Elsevier

Neuropsychologia

Volume 141, April 2020, 107418
Neuropsychologia

Relationships between multiple dimensions of executive functioning and resting-state networks in adults

https://doi.org/10.1016/j.neuropsychologia.2020.107418Get rights and content

Highlights

  • Karr et al. (2018) model was replicated in a unique sample.

  • EF performance demonstrated domain-specific relationships with unique resting state networks.

  • Inhibition was related to the left striatum and attentional control network.

  • Shifting was related to the pre- and postcentral gyri and speech and sensorimotor network.

Abstract

The current study sought to examine the functional connectivity of resting state networks (RSNs) as they relate to the individual domains of executive functioning (EF). Based on the Unity and Diversity model (Miyake et al., 2000), EF performance was captured using a three-factor model proposed by Karr et al. (2018), which includes inhibition, shifting, and fluency. Publicly available data was used from the Nathan Kline Institute -Rockland project was used. Of the 722 participants who completed the Delis-Kaplan Executive Function System (D-KEFS), which was used to measure EF performance, 269 of these individuals completed resting state fMRI scans. First, a confirmatory factory analysis replicated Karr et al. (2018) revealing three components: inhibition, shifting and fluency. Next, RSNs were identified across the sample using an Independent Components Analysis (ICA) and was compared to previously established intrinsic connectivity networks (Laird et al., 2011). Finally, dual regression was used to analyze the relationships between the functional connectivity of RSNs and EF performance, which indicated that RSNs were differentially associated with inhibition and shifting. Better inhibition was related to increased connectivity between the left striatum and the attentional control network. Better shifting performance was related to increased connectivity between the pre- and postcentral gyri and the speech and sensorimotor network. These results highlight individual differences within these RSNs that are unique to the literature, as non-EF confounds are mitigated within the current measurements of EF performance.

Introduction

Empirical studies of the relationships among executive functioning (EF) measures suggest that EF can be conceptualized within the Unity and Diversity model (Miyake et al., 2000) partitioned into three correlated but distinct factors. The first factor, inhibition, is the ability to perform a desired response while suppressing an automatic or prepotent response. Second, shifting is the ability to switch between cognitive processes and avoid proactive interference. Third, updating is the ability to incorporate new information into working memory processes (Miyake and Friedman, 2012). This model of EF has been widely studied across various samples of both younger and older adults using a number of different measures from cognitive and clinical psychology (Fisk and Sharp, 2004; Fournier-Vicente, Larigauderie, & Gaonac'h, 2008; Hedden and Yoon, 2006; Hull et al., 2008; Ito et al., 2015; Latzman and Markon, 2010).

Historically, executive dysfunction was believed to derive from damage to the frontal lobe (e.g., Friedman and Miyake, 2017; Stuss and Alexander, 2000), particularly disrupting goal-directed behaviors (Stuss, 2011), or controlled processing during novel tasks (Rabbitt, 1997). Earlier theories also suggest that the frontoparietal network activates in order to maintain goal-directed attention and therefore, makes it difficult to separate regions and networks when identifying individual task activation (Duncan and Owen, 2000). More recently, studies have demonstrated that the prefrontal cortex (PFC) coordinates the use of specific subregions to carry out individual EF processes (Friedman and Miyake, 2017; Kim et al., 2012; Rottschy et al., 2012). For example, task-based fMRI studies assessing inhibition performance have identified activation within inferior frontal/insula, dlPFC, insula, and inferior parietal. Additionally, activation is also noted within the medial frontal/anterior cingulate, left premotor cortex, and left precuneus (Nee et al., 2007). Shifting-specific tasks demonstrate activation within the inferior frontal junction and posterior parietal cortex, as well as within the frontopolar cortex (Kim et al., 2012). Given the correlation among EF domain, it is not surprising that Sylvester et al. (2003) noted similar activations between both inhibition- and shifting-specific tasks, within the bilateral superior parietal cortex, left dlPFC, and medial frontal cortex. Therefore, given the wide range of areas within the PFC activated during EF tasks, as well as activation throughout other regions of the brain, it has been suggested that the PFC may serve to synchronize activation of cortical and subcortical regions that are necessary for accomplishing specific EF tasks (Friedman and Miyake, 2017). Overall, these findings also provide support for the Unity and Diversity model, suggesting that the individual EF domains are both similar and unique at the cortical level.

