The interdisciplinary research conducted in the NET Lab is centered on answering the following key questions about the nature of internally-guided cognition:
1) What are the psychological and neural mechanisms underlying spontaneous and deliberate internally-guided cognition?
2) What are the factors that determine whether internal thoughts can be helpful versus unhelpful?
3) How are internally-guided processes regulated by top-down control mechanisms, and what is the role of executive function in mind-wandering?
4) How do internally-guided processes change throughout the lifespan?
5) How do internally-guided processes become dysfunctional in mental health disorders such as depression and anxiety?
6) Can we improve the quality of our internally-guided thoughts with interventions such as mindfulness meditation or exercise?
Andrews-Hanna, The Neuroscientist, 2012
Importantly, when people daydream on a day-to-day basis, the subsystems are activated together and form the basis of default network activity during passive states. Our research has linked activity throughout the default network to mind-wandering, and has shown that autobiographical thoughts about the past and future are associated with increased resting state functional connectivity in the medial temporal subsystem (Andrews-Hanna et al., J. Neurophysiology, 2010). At the University of Colorado Boulder, we recently illustrated the complexity of self-generated thoughts in daily life by delineating that such thoughts are comprised of multiple interacting component processes (Andrews-Hanna et al., Frontiers in Psychology - special issue on mind-wandering, 2013). We further revealed that participants’ thoughts are characterized by a positive constructive nature, providing support for their largely adaptive nature.
THE HETEROGENEOUS NATURE OF SELF-GENERATED THOUGHT
Andrews-Hanna et al., Frontiers in Psychology - Special Issue on Mind-Wandering, 2013
At the University of Colorado Boulder, we have been conducting several studies to further investigate these topics, including the adaptive nature of spontaneous thought, the precise functions of specific regions within the default network, and the neural predictors of compassionate behavior. See below for our model of the role of the dorsal medial subsystem and the aMPFC in compassionate behavior.
A NEUROPSYCHOLOGICAL MODEL OF COMPASSION
Ashar et al., Positive Psychology chapter, 2013
DYNAMIC CONTROL OF INTERNALLY-GUIDED COGNITION
Despite initial progress towards understanding the default network and internally-guided cognition, much less is known about how the network and its respective functions are regulated by top-down control. How are its functions enhanced, as when one plans his/her upcoming activities or reflects on past events? In contrast, how are its functions suppressed, as when one’s thoughts (i.e. “daydreams” or “ruminations”) have the potential to disrupt the task at hand or interfere with one’s well-being? Though a growing number of studies have begun to characterize the neural basis of top-down control of externally-directed attention, the mechanisms supporting the regulation of internally-guided cognition are far less understood. Multiple lines of work seek insight into these questions.
CHANGES IN INTERNALLY-GUIDED COGNITION ACROSS THE LIFESPAN
A long-standing interest of Dr. Jessica Andrews-Hanna involves understanding the neural underpinnings of cognitive change across the lifespan. What predicts whether one child is more impulsive than another, or whether an older adult suffers from more or less memory loss than his/her peers? In what ways are the structural and functional changes associated with cognitive development similar to those that accompany old age? How does the frequency and nature of spontaneous thoughts change across the lifespan?
At the University of Colorado, we showed that adolescents prefer a reactive approach, implementing control on a when-needed basis, selectively for difficult trials (Andrews-Hanna et al., PLOS One, 2011). Adults, on the other hand, prefer a preparatory “proactive” approach, implementing control in a sustained fashion across blocks of easy and difficult trials. Interestingly, individual differences in neural development of proactive control predicted cognitive, social, and prospective behavior in the real world.
DEVELOPMENTAL TRAJECTORIES OF PROACTIVE CONTROL IN ADOLESCENCE
We also extended these findings to delay discounting paradigms, revealing that hypoactivity of frontoparietal attention and default systems in adolescents predict short-sighted decisions (Banich et al., Psych. Addictive Behavior, 2013).
