Justine Megan Gatt
Neuroscience Research Australia, Sydney, NSW, Australia
School of Psychology, University of New South Wales, Sydney, NSW, Australia
Part 1 of this chapter series summarized the various ways mental well‐being has been conceptualized and measured to date, spanning the hedonic, eudaimonic, and composite well‐being perspectives. As discussed, research to date has mostly focused on defining the best ways to measure well‐being at the phenotypic level and its role in psychological health. In contrast, knowledge of the neuroscience mechanisms that underpin well‐being is still relatively lacking. In the current chapter, the core neural networks of threat, reward, and executive control and their role in mental well‐being will be discussed. Some evidence supporting these neural networks in mental well‐being is reviewed, and future developments for this work are highlighted.
Emotional and/or cognitive function are typically impaired in major mental disorders such as major depression and anxiety disorder (Williams, 2016a), and it is neural networks such as threat, reward, and executive control (both attention and cognitive control) that underscore some of the core processes of emotional and cognitive function (Williams, 2016b). These networks are therefore likely to be central to resilience to adversity (Berridge & Kringelbach, 2011; Feder, Nestler, & Charney, 2009), motivated goal‐directed behavior (Ernst & Fudge, 2009), and mental well‐being. Patterns of network activation are usually measured not only during the presentation of a specific task known to activate the regions of interest but also at rest in the absence of a task to measure “default mode” function (Greicius, Krasnow, Reiss, & Menon, 2003). It is also sometimes used as an alternative indicator of neural activity because it minimizes the typical limits in methodology that are sometimes associated with task‐activated paradigms (e.g., poor signal‐to‐noise ratio, practice effects, and habituation) (Greicius, 2008). Changes in these networks can be identified physiologically using tools such as functional magnetic resonance imaging (fMRI) and electroencephalography (EEG), each tapping into different spatial and temporal network properties, or behaviorally using neurocognitive performance tasks (Williams et al., 2008).
A breakdown in the structure and interactivity of one or more of these networks can lead to the onset of different psychiatric syndromes. For instance, patients with disorders like posttraumatic stress disorder and major depression have shown a cognitive bias toward stimuli that are negative and threatening over stimuli that are positive and rewarding (Williams, 2016a). Whereas patients with pleasure‐seeking disorders such as dependent drug use and compulsive disorders have demonstrated a bias toward stimuli that are pleasurable, combined with a lack of inhibitory behavioral control (Berridge & Kringelbach, 2011; Feder et al., 2009). Changes in neural activation have also been found at rest in different psychiatric disorders including depression, anxiety disorders, and schizophrenia (Fox & Greicius, 2010; Williams, 2016a). It could therefore be hypothesized that individuals who are flourishing will demonstrate patterns of neural activation contrary to patients with mood or anxiety disorders. It could also be predicted that an optimal state of well‐being would be associated with neural activity that is demonstrative of a bias in attention toward seeking reward over avoiding threat, in combination with superior executive control.
Humans and animals have an innate automatic survival instinct to detect threat in their environment. The experience of a potential threat triggers a cascade of neurophysiological changes commonly known as the “fight‐or‐flight” response. In the fight‐or‐flight state, the sensory cortex signals the amygdala (the emotional processing center), which then signals the hypothalamic–pituitary–adrenal (HPA) axis. These signals facilitate the response to stress. Following this feedforward response, reentrant feedback is then sent from the higher cortical areas, like the prefrontal cortex (PFC) and the hippocampus, to appraise the threat and then respond. The physiological response to the threat cue is determined by the interplay of the HPA with the autonomic nervous system (ANS) and its sympathetic and parasympathetic nervous system responses. The sympathetic nervous system triggers the fight‐or‐flight response, which includes the release of corticotropin‐releasing hormone (CRH), adrenocorticotropic hormone (ACTH), and cortisol. The consequent circulation of epinephrine through the bloodstream then triggers other physiological effects like an increase in heart rate, blood pressure, and blood glucose for the production of energy. Once the perceived threat is removed, the parasympathetic system is engaged via the vagus nerve of the nucleus ambiguus (which modulates heart rate oscillations, as measured by heart rate variability), thereby easing the effects of sympathetic activity and returning the body to a calm homeostatic state (Thayer & Lane, 2000).
