Subjective cognitive impairment differs by racial/ethnic groups and sociodemographic factors: The Multi‐Ethnic Study of Atherosclerosis and Alzheimer’s Disease (MESA‐MIND)

Abstract

AbstractBackgroundSubjective cognitive decline (SCD) is associated with objective cognitive decline, incident cognitive impairment, and AD pathology in mostly white populations. Little data are available to describe SCD and objective cognitive impairment in diverse populations.MethodThe Multi‐Ethnic Study of Atherosclerosis (MESA) began cognitive testing, including the Cognitive Abilities Screening Instrument (CASI), in 2010‐12 among four racial/ethnic groups: 38% white, 28% African‐American, 22% Hispanic, and 12% Chinese Americans across six field centers. MESA‐MIND and ancillary studies added repeated CASI by in‐person exam (2016‐2018) and the 12‐item Cognitive Change Index (CCI‐12) during the 2018 annual follow‐up call. Statistical analyses using the CCI‐12 item sum and two published thresholds (CCI‐12≥16 and CCI‐12≥23) established the point prevalence for SCD across racial/ethnic groups. Socio‐demographic factors were examined using multivariable logistic regression. Spearman correlations and general linear models related 5‐year subjective change (CCI‐12) and objective change in cognitive performance (change in CASI) over the same time period.ResultCCI‐12 data were available on 3,676 MESA participants (aged 60‐95) across all racial/ethnic groups and sites. Table 1 presents demographics, CCI‐12 scores, and CASI scores. The prevalence of SCD (CCI‐12≥16) across racial/ethnic groups ranged between 52% to 72%. There was a lower prevalence in African‐American and white and a higher prevalence in Chinese and Hispanic participants. In adjusted models, various socio‐demographic factors were significantly associated with a higher odds of reporting SCD (CCI‐12≥16), including: older age, female gender, Chinese language preference, diabetes, smoking, birth in the Western US or Mexico, and lacking health insurance coverage. African‐American ethnicity, higher education, and income greater than $75,000 were each associated with a lower reporting of SCD. CCI‐12 was weakly and inversely correlated with antecedent cognitive performance (2010‐12, n=3,200, Rho=‐0.16, p<0.01) and cognitive decline (n=1,800, Rho=‐0.06, p=0.01) over 6 years of follow‐up. In stratified analyses, associations between CCI‐12 and objective cognitive decline were observed in white participants only.ConclusionIn a large regionally and ethnically diverse cohort of older adults, SCD defined by subjective memory concerns on the CCI‐12 significantly differed by race/ethnicity and indicators of socioeconomic status, suggesting that cultural and institutional biases may influence reporting SCD in diverse populations.