Earlier work revealed that typical aging is linked to disruptions in white matter integrity as well as widespread disruption in functional coordination amongst regions within large-scale brain networks (including the default network and dorsal attention network), as well as between networks (Andrews-Hanna et al., Neuron, 2007). These age-associated structural and functional alterations predict cognitive impairment in memory, executive function, and processing speed. Interestingly, older individuals with fewer structural and functional alterations exhibit improved cognitive performance compared to their age-matched peers.
DISRUPTION OF LARGE-SCALE NETWORKS IN HEALTHY AGING
Andrews-Hanna et al., Neuron, 2007
NEUROCOGNITIVE BASIS OF DYSFUNCTIONAL THOUGHT
While much of our research provides support for the adaptive functions of internally-guided thought, the experience can also be associated with significant costs, disrupting task performance and hindering psychological well-being. An important direction of ongoing and future work in the NET lab involves understanding the factors that underlie individual variability in the consequences of self-generate thought. We recently proposed that the ability to regulate the content of one’s internal thoughts as well as the context during which such thoughts occur are two important factors that constrain the costs and benefits of the experience (Smallwood & Andrews-Hanna, Frontiers in Psychology, 2013; Andrews-Hanna, Smallwood & Spreng, ANYAS, 2014). Providing support for the content regulation hypothesis, we recently showed that different aspects of thought content predict different mental health constructs. For example, individuals who characterized their thoughts as more negative and more personally-significant exhibited elevated depressive symptoms, while those who characterized their thoughts as less specific exhibited elevated symptoms of rumination. Interestingly, individuals who were more mindful rated their thoughts as more positive, less personally-significant. These results are interesting, particularly in light of findings that mindfulness-based therapeutic interventions can improve symptoms of distress by training people to become less attached to their thoughts and view them for what they are rather than the literal truth.
THE CONTENT OF SELF-GENERATED THOUGHTS PREDICT DIFFERENT
MENTAL HEALTH CONSTRUCTS
INTERNALLY-GUIDED COGNITION AND THE BRAIN’S DEFAULT NETWORK
The default network (pictured below) is a set of midline and lateral regions that we suspect plays an important role in internally-directed, or “self-generated,” cognitive processes including self-referential decision making, theory of mind, memory, prospection, and valuation. Perhaps not surprisingly, this network also activates when our attention shifts away from the external world and our mind begins to wander (i.e. Fox et al., Neuroimage, 2015). Our work reveals that the default network comprises distinct subsystems that interact through common hubs (Andrews-Hanna et al., Neuron, 2010). We propose that a dorsal medial cortex subsystem supports meta-cognitive aspects of internally-guided cognition, allowing us to reflect upon our own and others’ mental states, while a medial temporal subsystem facilitates construction of imagined scenes based on memory. Finally, anterior medial prefrontal cortex and posterior cingulate cortex, important “hubs” within the network, might help us assess the personal significance of incoming information, allowing us to integrate our current mental states with prior conceptual and episodic knowledge into an overarching personal meaning. See our functional model in the picture below (Andrews-Hanna, The Neuroscientist, 2012; see also Andrews-Hanna, Smallwood & Spreng, ANYAS, 2014).
A FUNCTIONAL MODEL OF THE DEFAULT NETWORK
Andrews-Hanna et al., Frontiers in Psychology - Special Issue on Mind-Wandering, 2013
Ongoing work at the University of Colorado Boulder explores internally-guided thoughts and the default network in depression and social anxiety. My colleagues (especially Marina Lopez-Sola, Joanna Arch, Sona Dimidjian, and Tor Wager) and I have several studies underway related to these questions, so stay tuned!
Earlier work of Dr. Jessica Andrews-Hanna and colleagues involved exploring the structural and functional correlates of individual differences in cognitive performance and symptomatology in individuals with schizophrenia (Andrews et al., Am J Psychiatry, 2006). Later, she and her colleagues at Harvard Medical School and MIT used task-related fMRI and resting-state functional connectivity to reveal that individuals with schizophrenia rely more heavily on posterior cortical midline regions than controls while reflecting on one’s self. Group differences in task-related activity were also accompanied by alterations in resting-state functional connectivity (Holt et al., Biological Psychiatry, 2011). More recently, she began collaborating with Dr. Vijay Mittal’s lab to explore risk factors for schizophrenia by investigating functional networks in individuals at risk for psychosis.