Brain Signal Variability During Rest as a Neural Mechanism Underlying Cognitive Reserve

Abstract

Background: Resting-state brain signal variability has been found to vary with age and cognitive function. Neural flexibility has been suggested as a neural mechanism underlying cognitive reserve (CR), a construct that describes better than expected cognition given brain status. Thus, we examined the associations between age, resting-state brain signal variability, cognition, and CR. Method: Analysis was based on resting-state functional neuroimaging data from 470 participants (aged 20-80 years) from the Reference Ability Neural Networks and the CR studies. Brain signal variability was quantified for each brain region as the log-transformed standard deviation of the time-varying blood-oxygen-dependent (BOLD) signal. We then derived variability patterns related to age, perceptual speed, fluid reasoning, episodic memory, and vocabulary using Scaled Subprofile Modelling principal component analysis. To perform the formal test whether these patterns fulfill the requirements for CR, we examined whether they explained additional variance in cognition beyond brain status, age, sex, and education, or moderated the brain status-cognition relationship. We additionally stratified all regression models by age (cutoff: 60 years) and sex. Results: BOLD signal variability showed an age-related increase in subcortical/medial brain regions, and an age-related decrease in cortical regions. It also met the CR test for speed (standardized regression coefficient ({beta})=0.251, 95% confidence interval (CI): 0.118-0.384, pFDR<0.001), episodic memory ({beta}=0.344, CI: 0.200-0.489, pFDR<0.001), reasoning ({beta}=0.316, CI: 0.197-0.436, pFDR<0.001), and vocabulary ({beta}=0.270, CI: 0.167-0.373, pFDR<0.001). Associations were stronger in women for vocabulary and in young individuals for reasoning. Conclusions: BOLD signal variability plays a role in aging and cognition and underlies CR.