Stimuli of emotional expressions are frequently used in laboratories to measure a response to threat, particularly a fearful face relative to a neutral or happy face (Williams, Palmer, Liddell, Song, & Gordon, 2006). In stress‐related disorders, fear triggers an accentuated neural circuit for threat that involves key projections of the amygdala to the medial prefrontal cortex (MPFC) (Morris et al., 1996). Patients with lesions or damage to the amygdala show specific impairments in recognizing fearful facial expressions (Broks et al., 1998). Studies in clinical cohorts of anxiety disorder patients also show that fear, more than any other emotion, activates the amygdala (Morris et al., 1996), particularly to masked (or nonconscious) rather than unmasked (or conscious) fear stimuli (P. J. Whalen et al., 1998) and also to other forms of stimuli like an unpleasant picture versus a neutral or pleasant one (Irwin et al., 1996). In animal lesion studies, the amygdala is only activated when new aversive learning is acquired but not in later stages of fear processing or behavioral expressions of anxiety (Kalin & Shelton, 2000).
Negative mood states also asymmetrically activate the PFC. Negative affect or anxiety, induced by film, increases activation in the right side of the prefrontal and anterior temporal regions (Davidson, Marshall, Tomarken, & Henriques, 2000), particularly in the right inferior PFC and right medial orbital prefrontal cortex (OFC) (Rauch, Savage, Alpert, Fischman, & Jenike, 1997). Positron emission tomography (PET) studies of glucose metabolism during evoked affective paradigms show right‐sided increases in metabolic rate in anterior orbital, inferior frontal, middle, and superior frontal gyri during induced negative affect (Sutton et al., 1997). These patterns of asymmetry have been shown in clinical patients to phobic stimuli with increased activation in the right inferior PFC and right medial OFC (Rauch et al., 1997) and in patients with bilateral lesions of the ventromedial PFC (VMPFC) who found it difficult to anticipate future negative (or positive) consequences but not immediate punishment or reward (Bechara, Damasio, Damasio, & Anderson, 1994). Asymmetry may also assist an emotional response because individuals with greater left‐sided prefrontal activity show greater recovery after the offset of a stimulus that is negative (Larson, Sutton, & Davidson, 1998).
Together, these findings highlight the role of the fight‐or‐flight response for basic survival and how it can be accentuated toward increased acute and/or chronic stress in patients with psychiatric disorders. Neuroimaging studies need to examine the same relationships in individuals who have varied levels of mental well‐being. To better understand how the threat network functions in the flourishing well‐being state, we need to investigate the changes in the structural regions of the brain, how those regions may function differently, and also how they may connect differently in interaction with each other. Future studies could also examine whether these differences are more focal to a specific region (e.g., sensory cortex alone) or whether it is the magnitude of activation to the stressor or threat response recovery time rather than overt activation that differentiates levels of well‐being separate from levels of risk (Schuyler et al., 2014).
Reward, like food and social relationships, is also necessary for survival because it induces learning. Moreover, the prospect of reward also induces goal‐directed behaviors to achieve the reward and experience subjective well‐being and pleasure. The presence of reward encourages more frequent and intense behaviors, and the absence of reward can lead to behavioral extinction (Schultz, 2000). At the neural level, reward is processed by dopamine neurons of the ventral tegmental area (VTA) and the substantia nigra. The VTA projects to the nucleus accumbens (NAc) (of the basal ganglia) through to the amygdala and hippocampus. The VTA also projects to the dorsal and ventral PFC via the mesolimbic pathway, whereas the substantia nigra projects to the caudate and putamen via the nigrostriatal pathway (McClure, York, & Montague, 2004).
A number of neuroimaging studies have started to examine associations between well‐being and the reward system. For instance, sustained engagement toward positive images in the striatum and right dorsolateral prefrontal cortex (DLPFC) is positively associated with mental well‐being, whereas aggregate striatal activity over the scan session (how fMRI is normally coded) was not (Heller et al., 2013). Both psychological and subjective well‐being independently predict sustained striatal activity, but when combined into the same model, only psychological well‐being predicted sustained activity. Individuals with greater sustained striatal activity also had lower daily cortisol output. These same patterns were also found for the DLPFC. Together, sustained activity in the striatum and DLPFC has been shown to account for 29% of the variance in well‐being. The striatum helps to anticipate and process learning associated with reward and reinforcement (Haber & Knutson, 2010).