As an alternative to task-based activation studies, recent resting state functional connective MRI (rs-fcMRI) studies have examined the relationship between cognition, including EF, and cortical networks (Banich, 2009; Houdé et al., 2010; Reineberg et al., 2015; Reineberg, 2016). This approach measures brain activity while individuals are not engaged in goal-directed tasks, thus providing a unique perspective of network activation that is free from task-related instructions (Reineberg et al., 2018). Resting state analyses have demonstrated strong relationships with unique cortical networks that co-occur during task performances (Smith et al., 2009). In fact, Laird et al. (2011) discovered 20 intrinsic connectivity networks that are consistent during both resting and task-based neuroimaging. While identifying intrinsic connectivity networks was pioneered by Smith et al. (2009), Laird et al. (2011) extended this method by enriching the meta-data included in the analysis and by applying hierarchical clustering analysis to sort components into functionally related groupings. This approach has been viewed as providing a much more refined association of intrinsic connectivity networks with behaviors (Fox et al., 2014).

Previous work has compared RSNs to the Unity and Diversity model of EF (Miyake et al., 2000) and demonstrated both differences and similarities between cortical networks related to the individual domains of EF. For example, common EF, which is conceptualized as the variance shared across EF tasks (Miyake and Friedman, 2012; Friedman and Miyake, 2017), has demonstrated positive relationships with RSNs in the frontoparietal, somatomotor cortex (Reineberg et al., 2015), and default mode network (Reineberg et al., 2018), along with decreased connectivity within the dorsal attention network (Reineberg et al., 2018), including the precuneus, dorsolateral frontal pole, and pregenual cortex (Reineberg and Banich, 2016). During rs-fcMRI, inhibition performance has demonstrated positive relationships with the attentional control and front-parietal network (Cai & Leung, 2009; Liu et al., 2015), including left inferior frontal gyrus, left insula, ventral anterior cingulate cortex, and medial frontal gyrus (Liu et al., 2015). Shifting performance has been found to have positive relationships with the inferior frontal junction and parietal sulcus (Derrfuss et al., 2005; Reineberg and Banich, 2016; Wager et al., 2004), as well as connectivity between the angular gyrus and ventral attention RSNs (Reineberg et al., 2015, 2018; Reineberg, 2016). Lastly, the relationship between updating and other cortical networks is unclear (Friedman and Miyake, 2017), demonstrating both positive and negative relationships with the dlPFC (Frank et al., 2001; Hazy et al., 2007; Reineberg et al., 2015; Reineberg and Banich, 2016).

Relationships between RSNs and EF task performances vary within the current literature. One reason for this may derive from the use of different EF measures between studies. When quantifying or conceptualizing EF, it is important to consider the measures being used across studies because they can lead to major differences in results (Bennett and Miller, 2013). Inconsistent methods of measuring EF are a problem in the current literature that limits both research and how EF is utilized clinically (Alvarez and Emory, 2006; Nowrangi et al., 2014). For example, many studies will often use a single measure to assess EF, which may measure one or multiple domains of EF and also contain variance related to other non-EF constructs (i.e., test impurity) (Snyder et al., 2015). For example, previous studies using a single task to assess domain performance within their EF model (Reineberg et al., 2015; Reineberg, 2016) have identified different relationships between EF performance and RSNs compared to studies that have used multiple measures to assess EF domains (Reineberg et al., 2018). Although results are similar between these studies, it is less clear that the single measure studies are identifying task variance that reflects EF performance or non-EF variance that is required for successful task performance. This issue can interfere with identifying reasons for low performances on any individual EF measure.

One potential issue in using multiple measures to assess a single domain however, is that latent variables are often constructed using multiple scores from a single subtest, resulting in shared variance from both EF and non-EF constructs (Markon & Latzman, 2010). Karr et al. (2018) attempted to address some of these limitations within their three-factor EF model by using multiple measures from the Delis–Kaplan Executive Function System (D-KEFS; Delis et al., 2001), a widely cited measure of EF used in research and clinical practice (Rabin et al., 2016). For example, they identified a three-factor structure using multiple measures from the D-KEFS, where certain factors included multiple scores from similar subtests, which can increase the variance related to both EF and non-EF performance. In order to mitigate the non-EF variance, the authors regressed out method variance attributable to speed and language ability from similar subtests within a single factor, creating a “pure” measure of EF performance. Results of their study demonstrated a three-factor model from the D-KEFS, which coincided with two of the three EF domains from the Miyake model, including inhibition and shifting. Notably, the authors argue that, within the D-KEFS subtests, updating, or the act of assessing and retrieving information from long-term memory (Friedman & Miyake, 2017), may be better defined as fluency, given that the D-KEFS's subtests measure an ability to utilize working memory and lexicon to strategically access and recall information from their long-term memory. Additionally, although verbal fluency is considered a complex task tapping multiple abilities (Snyder et al., 2015), research has shown that it relates more strongly to updating compared to other aspects of executive functioning (e.g., Shao et al., 2014).

