0091 Sleep as a Mediator of the Relationship Between Exercise and Self-Reported Cognitive Function

Introduction

Insufficient sleep and sedentary behavior are associated with cognitive decline. How sleep and physical activity interact to influence cognition is not fully understood. This study sought to examine whether self-reported sleep and exercise impact subjective cognitive complaints across adulthood.

Methods

Self-report questionnaire data from 2,744 adults (μ=56.18 yrs/old; 63.8% F; μ=16.36 yrs/edu) in the University of California Irvine Consent-to-Contact Registry were analyzed. Multiple regression models, analysis of covariance, and ordinary least squares path analysis were conducted to investigate relationships among the Cognitive Function Instrument (CFI), Medical Outcome Study Sleep Scale (MOS-SS) Subscales, and exercise frequency. All models adjusted for education, sex, age, BMI, medical comorbidities, depression, cancer diagnosis, and antidepressant usage. Individuals taking medications known to affect sleep or with a diagnosis major psychiatric illness were excluded from analyses.

Results

All MOS-SS Subscales significantly predicted CFI score (all p<0.001). Individuals who exercise >3times/week had significantly lower scores on the CFI (p<0.05), Sleep Problems Index I & II (both p<0.05), Somnolence Subscale (p<0.01), and higher scores on the Sleep Adequacy Subscale (p=0.001). Mediation analyses revealed that all subscales, aside from the Sleep Disturbance Subscale, mediated the relationship between exercise frequency and CFI Score (Bootstrapped CI’s did not include zero).

Conclusion

More self-reported sleep disturbance and greater daytime sleepiness are associated with more subjective cognitive complaints. Individuals who exercise more frequently report lower daytime sleepiness and higher quality sleep. The effects of exercise frequency on cognitive complaints appear to be mediated by the impact of exercise on sleep. These results suggest that sleep health may be a crucial consideration when evaluating outcomes of exercise-based therapies aimed at delaying the onset of cognitive impairment.

Support

The C2C registry was made possible by a donation from HCP, Inc. and is supported by NIA AG016573 and NCATS UL1 TR001414.

0422 Apocalypse Tau: The Relationship Between Inflammaging and Local Sleep Disruption in Older Adults is Mediated by Tau Burden

Introduction

Chronic inflammation in aging is independently associated with tau burden and sleep disruption, though the mechanism linking inflammation with sleep disruption remains unknown. Recent evidence associates tau burden with deficits in local expression of sleep spindles and slow wave activity (SWA). Here we test the hypothesis that age-related central inflammation disrupts local sleep by influencing tau pathology.

Methods

Cognitively asymptomatic older adults from the Wisconsin Alzheimer’s Disease Research Center underwent overnight polysomnography with high-density electroencephalography (hdEEG; 256 channels) at the University of Wisconsin-Madison (n=33, 61.9±6.7 years, 23 female). EEG data were subjected to multitaper spectral analysis (0.5-40Hz) to yield topographic maps of SWA (SWA1:0.5-1Hz, SWA2:1-4.5Hz) and spindle (sigma1:11-13Hz; sigma2:13-16Hz) power during NREM sleep. Cerebrospinal fluid assay-based measurements of YKL-40 (indicating glial activation), phosphorylated tau (Ptau), and total tau (Ttau), were correlated with SWA and sigma topographical power employing Holm-Bonferroni correction. Multiple linear regression models were implemented controlling for age, apnea-hypopnea index (AHI), and sex at significant derivations. Finally, Sobel testing was employed to assess whether tau burden mediated YKL-40-sleep associations.

Results

Age was associated with YKL-40 (r=0.53, p=0.002), and YKL-40 was associated with both Ptau (r=0.66, p<0.001) and Ttau (r=0.68, p<0.001). Correlations between sigma2 activity and both Ptau and Ttau were detected at 14 derivations, 12 of which remained significant after controlling for age, sex, and AHI. YKL-40 was associated with sigma2 power (r=-0.39, p=0.025) across derivations expressing peak significance with tau. Sobel mediation analyses indicated that both Ptau (t=-2.15, p=0.031) and Ttau (t=-2.36, p=0.018) mediated the relationship between YKL-40 and sigma2 activity at these derivations. SWA was not associated with Ttau, Ptau, or YKL-40.

Conclusion

These results suggest that age-related increases in central glial activation may disrupt local expression of fast spindles by increasing tau burden, highlighting a potential role for chronic inflammation in sleep deficits observed in aging and Alzheimer’s disease.

