The Differential Impact of APOE ε4 on Normative Cognitive Aging in Research Samples from the United States and Korea

Abstract

Background: Normative cognitive aging and the influence of genetic risk factors on neurodegenerative disease processes differ between ethnic and racialized groups. The aims of this study were to examine differences in normative cognitive aging between research cohorts from the United States and Korea and evaluate the impact of APOE ε4. Method: We estimated normative cognitive aging patterns among non‐Hispanic white participants in United States (ADNI, N = 609; Agemean=73±6; % Female=54; Years of education = 16.7±2.5) and Korean (KBASE, N = 245; Agemean=68±8; % Female=50; Years of education = 12.3±4.6) cohorts who remained free of cognitive impairment over 4‐years of follow‐up using linear mixed effects models by age. Harmonized data were leveraged to estimate cognition by age across domains of memory, executive function, language, and visuospatial abilities. Linear mixed‐effects models with a random intercept by participant and fixed effect interactions of age with sex, education, APOE ε4 (excluding ε2/ε4 carriers), and cohort (ADNI vs. KBASE) were specified for each domain; a three‐way interaction between age, APOE ε4, and cohort was also specified. Result: Normative aging patterns demonstrated higher intercepts for executive functioning (β[95%CI]=0.18[0.10,0.26]) and visuospatial (β[95%CI]=0.15 [0.06,0.24]) domains in KBASE compared to ADNI, with a lower intercept for memory (β[95% CI]=‐0.27[‐0.35,‐0.19]), and no difference in the intercept for language (β[95%CI]=‐0.05[‐0.13,0.03]). Participants in KBASE exhibited steeper age‐related declines across all domains of cognition (βrange = ‐0.21,‐0.09). APOE ε4 carrier status was associated with lower memory (β[95%CI]=‐0.17[‐0.32,‐0.02]) and visuospatial (β[95%CI]=‐0.31[‐0.48, ‐0.14]) intercepts and a slower rate of age‐related decline for executive function (β[95%CI]=0.17[0.05,0.15]) in KBASE, but not ADNI (Table, Figure). Main effects of interest were broadly consistent in a covariate matched sample. Conclusion: Results show variability in normative cognitive aging patterns between research study participants in United States and Korean imaging studies. APOE ε4 was more influential in rates of decline among Korean participants. These findings underscore the importance of replicating findings, diversifying study populations, and of considering ethnic and racialized backgrounds when examining genetic risk and cognitive aging, as different populations may show distinct patterns of decline.