The present study attempted to build upon previous research by evaluating the relationships between the individual domains of EF and RSNs. Individual EF domains were assessed using multiple measures and implemented a previously developed scoring method (Karr et al., 2018), which aims to reduce non-EF variance (e.g., processing speed, verbal abilities). Overall, we hypothesize that this approach will identify more specific relationships between RSNs and the individual domains of performance-based EF by reducing non-EF-variance from similar tasks. Consistent with findings from previous research, it is anticipated that inhibition performance will demonstrate significant relationships with the attentional control and frontoparietal networks (Cai & Leung, 2009; Liu et al., 2015), while shifting performance will be related to somatosensory (Reineberg et al., 2015; Reineberg, 2016) and attention networks (Reineberg et al., 2015, 2018; Reineberg, 2016).

Section snippets

Participants

The enhanced Nathan Kline Institute (NKI)-Rockland project (Nooner et al., 2012) was designed to create a data repository to test existing and generate new hypotheses about psychiatric illness. Participant recruitment targeted Rockland County, New York, an area of the country with ethnic and economic diversity that parallels the United States, with the intent of increasing the generalizability of findings. Enrollment and follow-up of the NKI project is currently ongoing. For the purposes of

Results

Descriptive statistics and cognitive performance can be found in Table 1, Table 2, respectively. Results indicated that skewness and kurtosis were within normal limits, suggesting a normal distribution. No multivariate outliers were identified using Mahalanobis’ distance.

Discussion

Results of this study offer two major findings. First, neuroimaging results indicate distinct, yet related associations between the inhibition and shifting domains of EF within the resting brain. These findings provide further evidence that EF performance is related to variability within consistently identified RSNs. Specifically, improved inhibition was significantly related to increased connectivity of the left striatum with the attentional control network, whereas better shifting performance

Conclusion

Overall, the model used in the current study was designed to capture EF performance and to mitigate non-EF processes that often confound EF performance. In addition to constructing each factor using multiple measures (Snyder et al., 2015), processing speed and verbal abilities were controlled using a regression approach. These findings demonstrate that the “purer” EF domains captured by this model were uniquely related to different RSNs. While the results of this study demonstrate clear and

Ethics approval and consent to participate

All procedures performed were approved by the Institutional Review Board of the Louisiana State University. Institutional Review Board Approval was also obtained for this project at the Nathan Kline Institute (Phase I #226781 and Phase II #239708) and at Montclair State University (Phase I #000983 A and Phase II #000983 B). Written informed consent was obtained for all study participants. Written consent and assent was also obtained from minor/child participants and their legal guardian.

Availability of data and materials

The behavioral and imaging data in the current study are available from the International Neuroimaging Data Sharing Initiative (INDI) online database (http://fcon_1000.projects.nitrc.org/indi/pro/nki.html). The NKI institutional review board approved all procedures with regards to collecting and sharing data, and consent was obtained for each participant. Additional information about the image acquisition protocol can be found on the INDI website.

CRediT authorship contribution statement

Scott Roye: Conceptualization, Methodology, Formal analysis, Writing - original draft. Peter J. Castagna: Formal analysis. Matthew Calamia: Supervision, Writing - review & editing. Alyssa N. De Vito: Writing - review & editing, Investigation. Tae-Ho Lee: Supervision. Steven G. Greening: Supervision, Writing - review & editing.

Declaration of competing interest

The authors declare that they have no conflict of interest.

References (70)

  • M.T. Banich

    Executive Function: the Search for an Integrated Account. Current Directions in Psychological Science

    (2009)
  • I.S. Baron

    Delis-kaplan executive function system

    Child Neuropsychol.

    (2004)
  • C.F. Beckmann et al.

