OBJECTIVE AND SUBJECTIVE COGNITIVE DECLINE IN BLACK AMERICANS FROM THE HARVARD AGING BRAIN STUDY

Abstract

Subjective cognitive concerns (SCC) and episodic memory have been proposed as a means of identifying risk for Alzheimer's disease (AD). However, the longitudinal utility of these measures has not been well-explored for African-Americans, who have higher AD risk and typically present with an increased burden of white matter hyperintensities, which are implicated in executive function performance. We investigated whether an interaction between executive function and race contributes variance in predicting cognitive decline over a four-year period using data from the Harvard Aging Brain Study. We hypothesized that African-Americans would show more cognitive decline than Caucasians, but that African-Americans’ decline would be mediated by baseline executive function performance, rather than SCCs or baseline memory performance. We used a composite of three SCC questionnaires, the Preclinical Alzheimer's Cognitive Composite (PACC), and an executive function factor score. We studied 220 (60% female; African-American n = 33) clinically normal older adults over a four-year period. A hierarchical linear regression assessed the modifying role of race on the contributions of baseline measures of SCC, episodic memory, and executive function on PACC change scores was performed. Covariates included age, APOE status, MMSE, verbal IQ, and socioeconomic status. The second step included baseline measures of memory, SCC, and executive function. The third step, testing our hypothesis, included interactions between race and each of SCC, memory, and executive function. As predicted, African-Americans demonstrated steeper rates of decline over four years, relative to Caucasians (ß= -.178). No main effects of baseline SCC, memory, or executive function were significant (all ßs < ± .046). Critically, although the race*baseline memory (ß= .083, p = .310) and race*baseline SCC (ß= .023) interactions were not significant, there was a significant contribution of race*baseline executive function (ß= -.187). Follow-up analyses revealed a main effect of baseline executive function for African-Americans (ß= -.237), but not for Caucasians (ß= .044). Results suggest that baseline executive function performance is a uniquely important predictor of longitudinal change on the PACC for African-Americans. These results highlight the importance of examining racial influences on risk factors implicated in AD, particularly for high-risk populations such as African-Americans.