Left frontal hub connectivity enhances task‐related brain network segregation and cognition in aging: Implications for reserve: Neuroimaging / Normal brain aging. (7th December 2020)

• Record Type: Journal Article
• Authors: Frontzkowski, Lukas; Franzmeier, Nicolai; Neitzel, Julia; Luan, Ying; Habeck, Christian; Stern, Yaakov; Ewers, Michael
• Is Part Of: Alzheimer's & dementia. Volume 16(2020)Supplement 4
• Page Start: n/a
• Page End: n/a
• Publication Date: 2020-12-07
• Subjects: Alzheimer's disease -- Periodicals; Alzheimer Disease -- Periodicals; Dementia -- Periodicals; Démence; Maladie d'Alzheimer; Périodique électronique (Descripteur de forme); Ressource Internet (Descripteur de forme); 616.83
• DOI: 10.1002/alz.045788
• Languages: English
• ISSNs: 1552-5260

Abstract

Abstract: Background: Global connectivity of the left frontal cortex (LFC), a hub of the cognitive control network, is associated with higher fluid intelligence and relatively preserved cognition despite age‐ and Alzheimer's disease (AD)‐related brain changes, rendering LFC‐connectivity a candidate substrate of reserve (Franzmeier et al. Brain 2018). Yet, the mechanisms by which LFC‐connectivity supports cognition at the functional network level remain unclear. Functional network‐segregation within the brain guards against diffuse network organization and is a fundamental brain property central to efficient cognitive processes. Given that the control network, and in particular the LFC, is thought to orchestrate activity of other functional networks (Cole, Nat Neurosci, 2013), we hypothesized that the association between LFC‐connectivity and higher cognitive abilities such as executive function is mediated via enhanced network‐segregation. Methods: We applied 3 fMRI‐tasks per cognitive domain (i.e. fluid‐reasoning, memory, speed, vocabulary) in 255 non‐demented participants from the RANN study. We determined atlas‐based (Fig.1) whole‐brain functional connectivity during task‐fMRI using psycho‐physiological‐interaction analyses. Global LFC‐connectivity was assessed as mean task‐related LFC‐connectivity to the remaining ROIs. Network‐segregation was assessed as the difference of within‐network and between‐network connectivity normalized to within‐network connectivity (Chan et al., PNAS, 2014). For the three tasks per cognitive domain, we determined domain‐specific performance, LFC‐connectivity and network segregation scores via principal‐component analysis. Using linear regression and mediation analysis, we tested for each cognitive domain whether 1) higher LFC‐connectivity was associated with better cognition 2) and subsequently whether this association was mediated by network‐segregation. Results: Higher task‐related LFC‐connectivity was associated with better performance in fluid‐reasoning (β=0.178, p=0.003, Fig.2A) but not in other cognitive domains (p>0.05). Higher LFC‐connectivity during fluid‐reasoning tasks was associated with network‐segregation (β=0.308, p<0.001, Fig.2B), which was associated with fluid‐reasoning performance (β=0.352, p<0.001, Fig.2C). In mediation analyses, we found that LFC‐connectivity effects on fluid‐reasoning performance were mediated by network‐segregation (bootstrapped mediation effect: β=0.0981, 95%CI[0.0481;0.16], p<0.001). Conclusions: LFC‐connectivity supports higher fluid reasoning performance in normal aging via enhanced brain network segregation, suggesting a potential mechanism by which LFC‐connectivity supports reserve in AD.