African-American Race Predicts 1-Year Cognitive Decline Among Adults Without Moderate Dementia

Abstract

Previous literature shows conflicting conclusions about the association between race and cognitive decline, particularly in early impairment. In this study, we aimed to test whether race predicted 1-year change in Montreal Cognitive Assessment (MoCA) score among older adults without moderate-severe dementia. We secondarily explored whether multimorbidity, polypharmacy, depressed mood, antidepressant use, body composition, or frailty changed the association. We analyzed data (n=122) from predominantly African American (AfA, 78.7%) community-dwelling older adults from the south side of Chicago. Participants underwent baseline and 1-year MoCA testing. Age, gender, race, education, monthly income, co-morbidities (Charlson Comorbidity Index), medication use (<5 vs ≥5), depression (PHQ-2), proportion lean mass (DEXA), and the frailty phenotype (range 0-5) were collected at baseline. In a multivariate linear model, we regressed 1-year MoCA score on baseline MoCA score, race, and demographics and then evaluated the impact of each covariate added separately to the model on the race-cognition relationship. The mean MoCA score at baseline was 25.2+/-0.2 (range 18-30) and 41.0% of participants experienced ≥1 point MoCA decline at 1 year. After adjusting for demographics, AfAs experienced a greater 1-year MoCA decline (β= -1.3, p=0.04) compared to other races. The effect size was unchanged after adjusting for multimorbidity and polypharmacy (β= -1.3, p=0.04), attenuated slightly after adjusting for frailty (β= -1.2, p=0.06), depressed mood (β= -1.2, p=0.05), lean mass (β= -1.2, p=0.04), and attenuated notably after adjusting for antidepressant use (β= -1.0, p=0.11). Findings support the need to further explore racial differences in cognitive decline and potentially related anti-depressant underuse.