Many studies have also observed that positive mood states asymmetrically activate the PFC. For instance, left‐sided electrical brain activity in the prefrontal and anterior temporal regions is increased by reward or positive effect induced by film across EEG and fMRI methods (Davidson, Jackson, & Kalin, 2000; Tomarken, Davidson, Wheeler, & Kinney, 1992). Reward also activates the left dorsolateral prefrontal (DLPFC) and medial orbitofrontal (OFC) regions when examined in EEG studies using source‐localization LORETA techniques—specifically within the alpha 2 sub‐band (10.5–12 Hz) (Pizzagalli, Sherwood, Henriques, & Davidson, 2005). Some studies indicate that activity in the OFC reflects the amount of reward, whereas other studies indicate that activity in the DLPFC shows the amount of reward and the consequent response in behavior (Wallis & Miller, 2003). Neural connectivity between these regions may channel behavior because OFC activation has been found to peak at 80 ms prior to DLPFC activation (Pizzagalli et al., 2005). Moreover, during the inducement of positive affect, metabolic increases in left‐sided pre‐ and postcentral gyri have been confirmed by PET studies of glucose metabolism evoked during affective paradigms (Sutton et al., 1997). These compelling patterns of asymmetry have also been confirmed in association with self‐report measures of psychological and subjective well‐being in healthy older adults (Urry et al., 2004). It is predicted that these patterns of asymmetry reflect how the PFC is integral to the organization of emotional behavior (Levenson, 1994) and affective working memory (Watanabe, 1996). Even nonhuman primates with left‐sided prefrontal activation show lower levels of cortisol and CRH levels over time, compared with animals with higher right‐sided activation (Kalin, Shelton, & Davidson, 2000).
One possible mechanism for the effects on asymmetry is the inhibitory pathway of the MPFC to the amygdala, demonstrated in both rat extinction studies (Morgan, Romanski, & LeDoux, 1993) and human metabolic studies (Abercrombie et al., 1996). Subjects with increased left prefrontal activity may possess greater inhibitory control over the amygdala when exposed to different emotional stimuli (Davidson, Jackson, et al., 2000). However, some studies failed to produce these effects (Murphy, Nimmo‐Smith, & Lawrence, 2003). It is possible however that the motivational context of the stimuli may impact the effects as positive stimuli with stronger emotive tendencies (e.g., pictures of desserts) show greater left frontal activation when compared with affective pictures, which did not always cause significant shifts in cortical activity (Harmon‐Jones & Gable, 2009). Moreover, other studies suggest that the affective component is not central to determine activity in asymmetry but that it is how the subject is motivated that distinguishes this activity. “Approach motivation” (typically measured using stimuli that display or induce positive affect or anger) is associated with greater left frontal cortical activity, whereas “withdrawal motivation” (fear stimuli) is associated with greater right frontal activity (Harmon‐Jones, Gable, & Peterson, 2010).
Within the limbic region, studies suggest associations between reward processing and activity within the amygdala and posterior cingulate cortex (PCC). In particular, the amygdala has been shown to respond to stimuli that are appetitive as well as stimuli that are threatening (Hamann, Ely, Hoffman, & Kilts, 2002). Asymmetrical patterns in the amygdala may affect responses to stimuli that are positive; increased activity in the right amygdala may be specific to stimuli that are unpleasant or to the removal of a reward (such as losing money), whereas increased activity in the left amygdala occurs with the attainment of reward (winning money) (Zalla et al., 2000). Studies using event‐related potentials (ERPs) also highlight the role of the PCC in positive affect. When individuals were shown pleasant and unpleasant pictures of varied intensity, individuals with high extraversion (associated with optimism and high subjective well‐being) reacted more to pleasant pictures regardless of emotional intensity and were more resilient to mildly unpleasant pictures compared with individuals with low extraversion (Yuan et al., 2012). Source modeling suggests that this enhanced brain sensitivity to pleasant events and resistance to the impact of unpleasant events are generated by the bilateral PCC (Yuan et al., 2012), which, together with their collected regions (including fusiform gyrus, amygdala, and basal ganglia), are critical to regulate positive emotion (Amin, Constable, & Canli, 2004).
