Individual differences in habitual sleep quality and episodic memory network activity across the adult lifespan:
Difficulty utilizing cognitive control and the patterns of prefrontal cortical (PFC) activity underlying it is a major contributor to episodic memory decline in normal aging. Sleep quality is related to PFC integrity and its associated network connectivity. Normal aging has been shown to negatively affect this activity, sleep quality, and episodic memory performance but the relationship between these variables is unclear. The overarching goal of this project is to determine the extent to which age-related sleep disruptions contribute to those in PFC network activity during learning that, in turn, affect episodic memory performance. A mediation model is proposed in which poor sleep quality partially mediates the negative impact of healthy aging on episodic memory via reduced resolution of interference during learning, a major age-related cognitive control deficit. A diverse sample of 90 young, middle-aged, and older adult participants without sleep disorders, perform multiple episodic memory tasks that tax different interference resolution functions is measured to assess the generalizability of the negative impact of poor sleep. Episodic memory is assessed behaviorally and supporting brain activity measured with fMRI. In order to evaluate the specificity of the model to episodic memory, both episodic and item memory are tested. Memory is tested immediately following learning and after a 24-hour delay in order to separately assess the impact of sleep on new learning from consolidation. The influence of multiple confounds, including medication, naps, depression, and exercise is controlled through screening and statistical analyses. Objective measures of sleep are obtained using actigraphy, the best method for assessing habitual sleep patterns over multiple days for a large number of individuals naturally, in their homes, with minimal intrusion. Mediation results from average sleep estimates, the night immediately prior to memory testing, and night-to-night variability are compared in order to separate the impact of habitually poor sleep from variable or a poor night’s sleep. Univariate analyses are used to identify PFC regions that support resolution of interference defined by activity that is correlated on a trial-by-trial basis with multivariate pattern analysis (MVPA)-derived interference signals. Functional connectivity analyses are additionally applied to identify the networks affected by sleep quality.