    Probabilistic Independent Component Analysis for Functional Magnetic Resonance Imaging. IEEE Transactions on Medical Imaging

    (2004)
  • C.M. Bennett et al.

    fMRI reliability: influences of task and experimental design

    Cognit. Affect Behav. Neurosci.

    (2013)
  • R.L. Buckner

    The cerebellum and cognitive function: 25 years of insight from anatomy and neuroimaging

    Neuron

    (2013)
  • W. Cai et al.

    Cortical activity during manual response inhibition guided by color and orientation cues

    Brain Res.

    (2009)
  • M. Calamia et al.

    The robust reliability of neuropsychological measures: meta-analyses of test–retest correlations

    Clin. Neuropsychol.

    (2013)
  • D. Clewett et al.

    Increased functional coupling between the left fronto‐parietal network and anterior insula predicts steeper delay discounting in smokers

    Hum. Brain Mapp.

    (2014)
  • F. Collette et al.

    Exploring the Unity and Diversity of the Neural Substrates of Executive functioning. Human Brain Mapping

    (2005)
  • D.C. Delis et al.

    Reliability and validity of the delis-kaplan executive function system: an update

    J. Int. Neuropsychol. Soc.

    (2004)
  • D.C. Delis et al.

    Delis-Kaplan Executive Function System®(D-KEFS®): Examiner's Manual: Flexibility of Thinking, Concept Formation, Problem Solving, Planning, Creativity, Impluse Control, Inhibition

    (2001)
  • B.E. Depue et al.

    The organization of right prefrontal networks reveals common mechanisms of inhibitory regulation across cognitive, emotional, and motor processes

    Cerebral Cortex

    (2015)
  • J Derrfuss et al.

    Involvement of the inferior frontal junction in cognitive control: Meta‐analyses of switching and Stroop studies

    Human Brain Mapping

    (2005)
  • Nico U.F. Dosenbach et al.

    Distinct brain networks for adaptive and stable task control in humans

    Proceedings of the National Academy of Sciences

    (2007)
  • J. Duncan et al.

    Common Regions of the Human Frontal Lobe Recruited by Diverse Cognitive Demands. Trends in Neurosciences

    (2000)
  • J.E. Fisk et al.

    Age-related impairment in executive functioning: updating, inhibition, shifting, and access

    J. Clin. Exp. Neuropsychol.

    (2004)
  • S. Fournier-Vicente et al.

    More dissociations and interactions within central executive functioning: a comprehensive latent-variable analysis

    Acta Psychol.

    (2008)
  • P.T. Fox et al.

    Meta-analysis in human neuroimaging: computational modeling of large-scale databases

    Annu. Rev. Neurosci.

    (2014)
  • M.J. Frank et al.

    Interactions between Frontal Cortex and Basal Ganglia in Working Memory: A Computational Model

    (2001)
  • N.P. Friedman et al.

    Unity and Diversity of Executive Functions: Individual Differences as a Window on Cognitive Structure

    (2017)
  • T.E. Hazy et al.

    Towards an executive without a homunculus: computational models of the prefrontal cortex/basal ganglia system

    Phil. Trans. Biol. Sci.

    (2007)
  • T. Hedden et al.

    Individual differences in executive processing predict susceptibility to interference in verbal working memory

    Neuropsychology

    (2006)
  • S. Homack et al.

    Test review: Delis-Kaplan executive function system

    J. Clin. Exp. Neuropsychol.

    (2005)
  • O. Houdé et al.

    Mapping Numerical Processing, Reading, and Executive Functions in the Developing Brain: an fMRI Meta-Analysis of 52 Studies Including 842 Children

    (2010)
  • R. Hull et al.

    Executive Function in Older Adults: A Structural Equation Modeling Approach

    (2008)
  • Cited by (10)

    • Examining patterns of executive functioning across dimensions of psychopathology

      2022, Journal of Behavior Therapy and Experimental Psychiatry
      Citation Excerpt :

      This model has been widely studied across various samples of both younger and older adults, using various EF measures to quantify performance (e.g., Karr et al., 2018; Latzman & Markon, 2010). Further supporting the multifaceted nature of EF, neuroimaging studies have found unique relationships between the individual domains of EF and areas of the brain (Friedman & Miyake, 2017; Reineberg et al., 2018; Roye et al., 2020). The diverse nature of EF has clinical relevance across multiple psychological disorders, as it influences a range of functional capacities and has been implicated as a factor for improving treatment outcomes for various mental health diagnoses.

    View all citing articles on Scopus
    View full text