Physiological responses are also fundamental to the reward network, like the acceleration of the heart through the ANS, which is normally heightened with affect intensity but is not necessarily emotion specific (Watson, Wiese, Vaidya, & Tellegen, 1999). For instance, in some studies, increases in positive affect were associated with reduced ambulatory heart rate over the day (Steptoe, Wardle, & Marmot, 2005). In contrast, other studies reported increased cardiovascular function (higher heart rate and blood pressure) with increased positive affect (Schwartz, Warren, & Pickering, 1994), but this may be due to physical activation caused by interpersonal interactions or a greater engagement with challenge (Maier, Waldstein, & Synowski, 2003). Other studies have suggested an “undoing hypothesis”: that positive emotions may speed the recovery from cardiovascular reactivity caused by negative emotions, thereby returning the body to a normal level of activation (Fredrickson, Mancuso, Braniqan, & Tugade, 2000). In these experimental studies, Fredrickson showed that participants who viewed clips of positive emotion following a clip of fear had the fastest cardiovascular recovery compared with those who were showed neutral or sad clips (Fredrickson & Levenson, 1998). Fredrickson and her colleagues interpreted this effect according to their broaden‐and‐build theory, which states that positive emotions broaden one's “thought–action repertoire.” A thought–action repertoire describes acts that are triggered by specific thoughts or emotions. For example, joy creates the urge to play, whereas fear creates the urge to escape (Fredrickson, 2004). Therefore, positive emotions may help to move one from a thought–action repertoire that is narrow and triggered by emotions that are negative, quick, and direct and encourage escape, attack, or expel toward a thought–action repertoire that is broader and expands the scope of one's tendencies to act. These broader thought–action repertoires could be used to explore, engage in novel experiences, promote self‐growth, and expand mechanisms that can help one to cope with stress. Therefore, positive mindsets are thought to possess long‐term benefits because they can develop personal resources that are adaptive and resilient (Fredrickson, 2004).
In summary, these findings highlight the importance of reward to survival because it encourages learning and induces behaviors that are goal directed to attain (or maintain) higher levels of well‐being and pleasure. Evidence from neuroimaging and EEG studies has started to support the role of reward in well‐being. This evidence is in stark contrast to the lack of studies that examine the relationship between well‐being and threat. Evidence suggests several key regions in the reward network may be associated with increased flourishing. These include increases in striatal and DLPFC activity in response to reward, as well as asymmetric activation of the left PFC (in particular the DLPFC and OFC), both of which are associated with increased well‐being. The relationship between well‐being and patterns in asymmetric left PFC activation may reflect an increased inhibitory control over the amygdala when subjected to different emotions. If this is true, we may predict that individuals with elevated well‐being would also possess similar levels of inhibitory control over the amygdala when exposed to threat. The autonomic evidence similarly suggests that the impact of positive emotions on cardiovascular recovery times is modulated following threat exposure. Future neuroscience studies should explore associations between mental well‐being and both reward and threat networks to assess their differences and how they may interact.
Executive control is a broad term used to outline the core functions of working memory, response inhibition, and set shifting (Friedman et al., 2008). Executive control, also called the central executive network (CEN), coordinates mental activity to achieve goals and to focus attention on tasks. In contrast, when the brain is at rest and not focused on a task (e.g., when relaxing on a beach), the mind‐wandering or “default mode” network (DMN) is activated (Raichle et al., 2001). The circuits central to executive control include activity in the ventral and lateral prefrontal cortex (VLPFC and DLPFC), the inferior parietal cortex to tasks that are inhibitory or require sustained attention (e.g., the go/no‐go or n‐back working memory tasks), and reduce activation of the motor cortex to enable motor response control (Kelly et al., 2004).
