AUTHOR=Neufang Susanne , Akhrif Atae 

TITLE=Regional Hurst Exponent Reflects Impulsivity-Related Alterations in Fronto-Hippocampal Pathways Within the Waiting Impulsivity Network

JOURNAL=Frontiers in Physiology

VOLUME=11

YEAR=2020

URL=https://www.frontiersin.org/journals/physiology/articles/10.3389/fphys.2020.00827

DOI=10.3389/fphys.2020.00827

ISSN=1664-042X

ABSTRACT=<p>In general, the Hurst exponent. is used as a measure of long-term memory of time series. In previous neuroimaging studies, <italic>H</italic> has been introduced as one important parameter to define resting-state networks, reflecting upon global scale-free properties emerging from a network. <italic>H</italic> has been examined in the waiting impulsivity (WI) network in an earlier study. We found that alterations of <italic>H</italic> in the anterior cingulate cortex (<italic>H</italic><sup><italic>A</italic><italic>C</italic><italic>C</italic></sup>) and the nucleus accumbens (<italic>H</italic><sup><italic>N</italic><italic>A</italic><italic>c</italic><italic>c</italic></sup>) were lower in high impulsive (highIMP) compared to low impulsive (lowIMP) participants. Following up on those findings, we addressed the relation between altered fractality in <italic>H</italic><sup><italic>A</italic><italic>C</italic><italic>C</italic></sup> and <italic>H</italic><sup><italic>N</italic><italic>A</italic><italic>c</italic><italic>c</italic></sup> and brain activation and neural network connectivity. To do so, brain activation maps were calculated, and network connectivity was determined using the Dynamic Causal Modeling (DCM) approach. Finally, 1–<italic>H</italic> scores were determined to quantify the alterations of <italic>H</italic>. This way, the focus of the analyses was placed on the potential effects of alterations of <italic>H</italic> on neural network activation and connectivity. Correlation analyses between the alterations of <italic>H</italic><sup><italic>A</italic><italic>C</italic><italic>C</italic></sup>/<italic>H</italic><sup><italic>N</italic><italic>A</italic><italic>c</italic><italic>c</italic></sup> and activation maps and DCM estimates were performed. We found that the alterations of <italic>H</italic> predominantly correlated with fronto-hippocampal pathways and correlations were significant only in highIMP subjects. For example, alterations of <italic>H</italic><sup><italic>A</italic><italic>C</italic><italic>C</italic></sup> was associated with a decrease in neural activation in the right HC in combination with increased ACC-hippocampal connectivity. Alteration in<italic>H</italic><sup><italic>N</italic><italic>A</italic><italic>c</italic><italic>c</italic></sup>, in return, was related to an increase in bilateral prefrontal activation in combination with increased fronto-hippocampal connectivity. The findings, that the WI network was related to <italic>H</italic> alteration in highIMP subjects indicated that impulse control was not reduced <italic>per se</italic> but lacked consistency. Additionally, <italic>H</italic> has been used to describe long-term memory processes before, e.g., in capital markets, energy future prices, and human memory. Thus, current findings supported the relation of <italic>H</italic> toward memory processing even when further prominent cognitive functions were involved.</p>