Only within resting state network connectivity of the frontoparietal control and Cingulo Opercular networks predict useful field of view performance.

Background: The Useful Field of View (UFOV) test assesses divided attention/speed of processing and relates to independent activities of daily living. Cognitive training utilizing UFOV improves speed‐of‐processing and reduces dementia risk by 29%. Neural correlates of UFOV include key regions of “higher‐order” resting state networks (RSNs) involved in attention processing and dementia risk: Frontoparietal control network (FPCN), Cingulo Opercular network (CON), Default Mode network (DMN), and Dorsal Attention network (DAN). Reduced anti‐correlation between the DMN and task positive RSNs (FPCN, CON, and DAN) is reported in dementia progression. Assessing the relationship of UFOV performance with within‐ and between‐ RSN connectivity may suggest neural mechanisms of cognitive training and potential pathways for reduced dementia risk. We hypothesize that FPCN, CON, and DAN within network connectivity will predict UFOV performance, as these networks are involved in attention processing. We also predict that DMN‐CON anti‐correlation will predict UFOV performance, as this anti‐correlation is related to dementia progression. Method: 279 healthy older adults (M = 71.6, SD = 5.1) from a larger multi‐site clinical trial were recruited. UFOV performance was measured via Double Decision (DD) task from Posit Science Brain HQ. Participants correctly identified a central target, while simultaneously specifying location of a peripheral target among distractors. Average within‐network connectivity of FPCN, CON, DAN, and DMN, and between‐network connectivity of DMN‐CON, DMN‐FPCN, and DMN‐DAN were extracted from resting‐state functional magnetic resonance imaging through the CONN Toolbox v18b via SPM 12. Multiple linear regressions predicted DD performance from within‐ and between‐network values, controlling for age, sex, education, and scanner. All between‐network values were included in the same regression model. Result: Reduced CON (β=‐.134, p=.023) and FPCN (β=‐.126, p=.029) connectivity was associated with poorer DD performance. Less DMN‐CON anticorrelation demonstrated a trend for predicting better DD performance (β=‐.126, p=.060). Conclusion: Consistent with previous research, faster DD performance is associated with RSNs involved in executive functioning, attention, and processing speed. Although only trending, reduced DMN‐CON anticorrelation may be related to DD performance. This suggests that within network connectivity may contribute more to DD performance than between‐network connectivity involved in dementia progression in healthy older adults. [ABSTRACT FROM AUTHOR]