Understanding the factors contributing to optimal cognitive function throughout the aging process is essential to better understand successful cognitive aging. Processing speed is an age sensitive cognitive domain that usually declines early in the aging process; however, this cognitive skill is essential for other cognitive tasks and everyday functioning. Evaluating brain network interactions in cognitively healthy older adults can help us understand how brain characteristics variations affect cognitive functioning. Functional connections among groups of brain areas give insight into the brain’s organization, and the cognitive effects of aging may relate to this large-scale organization. To follow-up on our prior work, we sought to replicate our findings regarding network segregation’s relationship with processing speed. In order to address possible influences of node location or network membership we replicated the analysis across 4 different node sets.

Data were acquired as part of a multi-center study of 85+ cognitively normal individuals, the McKnight Brain Aging Registry (MBAR). For this analysis, we included 146 community-dwelling, cognitively unimpaired older adults, ages 85-99, who had undergone structural and BOLD resting state MRI scans and a battery of neuropsychological tests. Exploratory factor analysis identified the processing speed factor of interest. We preprocessed BOLD scans using fmriprep, Ciftify, and XCPEngine algorithms. We used 4 different sets of connectivity-based parcellation: 1)MBAR data used to define nodes and Power (2011) atlas used to determine node network membership, 2) Younger adults data used to define nodes (Chan 2014) and Power (2011) atlas used to determine node network membership, 3) Older adults data from a different study (Han 2018) used to define nodes and Power (2011) atlas used to determine node network membership, and 4) MBAR data used to define nodes and MBAR data based community detection used to determine node network membership.

Segregation (balance of within-network and between-network connections) was measured within the association system and three wellcharacterized networks: Default Mode Network (DMN), Cingulo-Opercular Network (CON), and Fronto-Parietal Network (FPN). Correlation between processing speed and association system and networks was performed for all 4 node sets.

We replicated prior work and found the segregation of both the cortical association system, the segregation of FPN and DMN had a consistent relationship with processing speed across all node sets (association system range of correlations: r=.294 to .342, FPN: r=.254 to .272, DMN: r=.263 to .273). Additionally, compared to parcellations created with older adults, the parcellation created based on younger individuals showed attenuated and less robust findings as those with older adults (association system r=.263, FPN r=.255, DMN r=.263).

This study shows that network segregation of the oldest-old brain is closely linked with processing speed and this relationship is replicable across different node sets created with varied datasets. This work adds to the growing body of knowledge about age-related dedifferentiation by demonstrating replicability and consistency of the finding that as essential cognitive skill, processing speed, is associated with differentiated functional networks even in very old individuals experiencing successful cognitive aging.

Exercise elicits a variety of physiological responses in the body. In the brain, exercise can modulate levels of neurotransmitters and other neurochemicals as well as sparking neurogenesis and structural changes. Downstream psychological effects of exercise include changes in mood and cognition. These changes vary depending on the type of exercise conducted (e.g., running versus strength training). While much is known about the effects of exercise in animals and adult humans, literature on the oldest-old (>= 85 years old) is sparse. The present cross-sectional study explores the relationship between exercise and cognition in the oldest-old.

The final sample includes 194 cognitively healthy participants (106 females, MoCA mean score = 24.75) aged 85 to 99 years old (mean = 88.48). Each participant completed the Community Healthy Activities Model Program for Seniors (CHAMPS) questionnaire and a cognitive battery comprising of the NIH Toolbox, digit coding, symbol search, verbal fluency, and Stroop task. Three groups (sedentary, cardio, and cardio plus strength training) were derived from responses on CHAMPS.

The cardio plus strength training group performed significantly better on the cognitive measures compared to the sedentary group. For two measures, digit coding and symbol search, the cardio plus strength training group also performed significantly better than the cardio group. Cardio group did not significantly differ from the sedentary group on the cognitive measures. All at p < 0.05 and adjusted for multiple comparisons.

Our findings suggest exercise in the oldest-old is linked with higher fluid intelligence and better performance on cognitive measures of processing speed and that there may be an additive effect of exercise types on cognition.