using ELISA methods for YKL‐40 levels and using Simoa platform for phosphorylated tau (p‐tau 181 , p‐tau 231 ), total‐tau, glial fibrillary acidic (GFAP), and neurofilament light chain (NfL) levels. ANCOVAs tested for a cross‐sectional association between racial identity and each plasma biomarker and stratified by cognitive impairment status with covariates for age, education (in years), sex assigned at birth, and APOE ε4 carrier status. Cox proportional‐hazards models were used for longitudinal analyses of the interaction effect between racial identity and plasma biomarkers in the risk for conversion from normal cognition (NC) to mild cognitive impairment (MCI) and reversion from MCI to NC. Results Baseline YKL‐40 was lower among Black self‐identified participants (adjusted mean diff = 0.37 SD, p <.001); this association was particularly salient within the MCI group (adjusted mean diff = 0.62 SD, p <.001). The longitudinal analysis revealed interaction effects between race and GFAP for risk for conversion from NC to MCI and reversion from MCI to NC. Among White participants, higher GFAP predicted increased risk for conversion from NC to MCI and lower levels predicted reversion. However, no significant effects were found among Black self‐identified participants; these findings might be limited by sample size (n = 18) resulting from the high occurrence of baseline MCI diagnoses and subsequent lack of conversion or reversion among Black self‐identified participants. Conclusion In the context of an unrepresentative sample, we found trivial differences in plasma biomarkers between Black self‐identified and White participants. Recruitment of Black self‐identified participants representative of the population and further exploration of SDOH is essential in understanding the disproportionate risk for dementia carried by Black self‐identified individuals.