Factors Related to Cognitive Resilience according to Education Level in Normal and Cognitively Impaired Individuals

Abstract

AbstractBackgroundCognitive resilience (CR) is a concept to explain the difference between pathologic burden and cognitive function. Understanding CR is important as increasing CR can be applied to dementia prevention strategies. CR is closely related to education and it was widely used as proxy for CR. However, using an education as proxy of CR may have limitation in reflecting the CR of low educated individuals. Related factors to CR can be different according to the educational level. However, factors related to low educated individuals has not been sufficiently investigated. Therefore, we aimed to investigate factors related to cognitive resilience in low‐ and highly educated individuals using residual approach.MethodWe collected 1,085 patients who underwent neuropsychological assessment and MRI from the Establishment of Precision Medicine Platform for Early Diagnosis of Mild Cognitive Impairment (PREMIER) Cohort (305 SCD, 639 MCI, 141 AD). To capture CR using residual approach, firstly, cognitive score was predicted on global cortical thickness, adjusted for age, sex, education, and disease stage. Then, individual differences between predicted and observed cognitive score were used as the measure of CR(CR_rsd). Next, we divided the participants into two groups based on their education level. To investigate the factors related to CR_rsd between two groups, we used multiple linear regression and logistic regression model after adjusted age, sex. Demographic, socioeconomic and psychiatric factors were evaluated.ResultMean age was 70.5 years and 65.3% were female. There were 286 patients in low educated group (26.4%). Depression, literacy, cognitive activity, and presence of diabetes were related to CR_rsd. Literacy was related to CR_rsd in both low‐ and high‐education group (low education group, p = 0.035; high education group, p = <0.001). Depression was associated with low CR_rsd only in high education group (p = 0.001). Cognitive activity was associated with high CR_rsd in both group (low education group, p = 0.042; high education group, p = <0.001). However, social activity was only related to high CR_rsd low education group (p = 0.044).ConclusionOur findings suggest that factors related to CR could be different according to the educational level.