The executive control network also processes and regulates emotion and often involves the same regions with overlapping functions. For instance, the DLPFC is used for attention (Corbetta, Patel, & Shulman, 2008) and working memory (Owen, McMillan, Laird, & Bullmore, 2005) and for regulating emotion (Kalisch, 2009). Various regions of the ACC are thought to respond differently to cues that involve cognition versus emotion. For example, the dorsal ACC is activated more during cognitive processing and top‐down/bottom‐up connections to the PFC (measured with the classical Stroop), whereas the ventral ACC and amygdala–hypothalamic circuits are activated more during emotional cognitive tasks (i.e., the emotional Stroop) (P. Whalen et al., 1998). The hippocampus regulates context, in particular memory that is declarative and context dependent (Davidson, 2000). Loss in the hippocampus is frequently reported in patients with anxiety or depressive disorder (Bremner et al., 2000), and the impact of this hippocampal loss is thought to affect contextual learning and memory where otherwise “normal” emotional responses are activated in inappropriate contexts. Another key region involved in higher‐order regulation is the insula. The insula has been observed in functions involving higher‐level cognitive control and attention, including a regulatory role in emotional and empathic processes (Menon & Uddin, 2010). The insula's sensitivity to salient events (together with the ACC) suggests that these regions are a “neural switchboard” switching between the default and the executive control modes to access attention and memory and guide behavior (Menon & Uddin, 2010). In wellness studies, positive correlations have been reported between the right insula and purpose in life, positive relations, and personal growth in a cross‐sectional cohort of healthy participants (Lewis, Kanai, Rees, & Bates, 2013). Other studies have also associated the right insula with self‐awareness (Craig, 2009), regulation of bodily states (Singer, Critchley, & Preuschoff, 2009), and agentic control (Lee & Reeve, 2012). However, many of these studies are correlational. Therefore, to predict outcomes in well‐being over time, future studies need to investigate how different regions activate and respond during tasks that involve emotional and executive control.
Studying how cognition interacts with emotion may also help us to understand the links between executive control, threat, and reward. Generally speaking, we may anticipate that people who are in a happy and relaxed mood would demonstrate better cognitive performance than someone who is angry or sad. Fredrickson's (Fredrickson, 2001) broaden‐and‐build theory suggests that positive emotions broaden an individual's thought–action repertoires and expand personal resources like cognitive functioning (e.g., executive control, knowledge) and psychological resources (e.g., resilience, creativity), which may lead to more enduring states of well‐being. These newly expanded resources can then be drawn on later in times of coping or other cognitive challenges. The broaden hypothesis has been supported in studies that induce positive affect because positive emotions broaden the scope of attention and the size of the thought–action repertoire relative to a neutral state (Fredrickson & Branigan, 2005). Positive emotions are associated with cognitive processes that are more creative, flexible, and expansive (Murray, Sujan, Hirt, & Sujan, 1990) and increase the preference for varied behavioral responses (Kahn & Isen, 1993). Positive emotions also improve persuasive communication at different stages of memory processing: at encoding, participants in a positive mood simplify their message and are persuaded equally by both weak and strong arguments, and at retrieval, participants in a positive mood use more simplified processing strategies by drawing on global rather than detailed evaluations of the message to assess validity (Bless, Mackie, & Schwarz, 1992). Positive affect facilitates cognitive flexibility toward specific tasks such as word association (Isen, Johnson, Mertz, & Robinson, 1985), word fluency (Green & Noice, 1988), and specific problem solving (Ashby, Isen, & Turken, 1999). Usually, increased cognitive flexibility is only detected when the situation is neutral or positive in emotional content or is minimally engaging (Ashby et al., 1999). Furthermore, the presence of positive affect can help a person to cope with negative events because defensive behaviors are reduced (Aspinwall, 1998). Because these studies have used positive affect that is low in approach motivation, it is also important to consider the role of cues that trigger approach motivation such as reproduction, social attachment, and the ingestion of food and water. In contrast, breadth of attention is reduced in an individual with positive affect and high approach motivation because of their attempt to achieve the desired goal (Gable & Harmon‐Jones, 2008).
In summary, this chapter has discussed how the three networks of threat, reward, and executive control may interact to influence well‐being. Neuroimaging studies of well‐being have started to emerge, most particularly for circuits of reward and for specific measures of subjective well‐being. Few studies have considered the direct links between well‐being and the circuits of threat or executive control. For threat, we predict that the response to stimuli that are negative would attenuate. Similarly for executive control, we predict that increased well‐being would be associated with more superior executive control processes that include working memory, response inhibition, and set shifting. Furthermore, we predict that elevated executive control would also make an integrative impact on the processing of emotion for threat and reward. Future neuroscience studies need to examine both the structural morphology of different brain regions using different imaging techniques (e.g., MRI, DTI) and how well different regions of the brain and their connections function during both task‐activated and resting paradigms. For a complete and composite view of both structure and function, a combination of functional imaging and electrophysiological techniques (i.e., fMRI and EEG) may also help to combine high spatial resolution with high temporal resolution.