Support

Supported by R56 AG052698, P50AG033514

0335 Frontal Expression of NREM Sleep Oscillations are Associated with Executive Function in Children and Adolescents

Introduction

The prefrontal cortex, an area known for executive functioning (including inhibition and self-monitoring) develops during childhood and adolescents, with a pattern of posterior to anterior brain development. Slow-wave activity (SWA) in NREM sleep, tracks brain development with high SWA power migrating from occipital to frontal region as brain maturation occurs. This pilot study aimed to examine whether slow wave topography is correlated with executive function in youth.

Methods

Seventeen healthy children and adolescents (ages 11-17; 10 females) underwent overnight polysomnography (PSG) with high-density electroencephalography (hdEEG). Behavior Rating Inventory of Executive Function (BRIEF) was administered to assess executive function. SWA (SWA1: 0.5-1 Hz; SWA2: 1-4.5 Hz) and spindle (slow sigma: 11-13 Hz; fast sigma: 13-16 Hz) activity was analyzed with spectral analysis using Welch’s method. BRIEF subscales of inhibition and monitor were correlated with SWA and sigma power across all derivations, with Holm-Bonferroni correction (126 channels). Significant derivations were then controlled for sex and self-reported Tanner stage using multiple regression

Results

BRIEF-Inhibition scale (i.e., ability to repress impulsivity) and SWA1 in anterior frontal derivations were negatively correlated (R2=0.58, p=0.047 corrected). BRIEF-Monitor scale (i.e., self-perception of one’s own behavior and interpersonal awareness) was negatively correlated with fast sigma in anterior frontal derivations (R2=0.65, p=0.013 corrected). These associations were significant after controlling for sex and Tanner stage.

Conclusion

These results support the hypothesis that NREM sleep oscillations are associated with executive function and reflect changes in neuroplasticity related to “back-to-front” brain maturation. Future longitudinal studies should combine multi-modal neuroimaging of brain structure and local sleep with comprehensive assessments of executive function to evaluate the possible link between local sleep and development of higher-order cognition in frontal brain regions in youth.

Support

NCATS grant #UL1TR001414 & PERC Systems Biology Fund

0325 Better Aerobic Fitness is Associated with Distinct Sleep Characteristics in Children and Adolescents - A Pilot Study

Introduction

Aerobic fitness facilitates brain synaptic plasticity, which influences global and local sleep expression. While it is known that sleep patterns/behavior and non-rapid eye movement (NREM) sleep slow wave activity (SWA) tracks brain maturation, little is known about how aerobic fitness and sleep interact during growth and development in children and adolescents. The aim of this pilot study was to characterize relationships among aerobic fitness, measures of global/local sleep expression, and habitual sleep patterns in children and adolescents. We hypothesized that greater aerobic fitness would be associated with better sleep quality, indicated by increased SWA.

Methods

Twenty healthy youth (11-17 years-old, 11 female) were evaluated during summer vacation (no school schedule constraints). Aerobic fitness (VO2peak) was measured using ramp-type progressive cycle ergometry, habitual sleep (i.e., sleep-time consistency and circadian activity patterns) was assessed with 7-day actigraphy, and ad lib sleep was evaluated during overnight polysomnography (PSG) with high-density electroencephalography (hdEEG; 128 channels). Spectral analysis was implemented to quantify SWA (0.5-4.5Hz). Data were analyzed using linear regression analyses and exploratory independent samples t-tests.

Results

Negative correlations were observed between VO2peak and sleep measures including sleep-time consistency (partial r=-0.53, p=0.045) and timing/acrophase of the circadian activity rhythm (partial r=-0.64, p=0.01) while controlling for sex and age. Additionally, after accounting for Tanner stage and sex, data demonstrated significant effects in SWA at frontal derivations (p=0.024) between low and high fitness levels at topographically specific and meaningful EEG derivations, e.g. over frontal cortex.

Conclusion

These results suggest that children and adolescents with greater fitness have less variability in sleep-times (improved sleep consistency), tend to have a more advanced circadian activity phase (i.e., go to sleep earlier), and express greater frontal SWA, supporting the hypothesis that fitness is associated with improved local and global sleep quality. Future research with larger samples is necessary to further evaluate these relationships, and to determine if interventions that improve fitness also improve sleep and related brain plasticity.

Support

NCATS grant #UL1TR001414 & PERC Systems Biology Fund