Polygenic risk score for Alzheimer’s disease and APOE ε4 status in relation to cognitive decline: a population‐based study of individuals followed over 16 years.

Abstract Background Polygenic risk scores for Alzheimer’s disease (AD‐PRS) have been associated with incident AD and dementia. However, few long‐term studies have examined the effect of AD‐PRSs, and the interaction with APOE ε4 status, on cognitive performance over time in the general population. Method A population‐based sample of 855 participants born 1930 were examined at age 70, 75, 79, and 85 years. AD‐PRSs that surpassed four P value thresholds ( P <5e‐8, P <1e‐5, P <1e‐3, P <1e‐1) were generated based on information from recent genome‐wide association studies. Cognitive performance was measured at all examinations using a composite global cognitive score based on five psychometric tests (Thurstone’s Picture Memory, immediate recall of 10 words, Block Design, PSIF [perceptual speed], and Mini‐Mental State Examination score). Linear mixed effect models with random intercepts and slopes were used to analyze the effect of AD‐PRSs and APOE ε4 status on cognitive performance. The models included sex*age, AD‐PRS*age, APOE ε4 status*age, and 10 principal components to correct for population stratification. Analyses were repeated in samples without prevalent and incident dementia. Result The interaction‐term 5e‐8 AD‐PRS*age (β=‐0.2, P =.03) and APOE ε4 status*age (β=‐0.2, P =.02) were associated with cognitive decline. The results remained in those without prevalent and incident dementia (the 5e‐8 AD‐PRS*age: β=‐0.2, P =.02, APOE ε4 status*age: β=‐0.2, P =.007). No association was observed between AD‐PRSs that surpassed P value thresholds >5e‐8 and cognitive performance over time (1e‐5 AD‐PRS*age: β=‐0.06, P =.08, 1e‐3 AD‐PRS*age: β=‐0.03, P =.3, and 1e‐1AD‐PRS*age: β=‐0.02, P =.6). Further, neither APOE ε4 status nor AD‐PRSs were associated with cognitive performance at baseline ( APOE ε4 status: β=‐0.9, P =.3, the 5e‐8 AD‐PRS: β=0.6, P =.5, 1e‐5AD‐PRS: β=0.5, P =.2, 1e‐3 AD‐PRS: β=0.5, P =.1, and 1e‐1 AD‐PRS: β=0.2, P =.6). Conclusion We found an effect of a 5e‐8 AD‐PRS and APOE ε4 status on cognitive decline in individuals from the general population followed over 16 years, regardless of dementia diagnosis. These results contribute with additional knowledge to identify individuals at increased risk of cognitive decline, which may be important to implement personalized prevention as early as possible.