In addition to more comprehensive neuroimaging assessments, future studies also need to incorporate phenotypic assessments of well‐being that encompass both subjective and psychological aspects of well‐being. The COMPAS‐W well‐being scale (Gatt, Burton, Schofield, Bryant, & Williams, 2014) is one such example that could be used in future studies as anall‐inclusive index of well‐being. This same measure could also be used in studies of resilience whereby the trajectory of well‐being scores and mental health over time could be tracked in individuals exposed to a childhood or adult trauma or adversity (e.g., the experience of neglect, abuse, war, natural disasters, personal illness or injury, and other major life events). Resilience defines the process of adaptive recovery from adversity and the ability to maintain optimal levels of well‐being in the face of dynamic challenges to psychological resources (APA, 2010). Therefore, while mental well‐being defines a state of positive mental health, resilience defines the process of returning to that state following adversity or trauma exposure. In this sense, we could predict that some of the same neural and biological mechanisms that underpin well‐being may also contribute to resilience. This does not however necessarily implicate that someone who is flourishing at any given point in time would also be resilient should they be exposed to serious adversity; to demonstrate resilience would necessitate the presence of additional factors that buffer against the trauma (e.g., genetic, environmental, or other socio‐ecological factors). The characterization of such predictive factors of resilience is still yet to be determined (see Alexander and Gatt, In Press for further discussion).
Many of the well‐being (or resilience) studies to date are cross‐sectional and/or correlational, which limits interpretation in terms of cause and effect. Longitudinal and experimental studies are required to better understand these underlying networks. For instance, causality could be more easily established in studies that consider whether interventions that improve executive function or other functions (albeit, via behavioral or cognitive training or neurofeedback strategies) also promote increased well‐being and resilience. The first step however is to understand the neural mechanisms of well‐being in order to develop strategies that optimize these features. Should evidence support the elementary networks of threat, reward, and executive control in well‐being, then it may be arguable that interventions would need to promote adaptive responses to threat and perceived stress, a positive outlook, autonomy, higher‐order cognitive capacities to master the regulation of emotion and behavior, and the ability to set goals and strive. When viewed as a whole, these strengths contribute to an overall sense of subjective and psychological well‐being (Gatt et al., 2014). This framework also complements Fredrickson's “broaden‐and‐build” theory that similarly suggests that “…positive emotions broaden individuals' thought‐action repertoires, enabling them to draw flexibly on higher‐level connections and wide‐than‐usual ranges of precepts, ideas, and action urges; broadened cognition in turn creates behavioral flexibility that over time builds personal resources, such as mindfulness, resilience, social closeness, and even physical health” (p. 3) (Garland et al., 2010). When mapping the trajectory of mental health outcomes in response to such health promotion strategies, or simply over time, it is important to keep in mind the modulating influence of basic demographic variables such as age and sex. The outcome of such studies can then help shape the types of interventions that will promote well‐being and resilience and the specific individuals to target, particularly in more vulnerable cohorts such as children and adolescents.
Understanding mental well‐being is paramount to psychological and psychiatric research and mental health promotion in the general population. Much work is needed to understand the neurobiological mechanisms of well‐being so that these essential qualities can be promoted in mental health programs in children and adults alike. This chapter has provided an outline of possible neural networks involved in well‐being and some evidence to date, as well as possible future developments to aid progress in mental health and mental illness research, policy, and treatment.
Dr. Justine Gatt is currently supported by a NHMRC CDF Fellowship APP1062495.
Dr. Justine Megan Gatt is a group leader and senior research scientist at Neuroscience Research Australia and at the School of Psychology at the University of New South Wales, Australia. Dr. Gatt leads a research program focusing on the neuroscience and gene–environment mechanisms of resilience and mental well‐being across adult and youth groups. This work incorporates various genetic, neuroimaging, and neurocognitive techniques and various methodological designs (twin/singleton studies, longitudinal, and intervention work), with the ultimate aim being to understand and promote optimal mental health in the general population.